system
stringclasses
7 values
input
stringlengths
10
94.5k
output
stringlengths
14
55.5k
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_0002ca4f_00000000-6_dr_model_poc.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2060: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2060: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z42__device_stub__Z15gpu_matrix_multPiS_S_iiiPiS_S_iii .type _Z42__device_stub__Z15gpu_matrix_multPiS_S_iiiPiS_S_iii, @function _Z42__device_stub__Z15gpu_matrix_multPiS_S_iiiPiS_S_iii: .LFB2082: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movq %rdi, 40(%rsp) movq %rsi, 32(%rsp) movq %rdx, 24(%rsp) movl %ecx, 20(%rsp) movl %r8d, 16(%rsp) movl %r9d, 12(%rsp) movq %fs:40, %rax movq %rax, 168(%rsp) xorl %eax, %eax leaq 40(%rsp), %rax movq %rax, 112(%rsp) leaq 32(%rsp), %rax movq %rax, 120(%rsp) leaq 24(%rsp), %rax movq %rax, 128(%rsp) leaq 20(%rsp), %rax movq %rax, 136(%rsp) leaq 16(%rsp), %rax movq %rax, 144(%rsp) leaq 12(%rsp), %rax movq %rax, 152(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) movl $1, 72(%rsp) movl $1, 76(%rsp) movl $1, 80(%rsp) movl $1, 84(%rsp) leaq 56(%rsp), %rcx leaq 48(%rsp), %rdx leaq 76(%rsp), %rsi leaq 64(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 168(%rsp), %rax subq %fs:40, %rax jne .L8 addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 56(%rsp) .cfi_def_cfa_offset 200 pushq 56(%rsp) .cfi_def_cfa_offset 208 leaq 128(%rsp), %r9 movq 92(%rsp), %rcx movl 100(%rsp), %r8d movq 80(%rsp), %rsi movl 88(%rsp), %edx leaq _Z15gpu_matrix_multPiS_S_iii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 192 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _Z42__device_stub__Z15gpu_matrix_multPiS_S_iiiPiS_S_iii, .-_Z42__device_stub__Z15gpu_matrix_multPiS_S_iiiPiS_S_iii .globl _Z15gpu_matrix_multPiS_S_iii .type _Z15gpu_matrix_multPiS_S_iii, @function _Z15gpu_matrix_multPiS_S_iii: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z42__device_stub__Z15gpu_matrix_multPiS_S_iiiPiS_S_iii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _Z15gpu_matrix_multPiS_S_iii, .-_Z15gpu_matrix_multPiS_S_iii .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC3: .string "Avg. Latency: %g ms :: Avg. Throughput: %g examples/sec\n" .text .globl main .type main, @function main: .LFB2057: .cfi_startproc endbr64 pushq %r12 .cfi_def_cfa_offset 16 .cfi_offset 12, -16 pushq %rbp .cfi_def_cfa_offset 24 .cfi_offset 6, -24 pushq %rbx .cfi_def_cfa_offset 32 .cfi_offset 3, -32 subq $240, %rsp .cfi_def_cfa_offset 272 movq %fs:40, %rax movq %rax, 232(%rsp) xorl %eax, %eax movl $3333, %edi call srand@PLT leaq 32(%rsp), %rdi movl $262144, %esi call cudaMallocHost@PLT leaq 40(%rsp), %rdi movl $1073741824, %esi call cudaMallocHost@PLT leaq 48(%rsp), %rdi movl $16384, %esi call cudaMallocHost@PLT leaq 56(%rsp), %rdi movl $16777216, %esi call cudaMallocHost@PLT leaq 64(%rsp), %rdi movl $4096, %esi call cudaMallocHost@PLT leaq 72(%rsp), %rdi movl $163840, %esi call cudaMallocHost@PLT leaq 80(%rsp), %rdi movl $40, %esi call cudaMallocHost@PLT movl $16384, %ebp .L12: leaq -16384(%rbp), %rbx .L13: call rand@PLT cltd shrl $22, %edx addl %edx, %eax andl $1023, %eax subl %edx, %eax movq 40(%rsp), %rdx movl %eax, (%rdx,%rbx) addq $4, %rbx cmpq %rbp, %rbx jne .L13 addq $16384, %rbp cmpq $1073758208, %rbp jne .L12 movl $4096, %ebp .L14: leaq -4096(%rbp), %rbx .L15: call rand@PLT cltd shrl $22, %edx addl %edx, %eax andl $1023, %eax subl %edx, %eax movq 56(%rsp), %rdx movl %eax, (%rdx,%rbx) addq $4, %rbx cmpq %rbp, %rbx jne .L15 addq $4096, %rbp cmpq $16781312, %rbp jne .L14 movl $40, %ebp jmp .L16 .L34: addq $40, %rbp cmpq $163880, %rbp je .L33 .L16: leaq -40(%rbp), %rbx .L17: call rand@PLT cltd shrl $22, %edx addl %edx, %eax andl $1023, %eax subl %edx, %eax movq 72(%rsp), %rdx movl %eax, (%rdx,%rbx) addq $4, %rbx cmpq %rbx, %rbp jne .L17 jmp .L34 .L33: leaq 88(%rsp), %rdi call cudaEventCreate@PLT leaq 96(%rsp), %rdi call cudaEventCreate@PLT leaq 104(%rsp), %rdi movl $262144, %esi call cudaMalloc@PLT leaq 112(%rsp), %rdi movl $1073741824, %esi call cudaMalloc@PLT leaq 120(%rsp), %rdi movl $16384, %esi call cudaMalloc@PLT leaq 128(%rsp), %rdi movl $16777216, %esi call cudaMalloc@PLT leaq 136(%rsp), %rdi movl $4096, %esi call cudaMalloc@PLT leaq 144(%rsp), %rdi movl $163840, %esi call cudaMalloc@PLT leaq 152(%rsp), %rdi movl $40, %esi call cudaMalloc@PLT movl $1, %ecx movl $1073741824, %edx movq 40(%rsp), %rsi movq 112(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movl $16777216, %edx movq 56(%rsp), %rsi movq 128(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movl $163840, %edx movq 72(%rsp), %rsi movq 144(%rsp), %rdi call cudaMemcpy@PLT movl $1024, %ebp movq $0x000000000, 8(%rsp) leaq 28(%rsp), %r12 jmp .L19 .L35: movl $4096, %r9d movl $65536, %r8d movl $1, %ecx movq 120(%rsp), %rdx movq 112(%rsp), %rsi movq 104(%rsp), %rdi call _Z42__device_stub__Z15gpu_matrix_multPiS_S_iiiPiS_S_iii jmp .L21 .L36: movl $1024, %r9d movl $4096, %r8d movl $1, %ecx movq 136(%rsp), %rdx movq 128(%rsp), %rsi movq 120(%rsp), %rdi call _Z42__device_stub__Z15gpu_matrix_multPiS_S_iiiPiS_S_iii jmp .L22 .L23: movl $2, %ecx movl $4096, %edx movq 152(%rsp), %rsi movq 80(%rsp), %rdi call cudaMemcpy@PLT call cudaThreadSynchronize@PLT movl $0, %esi movq 96(%rsp), %rdi call cudaEventRecord@PLT movq 96(%rsp), %rdi call cudaEventSynchronize@PLT movq 96(%rsp), %rdx movq 88(%rsp), %rsi movq %r12, %rdi call cudaEventElapsedTime@PLT pxor %xmm0, %xmm0 cvtss2sd 28(%rsp), %xmm0 addsd 8(%rsp), %xmm0 movsd %xmm0, 8(%rsp) subl $1, %ebp je .L24 .L19: movl $0, %ebx .L20: call rand@PLT cltd shrl $22, %edx addl %edx, %eax andl $1023, %eax subl %edx, %eax movq 32(%rsp), %rdx movl %eax, (%rdx,%rbx) addq $4, %rbx cmpq $262144, %rbx jne .L20 movl $0, %esi movq 88(%rsp), %rdi call cudaEventRecord@PLT movl $1, %ecx movl $262144, %edx movq 32(%rsp), %rsi movq 104(%rsp), %rdi call cudaMemcpy@PLT movl $256, 160(%rsp) movl $1, 164(%rsp) movl $16, 172(%rsp) movl $16, 176(%rsp) movl $0, %r9d movl $0, %r8d movq 172(%rsp), %rdx movl $1, %ecx movq 160(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L35 .L21: call cudaDeviceSynchronize@PLT movl $64, 184(%rsp) movl $1, 188(%rsp) movl $16, 196(%rsp) movl $16, 200(%rsp) movl $0, %r9d movl $0, %r8d movq 196(%rsp), %rdx movl $1, %ecx movq 184(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L36 .L22: movl $1, 208(%rsp) movl $1, 212(%rsp) movl $16, 220(%rsp) movl $16, 224(%rsp) movl $0, %r9d movl $0, %r8d movq 220(%rsp), %rdx movl $1, %ecx movq 208(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax jne .L23 movl $10, %r9d movl $4096, %r8d movl $1, %ecx movq 152(%rsp), %rdx movq 144(%rsp), %rsi movq 120(%rsp), %rdi call _Z42__device_stub__Z15gpu_matrix_multPiS_S_iiiPiS_S_iii jmp .L23 .L24: movsd 8(%rsp), %xmm3 movapd %xmm3, %xmm0 mulsd .LC2(%rip), %xmm0 movsd .LC1(%rip), %xmm1 divsd %xmm3, %xmm1 leaq .LC3(%rip), %rsi movl $2, %edi movl $2, %eax call __printf_chk@PLT movq 104(%rsp), %rdi call cudaFree@PLT movq 112(%rsp), %rdi call cudaFree@PLT movq 120(%rsp), %rdi call cudaFree@PLT movq 128(%rsp), %rdi call cudaFree@PLT movq 136(%rsp), %rdi call cudaFree@PLT movq 32(%rsp), %rdi call cudaFreeHost@PLT movq 40(%rsp), %rdi call cudaFreeHost@PLT movq 48(%rsp), %rdi call cudaFreeHost@PLT movq 56(%rsp), %rdi call cudaFreeHost@PLT movq 64(%rsp), %rdi call cudaFreeHost@PLT movq 232(%rsp), %rax subq %fs:40, %rax jne .L37 movl $0, %eax addq $240, %rsp .cfi_remember_state .cfi_def_cfa_offset 32 popq %rbx .cfi_def_cfa_offset 24 popq %rbp .cfi_def_cfa_offset 16 popq %r12 .cfi_def_cfa_offset 8 ret .L37: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size main, .-main .section .rodata.str1.1,"aMS",@progbits,1 .LC4: .string "_Z15gpu_matrix_multPiS_S_iii" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2085: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC4(%rip), %rdx movq %rdx, %rcx leaq _Z15gpu_matrix_multPiS_S_iii(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2085: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .section .rodata.cst8,"aM",@progbits,8 .align 8 .LC1: .long 0 .long 1093615616 .align 8 .LC2: .long 0 .long 1062207488 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "dr_model_poc.hip" .globl _Z30__device_stub__gpu_matrix_multPiS_S_iii # -- Begin function _Z30__device_stub__gpu_matrix_multPiS_S_iii .p2align 4, 0x90 .type _Z30__device_stub__gpu_matrix_multPiS_S_iii,@function _Z30__device_stub__gpu_matrix_multPiS_S_iii: # @_Z30__device_stub__gpu_matrix_multPiS_S_iii .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 88(%rsp) movq %rsi, 80(%rsp) movq %rdx, 72(%rsp) movl %ecx, 20(%rsp) movl %r8d, 16(%rsp) movl %r9d, 12(%rsp) leaq 88(%rsp), %rax movq %rax, 96(%rsp) leaq 80(%rsp), %rax movq %rax, 104(%rsp) leaq 72(%rsp), %rax movq %rax, 112(%rsp) leaq 20(%rsp), %rax movq %rax, 120(%rsp) leaq 16(%rsp), %rax movq %rax, 128(%rsp) leaq 12(%rsp), %rax movq %rax, 136(%rsp) leaq 56(%rsp), %rdi leaq 40(%rsp), %rsi leaq 32(%rsp), %rdx leaq 24(%rsp), %rcx callq __hipPopCallConfiguration movq 56(%rsp), %rsi movl 64(%rsp), %edx movq 40(%rsp), %rcx movl 48(%rsp), %r8d leaq 96(%rsp), %r9 movl $_Z15gpu_matrix_multPiS_S_iii, %edi pushq 24(%rsp) .cfi_adjust_cfa_offset 8 pushq 40(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $168, %rsp .cfi_adjust_cfa_offset -168 retq .Lfunc_end0: .size _Z30__device_stub__gpu_matrix_multPiS_S_iii, .Lfunc_end0-_Z30__device_stub__gpu_matrix_multPiS_S_iii .cfi_endproc # -- End function .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 # -- Begin function main .LCPI1_0: .quad 0x3f50000000000000 # double 9.765625E-4 .LCPI1_1: .quad 0x412f400000000000 # double 1024000 .text .globl main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 subq $280, %rsp # imm = 0x118 .cfi_def_cfa_offset 336 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl $3333, %edi # imm = 0xD05 callq srand leaq 144(%rsp), %rdi xorl %ebx, %ebx movl $262144, %esi # imm = 0x40000 xorl %edx, %edx callq hipHostMalloc leaq 136(%rsp), %rdi movl $1073741824, %esi # imm = 0x40000000 xorl %edx, %edx callq hipHostMalloc leaq 272(%rsp), %rdi movl $16384, %esi # imm = 0x4000 xorl %edx, %edx callq hipHostMalloc leaq 128(%rsp), %rdi movl $16777216, %esi # imm = 0x1000000 xorl %edx, %edx callq hipHostMalloc leaq 264(%rsp), %rdi movl $4096, %esi # imm = 0x1000 xorl %edx, %edx callq hipHostMalloc leaq 240(%rsp), %rdi movl $163840, %esi # imm = 0x28000 xorl %edx, %edx callq hipHostMalloc leaq 256(%rsp), %rdi movl $40, %esi xorl %edx, %edx callq hipHostMalloc xorl %r14d, %r14d .p2align 4, 0x90 .LBB1_1: # %.preheader162 # =>This Loop Header: Depth=1 # Child Loop BB1_2 Depth 2 xorl %r15d, %r15d .p2align 4, 0x90 .LBB1_2: # Parent Loop BB1_1 Depth=1 # => This Inner Loop Header: Depth=2 callq rand # kill: def $eax killed $eax def $rax leal 1023(%rax), %ecx testl %eax, %eax cmovnsl %eax, %ecx andl $-1024, %ecx # imm = 0xFC00 subl %ecx, %eax movq 136(%rsp), %rcx addq %rbx, %rcx movl %eax, (%rcx,%r15,4) incq %r15 cmpq $4096, %r15 # imm = 0x1000 jne .LBB1_2 # %bb.3: # in Loop: Header=BB1_1 Depth=1 incq %r14 addq $16384, %rbx # imm = 0x4000 cmpq $65536, %r14 # imm = 0x10000 jne .LBB1_1 # %bb.4: # %.preheader160.preheader xorl %ebx, %ebx xorl %r14d, %r14d .p2align 4, 0x90 .LBB1_5: # %.preheader160 # =>This Loop Header: Depth=1 # Child Loop BB1_6 Depth 2 xorl %r15d, %r15d .p2align 4, 0x90 .LBB1_6: # Parent Loop BB1_5 Depth=1 # => This Inner Loop Header: Depth=2 callq rand # kill: def $eax killed $eax def $rax leal 1023(%rax), %ecx testl %eax, %eax cmovnsl %eax, %ecx andl $-1024, %ecx # imm = 0xFC00 subl %ecx, %eax movq 128(%rsp), %rcx addq %rbx, %rcx movl %eax, (%rcx,%r15,4) incq %r15 cmpq $1024, %r15 # imm = 0x400 jne .LBB1_6 # %bb.7: # in Loop: Header=BB1_5 Depth=1 incq %r14 addq $4096, %rbx # imm = 0x1000 cmpq $4096, %r14 # imm = 0x1000 jne .LBB1_5 # %bb.8: # %.preheader158.preheader xorl %ebx, %ebx xorl %r14d, %r14d .p2align 4, 0x90 .LBB1_9: # %.preheader158 # =>This Loop Header: Depth=1 # Child Loop BB1_10 Depth 2 xorl %r15d, %r15d .p2align 4, 0x90 .LBB1_10: # Parent Loop BB1_9 Depth=1 # => This Inner Loop Header: Depth=2 callq rand # kill: def $eax killed $eax def $rax leal 1023(%rax), %ecx testl %eax, %eax cmovnsl %eax, %ecx andl $-1024, %ecx # imm = 0xFC00 subl %ecx, %eax movq 240(%rsp), %rcx addq %rbx, %rcx movl %eax, (%rcx,%r15,4) incq %r15 cmpq $10, %r15 jne .LBB1_10 # %bb.11: # in Loop: Header=BB1_9 Depth=1 incq %r14 addq $40, %rbx cmpq $4096, %r14 # imm = 0x1000 jne .LBB1_9 # %bb.12: movabsq $68719476752, %rbx # imm = 0x1000000010 movabsq $4294967297, %r14 # imm = 0x100000001 leaq 232(%rsp), %rdi callq hipEventCreate leaq 120(%rsp), %rdi callq hipEventCreate leaq 112(%rsp), %rdi movl $262144, %esi # imm = 0x40000 callq hipMalloc leaq 104(%rsp), %rdi movl $1073741824, %esi # imm = 0x40000000 callq hipMalloc leaq 88(%rsp), %rdi movl $16384, %esi # imm = 0x4000 callq hipMalloc leaq 96(%rsp), %rdi movl $16777216, %esi # imm = 0x1000000 callq hipMalloc leaq 224(%rsp), %rdi movl $4096, %esi # imm = 0x1000 callq hipMalloc leaq 216(%rsp), %rdi movl $163840, %esi # imm = 0x28000 callq hipMalloc leaq 208(%rsp), %rdi movl $40, %esi callq hipMalloc movq 104(%rsp), %rdi movq 136(%rsp), %rsi movl $1073741824, %edx # imm = 0x40000000 movl $1, %ecx callq hipMemcpy movq 96(%rsp), %rdi movq 128(%rsp), %rsi movl $16777216, %edx # imm = 0x1000000 movl $1, %ecx callq hipMemcpy movq 216(%rsp), %rdi movq 240(%rsp), %rsi movl $163840, %edx # imm = 0x28000 movl $1, %ecx callq hipMemcpy xorpd %xmm2, %xmm2 xorl %ebp, %ebp leaq 255(%r14), %r15 leaq 160(%rsp), %r12 leaq 63(%r14), %r13 jmp .LBB1_13 .p2align 4, 0x90 .LBB1_21: # in Loop: Header=BB1_13 Depth=1 movq 256(%rsp), %rdi movq 208(%rsp), %rsi movl $4096, %edx # imm = 0x1000 movl $2, %ecx callq hipMemcpy callq hipDeviceSynchronize movq 120(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movq 120(%rsp), %rdi callq hipEventSynchronize movq 232(%rsp), %rsi movq 120(%rsp), %rdx leaq 156(%rsp), %rdi callq hipEventElapsedTime incl %ebp movss 156(%rsp), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movsd 248(%rsp), %xmm2 # 8-byte Reload # xmm2 = mem[0],zero addsd %xmm0, %xmm2 cmpl $1024, %ebp # imm = 0x400 je .LBB1_22 .LBB1_13: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB1_14 Depth 2 movsd %xmm2, 248(%rsp) # 8-byte Spill xorl %r14d, %r14d .p2align 4, 0x90 .LBB1_14: # Parent Loop BB1_13 Depth=1 # => This Inner Loop Header: Depth=2 callq rand # kill: def $eax killed $eax def $rax leal 1023(%rax), %ecx testl %eax, %eax cmovnsl %eax, %ecx andl $-1024, %ecx # imm = 0xFC00 subl %ecx, %eax movq 144(%rsp), %rcx movl %eax, (%rcx,%r14,4) incq %r14 cmpq $65536, %r14 # imm = 0x10000 jne .LBB1_14 # %bb.15: # %.critedge # in Loop: Header=BB1_13 Depth=1 movq 232(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movq 112(%rsp), %rdi movq 144(%rsp), %rsi movl $262144, %edx # imm = 0x40000 movl $1, %ecx callq hipMemcpy movq %r15, %rdi movl $1, %esi movq %rbx, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_17 # %bb.16: # in Loop: Header=BB1_13 Depth=1 movq 112(%rsp), %rax movq 104(%rsp), %rcx movq 88(%rsp), %rdx movq %rax, 80(%rsp) movq %rcx, 72(%rsp) movq %rdx, 64(%rsp) movl $1, 12(%rsp) movl $65536, 8(%rsp) # imm = 0x10000 movl $4096, 4(%rsp) # imm = 0x1000 leaq 80(%rsp), %rax movq %rax, 160(%rsp) leaq 72(%rsp), %rax movq %rax, 168(%rsp) leaq 64(%rsp), %rax movq %rax, 176(%rsp) leaq 12(%rsp), %rax movq %rax, 184(%rsp) leaq 8(%rsp), %rax movq %rax, 192(%rsp) leaq 4(%rsp), %rax movq %rax, 200(%rsp) leaq 48(%rsp), %rdi leaq 32(%rsp), %rsi leaq 24(%rsp), %rdx leaq 16(%rsp), %rcx callq __hipPopCallConfiguration movq 48(%rsp), %rsi movl 56(%rsp), %edx movq 32(%rsp), %rcx movl 40(%rsp), %r8d movl $_Z15gpu_matrix_multPiS_S_iii, %edi movq %r12, %r9 pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_17: # in Loop: Header=BB1_13 Depth=1 callq hipDeviceSynchronize movq %r13, %rdi movl $1, %esi movq %rbx, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_19 # %bb.18: # in Loop: Header=BB1_13 Depth=1 movq 88(%rsp), %rax movq 96(%rsp), %rcx movq 224(%rsp), %rdx movq %rax, 80(%rsp) movq %rcx, 72(%rsp) movq %rdx, 64(%rsp) movl $1, 12(%rsp) movl $4096, 8(%rsp) # imm = 0x1000 movl $1024, 4(%rsp) # imm = 0x400 leaq 80(%rsp), %rax movq %rax, 160(%rsp) leaq 72(%rsp), %rax movq %rax, 168(%rsp) leaq 64(%rsp), %rax movq %rax, 176(%rsp) leaq 12(%rsp), %rax movq %rax, 184(%rsp) leaq 8(%rsp), %rax movq %rax, 192(%rsp) leaq 4(%rsp), %rax movq %rax, 200(%rsp) leaq 48(%rsp), %rdi leaq 32(%rsp), %rsi leaq 24(%rsp), %rdx leaq 16(%rsp), %rcx callq __hipPopCallConfiguration movq 48(%rsp), %rsi movl 56(%rsp), %edx movq 32(%rsp), %rcx movl 40(%rsp), %r8d movl $_Z15gpu_matrix_multPiS_S_iii, %edi movq %r12, %r9 pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_19: # in Loop: Header=BB1_13 Depth=1 movabsq $4294967297, %rdi # imm = 0x100000001 movl $1, %esi movq %rbx, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_21 # %bb.20: # in Loop: Header=BB1_13 Depth=1 movq 88(%rsp), %rax movq 216(%rsp), %rcx movq 208(%rsp), %rdx movq %rax, 80(%rsp) movq %rcx, 72(%rsp) movq %rdx, 64(%rsp) movl $1, 12(%rsp) movl $4096, 8(%rsp) # imm = 0x1000 movl $10, 4(%rsp) leaq 80(%rsp), %rax movq %rax, 160(%rsp) leaq 72(%rsp), %rax movq %rax, 168(%rsp) leaq 64(%rsp), %rax movq %rax, 176(%rsp) leaq 12(%rsp), %rax movq %rax, 184(%rsp) leaq 8(%rsp), %rax movq %rax, 192(%rsp) leaq 4(%rsp), %rax movq %rax, 200(%rsp) leaq 48(%rsp), %rdi leaq 32(%rsp), %rsi leaq 24(%rsp), %rdx leaq 16(%rsp), %rcx callq __hipPopCallConfiguration movq 48(%rsp), %rsi movl 56(%rsp), %edx movq 32(%rsp), %rcx movl 40(%rsp), %r8d movl $_Z15gpu_matrix_multPiS_S_iii, %edi movq %r12, %r9 pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 jmp .LBB1_21 .LBB1_22: movsd .LCPI1_0(%rip), %xmm0 # xmm0 = mem[0],zero mulsd %xmm2, %xmm0 movsd .LCPI1_1(%rip), %xmm1 # xmm1 = mem[0],zero divsd %xmm2, %xmm1 movl $.L.str, %edi movb $2, %al callq printf movq 112(%rsp), %rdi callq hipFree movq 104(%rsp), %rdi callq hipFree movq 88(%rsp), %rdi callq hipFree movq 96(%rsp), %rdi callq hipFree movq 224(%rsp), %rdi callq hipFree movq 144(%rsp), %rdi callq hipHostFree movq 136(%rsp), %rdi callq hipHostFree movq 272(%rsp), %rdi callq hipHostFree movq 128(%rsp), %rdi callq hipHostFree movq 264(%rsp), %rdi callq hipHostFree xorl %eax, %eax addq $280, %rsp # imm = 0x118 .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end1: .size main, .Lfunc_end1-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB2_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB2_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z15gpu_matrix_multPiS_S_iii, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end2: .size __hip_module_ctor, .Lfunc_end2-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB3_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB3_2: retq .Lfunc_end3: .size __hip_module_dtor, .Lfunc_end3-__hip_module_dtor .cfi_endproc # -- End function .type _Z15gpu_matrix_multPiS_S_iii,@object # @_Z15gpu_matrix_multPiS_S_iii .section .rodata,"a",@progbits .globl _Z15gpu_matrix_multPiS_S_iii .p2align 3, 0x0 _Z15gpu_matrix_multPiS_S_iii: .quad _Z30__device_stub__gpu_matrix_multPiS_S_iii .size _Z15gpu_matrix_multPiS_S_iii, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "Avg. Latency: %g ms :: Avg. Throughput: %g examples/sec\n" .size .L.str, 57 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z15gpu_matrix_multPiS_S_iii" .size .L__unnamed_1, 29 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z30__device_stub__gpu_matrix_multPiS_S_iii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z15gpu_matrix_multPiS_S_iii .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include "includes.h" __global__ void windowBlackman(float* idata, int length) { int tidx = threadIdx.x + blockIdx.x*blockDim.x; if (tidx < length) { idata[tidx] = 0.74 / 2 * -0.5 * cos(2 * PI_F*tidx / (length - 1)) + 0.16 / 2 * sin(4 * PI_F*tidx / (length - 1)); } }
.file "tmpxft_0007129d_00000000-6_windowBlackman.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2029: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2029: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z35__device_stub__Z14windowBlackmanPfiPfi .type _Z35__device_stub__Z14windowBlackmanPfiPfi, @function _Z35__device_stub__Z14windowBlackmanPfiPfi: .LFB2051: .cfi_startproc endbr64 subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 8(%rsp) movl %esi, 4(%rsp) movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax leaq 8(%rsp), %rax movq %rax, 80(%rsp) leaq 4(%rsp), %rax movq %rax, 88(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) movl $1, 40(%rsp) movl $1, 44(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) leaq 24(%rsp), %rcx leaq 16(%rsp), %rdx leaq 44(%rsp), %rsi leaq 32(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 104(%rsp), %rax subq %fs:40, %rax jne .L8 addq $120, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 24(%rsp) .cfi_def_cfa_offset 136 pushq 24(%rsp) .cfi_def_cfa_offset 144 leaq 96(%rsp), %r9 movq 60(%rsp), %rcx movl 68(%rsp), %r8d movq 48(%rsp), %rsi movl 56(%rsp), %edx leaq _Z14windowBlackmanPfi(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 128 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z35__device_stub__Z14windowBlackmanPfiPfi, .-_Z35__device_stub__Z14windowBlackmanPfiPfi .globl _Z14windowBlackmanPfi .type _Z14windowBlackmanPfi, @function _Z14windowBlackmanPfi: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z35__device_stub__Z14windowBlackmanPfiPfi addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z14windowBlackmanPfi, .-_Z14windowBlackmanPfi .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "_Z14windowBlackmanPfi" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2054: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC0(%rip), %rdx movq %rdx, %rcx leaq _Z14windowBlackmanPfi(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2054: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include "includes.h" __global__ void windowBlackman(float* idata, int length) { int tidx = threadIdx.x + blockIdx.x*blockDim.x; if (tidx < length) { idata[tidx] = 0.74 / 2 * -0.5 * cos(2 * PI_F*tidx / (length - 1)) + 0.16 / 2 * sin(4 * PI_F*tidx / (length - 1)); } }
#include <hip/hip_runtime.h> #include "includes.h" __global__ void windowBlackman(float* idata, int length) { int tidx = threadIdx.x + blockIdx.x*blockDim.x; if (tidx < length) { idata[tidx] = 0.74 / 2 * -0.5 * cos(2 * PI_F*tidx / (length - 1)) + 0.16 / 2 * sin(4 * PI_F*tidx / (length - 1)); } }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include "includes.h" __global__ void windowBlackman(float* idata, int length) { int tidx = threadIdx.x + blockIdx.x*blockDim.x; if (tidx < length) { idata[tidx] = 0.74 / 2 * -0.5 * cos(2 * PI_F*tidx / (length - 1)) + 0.16 / 2 * sin(4 * PI_F*tidx / (length - 1)); } }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z14windowBlackmanPfi .globl _Z14windowBlackmanPfi .p2align 8 .type _Z14windowBlackmanPfi,@function _Z14windowBlackmanPfi: s_clause 0x1 s_load_b32 s3, s[0:1], 0x1c s_load_b32 s2, s[0:1], 0x8 s_waitcnt lgkmcnt(0) s_and_b32 s3, s3, 0xffff s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_mad_u64_u32 v[1:2], null, s15, s3, v[0:1] s_mov_b32 s3, exec_lo v_cmpx_gt_i32_e64 s2, v1 s_cbranch_execz .LBB0_10 v_cvt_f32_i32_e32 v5, v1 s_add_i32 s2, s2, -1 s_mov_b32 s3, exec_lo v_cvt_f32_i32_e32 v4, s2 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_f32_e32 v0, 0x40c90fd0, v5 v_div_scale_f32 v2, null, v4, v4, v0 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_1) v_rcp_f32_e32 v3, v2 s_waitcnt_depctr 0xfff v_fma_f32 v6, -v2, v3, 1.0 v_fmac_f32_e32 v3, v6, v3 v_div_scale_f32 v6, vcc_lo, v0, v4, v0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_f32_e32 v7, v6, v3 v_fma_f32 v8, -v2, v7, v6 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmac_f32_e32 v7, v8, v3 v_fma_f32 v2, -v2, v7, v6 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_div_fmas_f32 v2, v2, v3, v7 v_div_fixup_f32 v0, v2, v4, v0 s_delay_alu instid0(VALU_DEP_1) v_cmpx_ngt_f32_e64 0x48000000, |v0| s_xor_b32 s4, exec_lo, s3 s_cbranch_execz .LBB0_3 v_dual_mov_b32 v7, 0 :: v_dual_and_b32 v12, 0x7fffffff, v0 s_mov_b32 s2, 0x7fffff s_delay_alu instid0(VALU_DEP_1) | instid1(SALU_CYCLE_1) v_and_or_b32 v15, v12, s2, 0x800000 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mad_u64_u32 v[2:3], null, v15, 0xfe5163ab, 0 v_mov_b32_e32 v6, v3 v_lshrrev_b32_e32 v3, 23, v12 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_nc_u32_e32 v3, 0xffffff88, v3 v_mad_u64_u32 v[8:9], null, v15, 0x3c439041, v[6:7] s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_cmp_lt_u32_e32 vcc_lo, 63, v3 v_cndmask_b32_e64 v13, 0, 0xffffffc0, vcc_lo v_add_nc_u32_e32 v3, v13, v3 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cmp_lt_u32_e64 s2, 31, v3 v_cndmask_b32_e64 v14, 0, 0xffffffe0, s2 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_dual_mov_b32 v6, v9 :: v_dual_add_nc_u32 v3, v14, v3 v_mad_u64_u32 v[9:10], null, v15, 0xdb629599, v[6:7] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_cmp_lt_u32_e64 s3, 31, v3 v_mov_b32_e32 v6, v10 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_cndmask_b32_e32 v2, v9, v2, vcc_lo v_mad_u64_u32 v[10:11], null, v15, 0xf534ddc0, v[6:7] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mov_b32_e32 v6, v11 v_mad_u64_u32 v[11:12], null, v15, 0xfc2757d1, v[6:7] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mov_b32_e32 v6, v12 v_mad_u64_u32 v[12:13], null, v15, 0x4e441529, v[6:7] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mov_b32_e32 v6, v13 v_mad_u64_u32 v[13:14], null, v15, 0xa2f9836e, v[6:7] v_cndmask_b32_e64 v6, 0, 0xffffffe0, s3 s_delay_alu instid0(VALU_DEP_4) | instskip(NEXT) | instid1(VALU_DEP_2) v_cndmask_b32_e32 v7, v12, v10, vcc_lo v_add_nc_u32_e32 v3, v6, v3 s_delay_alu instid0(VALU_DEP_4) | instskip(SKIP_1) | instid1(VALU_DEP_3) v_dual_cndmask_b32 v13, v13, v11 :: v_dual_cndmask_b32 v12, v14, v12 v_dual_cndmask_b32 v11, v11, v9 :: v_dual_cndmask_b32 v6, v10, v8 v_cmp_eq_u32_e32 vcc_lo, 0, v3 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_4) v_cndmask_b32_e64 v8, v13, v7, s2 v_cndmask_b32_e64 v10, v12, v13, s2 s_delay_alu instid0(VALU_DEP_4) | instskip(SKIP_1) | instid1(VALU_DEP_3) v_cndmask_b32_e64 v7, v7, v11, s2 v_sub_nc_u32_e32 v12, 32, v3 v_cndmask_b32_e64 v10, v10, v8, s3 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_cndmask_b32_e64 v8, v8, v7, s3 v_alignbit_b32 v13, v10, v8, v12 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_2) v_cndmask_b32_e32 v3, v13, v10, vcc_lo v_cndmask_b32_e64 v11, v11, v6, s2 v_cndmask_b32_e64 v2, v6, v2, s2 v_cndmask_b32_e64 v7, v7, v11, s3 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_cndmask_b32_e64 v2, v11, v2, s3 v_alignbit_b32 v9, v8, v7, v12 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_alignbit_b32 v11, v7, v2, v12 v_cndmask_b32_e32 v6, v9, v8, vcc_lo v_bfe_u32 v8, v3, 29, 1 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_cndmask_b32_e32 v7, v11, v7, vcc_lo v_alignbit_b32 v9, v3, v6, 30 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_sub_nc_u32_e32 v10, 0, v8 v_alignbit_b32 v6, v6, v7, 30 v_alignbit_b32 v2, v7, v2, 30 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_xor_b32_e32 v9, v9, v10 v_xor_b32_e32 v6, v6, v10 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_xor_b32_e32 v2, v2, v10 v_clz_i32_u32_e32 v11, v9 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_min_u32_e32 v11, 32, v11 v_sub_nc_u32_e32 v7, 31, v11 v_lshlrev_b32_e32 v13, 23, v11 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_3) | instid1(VALU_DEP_3) v_alignbit_b32 v9, v9, v6, v7 v_alignbit_b32 v2, v6, v2, v7 v_lshrrev_b32_e32 v7, 29, v3 v_lshrrev_b32_e32 v3, 30, v3 v_alignbit_b32 v6, v9, v2, 9 s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_4) v_lshlrev_b32_e32 v7, 31, v7 v_lshrrev_b32_e32 v9, 9, v9 v_add_nc_u32_e32 v3, v8, v3 s_delay_alu instid0(VALU_DEP_4) | instskip(NEXT) | instid1(VALU_DEP_4) v_clz_i32_u32_e32 v10, v6 v_or_b32_e32 v12, 0.5, v7 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_min_u32_e32 v10, 32, v10 v_sub_nc_u32_e32 v12, v12, v13 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_sub_nc_u32_e32 v14, 31, v10 v_alignbit_b32 v2, v6, v2, v14 s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_3) v_or_b32_e32 v6, v9, v12 v_add_lshl_u32 v9, v10, v11, 23 v_lshrrev_b32_e32 v2, 9, v2 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_mul_f32_e32 v10, 0x3fc90fda, v6 v_sub_nc_u32_e32 v2, v2, v9 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f32 v9, v6, 0x3fc90fda, -v10 v_add_nc_u32_e32 v2, 0x33000000, v2 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fmamk_f32 v6, v6, 0x33a22168, v9 v_or_b32_e32 v2, v2, v7 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmac_f32_e32 v6, 0x3fc90fda, v2 v_add_f32_e32 v2, v10, v6 .LBB0_3: s_and_not1_saveexec_b32 s2, s4 v_mul_f32_e64 v2, 0x3f22f983, |v0| s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_rndne_f32_e32 v3, v2 v_fma_f32 v2, v3, 0xbfc90fda, |v0| s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmamk_f32 v2, v3, 0xb3a22168, v2 v_fmamk_f32 v2, v3, 0xa7c234c4, v2 v_cvt_i32_f32_e32 v3, v3 s_or_b32 exec_lo, exec_lo, s2 v_mul_f32_e32 v5, 0x41490fd0, v5 s_mov_b32 s3, exec_lo s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_div_scale_f32 v6, null, v4, v4, v5 v_div_scale_f32 v9, vcc_lo, v5, v4, v5 v_rcp_f32_e32 v7, v6 s_waitcnt_depctr 0xfff v_fma_f32 v8, -v6, v7, 1.0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmac_f32_e32 v7, v8, v7 v_mul_f32_e32 v8, v9, v7 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f32 v10, -v6, v8, v9 v_fmac_f32_e32 v8, v10, v7 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f32 v6, -v6, v8, v9 v_div_fmas_f32 v6, v6, v7, v8 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_div_fixup_f32 v4, v6, v4, v5 v_and_b32_e32 v5, 0x7fffffff, v4 v_cmpx_ngt_f32_e64 0x48000000, |v4| s_xor_b32 s4, exec_lo, s3 s_cbranch_execz .LBB0_7 s_mov_b32 s2, 0x7fffff v_mov_b32_e32 v8, 0 v_and_or_b32 v16, v5, s2, 0x800000 v_lshrrev_b32_e32 v13, 23, v5 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_mad_u64_u32 v[6:7], null, v16, 0xfe5163ab, 0 v_add_nc_u32_e32 v14, 0xffffff88, v13 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_3) v_cmp_lt_u32_e32 vcc_lo, 63, v14 v_mad_u64_u32 v[9:10], null, v16, 0x3c439041, v[7:8] v_cndmask_b32_e64 v15, 0, 0xffffffc0, vcc_lo s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_mov_b32_e32 v7, v10 v_add_nc_u32_e32 v15, v15, v14 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_mad_u64_u32 v[10:11], null, v16, 0xdb629599, v[7:8] v_cmp_lt_u32_e64 s2, 31, v15 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_3) v_cndmask_b32_e64 v17, 0, 0xffffffe0, s2 v_dual_mov_b32 v7, v11 :: v_dual_cndmask_b32 v6, v10, v6 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_nc_u32_e32 v17, v17, v15 v_mad_u64_u32 v[11:12], null, v16, 0xf534ddc0, v[7:8] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_cmp_lt_u32_e64 s3, 31, v17 v_mov_b32_e32 v7, v12 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_cndmask_b32_e32 v9, v11, v9, vcc_lo v_mad_u64_u32 v[12:13], null, v16, 0xfc2757d1, v[7:8] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_cndmask_b32_e64 v6, v9, v6, s2 v_mov_b32_e32 v7, v13 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mad_u64_u32 v[13:14], null, v16, 0x4e441529, v[7:8] v_mov_b32_e32 v7, v14 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_mad_u64_u32 v[14:15], null, v16, 0xa2f9836e, v[7:8] v_cndmask_b32_e64 v7, 0, 0xffffffe0, s3 v_dual_cndmask_b32 v8, v13, v11 :: v_dual_add_nc_u32 v7, v7, v17 s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_3) v_dual_cndmask_b32 v14, v14, v12 :: v_dual_cndmask_b32 v13, v15, v13 v_cndmask_b32_e32 v12, v12, v10, vcc_lo v_cmp_eq_u32_e32 vcc_lo, 0, v7 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_4) v_cndmask_b32_e64 v11, v14, v8, s2 v_cndmask_b32_e64 v13, v13, v14, s2 s_delay_alu instid0(VALU_DEP_4) | instskip(SKIP_2) | instid1(VALU_DEP_4) v_cndmask_b32_e64 v8, v8, v12, s2 v_sub_nc_u32_e32 v14, 32, v7 v_cndmask_b32_e64 v12, v12, v9, s2 v_cndmask_b32_e64 v13, v13, v11, s3 s_delay_alu instid0(VALU_DEP_4) | instskip(NEXT) | instid1(VALU_DEP_3) v_cndmask_b32_e64 v11, v11, v8, s3 v_cndmask_b32_e64 v8, v8, v12, s3 v_cndmask_b32_e64 v6, v12, v6, s3 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_alignbit_b32 v15, v13, v11, v14 v_alignbit_b32 v10, v11, v8, v14 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_4) v_cndmask_b32_e32 v7, v15, v13, vcc_lo v_alignbit_b32 v13, v8, v6, v14 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_cndmask_b32_e32 v9, v10, v11, vcc_lo v_bfe_u32 v10, v7, 29, 1 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_cndmask_b32_e32 v8, v13, v8, vcc_lo v_alignbit_b32 v11, v7, v9, 30 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_sub_nc_u32_e32 v12, 0, v10 v_alignbit_b32 v9, v9, v8, 30 v_alignbit_b32 v6, v8, v6, 30 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_xor_b32_e32 v11, v11, v12 v_xor_b32_e32 v8, v9, v12 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_xor_b32_e32 v6, v6, v12 v_clz_i32_u32_e32 v13, v11 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_min_u32_e32 v13, 32, v13 v_sub_nc_u32_e32 v9, 31, v13 v_lshlrev_b32_e32 v15, 23, v13 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_2) | instid1(VALU_DEP_2) v_alignbit_b32 v11, v11, v8, v9 v_alignbit_b32 v6, v8, v6, v9 v_lshrrev_b32_e32 v9, 29, v7 v_alignbit_b32 v8, v11, v6, 9 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_3) v_lshlrev_b32_e32 v9, 31, v9 v_lshrrev_b32_e32 v11, 9, v11 v_clz_i32_u32_e32 v12, v8 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_or_b32_e32 v14, 0.5, v9 v_min_u32_e32 v12, 32, v12 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_sub_nc_u32_e32 v14, v14, v15 v_sub_nc_u32_e32 v16, 31, v12 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_3) v_alignbit_b32 v6, v8, v6, v16 v_or_b32_e32 v8, v11, v14 v_add_lshl_u32 v11, v12, v13, 23 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_lshrrev_b32_e32 v6, 9, v6 v_mul_f32_e32 v12, 0x3fc90fda, v8 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_sub_nc_u32_e32 v6, v6, v11 v_fma_f32 v11, v8, 0x3fc90fda, -v12 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_nc_u32_e32 v6, 0x33000000, v6 v_fmamk_f32 v8, v8, 0x33a22168, v11 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_or_b32_e32 v6, v6, v9 v_fmac_f32_e32 v8, 0x3fc90fda, v6 v_lshrrev_b32_e32 v7, 30, v7 s_delay_alu instid0(VALU_DEP_1) v_dual_add_f32 v6, v12, v8 :: v_dual_add_nc_u32 v7, v10, v7 .LBB0_7: s_and_not1_saveexec_b32 s2, s4 v_mul_f32_e64 v6, 0x3f22f983, |v4| s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_rndne_f32_e32 v7, v6 v_fma_f32 v6, v7, 0xbfc90fda, |v4| s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmamk_f32 v6, v7, 0xb3a22168, v6 v_fmamk_f32 v6, v7, 0xa7c234c4, v6 v_cvt_i32_f32_e32 v7, v7 s_or_b32 exec_lo, exec_lo, s2 v_dual_mul_f32 v8, v2, v2 :: v_dual_and_b32 v11, 1, v3 s_mov_b32 s2, 0xb94c1982 s_mov_b32 s3, 0x37d75334 v_xor_b32_e32 v5, v5, v4 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_3) | instid1(VALU_DEP_3) v_fmaak_f32 v9, s2, v8, 0x3c0881c4 v_cmp_eq_u32_e32 vcc_lo, 0, v11 s_load_b64 s[0:1], s[0:1], 0x0 v_lshlrev_b32_e32 v3, 30, v3 v_fmaak_f32 v9, v8, v9, 0xbe2aaa9d v_fmaak_f32 v10, s3, v8, 0xbab64f3b s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_and_b32_e32 v3, 0x80000000, v3 v_mul_f32_e32 v9, v8, v9 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_fmaak_f32 v10, v8, v10, 0x3d2aabf7 v_fmac_f32_e32 v2, v2, v9 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmaak_f32 v10, v8, v10, 0xbf000004 v_fma_f32 v8, v8, v10, 1.0 v_and_b32_e32 v10, 1, v7 v_lshlrev_b32_e32 v7, 30, v7 s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_3) v_cndmask_b32_e64 v2, -v2, v8, vcc_lo v_cmp_class_f32_e64 vcc_lo, v0, 0x1f8 v_dual_mul_f32 v8, v6, v6 :: v_dual_and_b32 v7, 0x80000000, v7 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_xor_b32_e32 v2, v3, v2 v_fmaak_f32 v9, s3, v8, 0xbab64f3b s_mov_b32 s3, 0xbfc7ae14 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_cndmask_b32_e32 v0, 0x7fc00000, v2, vcc_lo v_cmp_eq_u32_e32 vcc_lo, 0, v10 v_cvt_f64_f32_e32 v[2:3], v0 v_fmaak_f32 v0, s2, v8, 0x3c0881c4 s_mov_b32 s2, 0x7ae147ae s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmaak_f32 v0, v8, v0, 0xbe2aaa9d v_mul_f32_e32 v0, v8, v0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_dual_fmaak_f32 v9, v8, v9, 0x3d2aabf7 :: v_dual_fmac_f32 v6, v6, v0 v_fmaak_f32 v9, v8, v9, 0xbf000004 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f32 v0, v8, v9, 1.0 v_cndmask_b32_e32 v0, v0, v6, vcc_lo v_cmp_class_f32_e64 vcc_lo, v4, 0x1f8 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_xor3_b32 v0, v5, v7, v0 v_cndmask_b32_e32 v0, 0x7fc00000, v0, vcc_lo v_mul_f64 v[2:3], v[2:3], s[2:3] s_mov_b32 s3, 0x3fb47ae1 s_mov_b32 s2, 0x47ae147b s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_cvt_f64_f32_e32 v[4:5], v0 v_fma_f64 v[2:3], v[4:5], s[2:3], v[2:3] s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_cvt_f32_f64_e32 v3, v[2:3] v_ashrrev_i32_e32 v2, 31, v1 v_lshlrev_b64 v[0:1], 2, v[1:2] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v0, vcc_lo, s0, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo global_store_b32 v[0:1], v3, off .LBB0_10: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z14windowBlackmanPfi .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 272 .amdhsa_user_sgpr_count 15 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 0 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 0 .amdhsa_next_free_vgpr 18 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z14windowBlackmanPfi, .Lfunc_end0-_Z14windowBlackmanPfi .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .offset: 8 .size: 4 .value_kind: by_value - .offset: 16 .size: 4 .value_kind: hidden_block_count_x - .offset: 20 .size: 4 .value_kind: hidden_block_count_y - .offset: 24 .size: 4 .value_kind: hidden_block_count_z - .offset: 28 .size: 2 .value_kind: hidden_group_size_x - .offset: 30 .size: 2 .value_kind: hidden_group_size_y - .offset: 32 .size: 2 .value_kind: hidden_group_size_z - .offset: 34 .size: 2 .value_kind: hidden_remainder_x - .offset: 36 .size: 2 .value_kind: hidden_remainder_y - .offset: 38 .size: 2 .value_kind: hidden_remainder_z - .offset: 56 .size: 8 .value_kind: hidden_global_offset_x - .offset: 64 .size: 8 .value_kind: hidden_global_offset_y - .offset: 72 .size: 8 .value_kind: hidden_global_offset_z - .offset: 80 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 272 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z14windowBlackmanPfi .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z14windowBlackmanPfi.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 18 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <hip/hip_runtime.h> #include "includes.h" __global__ void windowBlackman(float* idata, int length) { int tidx = threadIdx.x + blockIdx.x*blockDim.x; if (tidx < length) { idata[tidx] = 0.74 / 2 * -0.5 * cos(2 * PI_F*tidx / (length - 1)) + 0.16 / 2 * sin(4 * PI_F*tidx / (length - 1)); } }
.text .file "windowBlackman.hip" .globl _Z29__device_stub__windowBlackmanPfi # -- Begin function _Z29__device_stub__windowBlackmanPfi .p2align 4, 0x90 .type _Z29__device_stub__windowBlackmanPfi,@function _Z29__device_stub__windowBlackmanPfi: # @_Z29__device_stub__windowBlackmanPfi .cfi_startproc # %bb.0: subq $88, %rsp .cfi_def_cfa_offset 96 movq %rdi, 56(%rsp) movl %esi, 4(%rsp) leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 4(%rsp), %rax movq %rax, 72(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 64(%rsp), %r9 movl $_Z14windowBlackmanPfi, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $104, %rsp .cfi_adjust_cfa_offset -104 retq .Lfunc_end0: .size _Z29__device_stub__windowBlackmanPfi, .Lfunc_end0-_Z29__device_stub__windowBlackmanPfi .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB1_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB1_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z14windowBlackmanPfi, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end1: .size __hip_module_ctor, .Lfunc_end1-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB2_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB2_2: retq .Lfunc_end2: .size __hip_module_dtor, .Lfunc_end2-__hip_module_dtor .cfi_endproc # -- End function .type _Z14windowBlackmanPfi,@object # @_Z14windowBlackmanPfi .section .rodata,"a",@progbits .globl _Z14windowBlackmanPfi .p2align 3, 0x0 _Z14windowBlackmanPfi: .quad _Z29__device_stub__windowBlackmanPfi .size _Z14windowBlackmanPfi, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z14windowBlackmanPfi" .size .L__unnamed_1, 22 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z29__device_stub__windowBlackmanPfi .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z14windowBlackmanPfi .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_0007129d_00000000-6_windowBlackman.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2029: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2029: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z35__device_stub__Z14windowBlackmanPfiPfi .type _Z35__device_stub__Z14windowBlackmanPfiPfi, @function _Z35__device_stub__Z14windowBlackmanPfiPfi: .LFB2051: .cfi_startproc endbr64 subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 8(%rsp) movl %esi, 4(%rsp) movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax leaq 8(%rsp), %rax movq %rax, 80(%rsp) leaq 4(%rsp), %rax movq %rax, 88(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) movl $1, 40(%rsp) movl $1, 44(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) leaq 24(%rsp), %rcx leaq 16(%rsp), %rdx leaq 44(%rsp), %rsi leaq 32(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 104(%rsp), %rax subq %fs:40, %rax jne .L8 addq $120, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 24(%rsp) .cfi_def_cfa_offset 136 pushq 24(%rsp) .cfi_def_cfa_offset 144 leaq 96(%rsp), %r9 movq 60(%rsp), %rcx movl 68(%rsp), %r8d movq 48(%rsp), %rsi movl 56(%rsp), %edx leaq _Z14windowBlackmanPfi(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 128 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z35__device_stub__Z14windowBlackmanPfiPfi, .-_Z35__device_stub__Z14windowBlackmanPfiPfi .globl _Z14windowBlackmanPfi .type _Z14windowBlackmanPfi, @function _Z14windowBlackmanPfi: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z35__device_stub__Z14windowBlackmanPfiPfi addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z14windowBlackmanPfi, .-_Z14windowBlackmanPfi .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "_Z14windowBlackmanPfi" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2054: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC0(%rip), %rdx movq %rdx, %rcx leaq _Z14windowBlackmanPfi(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2054: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "windowBlackman.hip" .globl _Z29__device_stub__windowBlackmanPfi # -- Begin function _Z29__device_stub__windowBlackmanPfi .p2align 4, 0x90 .type _Z29__device_stub__windowBlackmanPfi,@function _Z29__device_stub__windowBlackmanPfi: # @_Z29__device_stub__windowBlackmanPfi .cfi_startproc # %bb.0: subq $88, %rsp .cfi_def_cfa_offset 96 movq %rdi, 56(%rsp) movl %esi, 4(%rsp) leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 4(%rsp), %rax movq %rax, 72(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 64(%rsp), %r9 movl $_Z14windowBlackmanPfi, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $104, %rsp .cfi_adjust_cfa_offset -104 retq .Lfunc_end0: .size _Z29__device_stub__windowBlackmanPfi, .Lfunc_end0-_Z29__device_stub__windowBlackmanPfi .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB1_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB1_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z14windowBlackmanPfi, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end1: .size __hip_module_ctor, .Lfunc_end1-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB2_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB2_2: retq .Lfunc_end2: .size __hip_module_dtor, .Lfunc_end2-__hip_module_dtor .cfi_endproc # -- End function .type _Z14windowBlackmanPfi,@object # @_Z14windowBlackmanPfi .section .rodata,"a",@progbits .globl _Z14windowBlackmanPfi .p2align 3, 0x0 _Z14windowBlackmanPfi: .quad _Z29__device_stub__windowBlackmanPfi .size _Z14windowBlackmanPfi, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z14windowBlackmanPfi" .size .L__unnamed_1, 22 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z29__device_stub__windowBlackmanPfi .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z14windowBlackmanPfi .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include "includes.h" __global__ void convolutionRowGPU(double *h_Dst, double *h_Src, double *h_Filter, int imageW, int imageH, int filterR){ int k; double sum = 0; int ix = blockIdx.x * blockDim.x + threadIdx.x; int iy = blockIdx.y * blockDim.y + threadIdx.y; for (k = -filterR; k <= filterR; k++) { int d = ix + k; //edw einai to pou tha paei to filtro gi auto elegxei apo katw kai an to d einia ektos oriwn eikonas if (d >= 0 && d < imageW) { sum += h_Src[iy * imageW + d] * h_Filter[filterR - k]; } h_Dst[iy * imageW + ix] = sum; } }
.file "tmpxft_000e845a_00000000-6_convolutionRowGPU.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2029: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2029: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z44__device_stub__Z17convolutionRowGPUPdS_S_iiiPdS_S_iii .type _Z44__device_stub__Z17convolutionRowGPUPdS_S_iiiPdS_S_iii, @function _Z44__device_stub__Z17convolutionRowGPUPdS_S_iiiPdS_S_iii: .LFB2051: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movq %rdi, 40(%rsp) movq %rsi, 32(%rsp) movq %rdx, 24(%rsp) movl %ecx, 20(%rsp) movl %r8d, 16(%rsp) movl %r9d, 12(%rsp) movq %fs:40, %rax movq %rax, 168(%rsp) xorl %eax, %eax leaq 40(%rsp), %rax movq %rax, 112(%rsp) leaq 32(%rsp), %rax movq %rax, 120(%rsp) leaq 24(%rsp), %rax movq %rax, 128(%rsp) leaq 20(%rsp), %rax movq %rax, 136(%rsp) leaq 16(%rsp), %rax movq %rax, 144(%rsp) leaq 12(%rsp), %rax movq %rax, 152(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) movl $1, 72(%rsp) movl $1, 76(%rsp) movl $1, 80(%rsp) movl $1, 84(%rsp) leaq 56(%rsp), %rcx leaq 48(%rsp), %rdx leaq 76(%rsp), %rsi leaq 64(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 168(%rsp), %rax subq %fs:40, %rax jne .L8 addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 56(%rsp) .cfi_def_cfa_offset 200 pushq 56(%rsp) .cfi_def_cfa_offset 208 leaq 128(%rsp), %r9 movq 92(%rsp), %rcx movl 100(%rsp), %r8d movq 80(%rsp), %rsi movl 88(%rsp), %edx leaq _Z17convolutionRowGPUPdS_S_iii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 192 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z44__device_stub__Z17convolutionRowGPUPdS_S_iiiPdS_S_iii, .-_Z44__device_stub__Z17convolutionRowGPUPdS_S_iiiPdS_S_iii .globl _Z17convolutionRowGPUPdS_S_iii .type _Z17convolutionRowGPUPdS_S_iii, @function _Z17convolutionRowGPUPdS_S_iii: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z44__device_stub__Z17convolutionRowGPUPdS_S_iiiPdS_S_iii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z17convolutionRowGPUPdS_S_iii, .-_Z17convolutionRowGPUPdS_S_iii .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "_Z17convolutionRowGPUPdS_S_iii" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2054: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC0(%rip), %rdx movq %rdx, %rcx leaq _Z17convolutionRowGPUPdS_S_iii(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2054: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include "includes.h" __global__ void convolutionRowGPU(double *h_Dst, double *h_Src, double *h_Filter, int imageW, int imageH, int filterR){ int k; double sum = 0; int ix = blockIdx.x * blockDim.x + threadIdx.x; int iy = blockIdx.y * blockDim.y + threadIdx.y; for (k = -filterR; k <= filterR; k++) { int d = ix + k; //edw einai to pou tha paei to filtro gi auto elegxei apo katw kai an to d einia ektos oriwn eikonas if (d >= 0 && d < imageW) { sum += h_Src[iy * imageW + d] * h_Filter[filterR - k]; } h_Dst[iy * imageW + ix] = sum; } }
#include <hip/hip_runtime.h> #include "includes.h" __global__ void convolutionRowGPU(double *h_Dst, double *h_Src, double *h_Filter, int imageW, int imageH, int filterR){ int k; double sum = 0; int ix = blockIdx.x * blockDim.x + threadIdx.x; int iy = blockIdx.y * blockDim.y + threadIdx.y; for (k = -filterR; k <= filterR; k++) { int d = ix + k; //edw einai to pou tha paei to filtro gi auto elegxei apo katw kai an to d einia ektos oriwn eikonas if (d >= 0 && d < imageW) { sum += h_Src[iy * imageW + d] * h_Filter[filterR - k]; } h_Dst[iy * imageW + ix] = sum; } }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include "includes.h" __global__ void convolutionRowGPU(double *h_Dst, double *h_Src, double *h_Filter, int imageW, int imageH, int filterR){ int k; double sum = 0; int ix = blockIdx.x * blockDim.x + threadIdx.x; int iy = blockIdx.y * blockDim.y + threadIdx.y; for (k = -filterR; k <= filterR; k++) { int d = ix + k; //edw einai to pou tha paei to filtro gi auto elegxei apo katw kai an to d einia ektos oriwn eikonas if (d >= 0 && d < imageW) { sum += h_Src[iy * imageW + d] * h_Filter[filterR - k]; } h_Dst[iy * imageW + ix] = sum; } }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z17convolutionRowGPUPdS_S_iii .globl _Z17convolutionRowGPUPdS_S_iii .p2align 8 .type _Z17convolutionRowGPUPdS_S_iii,@function _Z17convolutionRowGPUPdS_S_iii: s_load_b32 s9, s[0:1], 0x20 s_waitcnt lgkmcnt(0) s_cmp_lt_i32 s9, 0 s_cbranch_scc1 .LBB0_5 s_clause 0x1 s_load_b32 s2, s[0:1], 0x34 s_load_b32 s8, s[0:1], 0x18 v_bfe_u32 v1, v0, 10, 10 v_and_b32_e32 v0, 0x3ff, v0 s_load_b128 s[4:7], s[0:1], 0x0 s_waitcnt lgkmcnt(0) s_lshr_b32 s3, s2, 16 s_and_b32 s2, s2, 0xffff v_mad_u64_u32 v[2:3], null, s15, s3, v[1:2] v_mad_u64_u32 v[3:4], null, s14, s2, v[0:1] s_load_b64 s[2:3], s[0:1], 0x10 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_mul_lo_u32 v4, v2, s8 v_subrev_nc_u32_e32 v5, s9, v3 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_nc_u32_e32 v0, v4, v3 v_ashrrev_i32_e32 v1, 31, v0 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_3) v_lshlrev_b64 v[6:7], 3, v[0:1] v_mov_b32_e32 v0, 0 v_mov_b32_e32 v1, 0 v_add_co_u32 v2, vcc_lo, s4, v6 v_mov_b32_e32 v6, 0 v_add_co_ci_u32_e32 v3, vcc_lo, s5, v7, vcc_lo s_lshl_b32 s4, s9, 1 s_set_inst_prefetch_distance 0x1 s_branch .LBB0_3 .p2align 6 .LBB0_2: s_or_b32 exec_lo, exec_lo, s0 v_add_nc_u32_e32 v5, 1, v5 s_add_i32 s4, s4, -1 global_store_b64 v[2:3], v[0:1], off s_cmp_lg_u32 s4, -1 s_cbranch_scc0 .LBB0_5 .LBB0_3: v_cmp_lt_i32_e32 vcc_lo, -1, v5 v_cmp_gt_i32_e64 s0, s8, v5 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_and_b32 s1, vcc_lo, s0 s_and_saveexec_b32 s0, s1 s_cbranch_execz .LBB0_2 v_add_nc_u32_e32 v7, v4, v5 s_ashr_i32 s5, s4, 31 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_4) | instid1(VALU_DEP_1) s_lshl_b64 s[10:11], s[4:5], 3 s_waitcnt lgkmcnt(0) s_add_u32 s10, s2, s10 v_ashrrev_i32_e32 v8, 31, v7 s_addc_u32 s11, s3, s11 v_lshlrev_b64 v[7:8], 3, v[7:8] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v7, vcc_lo, s6, v7 v_add_co_ci_u32_e32 v8, vcc_lo, s7, v8, vcc_lo global_load_b64 v[7:8], v[7:8], off global_load_b64 v[9:10], v6, s[10:11] s_waitcnt vmcnt(0) v_fma_f64 v[0:1], v[7:8], v[9:10], v[0:1] s_branch .LBB0_2 .LBB0_5: s_set_inst_prefetch_distance 0x2 s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z17convolutionRowGPUPdS_S_iii .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 296 .amdhsa_user_sgpr_count 14 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 1 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 1 .amdhsa_next_free_vgpr 11 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z17convolutionRowGPUPdS_S_iii, .Lfunc_end0-_Z17convolutionRowGPUPdS_S_iii .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 8 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .offset: 24 .size: 4 .value_kind: by_value - .offset: 28 .size: 4 .value_kind: by_value - .offset: 32 .size: 4 .value_kind: by_value - .offset: 40 .size: 4 .value_kind: hidden_block_count_x - .offset: 44 .size: 4 .value_kind: hidden_block_count_y - .offset: 48 .size: 4 .value_kind: hidden_block_count_z - .offset: 52 .size: 2 .value_kind: hidden_group_size_x - .offset: 54 .size: 2 .value_kind: hidden_group_size_y - .offset: 56 .size: 2 .value_kind: hidden_group_size_z - .offset: 58 .size: 2 .value_kind: hidden_remainder_x - .offset: 60 .size: 2 .value_kind: hidden_remainder_y - .offset: 62 .size: 2 .value_kind: hidden_remainder_z - .offset: 80 .size: 8 .value_kind: hidden_global_offset_x - .offset: 88 .size: 8 .value_kind: hidden_global_offset_y - .offset: 96 .size: 8 .value_kind: hidden_global_offset_z - .offset: 104 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 296 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z17convolutionRowGPUPdS_S_iii .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z17convolutionRowGPUPdS_S_iii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 11 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <hip/hip_runtime.h> #include "includes.h" __global__ void convolutionRowGPU(double *h_Dst, double *h_Src, double *h_Filter, int imageW, int imageH, int filterR){ int k; double sum = 0; int ix = blockIdx.x * blockDim.x + threadIdx.x; int iy = blockIdx.y * blockDim.y + threadIdx.y; for (k = -filterR; k <= filterR; k++) { int d = ix + k; //edw einai to pou tha paei to filtro gi auto elegxei apo katw kai an to d einia ektos oriwn eikonas if (d >= 0 && d < imageW) { sum += h_Src[iy * imageW + d] * h_Filter[filterR - k]; } h_Dst[iy * imageW + ix] = sum; } }
.text .file "convolutionRowGPU.hip" .globl _Z32__device_stub__convolutionRowGPUPdS_S_iii # -- Begin function _Z32__device_stub__convolutionRowGPUPdS_S_iii .p2align 4, 0x90 .type _Z32__device_stub__convolutionRowGPUPdS_S_iii,@function _Z32__device_stub__convolutionRowGPUPdS_S_iii: # @_Z32__device_stub__convolutionRowGPUPdS_S_iii .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 88(%rsp) movq %rsi, 80(%rsp) movq %rdx, 72(%rsp) movl %ecx, 20(%rsp) movl %r8d, 16(%rsp) movl %r9d, 12(%rsp) leaq 88(%rsp), %rax movq %rax, 96(%rsp) leaq 80(%rsp), %rax movq %rax, 104(%rsp) leaq 72(%rsp), %rax movq %rax, 112(%rsp) leaq 20(%rsp), %rax movq %rax, 120(%rsp) leaq 16(%rsp), %rax movq %rax, 128(%rsp) leaq 12(%rsp), %rax movq %rax, 136(%rsp) leaq 56(%rsp), %rdi leaq 40(%rsp), %rsi leaq 32(%rsp), %rdx leaq 24(%rsp), %rcx callq __hipPopCallConfiguration movq 56(%rsp), %rsi movl 64(%rsp), %edx movq 40(%rsp), %rcx movl 48(%rsp), %r8d leaq 96(%rsp), %r9 movl $_Z17convolutionRowGPUPdS_S_iii, %edi pushq 24(%rsp) .cfi_adjust_cfa_offset 8 pushq 40(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $168, %rsp .cfi_adjust_cfa_offset -168 retq .Lfunc_end0: .size _Z32__device_stub__convolutionRowGPUPdS_S_iii, .Lfunc_end0-_Z32__device_stub__convolutionRowGPUPdS_S_iii .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB1_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB1_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z17convolutionRowGPUPdS_S_iii, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end1: .size __hip_module_ctor, .Lfunc_end1-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB2_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB2_2: retq .Lfunc_end2: .size __hip_module_dtor, .Lfunc_end2-__hip_module_dtor .cfi_endproc # -- End function .type _Z17convolutionRowGPUPdS_S_iii,@object # @_Z17convolutionRowGPUPdS_S_iii .section .rodata,"a",@progbits .globl _Z17convolutionRowGPUPdS_S_iii .p2align 3, 0x0 _Z17convolutionRowGPUPdS_S_iii: .quad _Z32__device_stub__convolutionRowGPUPdS_S_iii .size _Z17convolutionRowGPUPdS_S_iii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z17convolutionRowGPUPdS_S_iii" .size .L__unnamed_1, 31 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z32__device_stub__convolutionRowGPUPdS_S_iii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z17convolutionRowGPUPdS_S_iii .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_000e845a_00000000-6_convolutionRowGPU.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2029: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2029: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z44__device_stub__Z17convolutionRowGPUPdS_S_iiiPdS_S_iii .type _Z44__device_stub__Z17convolutionRowGPUPdS_S_iiiPdS_S_iii, @function _Z44__device_stub__Z17convolutionRowGPUPdS_S_iiiPdS_S_iii: .LFB2051: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movq %rdi, 40(%rsp) movq %rsi, 32(%rsp) movq %rdx, 24(%rsp) movl %ecx, 20(%rsp) movl %r8d, 16(%rsp) movl %r9d, 12(%rsp) movq %fs:40, %rax movq %rax, 168(%rsp) xorl %eax, %eax leaq 40(%rsp), %rax movq %rax, 112(%rsp) leaq 32(%rsp), %rax movq %rax, 120(%rsp) leaq 24(%rsp), %rax movq %rax, 128(%rsp) leaq 20(%rsp), %rax movq %rax, 136(%rsp) leaq 16(%rsp), %rax movq %rax, 144(%rsp) leaq 12(%rsp), %rax movq %rax, 152(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) movl $1, 72(%rsp) movl $1, 76(%rsp) movl $1, 80(%rsp) movl $1, 84(%rsp) leaq 56(%rsp), %rcx leaq 48(%rsp), %rdx leaq 76(%rsp), %rsi leaq 64(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 168(%rsp), %rax subq %fs:40, %rax jne .L8 addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 56(%rsp) .cfi_def_cfa_offset 200 pushq 56(%rsp) .cfi_def_cfa_offset 208 leaq 128(%rsp), %r9 movq 92(%rsp), %rcx movl 100(%rsp), %r8d movq 80(%rsp), %rsi movl 88(%rsp), %edx leaq _Z17convolutionRowGPUPdS_S_iii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 192 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z44__device_stub__Z17convolutionRowGPUPdS_S_iiiPdS_S_iii, .-_Z44__device_stub__Z17convolutionRowGPUPdS_S_iiiPdS_S_iii .globl _Z17convolutionRowGPUPdS_S_iii .type _Z17convolutionRowGPUPdS_S_iii, @function _Z17convolutionRowGPUPdS_S_iii: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z44__device_stub__Z17convolutionRowGPUPdS_S_iiiPdS_S_iii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z17convolutionRowGPUPdS_S_iii, .-_Z17convolutionRowGPUPdS_S_iii .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "_Z17convolutionRowGPUPdS_S_iii" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2054: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC0(%rip), %rdx movq %rdx, %rcx leaq _Z17convolutionRowGPUPdS_S_iii(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2054: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "convolutionRowGPU.hip" .globl _Z32__device_stub__convolutionRowGPUPdS_S_iii # -- Begin function _Z32__device_stub__convolutionRowGPUPdS_S_iii .p2align 4, 0x90 .type _Z32__device_stub__convolutionRowGPUPdS_S_iii,@function _Z32__device_stub__convolutionRowGPUPdS_S_iii: # @_Z32__device_stub__convolutionRowGPUPdS_S_iii .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 88(%rsp) movq %rsi, 80(%rsp) movq %rdx, 72(%rsp) movl %ecx, 20(%rsp) movl %r8d, 16(%rsp) movl %r9d, 12(%rsp) leaq 88(%rsp), %rax movq %rax, 96(%rsp) leaq 80(%rsp), %rax movq %rax, 104(%rsp) leaq 72(%rsp), %rax movq %rax, 112(%rsp) leaq 20(%rsp), %rax movq %rax, 120(%rsp) leaq 16(%rsp), %rax movq %rax, 128(%rsp) leaq 12(%rsp), %rax movq %rax, 136(%rsp) leaq 56(%rsp), %rdi leaq 40(%rsp), %rsi leaq 32(%rsp), %rdx leaq 24(%rsp), %rcx callq __hipPopCallConfiguration movq 56(%rsp), %rsi movl 64(%rsp), %edx movq 40(%rsp), %rcx movl 48(%rsp), %r8d leaq 96(%rsp), %r9 movl $_Z17convolutionRowGPUPdS_S_iii, %edi pushq 24(%rsp) .cfi_adjust_cfa_offset 8 pushq 40(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $168, %rsp .cfi_adjust_cfa_offset -168 retq .Lfunc_end0: .size _Z32__device_stub__convolutionRowGPUPdS_S_iii, .Lfunc_end0-_Z32__device_stub__convolutionRowGPUPdS_S_iii .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB1_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB1_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z17convolutionRowGPUPdS_S_iii, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end1: .size __hip_module_ctor, .Lfunc_end1-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB2_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB2_2: retq .Lfunc_end2: .size __hip_module_dtor, .Lfunc_end2-__hip_module_dtor .cfi_endproc # -- End function .type _Z17convolutionRowGPUPdS_S_iii,@object # @_Z17convolutionRowGPUPdS_S_iii .section .rodata,"a",@progbits .globl _Z17convolutionRowGPUPdS_S_iii .p2align 3, 0x0 _Z17convolutionRowGPUPdS_S_iii: .quad _Z32__device_stub__convolutionRowGPUPdS_S_iii .size _Z17convolutionRowGPUPdS_S_iii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z17convolutionRowGPUPdS_S_iii" .size .L__unnamed_1, 31 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z32__device_stub__convolutionRowGPUPdS_S_iii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z17convolutionRowGPUPdS_S_iii .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
// // CrossCorrelation.cu // CrossCorrelation // // Created by Vivek Sridhar on 29/06/17. // Copyright © 2017 Vivek Sridhar. All rights reserved. // #include <iostream> #include <fstream> #include <sstream> #include <vector> #include <algorithm> #include <stdio.h> #include <stdlib.h> #include <dirent.h> template <typename T> std::string to_string(const T& value) { std::stringstream ss; ss << value; return ss.str(); } long factorial(long val) { std::cout << val << "\n"; long result = 1; for (long i = 1; i <= val; ++i) { result *= i; } return result; } long combination(long n, long r) { return (factorial(n)) / ((factorial(n - r)) * factorial(r)); } __global__ void kernel(float *x1, float *y1, float *x2, float *y2, float *res, int tau, int na_frames, long nElements) { int index = blockIdx.x * blockDim.x + threadIdx.x; if (tau < 0) { if (index >= -tau+na_frames) { res[index] = x1[index] * x2[index + tau] + y1[index] * y2[index + tau]; } else res[index] = 0.0; } else { if (index < nElements - tau) { res[index] = x1[index] * x2[index + tau] + y1[index] * y2[index + tau]; } else res[index] = 0.0; } } // total measurement points in the time series is defined by nElements #define M 1024 // number of threads per block #define fps 10 // frames per second of input video (used to determine tau) #define time 5 // time in seconds within which time delayed cross correlation is calculated (tau ranges from -time*fps to time*fps) #define n_inds 10 int na_frames = 0; // number of frames in the start with nas int scale = 1; // time window for analysis in seconds; varying this allows us to examine dynamics of leadership across varying timescales; setting scale larger than the entire time series or -1 gives aggregated statistics across the entire duration (otherwise, timescale of analysis is scale*fps) //const int pairs = combination(n_inds, 2); const bool aggregate = false; // this boolean decides whether you output a dynamic time variable leadership network or a static time aggregated network; scale is set to -1 if aggregate is true std::ofstream outputFile1; int main () { DIR *dir; FILE *pFile_x1; FILE *pFile_y1; FILE *pFile_x2; FILE *pFile_y2; long lSize; long nElements; struct dirent *file; float *d_x1, *d_y1, *d_x2, *d_y2, *d_res; float *x1, *y1, *x2, *y2, *res; size_t result_x1, result_y1, result_x2, result_y2; if (aggregate) scale = -1; std::vector<std::string> files; std::string directory = "/home/user/Documents/Vivek/cuda/DirectionalCorrelation/Data/Input/pigeons/10_birds/ffA3/cross_correlation/"; dir = opendir(directory.c_str()); int idx = 0; while ((file = readdir(dir)) != NULL) { if (file->d_name[0] == 'd') { files.push_back(file->d_name); ++idx; } } std::sort(files.begin(), files.begin()+2*n_inds); closedir(dir); // Open output file std::string filename_cc; if (scale != -1) filename_cc = "cross_correlation_01.csv"; else filename_cc = "avgd_cross_correlation.csv"; outputFile1.open(filename_cc.c_str()); // Output file headers if (aggregate || scale == -1) outputFile1 << "id1"<< ", " << "id2" << ", " << "tau" << ", " << "cc" << "\n"; else outputFile1 << "time" << ", " << "id1" << ", " << "id2" << ", " << "tau" << ", " << "cc" << "\n"; //files = {"dir_x00", "dir_x01", "dir_y00", "dir_y01"} for (int a = 0; a < n_inds; ++a) { for (int b = 0; b < n_inds; ++b) { if (b != a) { pFile_x1 = fopen ((directory + files[a]).c_str(), "rb"); pFile_y1 = fopen ((directory + files[a+n_inds]).c_str(), "rb"); pFile_x2 = fopen ((directory + files[b]).c_str(), "rb"); pFile_y2 = fopen ((directory + files[b+n_inds]).c_str(), "rb"); if (pFile_x1==NULL || pFile_y1==NULL || pFile_x2==NULL || pFile_y2==NULL) { fputs ("File error",stderr); exit (1); } // obtain file size fseek (pFile_x1 , 0 , SEEK_END); lSize = ftell (pFile_x1); rewind (pFile_x1); nElements = lSize / sizeof(float); // allocate memory to contain the whole file // device memory cudaMalloc((void **) &d_x1, lSize); cudaMalloc((void **) &d_y1, lSize); cudaMalloc((void **) &d_x2, lSize); cudaMalloc((void **) &d_y2, lSize); cudaMalloc((void **) &d_res, lSize); // host memory x1 = (float*) malloc(lSize); y1 = (float*) malloc(lSize); x2 = (float*) malloc(lSize); y2 = (float*) malloc(lSize); res = (float*) malloc(lSize); if (x1 == NULL || y1==NULL || x2==NULL || y2==NULL || res==NULL) { fputs ("Memory error",stderr); exit (2); } // copy the file into the respective float pointers result_x1 = fread (x1, sizeof(float), nElements, pFile_x1); result_y1 = fread (y1, sizeof(float), nElements, pFile_y1); result_x2 = fread (x2, sizeof(float), nElements, pFile_x2); result_y2 = fread (y2, sizeof(float), nElements, pFile_y2); if (result_x1 != nElements || result_y1 != nElements || result_x2 != nElements || result_y2 != nElements) { fputs ("Reading error",stderr); exit (3); } // the whole files are now loaded in the memory x1, y1, x2 and y2 respectively cudaMemcpy(d_x1, x1, lSize, cudaMemcpyHostToDevice); cudaMemcpy(d_y1, y1, lSize, cudaMemcpyHostToDevice); cudaMemcpy(d_x2, x2, lSize, cudaMemcpyHostToDevice); cudaMemcpy(d_y2, y2, lSize, cudaMemcpyHostToDevice); if (scale*fps > nElements) scale = -1; int tau_max[nElements - scale*fps]; float res_tmp[nElements - scale*fps]; float res_max[nElements - scale*fps]; std::fill_n(tau_max, nElements - scale*fps, 0); std::fill_n(res_tmp, nElements - scale*fps, 0.0); std::fill_n(res_max, nElements - scale*fps, -1.0); for (int tau = -time*fps; tau <= time*fps; ++tau) { kernel<<<(nElements + M - 1) / M, M>>>(d_x1, d_y1, d_x2, d_y2, d_res, tau, na_frames, nElements); cudaMemcpy(res, d_res, lSize, cudaMemcpyDeviceToHost); if (scale == -1) { float res_now = -1.0f; for (int i = na_frames; i < nElements; ++i) { if (res[i] != res[i]) std::cout << x1[i] << " " << y1[i] << " " << i << " " << tau << "\n"; // if nans res_now += res[i]; } outputFile1 << (to_string(files[a][5])).c_str() << (to_string(files[a][6])).c_str() << (to_string(files[a][7])).c_str() << ", " << (to_string(files[b][5])).c_str() << (to_string(files[b][6])).c_str() << (to_string(files[b][7])).c_str() << ", " << tau << ", " << res_now / nElements << "\n"; } else { std::fill_n(res_tmp, nElements - scale*fps, 0.0); for (int i = na_frames; i < nElements - scale*fps; ++i) { for (int j = i; j < i + scale*fps; ++j) { res_tmp[i] += res[j]; if (j == i + scale*fps - 1 && res_max[i] < res_tmp[i]) { res_max[i] = res_tmp[i]; tau_max[i] = tau; } } } } } if (scale != -1) { for (int t = 0; t < nElements - scale*fps; ++t) { outputFile1 << t + scale*fps/2 << ", " << (to_string(files[a][5])).c_str() << (to_string(files[a][6])).c_str() << (to_string(files[a][7])).c_str() << ", " << (to_string(files[b][5])).c_str() << (to_string(files[b][6])).c_str() << (to_string(files[b][7])).c_str() << ", " << tau_max[t] << ", " << res_max[t] / (scale*fps) << "\n"; } } fclose(pFile_x1); fclose(pFile_x2); fclose(pFile_y1); fclose(pFile_y2); cudaFree(d_x1); cudaFree(d_y1); cudaFree(d_x2); cudaFree(d_y2); cudaFree(d_res); free(x1); free(y1); free(x2); free(y2); } } } // terminate fclose(pFile_x1); fclose(pFile_y1); fclose(pFile_x2); fclose(pFile_y2); return 0; }
code for sm_80 Function : _Z6kernelPfS_S_S_S_iil .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ IMAD.MOV.U32 R1, RZ, RZ, c[0x0][0x28] ; /* 0x00000a00ff017624 */ /* 0x000fe400078e00ff */ /*0010*/ S2R R0, SR_CTAID.X ; /* 0x0000000000007919 */ /* 0x000e220000002500 */ /*0020*/ ISETP.LE.AND P0, PT, RZ, c[0x0][0x188], PT ; /* 0x00006200ff007a0c */ /* 0x000fe20003f03270 */ /*0030*/ ULDC.64 UR8, c[0x0][0x118] ; /* 0x0000460000087ab9 */ /* 0x000fe40000000a00 */ /*0040*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e240000002100 */ /*0050*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x001fd000078e0203 */ /*0060*/ @!P0 BRA 0x280 ; /* 0x0000021000008947 */ /* 0x000fea0003800000 */ /*0070*/ ULDC UR5, c[0x0][0x188] ; /* 0x0000620000057ab9 */ /* 0x000fe20000000800 */ /*0080*/ SHF.R.S32.HI R3, RZ, 0x1f, R0 ; /* 0x0000001fff037819 */ /* 0x000fe20000011400 */ /*0090*/ ULDC.64 UR6, c[0x0][0x190] ; /* 0x0000640000067ab9 */ /* 0x000fe40000000a00 */ /*00a0*/ USHF.R.S32.HI UR4, URZ, 0x1f, UR5 ; /* 0x0000001f3f047899 */ /* 0x000fe40008011405 */ /*00b0*/ UIADD3 UR5, UP0, -UR5, UR6, URZ ; /* 0x0000000605057290 */ /* 0x000fc8000ff1e13f */ /*00c0*/ UIADD3.X UR4, ~UR4, UR7, URZ, UP0, !UPT ; /* 0x0000000704047290 */ /* 0x000fe400087fe53f */ /*00d0*/ ISETP.LT.U32.AND P0, PT, R0, UR5, PT ; /* 0x0000000500007c0c */ /* 0x000fc8000bf01070 */ /*00e0*/ IMAD.U32 R2, RZ, RZ, UR4 ; /* 0x00000004ff027e24 */ /* 0x000fca000f8e00ff */ /*00f0*/ ISETP.GT.AND.EX P0, PT, R2, R3, PT, P0 ; /* 0x000000030200720c */ /* 0x000fda0003f04300 */ /*0100*/ @!P0 LEA R4, P1, R0, c[0x0][0x180], 0x2 ; /* 0x0000600000048a11 */ /* 0x000fc800078210ff */ /*0110*/ @!P0 LEA.HI.X R5, R0, c[0x0][0x184], R3, 0x2, P1 ; /* 0x0000610000058a11 */ /* 0x000fca00008f1403 */ /*0120*/ @!P0 STG.E [R4.64], RZ ; /* 0x000000ff04008986 */ /* 0x0001e2000c101908 */ /*0130*/ @!P0 EXIT ; /* 0x000000000000894d */ /* 0x000fea0003800000 */ /*0140*/ IMAD.SHL.U32 R10, R0.reuse, 0x4, RZ ; /* 0x00000004000a7824 */ /* 0x040fe200078e00ff */ /*0150*/ IADD3 R2, R0.reuse, c[0x0][0x188], RZ ; /* 0x0000620000027a10 */ /* 0x040fe20007ffe0ff */ /*0160*/ IMAD.MOV.U32 R7, RZ, RZ, 0x4 ; /* 0x00000004ff077424 */ /* 0x000fe200078e00ff */ /*0170*/ SHF.L.U64.HI R11, R0, 0x2, R3 ; /* 0x00000002000b7819 */ /* 0x000fe40000010203 */ /*0180*/ IADD3 R8, P1, R10, c[0x0][0x168], RZ ; /* 0x00005a000a087a10 */ /* 0x000fe20007f3e0ff */ /*0190*/ IMAD.WIDE R4, R2, R7.reuse, c[0x0][0x178] ; /* 0x00005e0002047625 */ /* 0x081fe200078e0207 */ /*01a0*/ IADD3 R6, P0, R10, c[0x0][0x160], RZ ; /* 0x000058000a067a10 */ /* 0x000fe40007f1e0ff */ /*01b0*/ IADD3.X R9, R11.reuse, c[0x0][0x16c], RZ, P1, !PT ; /* 0x00005b000b097a10 */ /* 0x040fe20000ffe4ff */ /*01c0*/ IMAD.WIDE R2, R2, R7, c[0x0][0x170] ; /* 0x00005c0002027625 */ /* 0x000fe200078e0207 */ /*01d0*/ IADD3.X R7, R11, c[0x0][0x164], RZ, P0, !PT ; /* 0x000059000b077a10 */ /* 0x000fe200007fe4ff */ /*01e0*/ LDG.E R4, [R4.64] ; /* 0x0000000804047981 */ /* 0x000ea8000c1e1900 */ /*01f0*/ LDG.E R9, [R8.64] ; /* 0x0000000808097981 */ /* 0x000ea8000c1e1900 */ /*0200*/ LDG.E R2, [R2.64] ; /* 0x0000000802027981 */ /* 0x000ee8000c1e1900 */ /*0210*/ LDG.E R7, [R6.64] ; /* 0x0000000806077981 */ /* 0x000ee2000c1e1900 */ /*0220*/ IADD3 R10, P0, R10, c[0x0][0x180], RZ ; /* 0x000060000a0a7a10 */ /* 0x000fc80007f1e0ff */ /*0230*/ IADD3.X R11, R11, c[0x0][0x184], RZ, P0, !PT ; /* 0x000061000b0b7a10 */ /* 0x000fe200007fe4ff */ /*0240*/ FMUL R0, R4, R9 ; /* 0x0000000904007220 */ /* 0x004fc80000400000 */ /*0250*/ FFMA R13, R2, R7, R0 ; /* 0x00000007020d7223 */ /* 0x008fca0000000000 */ /*0260*/ STG.E [R10.64], R13 ; /* 0x0000000d0a007986 */ /* 0x000fe2000c101908 */ /*0270*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0280*/ ULDC.64 UR4, c[0x0][0x188] ; /* 0x0000620000047ab9 */ /* 0x000fe20000000a00 */ /*0290*/ SHF.R.S32.HI R3, RZ, 0x1f, R0 ; /* 0x0000001fff037819 */ /* 0x000fe20000011400 */ /*02a0*/ UIADD3 UR4, -UR4, UR5, URZ ; /* 0x0000000504047290 */ /* 0x000fcc000fffe13f */ /*02b0*/ ISETP.GE.AND P0, PT, R0, UR4, PT ; /* 0x0000000400007c0c */ /* 0x000fda000bf06270 */ /*02c0*/ @!P0 BRA 0x410 ; /* 0x0000014000008947 */ /* 0x000fea0003800000 */ /*02d0*/ IMAD.SHL.U32 R10, R0.reuse, 0x4, RZ ; /* 0x00000004000a7824 */ /* 0x040fe200078e00ff */ /*02e0*/ IADD3 R2, R0.reuse, c[0x0][0x188], RZ ; /* 0x0000620000027a10 */ /* 0x040fe20007ffe0ff */ /*02f0*/ IMAD.MOV.U32 R5, RZ, RZ, 0x4 ; /* 0x00000004ff057424 */ /* 0x000fe200078e00ff */ /*0300*/ SHF.L.U64.HI R11, R0, 0x2, R3 ; /* 0x00000002000b7819 */ /* 0x000fe40000010203 */ /*0310*/ IADD3 R8, P1, R10, c[0x0][0x168], RZ ; /* 0x00005a000a087a10 */ /* 0x000fe20007f3e0ff */ /*0320*/ IMAD.WIDE R6, R2, R5.reuse, c[0x0][0x178] ; /* 0x00005e0002067625 */ /* 0x080fe200078e0205 */ /*0330*/ IADD3 R4, P0, R10, c[0x0][0x160], RZ ; /* 0x000058000a047a10 */ /* 0x000fe40007f1e0ff */ /*0340*/ IADD3.X R9, R11.reuse, c[0x0][0x16c], RZ, P1, !PT ; /* 0x00005b000b097a10 */ /* 0x040fe20000ffe4ff */ /*0350*/ IMAD.WIDE R2, R2, R5, c[0x0][0x170] ; /* 0x00005c0002027625 */ /* 0x000fe200078e0205 */ /*0360*/ IADD3.X R5, R11, c[0x0][0x164], RZ, P0, !PT ; /* 0x000059000b057a10 */ /* 0x000fe200007fe4ff */ /*0370*/ LDG.E R6, [R6.64] ; /* 0x0000000806067981 */ /* 0x000ea8000c1e1900 */ /*0380*/ LDG.E R9, [R8.64] ; /* 0x0000000808097981 */ /* 0x000ea8000c1e1900 */ /*0390*/ LDG.E R2, [R2.64] ; /* 0x0000000802027981 */ /* 0x000ee8000c1e1900 */ /*03a0*/ LDG.E R5, [R4.64] ; /* 0x0000000804057981 */ /* 0x000ee2000c1e1900 */ /*03b0*/ IADD3 R10, P0, R10, c[0x0][0x180], RZ ; /* 0x000060000a0a7a10 */ /* 0x000fc80007f1e0ff */ /*03c0*/ IADD3.X R11, R11, c[0x0][0x184], RZ, P0, !PT ; /* 0x000061000b0b7a10 */ /* 0x000fe200007fe4ff */ /*03d0*/ FMUL R0, R6, R9 ; /* 0x0000000906007220 */ /* 0x004fc80000400000 */ /*03e0*/ FFMA R13, R2, R5, R0 ; /* 0x00000005020d7223 */ /* 0x008fca0000000000 */ /*03f0*/ STG.E [R10.64], R13 ; /* 0x0000000d0a007986 */ /* 0x000fe2000c101908 */ /*0400*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0410*/ LEA R2, P0, R0, c[0x0][0x180], 0x2 ; /* 0x0000600000027a11 */ /* 0x000fc800078010ff */ /*0420*/ LEA.HI.X R3, R0, c[0x0][0x184], R3, 0x2, P0 ; /* 0x0000610000037a11 */ /* 0x000fca00000f1403 */ /*0430*/ STG.E [R2.64], RZ ; /* 0x000000ff02007986 */ /* 0x000fe2000c101908 */ /*0440*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0450*/ BRA 0x450; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0460*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0470*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0480*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0490*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
// // CrossCorrelation.cu // CrossCorrelation // // Created by Vivek Sridhar on 29/06/17. // Copyright © 2017 Vivek Sridhar. All rights reserved. // #include <iostream> #include <fstream> #include <sstream> #include <vector> #include <algorithm> #include <stdio.h> #include <stdlib.h> #include <dirent.h> template <typename T> std::string to_string(const T& value) { std::stringstream ss; ss << value; return ss.str(); } long factorial(long val) { std::cout << val << "\n"; long result = 1; for (long i = 1; i <= val; ++i) { result *= i; } return result; } long combination(long n, long r) { return (factorial(n)) / ((factorial(n - r)) * factorial(r)); } __global__ void kernel(float *x1, float *y1, float *x2, float *y2, float *res, int tau, int na_frames, long nElements) { int index = blockIdx.x * blockDim.x + threadIdx.x; if (tau < 0) { if (index >= -tau+na_frames) { res[index] = x1[index] * x2[index + tau] + y1[index] * y2[index + tau]; } else res[index] = 0.0; } else { if (index < nElements - tau) { res[index] = x1[index] * x2[index + tau] + y1[index] * y2[index + tau]; } else res[index] = 0.0; } } // total measurement points in the time series is defined by nElements #define M 1024 // number of threads per block #define fps 10 // frames per second of input video (used to determine tau) #define time 5 // time in seconds within which time delayed cross correlation is calculated (tau ranges from -time*fps to time*fps) #define n_inds 10 int na_frames = 0; // number of frames in the start with nas int scale = 1; // time window for analysis in seconds; varying this allows us to examine dynamics of leadership across varying timescales; setting scale larger than the entire time series or -1 gives aggregated statistics across the entire duration (otherwise, timescale of analysis is scale*fps) //const int pairs = combination(n_inds, 2); const bool aggregate = false; // this boolean decides whether you output a dynamic time variable leadership network or a static time aggregated network; scale is set to -1 if aggregate is true std::ofstream outputFile1; int main () { DIR *dir; FILE *pFile_x1; FILE *pFile_y1; FILE *pFile_x2; FILE *pFile_y2; long lSize; long nElements; struct dirent *file; float *d_x1, *d_y1, *d_x2, *d_y2, *d_res; float *x1, *y1, *x2, *y2, *res; size_t result_x1, result_y1, result_x2, result_y2; if (aggregate) scale = -1; std::vector<std::string> files; std::string directory = "/home/user/Documents/Vivek/cuda/DirectionalCorrelation/Data/Input/pigeons/10_birds/ffA3/cross_correlation/"; dir = opendir(directory.c_str()); int idx = 0; while ((file = readdir(dir)) != NULL) { if (file->d_name[0] == 'd') { files.push_back(file->d_name); ++idx; } } std::sort(files.begin(), files.begin()+2*n_inds); closedir(dir); // Open output file std::string filename_cc; if (scale != -1) filename_cc = "cross_correlation_01.csv"; else filename_cc = "avgd_cross_correlation.csv"; outputFile1.open(filename_cc.c_str()); // Output file headers if (aggregate || scale == -1) outputFile1 << "id1"<< ", " << "id2" << ", " << "tau" << ", " << "cc" << "\n"; else outputFile1 << "time" << ", " << "id1" << ", " << "id2" << ", " << "tau" << ", " << "cc" << "\n"; //files = {"dir_x00", "dir_x01", "dir_y00", "dir_y01"} for (int a = 0; a < n_inds; ++a) { for (int b = 0; b < n_inds; ++b) { if (b != a) { pFile_x1 = fopen ((directory + files[a]).c_str(), "rb"); pFile_y1 = fopen ((directory + files[a+n_inds]).c_str(), "rb"); pFile_x2 = fopen ((directory + files[b]).c_str(), "rb"); pFile_y2 = fopen ((directory + files[b+n_inds]).c_str(), "rb"); if (pFile_x1==NULL || pFile_y1==NULL || pFile_x2==NULL || pFile_y2==NULL) { fputs ("File error",stderr); exit (1); } // obtain file size fseek (pFile_x1 , 0 , SEEK_END); lSize = ftell (pFile_x1); rewind (pFile_x1); nElements = lSize / sizeof(float); // allocate memory to contain the whole file // device memory cudaMalloc((void **) &d_x1, lSize); cudaMalloc((void **) &d_y1, lSize); cudaMalloc((void **) &d_x2, lSize); cudaMalloc((void **) &d_y2, lSize); cudaMalloc((void **) &d_res, lSize); // host memory x1 = (float*) malloc(lSize); y1 = (float*) malloc(lSize); x2 = (float*) malloc(lSize); y2 = (float*) malloc(lSize); res = (float*) malloc(lSize); if (x1 == NULL || y1==NULL || x2==NULL || y2==NULL || res==NULL) { fputs ("Memory error",stderr); exit (2); } // copy the file into the respective float pointers result_x1 = fread (x1, sizeof(float), nElements, pFile_x1); result_y1 = fread (y1, sizeof(float), nElements, pFile_y1); result_x2 = fread (x2, sizeof(float), nElements, pFile_x2); result_y2 = fread (y2, sizeof(float), nElements, pFile_y2); if (result_x1 != nElements || result_y1 != nElements || result_x2 != nElements || result_y2 != nElements) { fputs ("Reading error",stderr); exit (3); } // the whole files are now loaded in the memory x1, y1, x2 and y2 respectively cudaMemcpy(d_x1, x1, lSize, cudaMemcpyHostToDevice); cudaMemcpy(d_y1, y1, lSize, cudaMemcpyHostToDevice); cudaMemcpy(d_x2, x2, lSize, cudaMemcpyHostToDevice); cudaMemcpy(d_y2, y2, lSize, cudaMemcpyHostToDevice); if (scale*fps > nElements) scale = -1; int tau_max[nElements - scale*fps]; float res_tmp[nElements - scale*fps]; float res_max[nElements - scale*fps]; std::fill_n(tau_max, nElements - scale*fps, 0); std::fill_n(res_tmp, nElements - scale*fps, 0.0); std::fill_n(res_max, nElements - scale*fps, -1.0); for (int tau = -time*fps; tau <= time*fps; ++tau) { kernel<<<(nElements + M - 1) / M, M>>>(d_x1, d_y1, d_x2, d_y2, d_res, tau, na_frames, nElements); cudaMemcpy(res, d_res, lSize, cudaMemcpyDeviceToHost); if (scale == -1) { float res_now = -1.0f; for (int i = na_frames; i < nElements; ++i) { if (res[i] != res[i]) std::cout << x1[i] << " " << y1[i] << " " << i << " " << tau << "\n"; // if nans res_now += res[i]; } outputFile1 << (to_string(files[a][5])).c_str() << (to_string(files[a][6])).c_str() << (to_string(files[a][7])).c_str() << ", " << (to_string(files[b][5])).c_str() << (to_string(files[b][6])).c_str() << (to_string(files[b][7])).c_str() << ", " << tau << ", " << res_now / nElements << "\n"; } else { std::fill_n(res_tmp, nElements - scale*fps, 0.0); for (int i = na_frames; i < nElements - scale*fps; ++i) { for (int j = i; j < i + scale*fps; ++j) { res_tmp[i] += res[j]; if (j == i + scale*fps - 1 && res_max[i] < res_tmp[i]) { res_max[i] = res_tmp[i]; tau_max[i] = tau; } } } } } if (scale != -1) { for (int t = 0; t < nElements - scale*fps; ++t) { outputFile1 << t + scale*fps/2 << ", " << (to_string(files[a][5])).c_str() << (to_string(files[a][6])).c_str() << (to_string(files[a][7])).c_str() << ", " << (to_string(files[b][5])).c_str() << (to_string(files[b][6])).c_str() << (to_string(files[b][7])).c_str() << ", " << tau_max[t] << ", " << res_max[t] / (scale*fps) << "\n"; } } fclose(pFile_x1); fclose(pFile_x2); fclose(pFile_y1); fclose(pFile_y2); cudaFree(d_x1); cudaFree(d_y1); cudaFree(d_x2); cudaFree(d_y2); cudaFree(d_res); free(x1); free(y1); free(x2); free(y2); } } } // terminate fclose(pFile_x1); fclose(pFile_y1); fclose(pFile_x2); fclose(pFile_y2); return 0; }
// // CrossCorrelation.cu // CrossCorrelation // // Created by Vivek Sridhar on 29/06/17. // Copyright © 2017 Vivek Sridhar. All rights reserved. // #include <hip/hip_runtime.h> #include <iostream> #include <fstream> #include <sstream> #include <vector> #include <algorithm> #include <stdio.h> #include <stdlib.h> #include <dirent.h> template <typename T> std::string to_string(const T& value) { std::stringstream ss; ss << value; return ss.str(); } long factorial(long val) { std::cout << val << "\n"; long result = 1; for (long i = 1; i <= val; ++i) { result *= i; } return result; } long combination(long n, long r) { return (factorial(n)) / ((factorial(n - r)) * factorial(r)); } __global__ void kernel(float *x1, float *y1, float *x2, float *y2, float *res, int tau, int na_frames, long nElements) { int index = blockIdx.x * blockDim.x + threadIdx.x; if (tau < 0) { if (index >= -tau+na_frames) { res[index] = x1[index] * x2[index + tau] + y1[index] * y2[index + tau]; } else res[index] = 0.0; } else { if (index < nElements - tau) { res[index] = x1[index] * x2[index + tau] + y1[index] * y2[index + tau]; } else res[index] = 0.0; } } // total measurement points in the time series is defined by nElements #define M 1024 // number of threads per block #define fps 10 // frames per second of input video (used to determine tau) #define time 5 // time in seconds within which time delayed cross correlation is calculated (tau ranges from -time*fps to time*fps) #define n_inds 10 int na_frames = 0; // number of frames in the start with nas int scale = 1; // time window for analysis in seconds; varying this allows us to examine dynamics of leadership across varying timescales; setting scale larger than the entire time series or -1 gives aggregated statistics across the entire duration (otherwise, timescale of analysis is scale*fps) //const int pairs = combination(n_inds, 2); const bool aggregate = false; // this boolean decides whether you output a dynamic time variable leadership network or a static time aggregated network; scale is set to -1 if aggregate is true std::ofstream outputFile1; int main () { DIR *dir; FILE *pFile_x1; FILE *pFile_y1; FILE *pFile_x2; FILE *pFile_y2; long lSize; long nElements; struct dirent *file; float *d_x1, *d_y1, *d_x2, *d_y2, *d_res; float *x1, *y1, *x2, *y2, *res; size_t result_x1, result_y1, result_x2, result_y2; if (aggregate) scale = -1; std::vector<std::string> files; std::string directory = "/home/user/Documents/Vivek/cuda/DirectionalCorrelation/Data/Input/pigeons/10_birds/ffA3/cross_correlation/"; dir = opendir(directory.c_str()); int idx = 0; while ((file = readdir(dir)) != NULL) { if (file->d_name[0] == 'd') { files.push_back(file->d_name); ++idx; } } std::sort(files.begin(), files.begin()+2*n_inds); closedir(dir); // Open output file std::string filename_cc; if (scale != -1) filename_cc = "cross_correlation_01.csv"; else filename_cc = "avgd_cross_correlation.csv"; outputFile1.open(filename_cc.c_str()); // Output file headers if (aggregate || scale == -1) outputFile1 << "id1"<< ", " << "id2" << ", " << "tau" << ", " << "cc" << "\n"; else outputFile1 << "time" << ", " << "id1" << ", " << "id2" << ", " << "tau" << ", " << "cc" << "\n"; //files = {"dir_x00", "dir_x01", "dir_y00", "dir_y01"} for (int a = 0; a < n_inds; ++a) { for (int b = 0; b < n_inds; ++b) { if (b != a) { pFile_x1 = fopen ((directory + files[a]).c_str(), "rb"); pFile_y1 = fopen ((directory + files[a+n_inds]).c_str(), "rb"); pFile_x2 = fopen ((directory + files[b]).c_str(), "rb"); pFile_y2 = fopen ((directory + files[b+n_inds]).c_str(), "rb"); if (pFile_x1==NULL || pFile_y1==NULL || pFile_x2==NULL || pFile_y2==NULL) { fputs ("File error",stderr); exit (1); } // obtain file size fseek (pFile_x1 , 0 , SEEK_END); lSize = ftell (pFile_x1); rewind (pFile_x1); nElements = lSize / sizeof(float); // allocate memory to contain the whole file // device memory hipMalloc((void **) &d_x1, lSize); hipMalloc((void **) &d_y1, lSize); hipMalloc((void **) &d_x2, lSize); hipMalloc((void **) &d_y2, lSize); hipMalloc((void **) &d_res, lSize); // host memory x1 = (float*) malloc(lSize); y1 = (float*) malloc(lSize); x2 = (float*) malloc(lSize); y2 = (float*) malloc(lSize); res = (float*) malloc(lSize); if (x1 == NULL || y1==NULL || x2==NULL || y2==NULL || res==NULL) { fputs ("Memory error",stderr); exit (2); } // copy the file into the respective float pointers result_x1 = fread (x1, sizeof(float), nElements, pFile_x1); result_y1 = fread (y1, sizeof(float), nElements, pFile_y1); result_x2 = fread (x2, sizeof(float), nElements, pFile_x2); result_y2 = fread (y2, sizeof(float), nElements, pFile_y2); if (result_x1 != nElements || result_y1 != nElements || result_x2 != nElements || result_y2 != nElements) { fputs ("Reading error",stderr); exit (3); } // the whole files are now loaded in the memory x1, y1, x2 and y2 respectively hipMemcpy(d_x1, x1, lSize, hipMemcpyHostToDevice); hipMemcpy(d_y1, y1, lSize, hipMemcpyHostToDevice); hipMemcpy(d_x2, x2, lSize, hipMemcpyHostToDevice); hipMemcpy(d_y2, y2, lSize, hipMemcpyHostToDevice); if (scale*fps > nElements) scale = -1; int tau_max[nElements - scale*fps]; float res_tmp[nElements - scale*fps]; float res_max[nElements - scale*fps]; std::fill_n(tau_max, nElements - scale*fps, 0); std::fill_n(res_tmp, nElements - scale*fps, 0.0); std::fill_n(res_max, nElements - scale*fps, -1.0); for (int tau = -time*fps; tau <= time*fps; ++tau) { kernel<<<(nElements + M - 1) / M, M>>>(d_x1, d_y1, d_x2, d_y2, d_res, tau, na_frames, nElements); hipMemcpy(res, d_res, lSize, hipMemcpyDeviceToHost); if (scale == -1) { float res_now = -1.0f; for (int i = na_frames; i < nElements; ++i) { if (res[i] != res[i]) std::cout << x1[i] << " " << y1[i] << " " << i << " " << tau << "\n"; // if nans res_now += res[i]; } outputFile1 << (to_string(files[a][5])).c_str() << (to_string(files[a][6])).c_str() << (to_string(files[a][7])).c_str() << ", " << (to_string(files[b][5])).c_str() << (to_string(files[b][6])).c_str() << (to_string(files[b][7])).c_str() << ", " << tau << ", " << res_now / nElements << "\n"; } else { std::fill_n(res_tmp, nElements - scale*fps, 0.0); for (int i = na_frames; i < nElements - scale*fps; ++i) { for (int j = i; j < i + scale*fps; ++j) { res_tmp[i] += res[j]; if (j == i + scale*fps - 1 && res_max[i] < res_tmp[i]) { res_max[i] = res_tmp[i]; tau_max[i] = tau; } } } } } if (scale != -1) { for (int t = 0; t < nElements - scale*fps; ++t) { outputFile1 << t + scale*fps/2 << ", " << (to_string(files[a][5])).c_str() << (to_string(files[a][6])).c_str() << (to_string(files[a][7])).c_str() << ", " << (to_string(files[b][5])).c_str() << (to_string(files[b][6])).c_str() << (to_string(files[b][7])).c_str() << ", " << tau_max[t] << ", " << res_max[t] / (scale*fps) << "\n"; } } fclose(pFile_x1); fclose(pFile_x2); fclose(pFile_y1); fclose(pFile_y2); hipFree(d_x1); hipFree(d_y1); hipFree(d_x2); hipFree(d_y2); hipFree(d_res); free(x1); free(y1); free(x2); free(y2); } } } // terminate fclose(pFile_x1); fclose(pFile_y1); fclose(pFile_x2); fclose(pFile_y2); return 0; }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
// // CrossCorrelation.cu // CrossCorrelation // // Created by Vivek Sridhar on 29/06/17. // Copyright © 2017 Vivek Sridhar. All rights reserved. // #include <hip/hip_runtime.h> #include <iostream> #include <fstream> #include <sstream> #include <vector> #include <algorithm> #include <stdio.h> #include <stdlib.h> #include <dirent.h> template <typename T> std::string to_string(const T& value) { std::stringstream ss; ss << value; return ss.str(); } long factorial(long val) { std::cout << val << "\n"; long result = 1; for (long i = 1; i <= val; ++i) { result *= i; } return result; } long combination(long n, long r) { return (factorial(n)) / ((factorial(n - r)) * factorial(r)); } __global__ void kernel(float *x1, float *y1, float *x2, float *y2, float *res, int tau, int na_frames, long nElements) { int index = blockIdx.x * blockDim.x + threadIdx.x; if (tau < 0) { if (index >= -tau+na_frames) { res[index] = x1[index] * x2[index + tau] + y1[index] * y2[index + tau]; } else res[index] = 0.0; } else { if (index < nElements - tau) { res[index] = x1[index] * x2[index + tau] + y1[index] * y2[index + tau]; } else res[index] = 0.0; } } // total measurement points in the time series is defined by nElements #define M 1024 // number of threads per block #define fps 10 // frames per second of input video (used to determine tau) #define time 5 // time in seconds within which time delayed cross correlation is calculated (tau ranges from -time*fps to time*fps) #define n_inds 10 int na_frames = 0; // number of frames in the start with nas int scale = 1; // time window for analysis in seconds; varying this allows us to examine dynamics of leadership across varying timescales; setting scale larger than the entire time series or -1 gives aggregated statistics across the entire duration (otherwise, timescale of analysis is scale*fps) //const int pairs = combination(n_inds, 2); const bool aggregate = false; // this boolean decides whether you output a dynamic time variable leadership network or a static time aggregated network; scale is set to -1 if aggregate is true std::ofstream outputFile1; int main () { DIR *dir; FILE *pFile_x1; FILE *pFile_y1; FILE *pFile_x2; FILE *pFile_y2; long lSize; long nElements; struct dirent *file; float *d_x1, *d_y1, *d_x2, *d_y2, *d_res; float *x1, *y1, *x2, *y2, *res; size_t result_x1, result_y1, result_x2, result_y2; if (aggregate) scale = -1; std::vector<std::string> files; std::string directory = "/home/user/Documents/Vivek/cuda/DirectionalCorrelation/Data/Input/pigeons/10_birds/ffA3/cross_correlation/"; dir = opendir(directory.c_str()); int idx = 0; while ((file = readdir(dir)) != NULL) { if (file->d_name[0] == 'd') { files.push_back(file->d_name); ++idx; } } std::sort(files.begin(), files.begin()+2*n_inds); closedir(dir); // Open output file std::string filename_cc; if (scale != -1) filename_cc = "cross_correlation_01.csv"; else filename_cc = "avgd_cross_correlation.csv"; outputFile1.open(filename_cc.c_str()); // Output file headers if (aggregate || scale == -1) outputFile1 << "id1"<< ", " << "id2" << ", " << "tau" << ", " << "cc" << "\n"; else outputFile1 << "time" << ", " << "id1" << ", " << "id2" << ", " << "tau" << ", " << "cc" << "\n"; //files = {"dir_x00", "dir_x01", "dir_y00", "dir_y01"} for (int a = 0; a < n_inds; ++a) { for (int b = 0; b < n_inds; ++b) { if (b != a) { pFile_x1 = fopen ((directory + files[a]).c_str(), "rb"); pFile_y1 = fopen ((directory + files[a+n_inds]).c_str(), "rb"); pFile_x2 = fopen ((directory + files[b]).c_str(), "rb"); pFile_y2 = fopen ((directory + files[b+n_inds]).c_str(), "rb"); if (pFile_x1==NULL || pFile_y1==NULL || pFile_x2==NULL || pFile_y2==NULL) { fputs ("File error",stderr); exit (1); } // obtain file size fseek (pFile_x1 , 0 , SEEK_END); lSize = ftell (pFile_x1); rewind (pFile_x1); nElements = lSize / sizeof(float); // allocate memory to contain the whole file // device memory hipMalloc((void **) &d_x1, lSize); hipMalloc((void **) &d_y1, lSize); hipMalloc((void **) &d_x2, lSize); hipMalloc((void **) &d_y2, lSize); hipMalloc((void **) &d_res, lSize); // host memory x1 = (float*) malloc(lSize); y1 = (float*) malloc(lSize); x2 = (float*) malloc(lSize); y2 = (float*) malloc(lSize); res = (float*) malloc(lSize); if (x1 == NULL || y1==NULL || x2==NULL || y2==NULL || res==NULL) { fputs ("Memory error",stderr); exit (2); } // copy the file into the respective float pointers result_x1 = fread (x1, sizeof(float), nElements, pFile_x1); result_y1 = fread (y1, sizeof(float), nElements, pFile_y1); result_x2 = fread (x2, sizeof(float), nElements, pFile_x2); result_y2 = fread (y2, sizeof(float), nElements, pFile_y2); if (result_x1 != nElements || result_y1 != nElements || result_x2 != nElements || result_y2 != nElements) { fputs ("Reading error",stderr); exit (3); } // the whole files are now loaded in the memory x1, y1, x2 and y2 respectively hipMemcpy(d_x1, x1, lSize, hipMemcpyHostToDevice); hipMemcpy(d_y1, y1, lSize, hipMemcpyHostToDevice); hipMemcpy(d_x2, x2, lSize, hipMemcpyHostToDevice); hipMemcpy(d_y2, y2, lSize, hipMemcpyHostToDevice); if (scale*fps > nElements) scale = -1; int tau_max[nElements - scale*fps]; float res_tmp[nElements - scale*fps]; float res_max[nElements - scale*fps]; std::fill_n(tau_max, nElements - scale*fps, 0); std::fill_n(res_tmp, nElements - scale*fps, 0.0); std::fill_n(res_max, nElements - scale*fps, -1.0); for (int tau = -time*fps; tau <= time*fps; ++tau) { kernel<<<(nElements + M - 1) / M, M>>>(d_x1, d_y1, d_x2, d_y2, d_res, tau, na_frames, nElements); hipMemcpy(res, d_res, lSize, hipMemcpyDeviceToHost); if (scale == -1) { float res_now = -1.0f; for (int i = na_frames; i < nElements; ++i) { if (res[i] != res[i]) std::cout << x1[i] << " " << y1[i] << " " << i << " " << tau << "\n"; // if nans res_now += res[i]; } outputFile1 << (to_string(files[a][5])).c_str() << (to_string(files[a][6])).c_str() << (to_string(files[a][7])).c_str() << ", " << (to_string(files[b][5])).c_str() << (to_string(files[b][6])).c_str() << (to_string(files[b][7])).c_str() << ", " << tau << ", " << res_now / nElements << "\n"; } else { std::fill_n(res_tmp, nElements - scale*fps, 0.0); for (int i = na_frames; i < nElements - scale*fps; ++i) { for (int j = i; j < i + scale*fps; ++j) { res_tmp[i] += res[j]; if (j == i + scale*fps - 1 && res_max[i] < res_tmp[i]) { res_max[i] = res_tmp[i]; tau_max[i] = tau; } } } } } if (scale != -1) { for (int t = 0; t < nElements - scale*fps; ++t) { outputFile1 << t + scale*fps/2 << ", " << (to_string(files[a][5])).c_str() << (to_string(files[a][6])).c_str() << (to_string(files[a][7])).c_str() << ", " << (to_string(files[b][5])).c_str() << (to_string(files[b][6])).c_str() << (to_string(files[b][7])).c_str() << ", " << tau_max[t] << ", " << res_max[t] / (scale*fps) << "\n"; } } fclose(pFile_x1); fclose(pFile_x2); fclose(pFile_y1); fclose(pFile_y2); hipFree(d_x1); hipFree(d_y1); hipFree(d_x2); hipFree(d_y2); hipFree(d_res); free(x1); free(y1); free(x2); free(y2); } } } // terminate fclose(pFile_x1); fclose(pFile_y1); fclose(pFile_x2); fclose(pFile_y2); return 0; }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z6kernelPfS_S_S_S_iil .globl _Z6kernelPfS_S_S_S_iil .p2align 8 .type _Z6kernelPfS_S_S_S_iil,@function _Z6kernelPfS_S_S_S_iil: s_clause 0x1 s_load_b32 s3, s[0:1], 0x44 s_load_b32 s2, s[0:1], 0x28 s_waitcnt lgkmcnt(0) s_and_b32 s3, s3, 0xffff s_cmp_gt_i32 s2, -1 v_mad_u64_u32 v[1:2], null, s15, s3, v[0:1] s_delay_alu instid0(VALU_DEP_1) v_ashrrev_i32_e32 v2, 31, v1 s_cbranch_scc0 .LBB0_2 s_load_b64 s[4:5], s[0:1], 0x30 s_mov_b32 s3, 0 s_waitcnt lgkmcnt(0) s_sub_u32 s4, s4, s2 s_subb_u32 s5, s5, 0 s_delay_alu instid0(SALU_CYCLE_1) v_cmp_gt_i64_e32 vcc_lo, s[4:5], v[1:2] s_and_b32 s4, vcc_lo, exec_lo s_cbranch_execz .LBB0_3 s_branch .LBB0_4 .LBB0_2: s_mov_b32 s3, -1 s_mov_b32 s4, 0 .LBB0_3: s_load_b32 s3, s[0:1], 0x2c s_and_not1_b32 s4, s4, exec_lo s_waitcnt lgkmcnt(0) s_sub_i32 s3, s3, s2 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_2) | instid1(SALU_CYCLE_1) v_cmp_le_i32_e32 vcc_lo, s3, v1 s_mov_b32 s3, 0 s_and_b32 s5, vcc_lo, exec_lo s_or_b32 s4, s4, s5 .LBB0_4: v_mov_b32_e32 v0, s3 s_and_saveexec_b32 s3, s4 s_cbranch_execz .LBB0_6 s_load_b256 s[4:11], s[0:1], 0x0 v_add_nc_u32_e32 v3, s2, v1 v_lshlrev_b64 v[5:6], 2, v[1:2] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v4, 31, v3 v_lshlrev_b64 v[3:4], 2, v[3:4] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_u32 v7, vcc_lo, s4, v5 v_add_co_ci_u32_e32 v8, vcc_lo, s5, v6, vcc_lo v_add_co_u32 v5, vcc_lo, s6, v5 v_add_co_ci_u32_e32 v6, vcc_lo, s7, v6, vcc_lo v_add_co_u32 v9, vcc_lo, s10, v3 v_add_co_ci_u32_e32 v10, vcc_lo, s11, v4, vcc_lo v_add_co_u32 v3, vcc_lo, s8, v3 v_add_co_ci_u32_e32 v4, vcc_lo, s9, v4, vcc_lo global_load_b32 v0, v[5:6], off global_load_b32 v5, v[9:10], off global_load_b32 v6, v[7:8], off global_load_b32 v3, v[3:4], off s_waitcnt vmcnt(2) v_mul_f32_e32 v0, v0, v5 s_waitcnt vmcnt(0) s_delay_alu instid0(VALU_DEP_1) v_fmac_f32_e32 v0, v6, v3 .LBB0_6: s_or_b32 exec_lo, exec_lo, s3 s_load_b64 s[0:1], s[0:1], 0x20 v_lshlrev_b64 v[1:2], 2, v[1:2] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v1, vcc_lo, s0, v1 v_add_co_ci_u32_e32 v2, vcc_lo, s1, v2, vcc_lo global_store_b32 v[1:2], v0, off s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z6kernelPfS_S_S_S_iil .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 312 .amdhsa_user_sgpr_count 15 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 0 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 0 .amdhsa_next_free_vgpr 11 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z6kernelPfS_S_S_S_iil, .Lfunc_end0-_Z6kernelPfS_S_S_S_iil .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 8 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 24 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 32 .size: 8 .value_kind: global_buffer - .offset: 40 .size: 4 .value_kind: by_value - .offset: 44 .size: 4 .value_kind: by_value - .offset: 48 .size: 8 .value_kind: by_value - .offset: 56 .size: 4 .value_kind: hidden_block_count_x - .offset: 60 .size: 4 .value_kind: hidden_block_count_y - .offset: 64 .size: 4 .value_kind: hidden_block_count_z - .offset: 68 .size: 2 .value_kind: hidden_group_size_x - .offset: 70 .size: 2 .value_kind: hidden_group_size_y - .offset: 72 .size: 2 .value_kind: hidden_group_size_z - .offset: 74 .size: 2 .value_kind: hidden_remainder_x - .offset: 76 .size: 2 .value_kind: hidden_remainder_y - .offset: 78 .size: 2 .value_kind: hidden_remainder_z - .offset: 96 .size: 8 .value_kind: hidden_global_offset_x - .offset: 104 .size: 8 .value_kind: hidden_global_offset_y - .offset: 112 .size: 8 .value_kind: hidden_global_offset_z - .offset: 120 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 312 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z6kernelPfS_S_S_S_iil .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z6kernelPfS_S_S_S_iil.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 11 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting from CUDA device assembly to AMD device assembly.
code for sm_80 Function : _Z6kernelPfS_S_S_S_iil .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ IMAD.MOV.U32 R1, RZ, RZ, c[0x0][0x28] ; /* 0x00000a00ff017624 */ /* 0x000fe400078e00ff */ /*0010*/ S2R R0, SR_CTAID.X ; /* 0x0000000000007919 */ /* 0x000e220000002500 */ /*0020*/ ISETP.LE.AND P0, PT, RZ, c[0x0][0x188], PT ; /* 0x00006200ff007a0c */ /* 0x000fe20003f03270 */ /*0030*/ ULDC.64 UR8, c[0x0][0x118] ; /* 0x0000460000087ab9 */ /* 0x000fe40000000a00 */ /*0040*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e240000002100 */ /*0050*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x001fd000078e0203 */ /*0060*/ @!P0 BRA 0x280 ; /* 0x0000021000008947 */ /* 0x000fea0003800000 */ /*0070*/ ULDC UR5, c[0x0][0x188] ; /* 0x0000620000057ab9 */ /* 0x000fe20000000800 */ /*0080*/ SHF.R.S32.HI R3, RZ, 0x1f, R0 ; /* 0x0000001fff037819 */ /* 0x000fe20000011400 */ /*0090*/ ULDC.64 UR6, c[0x0][0x190] ; /* 0x0000640000067ab9 */ /* 0x000fe40000000a00 */ /*00a0*/ USHF.R.S32.HI UR4, URZ, 0x1f, UR5 ; /* 0x0000001f3f047899 */ /* 0x000fe40008011405 */ /*00b0*/ UIADD3 UR5, UP0, -UR5, UR6, URZ ; /* 0x0000000605057290 */ /* 0x000fc8000ff1e13f */ /*00c0*/ UIADD3.X UR4, ~UR4, UR7, URZ, UP0, !UPT ; /* 0x0000000704047290 */ /* 0x000fe400087fe53f */ /*00d0*/ ISETP.LT.U32.AND P0, PT, R0, UR5, PT ; /* 0x0000000500007c0c */ /* 0x000fc8000bf01070 */ /*00e0*/ IMAD.U32 R2, RZ, RZ, UR4 ; /* 0x00000004ff027e24 */ /* 0x000fca000f8e00ff */ /*00f0*/ ISETP.GT.AND.EX P0, PT, R2, R3, PT, P0 ; /* 0x000000030200720c */ /* 0x000fda0003f04300 */ /*0100*/ @!P0 LEA R4, P1, R0, c[0x0][0x180], 0x2 ; /* 0x0000600000048a11 */ /* 0x000fc800078210ff */ /*0110*/ @!P0 LEA.HI.X R5, R0, c[0x0][0x184], R3, 0x2, P1 ; /* 0x0000610000058a11 */ /* 0x000fca00008f1403 */ /*0120*/ @!P0 STG.E [R4.64], RZ ; /* 0x000000ff04008986 */ /* 0x0001e2000c101908 */ /*0130*/ @!P0 EXIT ; /* 0x000000000000894d */ /* 0x000fea0003800000 */ /*0140*/ IMAD.SHL.U32 R10, R0.reuse, 0x4, RZ ; /* 0x00000004000a7824 */ /* 0x040fe200078e00ff */ /*0150*/ IADD3 R2, R0.reuse, c[0x0][0x188], RZ ; /* 0x0000620000027a10 */ /* 0x040fe20007ffe0ff */ /*0160*/ IMAD.MOV.U32 R7, RZ, RZ, 0x4 ; /* 0x00000004ff077424 */ /* 0x000fe200078e00ff */ /*0170*/ SHF.L.U64.HI R11, R0, 0x2, R3 ; /* 0x00000002000b7819 */ /* 0x000fe40000010203 */ /*0180*/ IADD3 R8, P1, R10, c[0x0][0x168], RZ ; /* 0x00005a000a087a10 */ /* 0x000fe20007f3e0ff */ /*0190*/ IMAD.WIDE R4, R2, R7.reuse, c[0x0][0x178] ; /* 0x00005e0002047625 */ /* 0x081fe200078e0207 */ /*01a0*/ IADD3 R6, P0, R10, c[0x0][0x160], RZ ; /* 0x000058000a067a10 */ /* 0x000fe40007f1e0ff */ /*01b0*/ IADD3.X R9, R11.reuse, c[0x0][0x16c], RZ, P1, !PT ; /* 0x00005b000b097a10 */ /* 0x040fe20000ffe4ff */ /*01c0*/ IMAD.WIDE R2, R2, R7, c[0x0][0x170] ; /* 0x00005c0002027625 */ /* 0x000fe200078e0207 */ /*01d0*/ IADD3.X R7, R11, c[0x0][0x164], RZ, P0, !PT ; /* 0x000059000b077a10 */ /* 0x000fe200007fe4ff */ /*01e0*/ LDG.E R4, [R4.64] ; /* 0x0000000804047981 */ /* 0x000ea8000c1e1900 */ /*01f0*/ LDG.E R9, [R8.64] ; /* 0x0000000808097981 */ /* 0x000ea8000c1e1900 */ /*0200*/ LDG.E R2, [R2.64] ; /* 0x0000000802027981 */ /* 0x000ee8000c1e1900 */ /*0210*/ LDG.E R7, [R6.64] ; /* 0x0000000806077981 */ /* 0x000ee2000c1e1900 */ /*0220*/ IADD3 R10, P0, R10, c[0x0][0x180], RZ ; /* 0x000060000a0a7a10 */ /* 0x000fc80007f1e0ff */ /*0230*/ IADD3.X R11, R11, c[0x0][0x184], RZ, P0, !PT ; /* 0x000061000b0b7a10 */ /* 0x000fe200007fe4ff */ /*0240*/ FMUL R0, R4, R9 ; /* 0x0000000904007220 */ /* 0x004fc80000400000 */ /*0250*/ FFMA R13, R2, R7, R0 ; /* 0x00000007020d7223 */ /* 0x008fca0000000000 */ /*0260*/ STG.E [R10.64], R13 ; /* 0x0000000d0a007986 */ /* 0x000fe2000c101908 */ /*0270*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0280*/ ULDC.64 UR4, c[0x0][0x188] ; /* 0x0000620000047ab9 */ /* 0x000fe20000000a00 */ /*0290*/ SHF.R.S32.HI R3, RZ, 0x1f, R0 ; /* 0x0000001fff037819 */ /* 0x000fe20000011400 */ /*02a0*/ UIADD3 UR4, -UR4, UR5, URZ ; /* 0x0000000504047290 */ /* 0x000fcc000fffe13f */ /*02b0*/ ISETP.GE.AND P0, PT, R0, UR4, PT ; /* 0x0000000400007c0c */ /* 0x000fda000bf06270 */ /*02c0*/ @!P0 BRA 0x410 ; /* 0x0000014000008947 */ /* 0x000fea0003800000 */ /*02d0*/ IMAD.SHL.U32 R10, R0.reuse, 0x4, RZ ; /* 0x00000004000a7824 */ /* 0x040fe200078e00ff */ /*02e0*/ IADD3 R2, R0.reuse, c[0x0][0x188], RZ ; /* 0x0000620000027a10 */ /* 0x040fe20007ffe0ff */ /*02f0*/ IMAD.MOV.U32 R5, RZ, RZ, 0x4 ; /* 0x00000004ff057424 */ /* 0x000fe200078e00ff */ /*0300*/ SHF.L.U64.HI R11, R0, 0x2, R3 ; /* 0x00000002000b7819 */ /* 0x000fe40000010203 */ /*0310*/ IADD3 R8, P1, R10, c[0x0][0x168], RZ ; /* 0x00005a000a087a10 */ /* 0x000fe20007f3e0ff */ /*0320*/ IMAD.WIDE R6, R2, R5.reuse, c[0x0][0x178] ; /* 0x00005e0002067625 */ /* 0x080fe200078e0205 */ /*0330*/ IADD3 R4, P0, R10, c[0x0][0x160], RZ ; /* 0x000058000a047a10 */ /* 0x000fe40007f1e0ff */ /*0340*/ IADD3.X R9, R11.reuse, c[0x0][0x16c], RZ, P1, !PT ; /* 0x00005b000b097a10 */ /* 0x040fe20000ffe4ff */ /*0350*/ IMAD.WIDE R2, R2, R5, c[0x0][0x170] ; /* 0x00005c0002027625 */ /* 0x000fe200078e0205 */ /*0360*/ IADD3.X R5, R11, c[0x0][0x164], RZ, P0, !PT ; /* 0x000059000b057a10 */ /* 0x000fe200007fe4ff */ /*0370*/ LDG.E R6, [R6.64] ; /* 0x0000000806067981 */ /* 0x000ea8000c1e1900 */ /*0380*/ LDG.E R9, [R8.64] ; /* 0x0000000808097981 */ /* 0x000ea8000c1e1900 */ /*0390*/ LDG.E R2, [R2.64] ; /* 0x0000000802027981 */ /* 0x000ee8000c1e1900 */ /*03a0*/ LDG.E R5, [R4.64] ; /* 0x0000000804057981 */ /* 0x000ee2000c1e1900 */ /*03b0*/ IADD3 R10, P0, R10, c[0x0][0x180], RZ ; /* 0x000060000a0a7a10 */ /* 0x000fc80007f1e0ff */ /*03c0*/ IADD3.X R11, R11, c[0x0][0x184], RZ, P0, !PT ; /* 0x000061000b0b7a10 */ /* 0x000fe200007fe4ff */ /*03d0*/ FMUL R0, R6, R9 ; /* 0x0000000906007220 */ /* 0x004fc80000400000 */ /*03e0*/ FFMA R13, R2, R5, R0 ; /* 0x00000005020d7223 */ /* 0x008fca0000000000 */ /*03f0*/ STG.E [R10.64], R13 ; /* 0x0000000d0a007986 */ /* 0x000fe2000c101908 */ /*0400*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0410*/ LEA R2, P0, R0, c[0x0][0x180], 0x2 ; /* 0x0000600000027a11 */ /* 0x000fc800078010ff */ /*0420*/ LEA.HI.X R3, R0, c[0x0][0x184], R3, 0x2, P0 ; /* 0x0000610000037a11 */ /* 0x000fca00000f1403 */ /*0430*/ STG.E [R2.64], RZ ; /* 0x000000ff02007986 */ /* 0x000fe2000c101908 */ /*0440*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0450*/ BRA 0x450; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0460*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0470*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0480*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0490*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z6kernelPfS_S_S_S_iil .globl _Z6kernelPfS_S_S_S_iil .p2align 8 .type _Z6kernelPfS_S_S_S_iil,@function _Z6kernelPfS_S_S_S_iil: s_clause 0x1 s_load_b32 s3, s[0:1], 0x44 s_load_b32 s2, s[0:1], 0x28 s_waitcnt lgkmcnt(0) s_and_b32 s3, s3, 0xffff s_cmp_gt_i32 s2, -1 v_mad_u64_u32 v[1:2], null, s15, s3, v[0:1] s_delay_alu instid0(VALU_DEP_1) v_ashrrev_i32_e32 v2, 31, v1 s_cbranch_scc0 .LBB0_2 s_load_b64 s[4:5], s[0:1], 0x30 s_mov_b32 s3, 0 s_waitcnt lgkmcnt(0) s_sub_u32 s4, s4, s2 s_subb_u32 s5, s5, 0 s_delay_alu instid0(SALU_CYCLE_1) v_cmp_gt_i64_e32 vcc_lo, s[4:5], v[1:2] s_and_b32 s4, vcc_lo, exec_lo s_cbranch_execz .LBB0_3 s_branch .LBB0_4 .LBB0_2: s_mov_b32 s3, -1 s_mov_b32 s4, 0 .LBB0_3: s_load_b32 s3, s[0:1], 0x2c s_and_not1_b32 s4, s4, exec_lo s_waitcnt lgkmcnt(0) s_sub_i32 s3, s3, s2 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_2) | instid1(SALU_CYCLE_1) v_cmp_le_i32_e32 vcc_lo, s3, v1 s_mov_b32 s3, 0 s_and_b32 s5, vcc_lo, exec_lo s_or_b32 s4, s4, s5 .LBB0_4: v_mov_b32_e32 v0, s3 s_and_saveexec_b32 s3, s4 s_cbranch_execz .LBB0_6 s_load_b256 s[4:11], s[0:1], 0x0 v_add_nc_u32_e32 v3, s2, v1 v_lshlrev_b64 v[5:6], 2, v[1:2] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v4, 31, v3 v_lshlrev_b64 v[3:4], 2, v[3:4] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_u32 v7, vcc_lo, s4, v5 v_add_co_ci_u32_e32 v8, vcc_lo, s5, v6, vcc_lo v_add_co_u32 v5, vcc_lo, s6, v5 v_add_co_ci_u32_e32 v6, vcc_lo, s7, v6, vcc_lo v_add_co_u32 v9, vcc_lo, s10, v3 v_add_co_ci_u32_e32 v10, vcc_lo, s11, v4, vcc_lo v_add_co_u32 v3, vcc_lo, s8, v3 v_add_co_ci_u32_e32 v4, vcc_lo, s9, v4, vcc_lo global_load_b32 v0, v[5:6], off global_load_b32 v5, v[9:10], off global_load_b32 v6, v[7:8], off global_load_b32 v3, v[3:4], off s_waitcnt vmcnt(2) v_mul_f32_e32 v0, v0, v5 s_waitcnt vmcnt(0) s_delay_alu instid0(VALU_DEP_1) v_fmac_f32_e32 v0, v6, v3 .LBB0_6: s_or_b32 exec_lo, exec_lo, s3 s_load_b64 s[0:1], s[0:1], 0x20 v_lshlrev_b64 v[1:2], 2, v[1:2] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v1, vcc_lo, s0, v1 v_add_co_ci_u32_e32 v2, vcc_lo, s1, v2, vcc_lo global_store_b32 v[1:2], v0, off s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z6kernelPfS_S_S_S_iil .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 312 .amdhsa_user_sgpr_count 15 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 0 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 0 .amdhsa_next_free_vgpr 11 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z6kernelPfS_S_S_S_iil, .Lfunc_end0-_Z6kernelPfS_S_S_S_iil .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 8 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 24 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 32 .size: 8 .value_kind: global_buffer - .offset: 40 .size: 4 .value_kind: by_value - .offset: 44 .size: 4 .value_kind: by_value - .offset: 48 .size: 8 .value_kind: by_value - .offset: 56 .size: 4 .value_kind: hidden_block_count_x - .offset: 60 .size: 4 .value_kind: hidden_block_count_y - .offset: 64 .size: 4 .value_kind: hidden_block_count_z - .offset: 68 .size: 2 .value_kind: hidden_group_size_x - .offset: 70 .size: 2 .value_kind: hidden_group_size_y - .offset: 72 .size: 2 .value_kind: hidden_group_size_z - .offset: 74 .size: 2 .value_kind: hidden_remainder_x - .offset: 76 .size: 2 .value_kind: hidden_remainder_y - .offset: 78 .size: 2 .value_kind: hidden_remainder_z - .offset: 96 .size: 8 .value_kind: hidden_global_offset_x - .offset: 104 .size: 8 .value_kind: hidden_global_offset_y - .offset: 112 .size: 8 .value_kind: hidden_global_offset_z - .offset: 120 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 312 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z6kernelPfS_S_S_S_iil .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z6kernelPfS_S_S_S_iil.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 11 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
#include <stdio.h> #include <stdlib.h> #define N 4096 #define block_Size 256 /* function to integrate, defined as a function on the GPU device */ __device__ float myfunction(float a) { return a*a+2.0*a + 3.0; } /* kernel function to compute the summation used in the trapezoidal rule for numerical integration */ __global__ void integratorKernel(float *a, float c, float deltaX) { int id = blockIdx.x * blockDim.x + threadIdx.x; float x = c + (float)id * deltaX; if (id<N) a[id] = myfunction(x)+myfunction(x+deltaX); } int main( int argc, char* argv[] ) { double end = 1.0, start = 0.0; // deltaX float deltaX = (end-start)/(double) N; // error code variable cudaError_t errorcode = cudaSuccess; // Size of the arrays in bytes int size = N*sizeof(float); // Allocate array on host and device float* a_h = (float *)malloc(size); float* a_d; if (( errorcode = cudaMalloc((void **)&a_d,size))!= cudaSuccess) { printf("cudaMalloc(): %s/n", cudaGetErrorString(errorcode)); exit(1); } // Do calculation on device int grid_Size = N/block_Size + ( N % block_Size == 0 ? 0:1); printf("blocks: %d\n", grid_Size); printf("block size: %d\n ", block_Size); integratorKernel <<< grid_Size, block_Size >>> (a_d, start, deltaX); // Copy results from device to host if((errorcode = cudaMemcpy(a_h, a_d, sizeof(float)*N, cudaMemcpyDeviceToHost)) !=cudaSuccess) { printf("cudaMemcpy(): %s\n", cudaGetErrorString(errorcode)); exit(1); } // Add up results float sum = 0.0; for(int i=0; i<N; i++) sum += a_h[i]; sum *= deltaX/2.0; printf("The integral is: %f\n", sum); // clean up free(a_h); cudaFree(a_d); return 0; }
code for sm_80 Function : _Z16integratorKernelPfff .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R0, SR_CTAID.X ; /* 0x0000000000007919 */ /* 0x000e280000002500 */ /*0020*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e240000002100 */ /*0030*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x001fe200078e0203 */ /*0040*/ MOV R3, c[0x0][0x16c] ; /* 0x00005b0000037a02 */ /* 0x000fc60000000f00 */ /*0050*/ I2F R2, R0 ; /* 0x0000000000027306 */ /* 0x000e220000201400 */ /*0060*/ ISETP.GT.AND P0, PT, R0, 0xfff, PT ; /* 0x00000fff0000780c */ /* 0x000fe20003f04270 */ /*0070*/ FFMA R10, R2, R3, c[0x0][0x168] ; /* 0x00005a00020a7623 */ /* 0x001fd80000000003 */ /*0080*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0090*/ FADD R12, R10.reuse, c[0x0][0x16c] ; /* 0x00005b000a0c7621 */ /* 0x040fe20000000000 */ /*00a0*/ F2F.F64.F32 R4, R10 ; /* 0x0000000a00047310 */ /* 0x000fe20000201800 */ /*00b0*/ FMUL R11, R10, R10 ; /* 0x0000000a0a0b7220 */ /* 0x000fe20000400000 */ /*00c0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*00d0*/ FMUL R13, R12, R12 ; /* 0x0000000c0c0d7220 */ /* 0x000fca0000400000 */ /*00e0*/ F2F.F64.F32 R2, R11 ; /* 0x0000000b00027310 */ /* 0x000e300000201800 */ /*00f0*/ F2F.F64.F32 R8, R12 ; /* 0x0000000c00087310 */ /* 0x000ff00000201800 */ /*0100*/ F2F.F64.F32 R6, R13 ; /* 0x0000000d00067310 */ /* 0x000e620000201800 */ /*0110*/ DFMA R2, R4, 2, R2 ; /* 0x400000000402782b */ /* 0x0010840000000002 */ /*0120*/ MOV R5, 0x4 ; /* 0x0000000400057802 */ /* 0x001fc80000000f00 */ /*0130*/ DADD R2, R2, 3 ; /* 0x4008000002027429 */ /* 0x004e220000000000 */ /*0140*/ IMAD.WIDE R4, R0, R5, c[0x0][0x160] ; /* 0x0000580000047625 */ /* 0x000fc600078e0205 */ /*0150*/ DFMA R6, R8, 2, R6 ; /* 0x400000000806782b */ /* 0x002e4c0000000006 */ /*0160*/ F2F.F32.F64 R2, R2 ; /* 0x0000000200027310 */ /* 0x001fe20000301000 */ /*0170*/ DADD R6, R6, 3 ; /* 0x4008000006067429 */ /* 0x002e140000000000 */ /*0180*/ F2F.F32.F64 R7, R6 ; /* 0x0000000600077310 */ /* 0x001e240000301000 */ /*0190*/ FADD R9, R2, R7 ; /* 0x0000000702097221 */ /* 0x001fca0000000000 */ /*01a0*/ STG.E [R4.64], R9 ; /* 0x0000000904007986 */ /* 0x000fe2000c101904 */ /*01b0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*01c0*/ BRA 0x1c0; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*01d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0200*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0210*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0220*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0230*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0240*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0250*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0260*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0270*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include <stdio.h> #include <stdlib.h> #define N 4096 #define block_Size 256 /* function to integrate, defined as a function on the GPU device */ __device__ float myfunction(float a) { return a*a+2.0*a + 3.0; } /* kernel function to compute the summation used in the trapezoidal rule for numerical integration */ __global__ void integratorKernel(float *a, float c, float deltaX) { int id = blockIdx.x * blockDim.x + threadIdx.x; float x = c + (float)id * deltaX; if (id<N) a[id] = myfunction(x)+myfunction(x+deltaX); } int main( int argc, char* argv[] ) { double end = 1.0, start = 0.0; // deltaX float deltaX = (end-start)/(double) N; // error code variable cudaError_t errorcode = cudaSuccess; // Size of the arrays in bytes int size = N*sizeof(float); // Allocate array on host and device float* a_h = (float *)malloc(size); float* a_d; if (( errorcode = cudaMalloc((void **)&a_d,size))!= cudaSuccess) { printf("cudaMalloc(): %s/n", cudaGetErrorString(errorcode)); exit(1); } // Do calculation on device int grid_Size = N/block_Size + ( N % block_Size == 0 ? 0:1); printf("blocks: %d\n", grid_Size); printf("block size: %d\n ", block_Size); integratorKernel <<< grid_Size, block_Size >>> (a_d, start, deltaX); // Copy results from device to host if((errorcode = cudaMemcpy(a_h, a_d, sizeof(float)*N, cudaMemcpyDeviceToHost)) !=cudaSuccess) { printf("cudaMemcpy(): %s\n", cudaGetErrorString(errorcode)); exit(1); } // Add up results float sum = 0.0; for(int i=0; i<N; i++) sum += a_h[i]; sum *= deltaX/2.0; printf("The integral is: %f\n", sum); // clean up free(a_h); cudaFree(a_d); return 0; }
.file "tmpxft_00121f50_00000000-6_trap.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2061: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2061: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z10myfunctionf .type _Z10myfunctionf, @function _Z10myfunctionf: .LFB2057: .cfi_startproc endbr64 pushq %rax .cfi_def_cfa_offset 16 popq %rax .cfi_def_cfa_offset 8 subq $24, %rsp .cfi_def_cfa_offset 32 movl $1, 12(%rsp) movl 12(%rsp), %edi call exit@PLT .cfi_endproc .LFE2057: .size _Z10myfunctionf, .-_Z10myfunctionf .globl _Z38__device_stub__Z16integratorKernelPfffPfff .type _Z38__device_stub__Z16integratorKernelPfffPfff, @function _Z38__device_stub__Z16integratorKernelPfffPfff: .LFB2083: .cfi_startproc endbr64 subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 8(%rsp) movss %xmm0, 4(%rsp) movss %xmm1, (%rsp) movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax leaq 8(%rsp), %rax movq %rax, 80(%rsp) leaq 4(%rsp), %rax movq %rax, 88(%rsp) movq %rsp, %rax movq %rax, 96(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) movl $1, 40(%rsp) movl $1, 44(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) leaq 24(%rsp), %rcx leaq 16(%rsp), %rdx leaq 44(%rsp), %rsi leaq 32(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L9 .L5: movq 104(%rsp), %rax subq %fs:40, %rax jne .L10 addq $120, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L9: .cfi_restore_state pushq 24(%rsp) .cfi_def_cfa_offset 136 pushq 24(%rsp) .cfi_def_cfa_offset 144 leaq 96(%rsp), %r9 movq 60(%rsp), %rcx movl 68(%rsp), %r8d movq 48(%rsp), %rsi movl 56(%rsp), %edx leaq _Z16integratorKernelPfff(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 128 jmp .L5 .L10: call __stack_chk_fail@PLT .cfi_endproc .LFE2083: .size _Z38__device_stub__Z16integratorKernelPfffPfff, .-_Z38__device_stub__Z16integratorKernelPfffPfff .globl _Z16integratorKernelPfff .type _Z16integratorKernelPfff, @function _Z16integratorKernelPfff: .LFB2084: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z38__device_stub__Z16integratorKernelPfffPfff addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2084: .size _Z16integratorKernelPfff, .-_Z16integratorKernelPfff .section .rodata.str1.1,"aMS",@progbits,1 .LC1: .string "cudaMalloc(): %s/n" .LC2: .string "blocks: %d\n" .LC3: .string "block size: %d\n " .LC5: .string "cudaMemcpy(): %s\n" .LC7: .string "The integral is: %f\n" .text .globl main .type main, @function main: .LFB2058: .cfi_startproc endbr64 pushq %rbx .cfi_def_cfa_offset 16 .cfi_offset 3, -16 subq $48, %rsp .cfi_def_cfa_offset 64 movq %fs:40, %rax movq %rax, 40(%rsp) xorl %eax, %eax movl $16384, %edi call malloc@PLT movq %rax, %rbx leaq 8(%rsp), %rdi movl $16384, %esi call cudaMalloc@PLT testl %eax, %eax jne .L21 movl $16, %edx leaq .LC2(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $256, %edx leaq .LC3(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $256, 28(%rsp) movl $1, 32(%rsp) movl $16, 16(%rsp) movl $1, 20(%rsp) movl $0, %r9d movl $0, %r8d movq 28(%rsp), %rdx movl $1, %ecx movq 16(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L22 .L15: movl $2, %ecx movl $16384, %edx movq 8(%rsp), %rsi movq %rbx, %rdi call cudaMemcpy@PLT testl %eax, %eax jne .L23 movq %rbx, %rax leaq 16384(%rbx), %rdx pxor %xmm0, %xmm0 .L17: addss (%rax), %xmm0 addq $4, %rax cmpq %rdx, %rax jne .L17 mulss .LC6(%rip), %xmm0 cvtss2sd %xmm0, %xmm0 leaq .LC7(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movq %rbx, %rdi call free@PLT movq 8(%rsp), %rdi call cudaFree@PLT movq 40(%rsp), %rax subq %fs:40, %rax jne .L24 movl $0, %eax addq $48, %rsp .cfi_remember_state .cfi_def_cfa_offset 16 popq %rbx .cfi_def_cfa_offset 8 ret .L21: .cfi_restore_state movl %eax, %edi call cudaGetErrorString@PLT movq %rax, %rdx leaq .LC1(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $1, %edi call exit@PLT .L22: movss .LC4(%rip), %xmm1 pxor %xmm0, %xmm0 movq 8(%rsp), %rdi call _Z38__device_stub__Z16integratorKernelPfffPfff jmp .L15 .L23: movl %eax, %edi call cudaGetErrorString@PLT movq %rax, %rdx leaq .LC5(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $1, %edi call exit@PLT .L24: call __stack_chk_fail@PLT .cfi_endproc .LFE2058: .size main, .-main .section .rodata.str1.1 .LC8: .string "_Z16integratorKernelPfff" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2086: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC8(%rip), %rdx movq %rdx, %rcx leaq _Z16integratorKernelPfff(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2086: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .section .rodata.cst4,"aM",@progbits,4 .align 4 .LC4: .long 964689920 .align 4 .LC6: .long 956301312 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include <stdio.h> #include <stdlib.h> #define N 4096 #define block_Size 256 /* function to integrate, defined as a function on the GPU device */ __device__ float myfunction(float a) { return a*a+2.0*a + 3.0; } /* kernel function to compute the summation used in the trapezoidal rule for numerical integration */ __global__ void integratorKernel(float *a, float c, float deltaX) { int id = blockIdx.x * blockDim.x + threadIdx.x; float x = c + (float)id * deltaX; if (id<N) a[id] = myfunction(x)+myfunction(x+deltaX); } int main( int argc, char* argv[] ) { double end = 1.0, start = 0.0; // deltaX float deltaX = (end-start)/(double) N; // error code variable cudaError_t errorcode = cudaSuccess; // Size of the arrays in bytes int size = N*sizeof(float); // Allocate array on host and device float* a_h = (float *)malloc(size); float* a_d; if (( errorcode = cudaMalloc((void **)&a_d,size))!= cudaSuccess) { printf("cudaMalloc(): %s/n", cudaGetErrorString(errorcode)); exit(1); } // Do calculation on device int grid_Size = N/block_Size + ( N % block_Size == 0 ? 0:1); printf("blocks: %d\n", grid_Size); printf("block size: %d\n ", block_Size); integratorKernel <<< grid_Size, block_Size >>> (a_d, start, deltaX); // Copy results from device to host if((errorcode = cudaMemcpy(a_h, a_d, sizeof(float)*N, cudaMemcpyDeviceToHost)) !=cudaSuccess) { printf("cudaMemcpy(): %s\n", cudaGetErrorString(errorcode)); exit(1); } // Add up results float sum = 0.0; for(int i=0; i<N; i++) sum += a_h[i]; sum *= deltaX/2.0; printf("The integral is: %f\n", sum); // clean up free(a_h); cudaFree(a_d); return 0; }
#include <hip/hip_runtime.h> #include <stdio.h> #include <stdlib.h> #define N 4096 #define block_Size 256 /* function to integrate, defined as a function on the GPU device */ __device__ float myfunction(float a) { return a*a+2.0*a + 3.0; } /* kernel function to compute the summation used in the trapezoidal rule for numerical integration */ __global__ void integratorKernel(float *a, float c, float deltaX) { int id = blockIdx.x * blockDim.x + threadIdx.x; float x = c + (float)id * deltaX; if (id<N) a[id] = myfunction(x)+myfunction(x+deltaX); } int main( int argc, char* argv[] ) { double end = 1.0, start = 0.0; // deltaX float deltaX = (end-start)/(double) N; // error code variable hipError_t errorcode = hipSuccess; // Size of the arrays in bytes int size = N*sizeof(float); // Allocate array on host and device float* a_h = (float *)malloc(size); float* a_d; if (( errorcode = hipMalloc((void **)&a_d,size))!= hipSuccess) { printf("cudaMalloc(): %s/n", hipGetErrorString(errorcode)); exit(1); } // Do calculation on device int grid_Size = N/block_Size + ( N % block_Size == 0 ? 0:1); printf("blocks: %d\n", grid_Size); printf("block size: %d\n ", block_Size); integratorKernel <<< grid_Size, block_Size >>> (a_d, start, deltaX); // Copy results from device to host if((errorcode = hipMemcpy(a_h, a_d, sizeof(float)*N, hipMemcpyDeviceToHost)) !=hipSuccess) { printf("cudaMemcpy(): %s\n", hipGetErrorString(errorcode)); exit(1); } // Add up results float sum = 0.0; for(int i=0; i<N; i++) sum += a_h[i]; sum *= deltaX/2.0; printf("The integral is: %f\n", sum); // clean up free(a_h); hipFree(a_d); return 0; }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include <stdio.h> #include <stdlib.h> #define N 4096 #define block_Size 256 /* function to integrate, defined as a function on the GPU device */ __device__ float myfunction(float a) { return a*a+2.0*a + 3.0; } /* kernel function to compute the summation used in the trapezoidal rule for numerical integration */ __global__ void integratorKernel(float *a, float c, float deltaX) { int id = blockIdx.x * blockDim.x + threadIdx.x; float x = c + (float)id * deltaX; if (id<N) a[id] = myfunction(x)+myfunction(x+deltaX); } int main( int argc, char* argv[] ) { double end = 1.0, start = 0.0; // deltaX float deltaX = (end-start)/(double) N; // error code variable hipError_t errorcode = hipSuccess; // Size of the arrays in bytes int size = N*sizeof(float); // Allocate array on host and device float* a_h = (float *)malloc(size); float* a_d; if (( errorcode = hipMalloc((void **)&a_d,size))!= hipSuccess) { printf("cudaMalloc(): %s/n", hipGetErrorString(errorcode)); exit(1); } // Do calculation on device int grid_Size = N/block_Size + ( N % block_Size == 0 ? 0:1); printf("blocks: %d\n", grid_Size); printf("block size: %d\n ", block_Size); integratorKernel <<< grid_Size, block_Size >>> (a_d, start, deltaX); // Copy results from device to host if((errorcode = hipMemcpy(a_h, a_d, sizeof(float)*N, hipMemcpyDeviceToHost)) !=hipSuccess) { printf("cudaMemcpy(): %s\n", hipGetErrorString(errorcode)); exit(1); } // Add up results float sum = 0.0; for(int i=0; i<N; i++) sum += a_h[i]; sum *= deltaX/2.0; printf("The integral is: %f\n", sum); // clean up free(a_h); hipFree(a_d); return 0; }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z16integratorKernelPfff .globl _Z16integratorKernelPfff .p2align 8 .type _Z16integratorKernelPfff,@function _Z16integratorKernelPfff: s_load_b32 s2, s[0:1], 0x1c s_waitcnt lgkmcnt(0) s_and_b32 s2, s2, 0xffff s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_mad_u64_u32 v[1:2], null, s15, s2, v[0:1] s_mov_b32 s2, exec_lo v_cmpx_gt_i32_e32 0x1000, v1 s_cbranch_execz .LBB0_2 s_load_b128 s[0:3], s[0:1], 0x0 v_cvt_f32_i32_e32 v0, v1 s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f32 v0, v0, s3, s2 v_add_f32_e32 v6, s3, v0 v_mul_f32_e32 v4, v0, v0 v_cvt_f64_f32_e32 v[2:3], v0 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_mul_f32_e32 v8, v6, v6 v_cvt_f64_f32_e32 v[4:5], v4 v_cvt_f64_f32_e32 v[6:7], v6 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_cvt_f64_f32_e32 v[8:9], v8 v_fma_f64 v[2:3], v[2:3], 2.0, v[4:5] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f64 v[4:5], v[6:7], 2.0, v[8:9] v_add_f64 v[2:3], v[2:3], 0x40080000 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[4:5], v[4:5], 0x40080000 v_cvt_f32_f64_e32 v3, v[2:3] v_ashrrev_i32_e32 v2, 31, v1 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_cvt_f32_f64_e32 v4, v[4:5] v_lshlrev_b64 v[0:1], 2, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v0, vcc_lo, s0, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo s_delay_alu instid0(VALU_DEP_4) v_add_f32_e32 v2, v3, v4 global_store_b32 v[0:1], v2, off .LBB0_2: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z16integratorKernelPfff .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 272 .amdhsa_user_sgpr_count 15 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 0 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 0 .amdhsa_next_free_vgpr 10 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z16integratorKernelPfff, .Lfunc_end0-_Z16integratorKernelPfff .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .offset: 8 .size: 4 .value_kind: by_value - .offset: 12 .size: 4 .value_kind: by_value - .offset: 16 .size: 4 .value_kind: hidden_block_count_x - .offset: 20 .size: 4 .value_kind: hidden_block_count_y - .offset: 24 .size: 4 .value_kind: hidden_block_count_z - .offset: 28 .size: 2 .value_kind: hidden_group_size_x - .offset: 30 .size: 2 .value_kind: hidden_group_size_y - .offset: 32 .size: 2 .value_kind: hidden_group_size_z - .offset: 34 .size: 2 .value_kind: hidden_remainder_x - .offset: 36 .size: 2 .value_kind: hidden_remainder_y - .offset: 38 .size: 2 .value_kind: hidden_remainder_z - .offset: 56 .size: 8 .value_kind: hidden_global_offset_x - .offset: 64 .size: 8 .value_kind: hidden_global_offset_y - .offset: 72 .size: 8 .value_kind: hidden_global_offset_z - .offset: 80 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 272 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z16integratorKernelPfff .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z16integratorKernelPfff.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 10 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <hip/hip_runtime.h> #include <stdio.h> #include <stdlib.h> #define N 4096 #define block_Size 256 /* function to integrate, defined as a function on the GPU device */ __device__ float myfunction(float a) { return a*a+2.0*a + 3.0; } /* kernel function to compute the summation used in the trapezoidal rule for numerical integration */ __global__ void integratorKernel(float *a, float c, float deltaX) { int id = blockIdx.x * blockDim.x + threadIdx.x; float x = c + (float)id * deltaX; if (id<N) a[id] = myfunction(x)+myfunction(x+deltaX); } int main( int argc, char* argv[] ) { double end = 1.0, start = 0.0; // deltaX float deltaX = (end-start)/(double) N; // error code variable hipError_t errorcode = hipSuccess; // Size of the arrays in bytes int size = N*sizeof(float); // Allocate array on host and device float* a_h = (float *)malloc(size); float* a_d; if (( errorcode = hipMalloc((void **)&a_d,size))!= hipSuccess) { printf("cudaMalloc(): %s/n", hipGetErrorString(errorcode)); exit(1); } // Do calculation on device int grid_Size = N/block_Size + ( N % block_Size == 0 ? 0:1); printf("blocks: %d\n", grid_Size); printf("block size: %d\n ", block_Size); integratorKernel <<< grid_Size, block_Size >>> (a_d, start, deltaX); // Copy results from device to host if((errorcode = hipMemcpy(a_h, a_d, sizeof(float)*N, hipMemcpyDeviceToHost)) !=hipSuccess) { printf("cudaMemcpy(): %s\n", hipGetErrorString(errorcode)); exit(1); } // Add up results float sum = 0.0; for(int i=0; i<N; i++) sum += a_h[i]; sum *= deltaX/2.0; printf("The integral is: %f\n", sum); // clean up free(a_h); hipFree(a_d); return 0; }
.text .file "trap.hip" .globl _Z31__device_stub__integratorKernelPfff # -- Begin function _Z31__device_stub__integratorKernelPfff .p2align 4, 0x90 .type _Z31__device_stub__integratorKernelPfff,@function _Z31__device_stub__integratorKernelPfff: # @_Z31__device_stub__integratorKernelPfff .cfi_startproc # %bb.0: subq $88, %rsp .cfi_def_cfa_offset 96 movq %rdi, 56(%rsp) movss %xmm0, 4(%rsp) movss %xmm1, (%rsp) leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 4(%rsp), %rax movq %rax, 72(%rsp) movq %rsp, %rax movq %rax, 80(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 64(%rsp), %r9 movl $_Z16integratorKernelPfff, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $104, %rsp .cfi_adjust_cfa_offset -104 retq .Lfunc_end0: .size _Z31__device_stub__integratorKernelPfff, .Lfunc_end0-_Z31__device_stub__integratorKernelPfff .cfi_endproc # -- End function .section .rodata.cst4,"aM",@progbits,4 .p2align 2, 0x0 # -- Begin function main .LCPI1_0: .long 0x39000000 # float 1.22070313E-4 .text .globl main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rbx .cfi_def_cfa_offset 16 subq $112, %rsp .cfi_def_cfa_offset 128 .cfi_offset %rbx, -16 movl $16384, %edi # imm = 0x4000 callq malloc movq %rax, %rbx leaq 8(%rsp), %rdi movl $16384, %esi # imm = 0x4000 callq hipMalloc testl %eax, %eax jne .LBB1_1 # %bb.3: movl $.L.str.1, %edi movl $16, %esi xorl %eax, %eax callq printf movl $.L.str.2, %edi movl $256, %esi # imm = 0x100 xorl %eax, %eax callq printf movabsq $4294967312, %rdi # imm = 0x100000010 leaq 240(%rdi), %rdx movl $1, %esi movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_5 # %bb.4: movq 8(%rsp), %rax movq %rax, 72(%rsp) movl $0, 20(%rsp) movl $964689920, 16(%rsp) # imm = 0x39800000 leaq 72(%rsp), %rax movq %rax, 80(%rsp) leaq 20(%rsp), %rax movq %rax, 88(%rsp) leaq 16(%rsp), %rax movq %rax, 96(%rsp) leaq 56(%rsp), %rdi leaq 40(%rsp), %rsi leaq 32(%rsp), %rdx leaq 24(%rsp), %rcx callq __hipPopCallConfiguration movq 56(%rsp), %rsi movl 64(%rsp), %edx movq 40(%rsp), %rcx movl 48(%rsp), %r8d leaq 80(%rsp), %r9 movl $_Z16integratorKernelPfff, %edi pushq 24(%rsp) .cfi_adjust_cfa_offset 8 pushq 40(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_5: movq 8(%rsp), %rsi movl $16384, %edx # imm = 0x4000 movq %rbx, %rdi movl $2, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_9 # %bb.6: # %.preheader.preheader xorps %xmm0, %xmm0 xorl %eax, %eax .p2align 4, 0x90 .LBB1_7: # %.preheader # =>This Inner Loop Header: Depth=1 addss (%rbx,%rax,4), %xmm0 incq %rax cmpq $4096, %rax # imm = 0x1000 jne .LBB1_7 # %bb.8: mulss .LCPI1_0(%rip), %xmm0 cvtss2sd %xmm0, %xmm0 movl $.L.str.4, %edi movb $1, %al callq printf movq %rbx, %rdi callq free movq 8(%rsp), %rdi callq hipFree xorl %eax, %eax addq $112, %rsp .cfi_def_cfa_offset 16 popq %rbx .cfi_def_cfa_offset 8 retq .LBB1_1: .cfi_def_cfa_offset 128 movl %eax, %edi callq hipGetErrorString movl $.L.str, %edi jmp .LBB1_2 .LBB1_9: movl %eax, %edi callq hipGetErrorString movl $.L.str.3, %edi .LBB1_2: movq %rax, %rsi xorl %eax, %eax callq printf movl $1, %edi callq exit .Lfunc_end1: .size main, .Lfunc_end1-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB2_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB2_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z16integratorKernelPfff, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end2: .size __hip_module_ctor, .Lfunc_end2-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB3_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB3_2: retq .Lfunc_end3: .size __hip_module_dtor, .Lfunc_end3-__hip_module_dtor .cfi_endproc # -- End function .type _Z16integratorKernelPfff,@object # @_Z16integratorKernelPfff .section .rodata,"a",@progbits .globl _Z16integratorKernelPfff .p2align 3, 0x0 _Z16integratorKernelPfff: .quad _Z31__device_stub__integratorKernelPfff .size _Z16integratorKernelPfff, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "cudaMalloc(): %s/n" .size .L.str, 19 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "blocks: %d\n" .size .L.str.1, 12 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "block size: %d\n " .size .L.str.2, 17 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "cudaMemcpy(): %s\n" .size .L.str.3, 18 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "The integral is: %f\n" .size .L.str.4, 21 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z16integratorKernelPfff" .size .L__unnamed_1, 25 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z31__device_stub__integratorKernelPfff .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z16integratorKernelPfff .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA device assembly to AMD device assembly.
code for sm_80 Function : _Z16integratorKernelPfff .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R0, SR_CTAID.X ; /* 0x0000000000007919 */ /* 0x000e280000002500 */ /*0020*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e240000002100 */ /*0030*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x001fe200078e0203 */ /*0040*/ MOV R3, c[0x0][0x16c] ; /* 0x00005b0000037a02 */ /* 0x000fc60000000f00 */ /*0050*/ I2F R2, R0 ; /* 0x0000000000027306 */ /* 0x000e220000201400 */ /*0060*/ ISETP.GT.AND P0, PT, R0, 0xfff, PT ; /* 0x00000fff0000780c */ /* 0x000fe20003f04270 */ /*0070*/ FFMA R10, R2, R3, c[0x0][0x168] ; /* 0x00005a00020a7623 */ /* 0x001fd80000000003 */ /*0080*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0090*/ FADD R12, R10.reuse, c[0x0][0x16c] ; /* 0x00005b000a0c7621 */ /* 0x040fe20000000000 */ /*00a0*/ F2F.F64.F32 R4, R10 ; /* 0x0000000a00047310 */ /* 0x000fe20000201800 */ /*00b0*/ FMUL R11, R10, R10 ; /* 0x0000000a0a0b7220 */ /* 0x000fe20000400000 */ /*00c0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*00d0*/ FMUL R13, R12, R12 ; /* 0x0000000c0c0d7220 */ /* 0x000fca0000400000 */ /*00e0*/ F2F.F64.F32 R2, R11 ; /* 0x0000000b00027310 */ /* 0x000e300000201800 */ /*00f0*/ F2F.F64.F32 R8, R12 ; /* 0x0000000c00087310 */ /* 0x000ff00000201800 */ /*0100*/ F2F.F64.F32 R6, R13 ; /* 0x0000000d00067310 */ /* 0x000e620000201800 */ /*0110*/ DFMA R2, R4, 2, R2 ; /* 0x400000000402782b */ /* 0x0010840000000002 */ /*0120*/ MOV R5, 0x4 ; /* 0x0000000400057802 */ /* 0x001fc80000000f00 */ /*0130*/ DADD R2, R2, 3 ; /* 0x4008000002027429 */ /* 0x004e220000000000 */ /*0140*/ IMAD.WIDE R4, R0, R5, c[0x0][0x160] ; /* 0x0000580000047625 */ /* 0x000fc600078e0205 */ /*0150*/ DFMA R6, R8, 2, R6 ; /* 0x400000000806782b */ /* 0x002e4c0000000006 */ /*0160*/ F2F.F32.F64 R2, R2 ; /* 0x0000000200027310 */ /* 0x001fe20000301000 */ /*0170*/ DADD R6, R6, 3 ; /* 0x4008000006067429 */ /* 0x002e140000000000 */ /*0180*/ F2F.F32.F64 R7, R6 ; /* 0x0000000600077310 */ /* 0x001e240000301000 */ /*0190*/ FADD R9, R2, R7 ; /* 0x0000000702097221 */ /* 0x001fca0000000000 */ /*01a0*/ STG.E [R4.64], R9 ; /* 0x0000000904007986 */ /* 0x000fe2000c101904 */ /*01b0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*01c0*/ BRA 0x1c0; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*01d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0200*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0210*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0220*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0230*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0240*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0250*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0260*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0270*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z16integratorKernelPfff .globl _Z16integratorKernelPfff .p2align 8 .type _Z16integratorKernelPfff,@function _Z16integratorKernelPfff: s_load_b32 s2, s[0:1], 0x1c s_waitcnt lgkmcnt(0) s_and_b32 s2, s2, 0xffff s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_mad_u64_u32 v[1:2], null, s15, s2, v[0:1] s_mov_b32 s2, exec_lo v_cmpx_gt_i32_e32 0x1000, v1 s_cbranch_execz .LBB0_2 s_load_b128 s[0:3], s[0:1], 0x0 v_cvt_f32_i32_e32 v0, v1 s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f32 v0, v0, s3, s2 v_add_f32_e32 v6, s3, v0 v_mul_f32_e32 v4, v0, v0 v_cvt_f64_f32_e32 v[2:3], v0 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_mul_f32_e32 v8, v6, v6 v_cvt_f64_f32_e32 v[4:5], v4 v_cvt_f64_f32_e32 v[6:7], v6 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_cvt_f64_f32_e32 v[8:9], v8 v_fma_f64 v[2:3], v[2:3], 2.0, v[4:5] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f64 v[4:5], v[6:7], 2.0, v[8:9] v_add_f64 v[2:3], v[2:3], 0x40080000 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[4:5], v[4:5], 0x40080000 v_cvt_f32_f64_e32 v3, v[2:3] v_ashrrev_i32_e32 v2, 31, v1 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_cvt_f32_f64_e32 v4, v[4:5] v_lshlrev_b64 v[0:1], 2, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v0, vcc_lo, s0, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo s_delay_alu instid0(VALU_DEP_4) v_add_f32_e32 v2, v3, v4 global_store_b32 v[0:1], v2, off .LBB0_2: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z16integratorKernelPfff .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 272 .amdhsa_user_sgpr_count 15 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 0 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 0 .amdhsa_next_free_vgpr 10 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z16integratorKernelPfff, .Lfunc_end0-_Z16integratorKernelPfff .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .offset: 8 .size: 4 .value_kind: by_value - .offset: 12 .size: 4 .value_kind: by_value - .offset: 16 .size: 4 .value_kind: hidden_block_count_x - .offset: 20 .size: 4 .value_kind: hidden_block_count_y - .offset: 24 .size: 4 .value_kind: hidden_block_count_z - .offset: 28 .size: 2 .value_kind: hidden_group_size_x - .offset: 30 .size: 2 .value_kind: hidden_group_size_y - .offset: 32 .size: 2 .value_kind: hidden_group_size_z - .offset: 34 .size: 2 .value_kind: hidden_remainder_x - .offset: 36 .size: 2 .value_kind: hidden_remainder_y - .offset: 38 .size: 2 .value_kind: hidden_remainder_z - .offset: 56 .size: 8 .value_kind: hidden_global_offset_x - .offset: 64 .size: 8 .value_kind: hidden_global_offset_y - .offset: 72 .size: 8 .value_kind: hidden_global_offset_z - .offset: 80 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 272 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z16integratorKernelPfff .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z16integratorKernelPfff.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 10 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_00121f50_00000000-6_trap.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2061: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2061: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z10myfunctionf .type _Z10myfunctionf, @function _Z10myfunctionf: .LFB2057: .cfi_startproc endbr64 pushq %rax .cfi_def_cfa_offset 16 popq %rax .cfi_def_cfa_offset 8 subq $24, %rsp .cfi_def_cfa_offset 32 movl $1, 12(%rsp) movl 12(%rsp), %edi call exit@PLT .cfi_endproc .LFE2057: .size _Z10myfunctionf, .-_Z10myfunctionf .globl _Z38__device_stub__Z16integratorKernelPfffPfff .type _Z38__device_stub__Z16integratorKernelPfffPfff, @function _Z38__device_stub__Z16integratorKernelPfffPfff: .LFB2083: .cfi_startproc endbr64 subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 8(%rsp) movss %xmm0, 4(%rsp) movss %xmm1, (%rsp) movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax leaq 8(%rsp), %rax movq %rax, 80(%rsp) leaq 4(%rsp), %rax movq %rax, 88(%rsp) movq %rsp, %rax movq %rax, 96(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) movl $1, 40(%rsp) movl $1, 44(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) leaq 24(%rsp), %rcx leaq 16(%rsp), %rdx leaq 44(%rsp), %rsi leaq 32(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L9 .L5: movq 104(%rsp), %rax subq %fs:40, %rax jne .L10 addq $120, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L9: .cfi_restore_state pushq 24(%rsp) .cfi_def_cfa_offset 136 pushq 24(%rsp) .cfi_def_cfa_offset 144 leaq 96(%rsp), %r9 movq 60(%rsp), %rcx movl 68(%rsp), %r8d movq 48(%rsp), %rsi movl 56(%rsp), %edx leaq _Z16integratorKernelPfff(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 128 jmp .L5 .L10: call __stack_chk_fail@PLT .cfi_endproc .LFE2083: .size _Z38__device_stub__Z16integratorKernelPfffPfff, .-_Z38__device_stub__Z16integratorKernelPfffPfff .globl _Z16integratorKernelPfff .type _Z16integratorKernelPfff, @function _Z16integratorKernelPfff: .LFB2084: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z38__device_stub__Z16integratorKernelPfffPfff addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2084: .size _Z16integratorKernelPfff, .-_Z16integratorKernelPfff .section .rodata.str1.1,"aMS",@progbits,1 .LC1: .string "cudaMalloc(): %s/n" .LC2: .string "blocks: %d\n" .LC3: .string "block size: %d\n " .LC5: .string "cudaMemcpy(): %s\n" .LC7: .string "The integral is: %f\n" .text .globl main .type main, @function main: .LFB2058: .cfi_startproc endbr64 pushq %rbx .cfi_def_cfa_offset 16 .cfi_offset 3, -16 subq $48, %rsp .cfi_def_cfa_offset 64 movq %fs:40, %rax movq %rax, 40(%rsp) xorl %eax, %eax movl $16384, %edi call malloc@PLT movq %rax, %rbx leaq 8(%rsp), %rdi movl $16384, %esi call cudaMalloc@PLT testl %eax, %eax jne .L21 movl $16, %edx leaq .LC2(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $256, %edx leaq .LC3(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $256, 28(%rsp) movl $1, 32(%rsp) movl $16, 16(%rsp) movl $1, 20(%rsp) movl $0, %r9d movl $0, %r8d movq 28(%rsp), %rdx movl $1, %ecx movq 16(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L22 .L15: movl $2, %ecx movl $16384, %edx movq 8(%rsp), %rsi movq %rbx, %rdi call cudaMemcpy@PLT testl %eax, %eax jne .L23 movq %rbx, %rax leaq 16384(%rbx), %rdx pxor %xmm0, %xmm0 .L17: addss (%rax), %xmm0 addq $4, %rax cmpq %rdx, %rax jne .L17 mulss .LC6(%rip), %xmm0 cvtss2sd %xmm0, %xmm0 leaq .LC7(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movq %rbx, %rdi call free@PLT movq 8(%rsp), %rdi call cudaFree@PLT movq 40(%rsp), %rax subq %fs:40, %rax jne .L24 movl $0, %eax addq $48, %rsp .cfi_remember_state .cfi_def_cfa_offset 16 popq %rbx .cfi_def_cfa_offset 8 ret .L21: .cfi_restore_state movl %eax, %edi call cudaGetErrorString@PLT movq %rax, %rdx leaq .LC1(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $1, %edi call exit@PLT .L22: movss .LC4(%rip), %xmm1 pxor %xmm0, %xmm0 movq 8(%rsp), %rdi call _Z38__device_stub__Z16integratorKernelPfffPfff jmp .L15 .L23: movl %eax, %edi call cudaGetErrorString@PLT movq %rax, %rdx leaq .LC5(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $1, %edi call exit@PLT .L24: call __stack_chk_fail@PLT .cfi_endproc .LFE2058: .size main, .-main .section .rodata.str1.1 .LC8: .string "_Z16integratorKernelPfff" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2086: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC8(%rip), %rdx movq %rdx, %rcx leaq _Z16integratorKernelPfff(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2086: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .section .rodata.cst4,"aM",@progbits,4 .align 4 .LC4: .long 964689920 .align 4 .LC6: .long 956301312 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "trap.hip" .globl _Z31__device_stub__integratorKernelPfff # -- Begin function _Z31__device_stub__integratorKernelPfff .p2align 4, 0x90 .type _Z31__device_stub__integratorKernelPfff,@function _Z31__device_stub__integratorKernelPfff: # @_Z31__device_stub__integratorKernelPfff .cfi_startproc # %bb.0: subq $88, %rsp .cfi_def_cfa_offset 96 movq %rdi, 56(%rsp) movss %xmm0, 4(%rsp) movss %xmm1, (%rsp) leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 4(%rsp), %rax movq %rax, 72(%rsp) movq %rsp, %rax movq %rax, 80(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 64(%rsp), %r9 movl $_Z16integratorKernelPfff, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $104, %rsp .cfi_adjust_cfa_offset -104 retq .Lfunc_end0: .size _Z31__device_stub__integratorKernelPfff, .Lfunc_end0-_Z31__device_stub__integratorKernelPfff .cfi_endproc # -- End function .section .rodata.cst4,"aM",@progbits,4 .p2align 2, 0x0 # -- Begin function main .LCPI1_0: .long 0x39000000 # float 1.22070313E-4 .text .globl main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rbx .cfi_def_cfa_offset 16 subq $112, %rsp .cfi_def_cfa_offset 128 .cfi_offset %rbx, -16 movl $16384, %edi # imm = 0x4000 callq malloc movq %rax, %rbx leaq 8(%rsp), %rdi movl $16384, %esi # imm = 0x4000 callq hipMalloc testl %eax, %eax jne .LBB1_1 # %bb.3: movl $.L.str.1, %edi movl $16, %esi xorl %eax, %eax callq printf movl $.L.str.2, %edi movl $256, %esi # imm = 0x100 xorl %eax, %eax callq printf movabsq $4294967312, %rdi # imm = 0x100000010 leaq 240(%rdi), %rdx movl $1, %esi movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_5 # %bb.4: movq 8(%rsp), %rax movq %rax, 72(%rsp) movl $0, 20(%rsp) movl $964689920, 16(%rsp) # imm = 0x39800000 leaq 72(%rsp), %rax movq %rax, 80(%rsp) leaq 20(%rsp), %rax movq %rax, 88(%rsp) leaq 16(%rsp), %rax movq %rax, 96(%rsp) leaq 56(%rsp), %rdi leaq 40(%rsp), %rsi leaq 32(%rsp), %rdx leaq 24(%rsp), %rcx callq __hipPopCallConfiguration movq 56(%rsp), %rsi movl 64(%rsp), %edx movq 40(%rsp), %rcx movl 48(%rsp), %r8d leaq 80(%rsp), %r9 movl $_Z16integratorKernelPfff, %edi pushq 24(%rsp) .cfi_adjust_cfa_offset 8 pushq 40(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_5: movq 8(%rsp), %rsi movl $16384, %edx # imm = 0x4000 movq %rbx, %rdi movl $2, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_9 # %bb.6: # %.preheader.preheader xorps %xmm0, %xmm0 xorl %eax, %eax .p2align 4, 0x90 .LBB1_7: # %.preheader # =>This Inner Loop Header: Depth=1 addss (%rbx,%rax,4), %xmm0 incq %rax cmpq $4096, %rax # imm = 0x1000 jne .LBB1_7 # %bb.8: mulss .LCPI1_0(%rip), %xmm0 cvtss2sd %xmm0, %xmm0 movl $.L.str.4, %edi movb $1, %al callq printf movq %rbx, %rdi callq free movq 8(%rsp), %rdi callq hipFree xorl %eax, %eax addq $112, %rsp .cfi_def_cfa_offset 16 popq %rbx .cfi_def_cfa_offset 8 retq .LBB1_1: .cfi_def_cfa_offset 128 movl %eax, %edi callq hipGetErrorString movl $.L.str, %edi jmp .LBB1_2 .LBB1_9: movl %eax, %edi callq hipGetErrorString movl $.L.str.3, %edi .LBB1_2: movq %rax, %rsi xorl %eax, %eax callq printf movl $1, %edi callq exit .Lfunc_end1: .size main, .Lfunc_end1-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB2_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB2_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z16integratorKernelPfff, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end2: .size __hip_module_ctor, .Lfunc_end2-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB3_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB3_2: retq .Lfunc_end3: .size __hip_module_dtor, .Lfunc_end3-__hip_module_dtor .cfi_endproc # -- End function .type _Z16integratorKernelPfff,@object # @_Z16integratorKernelPfff .section .rodata,"a",@progbits .globl _Z16integratorKernelPfff .p2align 3, 0x0 _Z16integratorKernelPfff: .quad _Z31__device_stub__integratorKernelPfff .size _Z16integratorKernelPfff, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "cudaMalloc(): %s/n" .size .L.str, 19 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "blocks: %d\n" .size .L.str.1, 12 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "block size: %d\n " .size .L.str.2, 17 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "cudaMemcpy(): %s\n" .size .L.str.3, 18 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "The integral is: %f\n" .size .L.str.4, 21 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z16integratorKernelPfff" .size .L__unnamed_1, 25 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z31__device_stub__integratorKernelPfff .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z16integratorKernelPfff .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
#include <stdio.h> #include <stdlib.h> #include <time.h> #include <inttypes.h> #include <cuda.h> #include <cuda_runtime.h> #define BLOCK_WIDTH 32 #define TAILLE 2048 #define gettime(t) clock_gettime(CLOCK_MONOTONIC_RAW, t) #define get_sub_seconde(t) (1e-9*(double)t.tv_nsec) /** return time in second */ double get_elapsedtime(void) { struct timespec st; int err = gettime(&st); if (err !=0) return 0; return (double)st.tv_sec + get_sub_seconde(st); } void init(double** A, double** B, double** C, int size) { int i = 0, j = 0; srand(2019); for(i = 0; i < size; i++) { for(j = 0; j < size; j++) { A[i][j] = rand(); B[i][j] = rand(); C[i][j] = 0.0; } } } void mult(double** A, double** B, double** C, int size) { int i = 0, j = 0, k = 0; for(i = 0; i < size; i++) { for(j = 0; j < size; j++) { double sum = 0.; for(k = 0; k < size; k++) { sum += A[i][k] * B[k][j]; } C[i][j] = sum; } } } // QUESTION 4 __global__ void MulMatrixKernel(double* A, double* B, double* C, int N) { __shared__ double share_A[BLOCK_WIDTH][BLOCK_WIDTH]; __shared__ double share_B[BLOCK_WIDTH][BLOCK_WIDTH]; double sum = 0; int col = threadIdx.x + blockDim.x * blockIdx.x; int line = threadIdx.y + blockDim.y * blockIdx.y; for(int tile = 0; tile < gridDim.x; tile++) { share_A[threadIdx.x][threadIdx.y] = A[line * N + tile * blockDim.x + threadIdx.x]; share_B[threadIdx.x][threadIdx.y] = B[(tile*blockIdx.y + threadIdx.y) * N + col]; __syncthreads(); for(int i = 0; i < BLOCK_WIDTH; i++) { sum += share_A[i][threadIdx.y]*share_B[threadIdx.x][i]; } __syncthreads(); } C[line * N + col] = sum; } // FIN QUESTION 4 int main(int argc, char** argv){ int N, i; double *A_data; double *B_data; double *C_data; double **A; double **B; double **C; double t0 = 0., t1 = 0., duration = 0.; N = (argc < 2)?TAILLE:atoi(argv[1]); fprintf(stdout, "Matrix Multiplication\n Size: %dx%d\n", N, N); // Memory allocation A_data = (double*) malloc(sizeof(double) * N * N); B_data = (double*) malloc(sizeof(double) * N * N); C_data = (double*) malloc(sizeof(double) * N * N); A = (double**) malloc(sizeof(double *) * N); B = (double**) malloc(sizeof(double *) * N); C = (double**) malloc(sizeof(double *) * N); for(i = 0; i < N; i++) { A[i] = &A_data[i * N]; B[i] = &B_data[i * N]; C[i] = &C_data[i * N]; } // Value initialization init(A, B, C, N); // QUESTION 8 cudaEvent_t start, stop; cudaEventCreate(&start); cudaEventCreate(&stop); //FIN QUESTION 8 // QUESTION 1 double *d_A, *d_B, *d_C; cudaMalloc(&d_A, sizeof(double) * N * N); cudaMalloc(&d_B, sizeof(double) * N * N); cudaMalloc(&d_C, sizeof(double) * N * N); // FIN QUESTION 1 // QUESTION 2 cudaMemcpy(d_A, A_data, sizeof(double) * N * N, cudaMemcpyHostToDevice); cudaMemcpy(d_B, B_data, sizeof(double) * N * N, cudaMemcpyHostToDevice); cudaMemcpy(d_C, C_data, sizeof(double) * N * N, cudaMemcpyHostToDevice); // FIN QUESTION 2 // QUESTION 3 int nbBlocks = N / BLOCK_WIDTH; if(N % BLOCK_WIDTH) nbBlocks++; dim3 gridSize(nbBlocks, nbBlocks); dim3 blockSize(BLOCK_WIDTH, BLOCK_WIDTH); // FIN QUESTION 3 // QUESTION 4 cudaEventRecord(start); // QUESTION 8 MulMatrixKernel<<<gridSize, blockSize>>>(d_A, d_B, d_C, N); cudaEventRecord(stop); // QUESTION 8 // FIN QUESTION 4 // QUESTION 5 cudaMemcpy(C_data, d_C, sizeof(double) * N * N, cudaMemcpyDeviceToHost); // FIN QUESTION 5 // QUESTION 8 cudaEventSynchronize(stop); uint64_t nb_op = N * N * N; float milliseconds = 0; cudaEventElapsedTime(&milliseconds, start, stop); printf("Matrice %dx%d\n\tTemps: %f s\n\tMFlops: %.2f\n", N, N, milliseconds/1000, (nb_op / (milliseconds/1000))*1E-6); // FIN QUESTION 8 // Compute multiplication t0 = get_elapsedtime(); // mult(A, B, C, N); t1 = get_elapsedtime(); // Pretty print duration = (t1 - t0); fprintf(stdout, "Performance results: \n"); fprintf(stdout, " Time: %lf s\n", duration); fprintf(stdout, " MFlops: %.2f\n", (nb_op / duration)*1E-6); return 0; }
code for sm_80 Function : _Z15MulMatrixKernelPdS_S_i .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R2, SR_CTAID.X ; /* 0x0000000000027919 */ /* 0x000e220000002500 */ /*0020*/ ISETP.NE.AND P0, PT, RZ, c[0x0][0xc], PT ; /* 0x00000300ff007a0c */ /* 0x000fe20003f05270 */ /*0030*/ HFMA2.MMA R17, -RZ, RZ, 0, 4.76837158203125e-07 ; /* 0x00000008ff117435 */ /* 0x000fe200000001ff */ /*0040*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*0050*/ S2R R9, SR_TID.X ; /* 0x0000000000097919 */ /* 0x000e220000002100 */ /*0060*/ CS2R R22, SRZ ; /* 0x0000000000167805 */ /* 0x000fc6000001ff00 */ /*0070*/ S2R R0, SR_CTAID.Y ; /* 0x0000000000007919 */ /* 0x000e680000002600 */ /*0080*/ S2R R3, SR_TID.Y ; /* 0x0000000000037919 */ /* 0x000e620000002200 */ /*0090*/ IMAD R2, R2, c[0x0][0x0], R9 ; /* 0x0000000002027a24 */ /* 0x001fe400078e0209 */ /*00a0*/ IMAD R5, R0, c[0x0][0x4], R3 ; /* 0x0000010000057a24 */ /* 0x002fc800078e0203 */ /*00b0*/ IMAD R16, R5, c[0x0][0x178], R2 ; /* 0x00005e0005107a24 */ /* 0x000fc800078e0202 */ /*00c0*/ IMAD.WIDE R16, R16, R17, c[0x0][0x170] ; /* 0x00005c0010107625 */ /* 0x000fe200078e0211 */ /*00d0*/ @!P0 BRA 0x730 ; /* 0x0000065000008947 */ /* 0x000fea0003800000 */ /*00e0*/ SHF.L.U32 R6, R3, 0x3, RZ ; /* 0x0000000303067819 */ /* 0x000fe200000006ff */ /*00f0*/ IMAD R4, R5, c[0x0][0x178], R9 ; /* 0x00005e0005047a24 */ /* 0x000fe200078e0209 */ /*0100*/ SHF.L.U32 R7, R9, 0x8, RZ ; /* 0x0000000809077819 */ /* 0x000fe200000006ff */ /*0110*/ IMAD R5, R3, c[0x0][0x178], R2 ; /* 0x00005e0003057a24 */ /* 0x000fe200078e0202 */ /*0120*/ LEA R6, R9, R6, 0x8 ; /* 0x0000000609067211 */ /* 0x000fe200078e40ff */ /*0130*/ CS2R R22, SRZ ; /* 0x0000000000167805 */ /* 0x000fe2000001ff00 */ /*0140*/ MOV R2, RZ ; /* 0x000000ff00027202 */ /* 0x000fe40000000f00 */ /*0150*/ HFMA2.MMA R28, -RZ, RZ, 0, 4.76837158203125e-07 ; /* 0x00000008ff1c7435 */ /* 0x000fd400000001ff */ /*0160*/ IMAD.WIDE.U32 R26, R4, R28, c[0x0][0x160] ; /* 0x00005800041a7625 */ /* 0x000fc800078e001c */ /*0170*/ IMAD.WIDE.U32 R28, R5, R28, c[0x0][0x168] ; /* 0x00005a00051c7625 */ /* 0x000fe400078e001c */ /*0180*/ LDG.E.64 R26, [R26.64] ; /* 0x000000041a1a7981 */ /* 0x000ea8000c1e1b00 */ /*0190*/ LDG.E.64 R28, [R28.64] ; /* 0x000000041c1c7981 */ /* 0x000ee2000c1e1b00 */ /*01a0*/ IADD3 R2, R2, 0x1, RZ ; /* 0x0000000102027810 */ /* 0x000fe20007ffe0ff */ /*01b0*/ IMAD R5, R0, c[0x0][0x178], R5 ; /* 0x00005e0000057a24 */ /* 0x000fe200078e0205 */ /*01c0*/ IADD3 R4, R4, c[0x0][0x0], RZ ; /* 0x0000000004047a10 */ /* 0x000fe40007ffe0ff */ /*01d0*/ ISETP.GE.U32.AND P0, PT, R2, c[0x0][0xc], PT ; /* 0x0000030002007a0c */ /* 0x000fe20003f06070 */ /*01e0*/ STS.64 [R6], R26 ; /* 0x0000001a06007388 */ /* 0x004fe80000000a00 */ /*01f0*/ STS.64 [R6+0x2000], R28 ; /* 0x0020001c06007388 */ /* 0x008fe80000000a00 */ /*0200*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0210*/ LDS.64 R18, [R3.X8] ; /* 0x0000000003127984 */ /* 0x000fe80000008a00 */ /*0220*/ LDS.128 R12, [R7+0x2000] ; /* 0x00200000070c7984 */ /* 0x000e280000000c00 */ /*0230*/ LDS.64 R24, [R3.X8+0x100] ; /* 0x0001000003187984 */ /* 0x000e680000008a00 */ /*0240*/ LDS.64 R20, [R3.X8+0x200] ; /* 0x0002000003147984 */ /* 0x000fe80000008a00 */ /*0250*/ LDS.128 R8, [R7+0x2010] ; /* 0x0020100007087984 */ /* 0x000ea20000000c00 */ /*0260*/ DFMA R12, R12, R18, R22 ; /* 0x000000120c0c722b */ /* 0x0010460000000016 */ /*0270*/ LDS.64 R22, [R3.X8+0x300] ; /* 0x0003000003167984 */ /* 0x001e260000008a00 */ /*0280*/ DFMA R24, R24, R14, R12 ; /* 0x0000000e1818722b */ /* 0x0022a2000000000c */ /*0290*/ LDS.64 R18, [R3.X8+0x400] ; /* 0x0004000003127984 */ /* 0x000fe80000008a00 */ /*02a0*/ LDS.128 R12, [R7+0x2020] ; /* 0x00202000070c7984 */ /* 0x002e620000000c00 */ /*02b0*/ DFMA R24, R8, R20, R24 ; /* 0x000000140818722b */ /* 0x0044060000000018 */ /*02c0*/ LDS.64 R20, [R3.X8+0x500] ; /* 0x0005000003147984 */ /* 0x004ea60000008a00 */ /*02d0*/ DFMA R24, R22, R10, R24 ; /* 0x0000000a1618722b */ /* 0x0010640000000018 */ /*02e0*/ LDS.64 R22, [R3.X8+0x600] ; /* 0x0006000003167984 */ /* 0x001fe80000008a00 */ /*02f0*/ LDS.128 R8, [R7+0x2030] ; /* 0x0020300007087984 */ /* 0x000e220000000c00 */ /*0300*/ DFMA R24, R12, R18, R24 ; /* 0x000000120c18722b */ /* 0x0022860000000018 */ /*0310*/ LDS.64 R18, [R3.X8+0x700] ; /* 0x0007000003127984 */ /* 0x002e660000008a00 */ /*0320*/ DFMA R24, R20, R14, R24 ; /* 0x0000000e1418722b */ /* 0x0044240000000018 */ /*0330*/ LDS.64 R20, [R3.X8+0x800] ; /* 0x0008000003147984 */ /* 0x004fe80000008a00 */ /*0340*/ LDS.128 R12, [R7+0x2040] ; /* 0x00204000070c7984 */ /* 0x000ea20000000c00 */ /*0350*/ DFMA R24, R8, R22, R24 ; /* 0x000000160818722b */ /* 0x0010460000000018 */ /*0360*/ LDS.64 R22, [R3.X8+0x900] ; /* 0x0009000003167984 */ /* 0x001e260000008a00 */ /*0370*/ DFMA R24, R18, R10, R24 ; /* 0x0000000a1218722b */ /* 0x0022a40000000018 */ /*0380*/ LDS.64 R18, [R3.X8+0xa00] ; /* 0x000a000003127984 */ /* 0x002fe80000008a00 */ /*0390*/ LDS.128 R8, [R7+0x2050] ; /* 0x0020500007087984 */ /* 0x000e620000000c00 */ /*03a0*/ DFMA R24, R12, R20, R24 ; /* 0x000000140c18722b */ /* 0x0044060000000018 */ /*03b0*/ LDS.64 R20, [R3.X8+0xb00] ; /* 0x000b000003147984 */ /* 0x004ea60000008a00 */ /*03c0*/ DFMA R24, R22, R14, R24 ; /* 0x0000000e1618722b */ /* 0x0010640000000018 */ /*03d0*/ LDS.64 R22, [R3.X8+0xc00] ; /* 0x000c000003167984 */ /* 0x001fe80000008a00 */ /*03e0*/ LDS.128 R12, [R7+0x2060] ; /* 0x00206000070c7984 */ /* 0x000e220000000c00 */ /*03f0*/ DFMA R24, R8, R18, R24 ; /* 0x000000120818722b */ /* 0x0022860000000018 */ /*0400*/ LDS.64 R18, [R3.X8+0xd00] ; /* 0x000d000003127984 */ /* 0x002e660000008a00 */ /*0410*/ DFMA R24, R20, R10, R24 ; /* 0x0000000a1418722b */ /* 0x0044240000000018 */ /*0420*/ LDS.64 R20, [R3.X8+0xe00] ; /* 0x000e000003147984 */ /* 0x004fe80000008a00 */ /*0430*/ LDS.128 R8, [R7+0x2070] ; /* 0x0020700007087984 */ /* 0x000ea20000000c00 */ /*0440*/ DFMA R24, R12, R22, R24 ; /* 0x000000160c18722b */ /* 0x0010460000000018 */ /*0450*/ LDS.64 R22, [R3.X8+0xf00] ; /* 0x000f000003167984 */ /* 0x001e260000008a00 */ /*0460*/ DFMA R24, R18, R14, R24 ; /* 0x0000000e1218722b */ /* 0x0022a40000000018 */ /*0470*/ LDS.64 R18, [R3.X8+0x1000] ; /* 0x0010000003127984 */ /* 0x002fe80000008a00 */ /*0480*/ LDS.128 R12, [R7+0x2080] ; /* 0x00208000070c7984 */ /* 0x000e620000000c00 */ /*0490*/ DFMA R24, R8, R20, R24 ; /* 0x000000140818722b */ /* 0x0044060000000018 */ /*04a0*/ LDS.64 R20, [R3.X8+0x1100] ; /* 0x0011000003147984 */ /* 0x004ea60000008a00 */ /*04b0*/ DFMA R24, R22, R10, R24 ; /* 0x0000000a1618722b */ /* 0x0010640000000018 */ /*04c0*/ LDS.64 R22, [R3.X8+0x1200] ; /* 0x0012000003167984 */ /* 0x001fe80000008a00 */ /*04d0*/ LDS.128 R8, [R7+0x2090] ; /* 0x0020900007087984 */ /* 0x000e220000000c00 */ /*04e0*/ DFMA R24, R12, R18, R24 ; /* 0x000000120c18722b */ /* 0x0022860000000018 */ /*04f0*/ LDS.64 R18, [R3.X8+0x1300] ; /* 0x0013000003127984 */ /* 0x002e660000008a00 */ /*0500*/ DFMA R24, R20, R14, R24 ; /* 0x0000000e1418722b */ /* 0x0044240000000018 */ /*0510*/ LDS.64 R20, [R3.X8+0x1400] ; /* 0x0014000003147984 */ /* 0x004fe80000008a00 */ /*0520*/ LDS.128 R12, [R7+0x20a0] ; /* 0x0020a000070c7984 */ /* 0x000ea20000000c00 */ /*0530*/ DFMA R24, R8, R22, R24 ; /* 0x000000160818722b */ /* 0x0010460000000018 */ /*0540*/ LDS.64 R22, [R3.X8+0x1500] ; /* 0x0015000003167984 */ /* 0x001e260000008a00 */ /*0550*/ DFMA R24, R18, R10, R24 ; /* 0x0000000a1218722b */ /* 0x0022a40000000018 */ /*0560*/ LDS.64 R18, [R3.X8+0x1600] ; /* 0x0016000003127984 */ /* 0x002fe80000008a00 */ /*0570*/ LDS.128 R8, [R7+0x20b0] ; /* 0x0020b00007087984 */ /* 0x000e620000000c00 */ /*0580*/ DFMA R24, R12, R20, R24 ; /* 0x000000140c18722b */ /* 0x0044060000000018 */ /*0590*/ LDS.64 R20, [R3.X8+0x1700] ; /* 0x0017000003147984 */ /* 0x004ea60000008a00 */ /*05a0*/ DFMA R24, R22, R14, R24 ; /* 0x0000000e1618722b */ /* 0x0010640000000018 */ /*05b0*/ LDS.64 R22, [R3.X8+0x1800] ; /* 0x0018000003167984 */ /* 0x001fe80000008a00 */ /*05c0*/ LDS.128 R12, [R7+0x20c0] ; /* 0x0020c000070c7984 */ /* 0x000e220000000c00 */ /*05d0*/ DFMA R24, R8, R18, R24 ; /* 0x000000120818722b */ /* 0x0022860000000018 */ /*05e0*/ LDS.64 R18, [R3.X8+0x1900] ; /* 0x0019000003127984 */ /* 0x002e660000008a00 */ /*05f0*/ DFMA R24, R20, R10, R24 ; /* 0x0000000a1418722b */ /* 0x0044240000000018 */ /*0600*/ LDS.64 R20, [R3.X8+0x1a00] ; /* 0x001a000003147984 */ /* 0x004fe80000008a00 */ /*0610*/ LDS.128 R8, [R7+0x20d0] ; /* 0x0020d00007087984 */ /* 0x000ea20000000c00 */ /*0620*/ DFMA R24, R12, R22, R24 ; /* 0x000000160c18722b */ /* 0x0010460000000018 */ /*0630*/ LDS.64 R22, [R3.X8+0x1b00] ; /* 0x001b000003167984 */ /* 0x001e260000008a00 */ /*0640*/ DFMA R24, R18, R14, R24 ; /* 0x0000000e1218722b */ /* 0x0022a40000000018 */ /*0650*/ LDS.64 R18, [R3.X8+0x1c00] ; /* 0x001c000003127984 */ /* 0x002fe80000008a00 */ /*0660*/ LDS.128 R12, [R7+0x20e0] ; /* 0x0020e000070c7984 */ /* 0x000e620000000c00 */ /*0670*/ DFMA R24, R8, R20, R24 ; /* 0x000000140818722b */ /* 0x0044060000000018 */ /*0680*/ LDS.64 R20, [R3.X8+0x1d00] ; /* 0x001d000003147984 */ /* 0x004ea60000008a00 */ /*0690*/ DFMA R24, R22, R10, R24 ; /* 0x0000000a1618722b */ /* 0x0010640000000018 */ /*06a0*/ LDS.64 R22, [R3.X8+0x1e00] ; /* 0x001e000003167984 */ /* 0x001fe80000008a00 */ /*06b0*/ LDS.128 R8, [R7+0x20f0] ; /* 0x0020f00007087984 */ /* 0x000e220000000c00 */ /*06c0*/ DFMA R18, R12, R18, R24 ; /* 0x000000120c12722b */ /* 0x0022860000000018 */ /*06d0*/ LDS.64 R12, [R3.X8+0x1f00] ; /* 0x001f0000030c7984 */ /* 0x002e660000008a00 */ /*06e0*/ DFMA R14, R20, R14, R18 ; /* 0x0000000e140e722b */ /* 0x004e0c0000000012 */ /*06f0*/ DFMA R22, R8, R22, R14 ; /* 0x000000160816722b */ /* 0x001e4c000000000e */ /*0700*/ DFMA R22, R12, R10, R22 ; /* 0x0000000a0c16722b */ /* 0x0020620000000016 */ /*0710*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0720*/ @!P0 BRA 0x150 ; /* 0xfffffa2000008947 */ /* 0x003fea000383ffff */ /*0730*/ STG.E.64 [R16.64], R22 ; /* 0x0000001610007986 */ /* 0x000fe2000c101b04 */ /*0740*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0750*/ BRA 0x750; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0760*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0770*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0780*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0790*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*07a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*07b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*07c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*07d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*07e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*07f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include <stdio.h> #include <stdlib.h> #include <time.h> #include <inttypes.h> #include <cuda.h> #include <cuda_runtime.h> #define BLOCK_WIDTH 32 #define TAILLE 2048 #define gettime(t) clock_gettime(CLOCK_MONOTONIC_RAW, t) #define get_sub_seconde(t) (1e-9*(double)t.tv_nsec) /** return time in second */ double get_elapsedtime(void) { struct timespec st; int err = gettime(&st); if (err !=0) return 0; return (double)st.tv_sec + get_sub_seconde(st); } void init(double** A, double** B, double** C, int size) { int i = 0, j = 0; srand(2019); for(i = 0; i < size; i++) { for(j = 0; j < size; j++) { A[i][j] = rand(); B[i][j] = rand(); C[i][j] = 0.0; } } } void mult(double** A, double** B, double** C, int size) { int i = 0, j = 0, k = 0; for(i = 0; i < size; i++) { for(j = 0; j < size; j++) { double sum = 0.; for(k = 0; k < size; k++) { sum += A[i][k] * B[k][j]; } C[i][j] = sum; } } } // QUESTION 4 __global__ void MulMatrixKernel(double* A, double* B, double* C, int N) { __shared__ double share_A[BLOCK_WIDTH][BLOCK_WIDTH]; __shared__ double share_B[BLOCK_WIDTH][BLOCK_WIDTH]; double sum = 0; int col = threadIdx.x + blockDim.x * blockIdx.x; int line = threadIdx.y + blockDim.y * blockIdx.y; for(int tile = 0; tile < gridDim.x; tile++) { share_A[threadIdx.x][threadIdx.y] = A[line * N + tile * blockDim.x + threadIdx.x]; share_B[threadIdx.x][threadIdx.y] = B[(tile*blockIdx.y + threadIdx.y) * N + col]; __syncthreads(); for(int i = 0; i < BLOCK_WIDTH; i++) { sum += share_A[i][threadIdx.y]*share_B[threadIdx.x][i]; } __syncthreads(); } C[line * N + col] = sum; } // FIN QUESTION 4 int main(int argc, char** argv){ int N, i; double *A_data; double *B_data; double *C_data; double **A; double **B; double **C; double t0 = 0., t1 = 0., duration = 0.; N = (argc < 2)?TAILLE:atoi(argv[1]); fprintf(stdout, "Matrix Multiplication\n Size: %dx%d\n", N, N); // Memory allocation A_data = (double*) malloc(sizeof(double) * N * N); B_data = (double*) malloc(sizeof(double) * N * N); C_data = (double*) malloc(sizeof(double) * N * N); A = (double**) malloc(sizeof(double *) * N); B = (double**) malloc(sizeof(double *) * N); C = (double**) malloc(sizeof(double *) * N); for(i = 0; i < N; i++) { A[i] = &A_data[i * N]; B[i] = &B_data[i * N]; C[i] = &C_data[i * N]; } // Value initialization init(A, B, C, N); // QUESTION 8 cudaEvent_t start, stop; cudaEventCreate(&start); cudaEventCreate(&stop); //FIN QUESTION 8 // QUESTION 1 double *d_A, *d_B, *d_C; cudaMalloc(&d_A, sizeof(double) * N * N); cudaMalloc(&d_B, sizeof(double) * N * N); cudaMalloc(&d_C, sizeof(double) * N * N); // FIN QUESTION 1 // QUESTION 2 cudaMemcpy(d_A, A_data, sizeof(double) * N * N, cudaMemcpyHostToDevice); cudaMemcpy(d_B, B_data, sizeof(double) * N * N, cudaMemcpyHostToDevice); cudaMemcpy(d_C, C_data, sizeof(double) * N * N, cudaMemcpyHostToDevice); // FIN QUESTION 2 // QUESTION 3 int nbBlocks = N / BLOCK_WIDTH; if(N % BLOCK_WIDTH) nbBlocks++; dim3 gridSize(nbBlocks, nbBlocks); dim3 blockSize(BLOCK_WIDTH, BLOCK_WIDTH); // FIN QUESTION 3 // QUESTION 4 cudaEventRecord(start); // QUESTION 8 MulMatrixKernel<<<gridSize, blockSize>>>(d_A, d_B, d_C, N); cudaEventRecord(stop); // QUESTION 8 // FIN QUESTION 4 // QUESTION 5 cudaMemcpy(C_data, d_C, sizeof(double) * N * N, cudaMemcpyDeviceToHost); // FIN QUESTION 5 // QUESTION 8 cudaEventSynchronize(stop); uint64_t nb_op = N * N * N; float milliseconds = 0; cudaEventElapsedTime(&milliseconds, start, stop); printf("Matrice %dx%d\n\tTemps: %f s\n\tMFlops: %.2f\n", N, N, milliseconds/1000, (nb_op / (milliseconds/1000))*1E-6); // FIN QUESTION 8 // Compute multiplication t0 = get_elapsedtime(); // mult(A, B, C, N); t1 = get_elapsedtime(); // Pretty print duration = (t1 - t0); fprintf(stdout, "Performance results: \n"); fprintf(stdout, " Time: %lf s\n", duration); fprintf(stdout, " MFlops: %.2f\n", (nb_op / duration)*1E-6); return 0; }
.file "tmpxft_001155b8_00000000-6_dgemm_8.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2063: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2063: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z15get_elapsedtimev .type _Z15get_elapsedtimev, @function _Z15get_elapsedtimev: .LFB2057: .cfi_startproc endbr64 subq $40, %rsp .cfi_def_cfa_offset 48 movq %fs:40, %rax movq %rax, 24(%rsp) xorl %eax, %eax movq %rsp, %rsi movl $4, %edi call clock_gettime@PLT pxor %xmm0, %xmm0 testl %eax, %eax jne .L3 pxor %xmm0, %xmm0 cvtsi2sdq 8(%rsp), %xmm0 mulsd .LC1(%rip), %xmm0 pxor %xmm1, %xmm1 cvtsi2sdq (%rsp), %xmm1 addsd %xmm1, %xmm0 .L3: movq 24(%rsp), %rax subq %fs:40, %rax jne .L8 addq $40, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L8: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size _Z15get_elapsedtimev, .-_Z15get_elapsedtimev .globl _Z4initPPdS0_S0_i .type _Z4initPPdS0_S0_i, @function _Z4initPPdS0_S0_i: .LFB2058: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $8, %rsp .cfi_def_cfa_offset 64 movq %rdi, %r15 movq %rsi, %r13 movq %rdx, %r12 movl %ecx, %ebx movl $2019, %edi call srand@PLT testl %ebx, %ebx jle .L9 movq %r15, %rbp movslq %ebx, %rbx leaq 0(,%rbx,8), %r14 addq %r14, %r15 .L11: movl $0, %ebx .L12: call rand@PLT movq 0(%rbp), %rdx pxor %xmm0, %xmm0 cvtsi2sdl %eax, %xmm0 movsd %xmm0, (%rdx,%rbx) call rand@PLT movq 0(%r13), %rdx pxor %xmm0, %xmm0 cvtsi2sdl %eax, %xmm0 movsd %xmm0, (%rdx,%rbx) movq (%r12), %rax movq $0x000000000, (%rax,%rbx) addq $8, %rbx cmpq %r14, %rbx jne .L12 addq $8, %rbp addq $8, %r13 addq $8, %r12 cmpq %r15, %rbp jne .L11 .L9: addq $8, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2058: .size _Z4initPPdS0_S0_i, .-_Z4initPPdS0_S0_i .globl _Z4multPPdS0_S0_i .type _Z4multPPdS0_S0_i, @function _Z4multPPdS0_S0_i: .LFB2059: .cfi_startproc endbr64 movq %rdi, %r10 movq %rsi, %r8 movq %rdx, %r11 movl %ecx, %eax testl %ecx, %ecx jle .L15 cltq leaq 0(,%rax,8), %rsi movl $0, %r9d .L17: movl $0, %ecx .L20: movq (%r10,%r9), %rdi movl $0, %eax pxor %xmm1, %xmm1 .L18: movq (%r8,%rax), %rdx movsd (%rdx,%rcx), %xmm0 mulsd (%rdi,%rax), %xmm0 addsd %xmm0, %xmm1 addq $8, %rax cmpq %rsi, %rax jne .L18 movq (%r11,%r9), %rax movsd %xmm1, (%rax,%rcx) addq $8, %rcx cmpq %rsi, %rcx jne .L20 addq $8, %r9 cmpq %rsi, %r9 jne .L17 .L15: ret .cfi_endproc .LFE2059: .size _Z4multPPdS0_S0_i, .-_Z4multPPdS0_S0_i .globl _Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i .type _Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i, @function _Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i: .LFB2085: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 24(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movl %ecx, 4(%rsp) movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax leaq 24(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) leaq 4(%rsp), %rax movq %rax, 120(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 56(%rsp) movl $1, 60(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) leaq 40(%rsp), %rcx leaq 32(%rsp), %rdx leaq 60(%rsp), %rsi leaq 48(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L26 .L22: movq 136(%rsp), %rax subq %fs:40, %rax jne .L27 addq $152, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L26: .cfi_restore_state pushq 40(%rsp) .cfi_def_cfa_offset 168 pushq 40(%rsp) .cfi_def_cfa_offset 176 leaq 112(%rsp), %r9 movq 76(%rsp), %rcx movl 84(%rsp), %r8d movq 64(%rsp), %rsi movl 72(%rsp), %edx leaq _Z15MulMatrixKernelPdS_S_i(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L22 .L27: call __stack_chk_fail@PLT .cfi_endproc .LFE2085: .size _Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i, .-_Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i .globl _Z15MulMatrixKernelPdS_S_i .type _Z15MulMatrixKernelPdS_S_i, @function _Z15MulMatrixKernelPdS_S_i: .LFB2086: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2086: .size _Z15MulMatrixKernelPdS_S_i, .-_Z15MulMatrixKernelPdS_S_i .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC2: .string "Matrix Multiplication\n Size: %dx%d\n" .align 8 .LC6: .string "Matrice %dx%d\n\tTemps: %f s\n\tMFlops: %.2f\n" .section .rodata.str1.1,"aMS",@progbits,1 .LC7: .string "Performance results: \n" .LC8: .string " Time: %lf s\n" .LC9: .string " MFlops: %.2f\n" .text .globl main .type main, @function main: .LFB2060: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $120, %rsp .cfi_def_cfa_offset 176 movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax cmpl $1, %edi jg .L45 movl $2048, %r8d movl $2048, %ecx leaq .LC2(%rip), %rdx movl $2, %esi movq stdout(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movl $33554432, %edi call malloc@PLT movq %rax, 8(%rsp) movl $33554432, %edi call malloc@PLT movq %rax, 16(%rsp) movl $33554432, %edi call malloc@PLT movq %rax, %r15 movl $16384, %edi call malloc@PLT movq %rax, %r12 movl $16384, %edi call malloc@PLT movq %rax, %rbp movl $16384, %edi call malloc@PLT movq %rax, %rdx movl $2048, %r13d movl $16384, %ebx movl $33554432, %r14d .L40: movq 8(%rsp), %r8 movq 16(%rsp), %rdi movq %r15, %rsi movl $0, %ecx .L33: movq %r8, (%r12,%rcx) movq %rdi, 0(%rbp,%rcx) movq %rsi, (%rdx,%rcx) addq %rbx, %r8 addq %rbx, %rdi addq %rbx, %rsi addq $8, %rcx cmpq %rbx, %rcx jne .L33 .L32: movl %r13d, %ecx movq %rbp, %rsi movq %r12, %rdi call _Z4initPPdS0_S0_i leaq 40(%rsp), %rdi call cudaEventCreate@PLT leaq 48(%rsp), %rdi call cudaEventCreate@PLT leaq 56(%rsp), %rdi movq %r14, %rsi call cudaMalloc@PLT leaq 64(%rsp), %rdi movq %r14, %rsi call cudaMalloc@PLT leaq 72(%rsp), %rdi movq %r14, %rsi call cudaMalloc@PLT movl $1, %ecx movq %r14, %rdx movq 8(%rsp), %rsi movq 56(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movq %r14, %rdx movq 16(%rsp), %rsi movq 64(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movq %r14, %rdx movq %r15, %rsi movq 72(%rsp), %rdi call cudaMemcpy@PLT leal 31(%r13), %eax testl %r13d, %r13d cmovns %r13d, %eax sarl $5, %eax movl %r13d, %edx andl $31, %edx cmpl $1, %edx sbbl $-1, %eax movl %eax, 80(%rsp) movl %eax, 84(%rsp) movl $1, 88(%rsp) movl $32, 92(%rsp) movl $32, 96(%rsp) movl $1, 100(%rsp) movl $0, %esi movq 40(%rsp), %rdi call cudaEventRecord@PLT movl 100(%rsp), %ecx movl $0, %r9d movl $0, %r8d movq 92(%rsp), %rdx movq 80(%rsp), %rdi movl 88(%rsp), %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L46 .L35: movl $0, %esi movq 48(%rsp), %rdi call cudaEventRecord@PLT movl $2, %ecx movq %r14, %rdx movq 72(%rsp), %rsi movq %r15, %rdi call cudaMemcpy@PLT movq 48(%rsp), %rdi call cudaEventSynchronize@PLT movl %r13d, %ebx imull %r13d, %ebx imull %r13d, %ebx movslq %ebx, %rbx movl $0x00000000, 36(%rsp) leaq 36(%rsp), %rdi movq 48(%rsp), %rdx movq 40(%rsp), %rsi call cudaEventElapsedTime@PLT movss 36(%rsp), %xmm0 divss .LC4(%rip), %xmm0 testq %rbx, %rbx js .L36 pxor %xmm1, %xmm1 cvtsi2ssq %rbx, %xmm1 .L37: divss %xmm0, %xmm1 cvtss2sd %xmm1, %xmm1 cvtss2sd %xmm0, %xmm0 mulsd .LC5(%rip), %xmm1 movl %r13d, %ecx movl %r13d, %edx leaq .LC6(%rip), %rsi movl $2, %edi movl $2, %eax call __printf_chk@PLT call _Z15get_elapsedtimev movsd %xmm0, 8(%rsp) call _Z15get_elapsedtimev movapd %xmm0, %xmm2 subsd 8(%rsp), %xmm2 movsd %xmm2, 8(%rsp) leaq .LC7(%rip), %rdx movl $2, %esi movq stdout(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movsd 8(%rsp), %xmm0 leaq .LC8(%rip), %rdx movl $2, %esi movq stdout(%rip), %rdi movl $1, %eax call __fprintf_chk@PLT testq %rbx, %rbx js .L38 pxor %xmm0, %xmm0 cvtsi2sdq %rbx, %xmm0 .L39: divsd 8(%rsp), %xmm0 mulsd .LC5(%rip), %xmm0 leaq .LC9(%rip), %rdx movl $2, %esi movq stdout(%rip), %rdi movl $1, %eax call __fprintf_chk@PLT movq 104(%rsp), %rax subq %fs:40, %rax jne .L47 movl $0, %eax addq $120, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L45: .cfi_restore_state movq 8(%rsi), %rdi movl $10, %edx movl $0, %esi call __isoc23_strtol@PLT movq %rax, %rbx movl %eax, %r13d movl %eax, %r8d movl %eax, %ecx leaq .LC2(%rip), %rdx movl $2, %esi movq stdout(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movq %rbx, 24(%rsp) movslq 24(%rsp), %rbx movq %rbx, %r14 imulq %rbx, %r14 salq $3, %r14 movq %r14, %rdi call malloc@PLT movq %rax, 8(%rsp) movq %r14, %rdi call malloc@PLT movq %rax, 16(%rsp) movq %r14, %rdi call malloc@PLT movq %rax, %r15 salq $3, %rbx movq %rbx, %rdi call malloc@PLT movq %rax, %r12 movq %rbx, %rdi call malloc@PLT movq %rax, %rbp movq %rbx, %rdi call malloc@PLT movq %rax, %rdx cmpl $0, 24(%rsp) jg .L40 jmp .L32 .L46: movl %r13d, %ecx movq 72(%rsp), %rdx movq 64(%rsp), %rsi movq 56(%rsp), %rdi call _Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i jmp .L35 .L36: movq %rbx, %rax shrq %rax movq %rbx, %rdx andl $1, %edx orq %rdx, %rax pxor %xmm1, %xmm1 cvtsi2ssq %rax, %xmm1 addss %xmm1, %xmm1 jmp .L37 .L38: movq %rbx, %rax shrq %rax andl $1, %ebx orq %rbx, %rax pxor %xmm0, %xmm0 cvtsi2sdq %rax, %xmm0 addsd %xmm0, %xmm0 jmp .L39 .L47: call __stack_chk_fail@PLT .cfi_endproc .LFE2060: .size main, .-main .section .rodata.str1.1 .LC10: .string "_Z15MulMatrixKernelPdS_S_i" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2088: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC10(%rip), %rdx movq %rdx, %rcx leaq _Z15MulMatrixKernelPdS_S_i(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2088: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .section .rodata.cst8,"aM",@progbits,8 .align 8 .LC1: .long -400107883 .long 1041313291 .section .rodata.cst4,"aM",@progbits,4 .align 4 .LC4: .long 1148846080 .section .rodata.cst8 .align 8 .LC5: .long -1598689907 .long 1051772663 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include <stdio.h> #include <stdlib.h> #include <time.h> #include <inttypes.h> #include <cuda.h> #include <cuda_runtime.h> #define BLOCK_WIDTH 32 #define TAILLE 2048 #define gettime(t) clock_gettime(CLOCK_MONOTONIC_RAW, t) #define get_sub_seconde(t) (1e-9*(double)t.tv_nsec) /** return time in second */ double get_elapsedtime(void) { struct timespec st; int err = gettime(&st); if (err !=0) return 0; return (double)st.tv_sec + get_sub_seconde(st); } void init(double** A, double** B, double** C, int size) { int i = 0, j = 0; srand(2019); for(i = 0; i < size; i++) { for(j = 0; j < size; j++) { A[i][j] = rand(); B[i][j] = rand(); C[i][j] = 0.0; } } } void mult(double** A, double** B, double** C, int size) { int i = 0, j = 0, k = 0; for(i = 0; i < size; i++) { for(j = 0; j < size; j++) { double sum = 0.; for(k = 0; k < size; k++) { sum += A[i][k] * B[k][j]; } C[i][j] = sum; } } } // QUESTION 4 __global__ void MulMatrixKernel(double* A, double* B, double* C, int N) { __shared__ double share_A[BLOCK_WIDTH][BLOCK_WIDTH]; __shared__ double share_B[BLOCK_WIDTH][BLOCK_WIDTH]; double sum = 0; int col = threadIdx.x + blockDim.x * blockIdx.x; int line = threadIdx.y + blockDim.y * blockIdx.y; for(int tile = 0; tile < gridDim.x; tile++) { share_A[threadIdx.x][threadIdx.y] = A[line * N + tile * blockDim.x + threadIdx.x]; share_B[threadIdx.x][threadIdx.y] = B[(tile*blockIdx.y + threadIdx.y) * N + col]; __syncthreads(); for(int i = 0; i < BLOCK_WIDTH; i++) { sum += share_A[i][threadIdx.y]*share_B[threadIdx.x][i]; } __syncthreads(); } C[line * N + col] = sum; } // FIN QUESTION 4 int main(int argc, char** argv){ int N, i; double *A_data; double *B_data; double *C_data; double **A; double **B; double **C; double t0 = 0., t1 = 0., duration = 0.; N = (argc < 2)?TAILLE:atoi(argv[1]); fprintf(stdout, "Matrix Multiplication\n Size: %dx%d\n", N, N); // Memory allocation A_data = (double*) malloc(sizeof(double) * N * N); B_data = (double*) malloc(sizeof(double) * N * N); C_data = (double*) malloc(sizeof(double) * N * N); A = (double**) malloc(sizeof(double *) * N); B = (double**) malloc(sizeof(double *) * N); C = (double**) malloc(sizeof(double *) * N); for(i = 0; i < N; i++) { A[i] = &A_data[i * N]; B[i] = &B_data[i * N]; C[i] = &C_data[i * N]; } // Value initialization init(A, B, C, N); // QUESTION 8 cudaEvent_t start, stop; cudaEventCreate(&start); cudaEventCreate(&stop); //FIN QUESTION 8 // QUESTION 1 double *d_A, *d_B, *d_C; cudaMalloc(&d_A, sizeof(double) * N * N); cudaMalloc(&d_B, sizeof(double) * N * N); cudaMalloc(&d_C, sizeof(double) * N * N); // FIN QUESTION 1 // QUESTION 2 cudaMemcpy(d_A, A_data, sizeof(double) * N * N, cudaMemcpyHostToDevice); cudaMemcpy(d_B, B_data, sizeof(double) * N * N, cudaMemcpyHostToDevice); cudaMemcpy(d_C, C_data, sizeof(double) * N * N, cudaMemcpyHostToDevice); // FIN QUESTION 2 // QUESTION 3 int nbBlocks = N / BLOCK_WIDTH; if(N % BLOCK_WIDTH) nbBlocks++; dim3 gridSize(nbBlocks, nbBlocks); dim3 blockSize(BLOCK_WIDTH, BLOCK_WIDTH); // FIN QUESTION 3 // QUESTION 4 cudaEventRecord(start); // QUESTION 8 MulMatrixKernel<<<gridSize, blockSize>>>(d_A, d_B, d_C, N); cudaEventRecord(stop); // QUESTION 8 // FIN QUESTION 4 // QUESTION 5 cudaMemcpy(C_data, d_C, sizeof(double) * N * N, cudaMemcpyDeviceToHost); // FIN QUESTION 5 // QUESTION 8 cudaEventSynchronize(stop); uint64_t nb_op = N * N * N; float milliseconds = 0; cudaEventElapsedTime(&milliseconds, start, stop); printf("Matrice %dx%d\n\tTemps: %f s\n\tMFlops: %.2f\n", N, N, milliseconds/1000, (nb_op / (milliseconds/1000))*1E-6); // FIN QUESTION 8 // Compute multiplication t0 = get_elapsedtime(); // mult(A, B, C, N); t1 = get_elapsedtime(); // Pretty print duration = (t1 - t0); fprintf(stdout, "Performance results: \n"); fprintf(stdout, " Time: %lf s\n", duration); fprintf(stdout, " MFlops: %.2f\n", (nb_op / duration)*1E-6); return 0; }
#include <stdio.h> #include <stdlib.h> #include <time.h> #include <inttypes.h> #include <hip/hip_runtime.h> #define BLOCK_WIDTH 32 #define TAILLE 2048 #define gettime(t) clock_gettime(CLOCK_MONOTONIC_RAW, t) #define get_sub_seconde(t) (1e-9*(double)t.tv_nsec) /** return time in second */ double get_elapsedtime(void) { struct timespec st; int err = gettime(&st); if (err !=0) return 0; return (double)st.tv_sec + get_sub_seconde(st); } void init(double** A, double** B, double** C, int size) { int i = 0, j = 0; srand(2019); for(i = 0; i < size; i++) { for(j = 0; j < size; j++) { A[i][j] = rand(); B[i][j] = rand(); C[i][j] = 0.0; } } } void mult(double** A, double** B, double** C, int size) { int i = 0, j = 0, k = 0; for(i = 0; i < size; i++) { for(j = 0; j < size; j++) { double sum = 0.; for(k = 0; k < size; k++) { sum += A[i][k] * B[k][j]; } C[i][j] = sum; } } } // QUESTION 4 __global__ void MulMatrixKernel(double* A, double* B, double* C, int N) { __shared__ double share_A[BLOCK_WIDTH][BLOCK_WIDTH]; __shared__ double share_B[BLOCK_WIDTH][BLOCK_WIDTH]; double sum = 0; int col = threadIdx.x + blockDim.x * blockIdx.x; int line = threadIdx.y + blockDim.y * blockIdx.y; for(int tile = 0; tile < gridDim.x; tile++) { share_A[threadIdx.x][threadIdx.y] = A[line * N + tile * blockDim.x + threadIdx.x]; share_B[threadIdx.x][threadIdx.y] = B[(tile*blockIdx.y + threadIdx.y) * N + col]; __syncthreads(); for(int i = 0; i < BLOCK_WIDTH; i++) { sum += share_A[i][threadIdx.y]*share_B[threadIdx.x][i]; } __syncthreads(); } C[line * N + col] = sum; } // FIN QUESTION 4 int main(int argc, char** argv){ int N, i; double *A_data; double *B_data; double *C_data; double **A; double **B; double **C; double t0 = 0., t1 = 0., duration = 0.; N = (argc < 2)?TAILLE:atoi(argv[1]); fprintf(stdout, "Matrix Multiplication\n Size: %dx%d\n", N, N); // Memory allocation A_data = (double*) malloc(sizeof(double) * N * N); B_data = (double*) malloc(sizeof(double) * N * N); C_data = (double*) malloc(sizeof(double) * N * N); A = (double**) malloc(sizeof(double *) * N); B = (double**) malloc(sizeof(double *) * N); C = (double**) malloc(sizeof(double *) * N); for(i = 0; i < N; i++) { A[i] = &A_data[i * N]; B[i] = &B_data[i * N]; C[i] = &C_data[i * N]; } // Value initialization init(A, B, C, N); // QUESTION 8 hipEvent_t start, stop; hipEventCreate(&start); hipEventCreate(&stop); //FIN QUESTION 8 // QUESTION 1 double *d_A, *d_B, *d_C; hipMalloc(&d_A, sizeof(double) * N * N); hipMalloc(&d_B, sizeof(double) * N * N); hipMalloc(&d_C, sizeof(double) * N * N); // FIN QUESTION 1 // QUESTION 2 hipMemcpy(d_A, A_data, sizeof(double) * N * N, hipMemcpyHostToDevice); hipMemcpy(d_B, B_data, sizeof(double) * N * N, hipMemcpyHostToDevice); hipMemcpy(d_C, C_data, sizeof(double) * N * N, hipMemcpyHostToDevice); // FIN QUESTION 2 // QUESTION 3 int nbBlocks = N / BLOCK_WIDTH; if(N % BLOCK_WIDTH) nbBlocks++; dim3 gridSize(nbBlocks, nbBlocks); dim3 blockSize(BLOCK_WIDTH, BLOCK_WIDTH); // FIN QUESTION 3 // QUESTION 4 hipEventRecord(start); // QUESTION 8 MulMatrixKernel<<<gridSize, blockSize>>>(d_A, d_B, d_C, N); hipEventRecord(stop); // QUESTION 8 // FIN QUESTION 4 // QUESTION 5 hipMemcpy(C_data, d_C, sizeof(double) * N * N, hipMemcpyDeviceToHost); // FIN QUESTION 5 // QUESTION 8 hipEventSynchronize(stop); uint64_t nb_op = N * N * N; float milliseconds = 0; hipEventElapsedTime(&milliseconds, start, stop); printf("Matrice %dx%d\n\tTemps: %f s\n\tMFlops: %.2f\n", N, N, milliseconds/1000, (nb_op / (milliseconds/1000))*1E-6); // FIN QUESTION 8 // Compute multiplication t0 = get_elapsedtime(); // mult(A, B, C, N); t1 = get_elapsedtime(); // Pretty print duration = (t1 - t0); fprintf(stdout, "Performance results: \n"); fprintf(stdout, " Time: %lf s\n", duration); fprintf(stdout, " MFlops: %.2f\n", (nb_op / duration)*1E-6); return 0; }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <stdio.h> #include <stdlib.h> #include <time.h> #include <inttypes.h> #include <hip/hip_runtime.h> #define BLOCK_WIDTH 32 #define TAILLE 2048 #define gettime(t) clock_gettime(CLOCK_MONOTONIC_RAW, t) #define get_sub_seconde(t) (1e-9*(double)t.tv_nsec) /** return time in second */ double get_elapsedtime(void) { struct timespec st; int err = gettime(&st); if (err !=0) return 0; return (double)st.tv_sec + get_sub_seconde(st); } void init(double** A, double** B, double** C, int size) { int i = 0, j = 0; srand(2019); for(i = 0; i < size; i++) { for(j = 0; j < size; j++) { A[i][j] = rand(); B[i][j] = rand(); C[i][j] = 0.0; } } } void mult(double** A, double** B, double** C, int size) { int i = 0, j = 0, k = 0; for(i = 0; i < size; i++) { for(j = 0; j < size; j++) { double sum = 0.; for(k = 0; k < size; k++) { sum += A[i][k] * B[k][j]; } C[i][j] = sum; } } } // QUESTION 4 __global__ void MulMatrixKernel(double* A, double* B, double* C, int N) { __shared__ double share_A[BLOCK_WIDTH][BLOCK_WIDTH]; __shared__ double share_B[BLOCK_WIDTH][BLOCK_WIDTH]; double sum = 0; int col = threadIdx.x + blockDim.x * blockIdx.x; int line = threadIdx.y + blockDim.y * blockIdx.y; for(int tile = 0; tile < gridDim.x; tile++) { share_A[threadIdx.x][threadIdx.y] = A[line * N + tile * blockDim.x + threadIdx.x]; share_B[threadIdx.x][threadIdx.y] = B[(tile*blockIdx.y + threadIdx.y) * N + col]; __syncthreads(); for(int i = 0; i < BLOCK_WIDTH; i++) { sum += share_A[i][threadIdx.y]*share_B[threadIdx.x][i]; } __syncthreads(); } C[line * N + col] = sum; } // FIN QUESTION 4 int main(int argc, char** argv){ int N, i; double *A_data; double *B_data; double *C_data; double **A; double **B; double **C; double t0 = 0., t1 = 0., duration = 0.; N = (argc < 2)?TAILLE:atoi(argv[1]); fprintf(stdout, "Matrix Multiplication\n Size: %dx%d\n", N, N); // Memory allocation A_data = (double*) malloc(sizeof(double) * N * N); B_data = (double*) malloc(sizeof(double) * N * N); C_data = (double*) malloc(sizeof(double) * N * N); A = (double**) malloc(sizeof(double *) * N); B = (double**) malloc(sizeof(double *) * N); C = (double**) malloc(sizeof(double *) * N); for(i = 0; i < N; i++) { A[i] = &A_data[i * N]; B[i] = &B_data[i * N]; C[i] = &C_data[i * N]; } // Value initialization init(A, B, C, N); // QUESTION 8 hipEvent_t start, stop; hipEventCreate(&start); hipEventCreate(&stop); //FIN QUESTION 8 // QUESTION 1 double *d_A, *d_B, *d_C; hipMalloc(&d_A, sizeof(double) * N * N); hipMalloc(&d_B, sizeof(double) * N * N); hipMalloc(&d_C, sizeof(double) * N * N); // FIN QUESTION 1 // QUESTION 2 hipMemcpy(d_A, A_data, sizeof(double) * N * N, hipMemcpyHostToDevice); hipMemcpy(d_B, B_data, sizeof(double) * N * N, hipMemcpyHostToDevice); hipMemcpy(d_C, C_data, sizeof(double) * N * N, hipMemcpyHostToDevice); // FIN QUESTION 2 // QUESTION 3 int nbBlocks = N / BLOCK_WIDTH; if(N % BLOCK_WIDTH) nbBlocks++; dim3 gridSize(nbBlocks, nbBlocks); dim3 blockSize(BLOCK_WIDTH, BLOCK_WIDTH); // FIN QUESTION 3 // QUESTION 4 hipEventRecord(start); // QUESTION 8 MulMatrixKernel<<<gridSize, blockSize>>>(d_A, d_B, d_C, N); hipEventRecord(stop); // QUESTION 8 // FIN QUESTION 4 // QUESTION 5 hipMemcpy(C_data, d_C, sizeof(double) * N * N, hipMemcpyDeviceToHost); // FIN QUESTION 5 // QUESTION 8 hipEventSynchronize(stop); uint64_t nb_op = N * N * N; float milliseconds = 0; hipEventElapsedTime(&milliseconds, start, stop); printf("Matrice %dx%d\n\tTemps: %f s\n\tMFlops: %.2f\n", N, N, milliseconds/1000, (nb_op / (milliseconds/1000))*1E-6); // FIN QUESTION 8 // Compute multiplication t0 = get_elapsedtime(); // mult(A, B, C, N); t1 = get_elapsedtime(); // Pretty print duration = (t1 - t0); fprintf(stdout, "Performance results: \n"); fprintf(stdout, " Time: %lf s\n", duration); fprintf(stdout, " MFlops: %.2f\n", (nb_op / duration)*1E-6); return 0; }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z15MulMatrixKernelPdS_S_i .globl _Z15MulMatrixKernelPdS_S_i .p2align 8 .type _Z15MulMatrixKernelPdS_S_i,@function _Z15MulMatrixKernelPdS_S_i: s_clause 0x2 s_load_b32 s4, s[0:1], 0x2c s_load_b32 s3, s[0:1], 0x20 s_load_b32 s2, s[0:1], 0x18 v_and_b32_e32 v4, 0x3ff, v0 v_bfe_u32 v3, v0, 10, 10 s_mov_b32 s9, 0 s_waitcnt lgkmcnt(0) s_and_b32 s8, s4, 0xffff s_lshr_b32 s4, s4, 16 v_mad_u64_u32 v[0:1], null, s14, s8, v[4:5] v_mad_u64_u32 v[1:2], null, s15, s4, v[3:4] s_cmp_eq_u32 s3, 0 s_cbranch_scc1 .LBB0_5 s_load_b128 s[4:7], s[0:1], 0x0 v_dual_mov_b32 v10, 0 :: v_dual_lshlrev_b32 v5, 8, v4 v_lshlrev_b32_e32 v2, 3, v3 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_nc_u32_e32 v8, 0x2000, v5 v_mad_u64_u32 v[6:7], null, v1, s2, v[4:5] s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_4) v_dual_mov_b32 v4, 0 :: v_dual_add_nc_u32 v7, v5, v2 v_mov_b32_e32 v5, 0 v_add_nc_u32_e32 v9, v8, v2 s_set_inst_prefetch_distance 0x1 .p2align 6 .LBB0_2: s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_3) | instid1(VALU_DEP_3) v_mad_u64_u32 v[11:12], null, s9, s15, v[3:4] v_mad_u64_u32 v[12:13], null, s9, s8, v[6:7] v_mov_b32_e32 v13, v10 s_mov_b32 s10, 0 v_mad_u64_u32 v[14:15], null, v11, s2, v[0:1] v_mov_b32_e32 v15, v10 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_lshlrev_b64 v[11:12], 3, v[12:13] v_lshlrev_b64 v[13:14], 3, v[14:15] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_co_u32 v11, vcc_lo, s4, v11 v_add_co_ci_u32_e32 v12, vcc_lo, s5, v12, vcc_lo s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_u32 v13, vcc_lo, s6, v13 v_add_co_ci_u32_e32 v14, vcc_lo, s7, v14, vcc_lo global_load_b64 v[15:16], v[11:12], off global_load_b64 v[12:13], v[13:14], off v_mov_b32_e32 v11, v2 s_waitcnt vmcnt(1) ds_store_b64 v7, v[15:16] s_waitcnt vmcnt(0) ds_store_b64 v9, v[12:13] s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv .LBB0_3: v_add_nc_u32_e32 v14, s10, v8 s_add_i32 s10, s10, 8 ds_load_b64 v[12:13], v11 ds_load_b64 v[14:15], v14 v_add_nc_u32_e32 v11, 0x100, v11 s_cmpk_eq_i32 s10, 0x100 s_waitcnt lgkmcnt(0) v_fma_f64 v[4:5], v[12:13], v[14:15], v[4:5] s_cbranch_scc0 .LBB0_3 s_add_i32 s9, s9, 1 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_eq_u32 s9, s3 s_barrier buffer_gl0_inv s_cbranch_scc0 .LBB0_2 s_branch .LBB0_6 .LBB0_5: v_mov_b32_e32 v4, 0 v_mov_b32_e32 v5, 0 .LBB0_6: s_set_inst_prefetch_distance 0x2 s_load_b64 s[0:1], s[0:1], 0x10 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_mad_u64_u32 v[2:3], null, v1, s2, v[0:1] v_ashrrev_i32_e32 v3, 31, v2 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[0:1], 3, v[2:3] s_waitcnt lgkmcnt(0) v_add_co_u32 v0, vcc_lo, s0, v0 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo global_store_b64 v[0:1], v[4:5], off s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z15MulMatrixKernelPdS_S_i .amdhsa_group_segment_fixed_size 16384 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 288 .amdhsa_user_sgpr_count 14 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 1 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 1 .amdhsa_next_free_vgpr 17 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z15MulMatrixKernelPdS_S_i, .Lfunc_end0-_Z15MulMatrixKernelPdS_S_i .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 8 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .offset: 24 .size: 4 .value_kind: by_value - .offset: 32 .size: 4 .value_kind: hidden_block_count_x - .offset: 36 .size: 4 .value_kind: hidden_block_count_y - .offset: 40 .size: 4 .value_kind: hidden_block_count_z - .offset: 44 .size: 2 .value_kind: hidden_group_size_x - .offset: 46 .size: 2 .value_kind: hidden_group_size_y - .offset: 48 .size: 2 .value_kind: hidden_group_size_z - .offset: 50 .size: 2 .value_kind: hidden_remainder_x - .offset: 52 .size: 2 .value_kind: hidden_remainder_y - .offset: 54 .size: 2 .value_kind: hidden_remainder_z - .offset: 72 .size: 8 .value_kind: hidden_global_offset_x - .offset: 80 .size: 8 .value_kind: hidden_global_offset_y - .offset: 88 .size: 8 .value_kind: hidden_global_offset_z - .offset: 96 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 16384 .kernarg_segment_align: 8 .kernarg_segment_size: 288 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z15MulMatrixKernelPdS_S_i .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z15MulMatrixKernelPdS_S_i.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 17 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <stdio.h> #include <stdlib.h> #include <time.h> #include <inttypes.h> #include <hip/hip_runtime.h> #define BLOCK_WIDTH 32 #define TAILLE 2048 #define gettime(t) clock_gettime(CLOCK_MONOTONIC_RAW, t) #define get_sub_seconde(t) (1e-9*(double)t.tv_nsec) /** return time in second */ double get_elapsedtime(void) { struct timespec st; int err = gettime(&st); if (err !=0) return 0; return (double)st.tv_sec + get_sub_seconde(st); } void init(double** A, double** B, double** C, int size) { int i = 0, j = 0; srand(2019); for(i = 0; i < size; i++) { for(j = 0; j < size; j++) { A[i][j] = rand(); B[i][j] = rand(); C[i][j] = 0.0; } } } void mult(double** A, double** B, double** C, int size) { int i = 0, j = 0, k = 0; for(i = 0; i < size; i++) { for(j = 0; j < size; j++) { double sum = 0.; for(k = 0; k < size; k++) { sum += A[i][k] * B[k][j]; } C[i][j] = sum; } } } // QUESTION 4 __global__ void MulMatrixKernel(double* A, double* B, double* C, int N) { __shared__ double share_A[BLOCK_WIDTH][BLOCK_WIDTH]; __shared__ double share_B[BLOCK_WIDTH][BLOCK_WIDTH]; double sum = 0; int col = threadIdx.x + blockDim.x * blockIdx.x; int line = threadIdx.y + blockDim.y * blockIdx.y; for(int tile = 0; tile < gridDim.x; tile++) { share_A[threadIdx.x][threadIdx.y] = A[line * N + tile * blockDim.x + threadIdx.x]; share_B[threadIdx.x][threadIdx.y] = B[(tile*blockIdx.y + threadIdx.y) * N + col]; __syncthreads(); for(int i = 0; i < BLOCK_WIDTH; i++) { sum += share_A[i][threadIdx.y]*share_B[threadIdx.x][i]; } __syncthreads(); } C[line * N + col] = sum; } // FIN QUESTION 4 int main(int argc, char** argv){ int N, i; double *A_data; double *B_data; double *C_data; double **A; double **B; double **C; double t0 = 0., t1 = 0., duration = 0.; N = (argc < 2)?TAILLE:atoi(argv[1]); fprintf(stdout, "Matrix Multiplication\n Size: %dx%d\n", N, N); // Memory allocation A_data = (double*) malloc(sizeof(double) * N * N); B_data = (double*) malloc(sizeof(double) * N * N); C_data = (double*) malloc(sizeof(double) * N * N); A = (double**) malloc(sizeof(double *) * N); B = (double**) malloc(sizeof(double *) * N); C = (double**) malloc(sizeof(double *) * N); for(i = 0; i < N; i++) { A[i] = &A_data[i * N]; B[i] = &B_data[i * N]; C[i] = &C_data[i * N]; } // Value initialization init(A, B, C, N); // QUESTION 8 hipEvent_t start, stop; hipEventCreate(&start); hipEventCreate(&stop); //FIN QUESTION 8 // QUESTION 1 double *d_A, *d_B, *d_C; hipMalloc(&d_A, sizeof(double) * N * N); hipMalloc(&d_B, sizeof(double) * N * N); hipMalloc(&d_C, sizeof(double) * N * N); // FIN QUESTION 1 // QUESTION 2 hipMemcpy(d_A, A_data, sizeof(double) * N * N, hipMemcpyHostToDevice); hipMemcpy(d_B, B_data, sizeof(double) * N * N, hipMemcpyHostToDevice); hipMemcpy(d_C, C_data, sizeof(double) * N * N, hipMemcpyHostToDevice); // FIN QUESTION 2 // QUESTION 3 int nbBlocks = N / BLOCK_WIDTH; if(N % BLOCK_WIDTH) nbBlocks++; dim3 gridSize(nbBlocks, nbBlocks); dim3 blockSize(BLOCK_WIDTH, BLOCK_WIDTH); // FIN QUESTION 3 // QUESTION 4 hipEventRecord(start); // QUESTION 8 MulMatrixKernel<<<gridSize, blockSize>>>(d_A, d_B, d_C, N); hipEventRecord(stop); // QUESTION 8 // FIN QUESTION 4 // QUESTION 5 hipMemcpy(C_data, d_C, sizeof(double) * N * N, hipMemcpyDeviceToHost); // FIN QUESTION 5 // QUESTION 8 hipEventSynchronize(stop); uint64_t nb_op = N * N * N; float milliseconds = 0; hipEventElapsedTime(&milliseconds, start, stop); printf("Matrice %dx%d\n\tTemps: %f s\n\tMFlops: %.2f\n", N, N, milliseconds/1000, (nb_op / (milliseconds/1000))*1E-6); // FIN QUESTION 8 // Compute multiplication t0 = get_elapsedtime(); // mult(A, B, C, N); t1 = get_elapsedtime(); // Pretty print duration = (t1 - t0); fprintf(stdout, "Performance results: \n"); fprintf(stdout, " Time: %lf s\n", duration); fprintf(stdout, " MFlops: %.2f\n", (nb_op / duration)*1E-6); return 0; }
.text .file "dgemm_8.hip" .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 # -- Begin function _Z15get_elapsedtimev .LCPI0_0: .quad 0x3e112e0be826d695 # double 1.0000000000000001E-9 .text .globl _Z15get_elapsedtimev .p2align 4, 0x90 .type _Z15get_elapsedtimev,@function _Z15get_elapsedtimev: # @_Z15get_elapsedtimev .cfi_startproc # %bb.0: subq $24, %rsp .cfi_def_cfa_offset 32 leaq 8(%rsp), %rsi movl $4, %edi callq clock_gettime xorps %xmm0, %xmm0 testl %eax, %eax je .LBB0_1 # %bb.2: addq $24, %rsp .cfi_def_cfa_offset 8 retq .LBB0_1: .cfi_def_cfa_offset 32 cvtsi2sdq 8(%rsp), %xmm1 xorps %xmm0, %xmm0 cvtsi2sdq 16(%rsp), %xmm0 mulsd .LCPI0_0(%rip), %xmm0 addsd %xmm1, %xmm0 addq $24, %rsp .cfi_def_cfa_offset 8 retq .Lfunc_end0: .size _Z15get_elapsedtimev, .Lfunc_end0-_Z15get_elapsedtimev .cfi_endproc # -- End function .globl _Z4initPPdS0_S0_i # -- Begin function _Z4initPPdS0_S0_i .p2align 4, 0x90 .type _Z4initPPdS0_S0_i,@function _Z4initPPdS0_S0_i: # @_Z4initPPdS0_S0_i .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 pushq %rax .cfi_def_cfa_offset 64 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl %ecx, %ebp movq %rdx, %rbx movq %rsi, %r14 movq %rdi, %r15 movl $2019, %edi # imm = 0x7E3 callq srand testl %ebp, %ebp jle .LBB1_5 # %bb.1: # %.preheader.lr.ph movl %ebp, %r12d xorl %r13d, %r13d .p2align 4, 0x90 .LBB1_2: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB1_3 Depth 2 xorl %ebp, %ebp .p2align 4, 0x90 .LBB1_3: # Parent Loop BB1_2 Depth=1 # => This Inner Loop Header: Depth=2 callq rand xorps %xmm0, %xmm0 cvtsi2sd %eax, %xmm0 movq (%r15,%r13,8), %rax movsd %xmm0, (%rax,%rbp,8) callq rand xorps %xmm0, %xmm0 cvtsi2sd %eax, %xmm0 movq (%r14,%r13,8), %rax movsd %xmm0, (%rax,%rbp,8) movq (%rbx,%r13,8), %rax movq $0, (%rax,%rbp,8) incq %rbp cmpq %rbp, %r12 jne .LBB1_3 # %bb.4: # %._crit_edge # in Loop: Header=BB1_2 Depth=1 incq %r13 cmpq %r12, %r13 jne .LBB1_2 .LBB1_5: # %._crit_edge18 addq $8, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end1: .size _Z4initPPdS0_S0_i, .Lfunc_end1-_Z4initPPdS0_S0_i .cfi_endproc # -- End function .globl _Z4multPPdS0_S0_i # -- Begin function _Z4multPPdS0_S0_i .p2align 4, 0x90 .type _Z4multPPdS0_S0_i,@function _Z4multPPdS0_S0_i: # @_Z4multPPdS0_S0_i .cfi_startproc # %bb.0: testl %ecx, %ecx jle .LBB2_8 # %bb.1: # %.preheader23.lr.ph pushq %rbx .cfi_def_cfa_offset 16 .cfi_offset %rbx, -16 movl %ecx, %eax xorl %ecx, %ecx .p2align 4, 0x90 .LBB2_2: # %.preheader23 # =>This Loop Header: Depth=1 # Child Loop BB2_3 Depth 2 # Child Loop BB2_4 Depth 3 movq (%rdx,%rcx,8), %r8 movq (%rdi,%rcx,8), %r9 xorl %r10d, %r10d .p2align 4, 0x90 .LBB2_3: # %.preheader # Parent Loop BB2_2 Depth=1 # => This Loop Header: Depth=2 # Child Loop BB2_4 Depth 3 xorpd %xmm0, %xmm0 xorl %r11d, %r11d .p2align 4, 0x90 .LBB2_4: # Parent Loop BB2_2 Depth=1 # Parent Loop BB2_3 Depth=2 # => This Inner Loop Header: Depth=3 movsd (%r9,%r11,8), %xmm1 # xmm1 = mem[0],zero movq (%rsi,%r11,8), %rbx mulsd (%rbx,%r10,8), %xmm1 addsd %xmm1, %xmm0 incq %r11 cmpq %r11, %rax jne .LBB2_4 # %bb.5: # %._crit_edge # in Loop: Header=BB2_3 Depth=2 movsd %xmm0, (%r8,%r10,8) incq %r10 cmpq %rax, %r10 jne .LBB2_3 # %bb.6: # %._crit_edge27 # in Loop: Header=BB2_2 Depth=1 incq %rcx cmpq %rax, %rcx jne .LBB2_2 # %bb.7: popq %rbx .cfi_def_cfa_offset 8 .cfi_restore %rbx .LBB2_8: # %._crit_edge29 retq .Lfunc_end2: .size _Z4multPPdS0_S0_i, .Lfunc_end2-_Z4multPPdS0_S0_i .cfi_endproc # -- End function .globl _Z30__device_stub__MulMatrixKernelPdS_S_i # -- Begin function _Z30__device_stub__MulMatrixKernelPdS_S_i .p2align 4, 0x90 .type _Z30__device_stub__MulMatrixKernelPdS_S_i,@function _Z30__device_stub__MulMatrixKernelPdS_S_i: # @_Z30__device_stub__MulMatrixKernelPdS_S_i .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movq %rdx, 56(%rsp) movl %ecx, 4(%rsp) leaq 72(%rsp), %rax movq %rax, 80(%rsp) leaq 64(%rsp), %rax movq %rax, 88(%rsp) leaq 56(%rsp), %rax movq %rax, 96(%rsp) leaq 4(%rsp), %rax movq %rax, 104(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 80(%rsp), %r9 movl $_Z15MulMatrixKernelPdS_S_i, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $136, %rsp .cfi_adjust_cfa_offset -136 retq .Lfunc_end3: .size _Z30__device_stub__MulMatrixKernelPdS_S_i, .Lfunc_end3-_Z30__device_stub__MulMatrixKernelPdS_S_i .cfi_endproc # -- End function .section .rodata.cst4,"aM",@progbits,4 .p2align 2, 0x0 # -- Begin function main .LCPI4_0: .long 0x447a0000 # float 1000 .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 .LCPI4_1: .quad 0x3eb0c6f7a0b5ed8d # double 9.9999999999999995E-7 .LCPI4_2: .quad 0x3e112e0be826d695 # double 1.0000000000000001E-9 .section .rodata.cst16,"aM",@progbits,16 .p2align 4, 0x0 .LCPI4_3: .long 1127219200 # 0x43300000 .long 1160773632 # 0x45300000 .long 0 # 0x0 .long 0 # 0x0 .LCPI4_4: .quad 0x4330000000000000 # double 4503599627370496 .quad 0x4530000000000000 # double 1.9342813113834067E+25 .text .globl main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 subq $200, %rsp .cfi_def_cfa_offset 256 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl $2048, %r14d # imm = 0x800 cmpl $2, %edi jl .LBB4_2 # %bb.1: movq 8(%rsi), %rdi xorl %esi, %esi movl $10, %edx callq __isoc23_strtol movq %rax, %r14 .LBB4_2: movq stdout(%rip), %rdi movl $.L.str, %esi movl %r14d, %edx movl %r14d, %ecx xorl %eax, %eax callq fprintf movslq %r14d, %rbx leaq (,%rbx,8), %r12 movq %r12, %r15 imulq %rbx, %r15 movq %r15, %rdi callq malloc movq %rax, 56(%rsp) # 8-byte Spill movq %r15, %rdi callq malloc movq %rax, 64(%rsp) # 8-byte Spill movq %r15, 136(%rsp) # 8-byte Spill movq %r15, %rdi callq malloc movq %rax, 16(%rsp) # 8-byte Spill movq %r12, %rdi callq malloc movq %rax, %rbp movq %r12, %rdi callq malloc movq %rax, %r15 movq %r12, %rdi callq malloc movq %rax, %r12 movl %r14d, %r13d testl %ebx, %ebx jle .LBB4_5 # %bb.3: # %.lr.ph.preheader leaq (,%r13,8), %rax xorl %ecx, %ecx movq 56(%rsp), %rdx # 8-byte Reload movq 64(%rsp), %rsi # 8-byte Reload movq 16(%rsp), %rdi # 8-byte Reload .p2align 4, 0x90 .LBB4_4: # %.lr.ph # =>This Inner Loop Header: Depth=1 movq %rdx, (%rbp,%rcx,8) movq %rsi, (%r15,%rcx,8) movq %rdi, (%r12,%rcx,8) incq %rcx addq %rax, %rdi addq %rax, %rsi addq %rax, %rdx cmpq %rcx, %r13 jne .LBB4_4 .LBB4_5: # %._crit_edge movl $2019, %edi # imm = 0x7E3 callq srand movq %r14, 8(%rsp) # 8-byte Spill testl %r14d, %r14d jle .LBB4_10 # %bb.6: # %.preheader.lr.ph.i xorl %ebx, %ebx .p2align 4, 0x90 .LBB4_7: # %.preheader.i # =>This Loop Header: Depth=1 # Child Loop BB4_8 Depth 2 xorl %r14d, %r14d .p2align 4, 0x90 .LBB4_8: # Parent Loop BB4_7 Depth=1 # => This Inner Loop Header: Depth=2 callq rand xorps %xmm0, %xmm0 cvtsi2sd %eax, %xmm0 movq (%rbp,%rbx,8), %rax movsd %xmm0, (%rax,%r14,8) callq rand xorps %xmm0, %xmm0 cvtsi2sd %eax, %xmm0 movq (%r15,%rbx,8), %rax movsd %xmm0, (%rax,%r14,8) movq (%r12,%rbx,8), %rax movq $0, (%rax,%r14,8) incq %r14 cmpq %r14, %r13 jne .LBB4_8 # %bb.9: # %._crit_edge.i # in Loop: Header=BB4_7 Depth=1 incq %rbx cmpq %r13, %rbx jne .LBB4_7 .LBB4_10: # %_Z4initPPdS0_S0_i.exit leaq 88(%rsp), %rdi callq hipEventCreate leaq 32(%rsp), %rdi callq hipEventCreate leaq 80(%rsp), %rdi movq 136(%rsp), %rbx # 8-byte Reload movq %rbx, %rsi callq hipMalloc leaq 72(%rsp), %rdi movq %rbx, %rsi callq hipMalloc leaq 24(%rsp), %rdi movq %rbx, %rsi callq hipMalloc movq 80(%rsp), %rdi movq 56(%rsp), %rsi # 8-byte Reload movq %rbx, %rdx movl $1, %ecx callq hipMemcpy movq 72(%rsp), %rdi movq 64(%rsp), %rsi # 8-byte Reload movq %rbx, %rdx movl $1, %ecx callq hipMemcpy movq 24(%rsp), %rdi movq 16(%rsp), %r12 # 8-byte Reload movq %r12, %rsi movq %rbx, %rdx movl $1, %ecx callq hipMemcpy movq 8(%rsp), %r14 # 8-byte Reload leal 31(%r14), %eax testl %r14d, %r14d cmovnsl %r14d, %eax sarl $5, %eax movl %r14d, %ecx andl $31, %ecx cmpl $1, %ecx sbbl $-1, %eax movq %rax, %r15 shlq $32, %r15 orq %rax, %r15 movq 88(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movabsq $137438953504, %rdx # imm = 0x2000000020 movq %r15, %rdi movl $1, %esi movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB4_12 # %bb.11: movq 80(%rsp), %rax movq 72(%rsp), %rcx movq 24(%rsp), %rdx movq %rax, 192(%rsp) movq %rcx, 184(%rsp) movq %rdx, 176(%rsp) movl %r14d, 52(%rsp) leaq 192(%rsp), %rax movq %rax, 96(%rsp) leaq 184(%rsp), %rax movq %rax, 104(%rsp) leaq 176(%rsp), %rax movq %rax, 112(%rsp) leaq 52(%rsp), %rax movq %rax, 120(%rsp) leaq 40(%rsp), %rdi leaq 160(%rsp), %rsi leaq 152(%rsp), %rdx leaq 144(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 160(%rsp), %rcx movl 168(%rsp), %r8d leaq 96(%rsp), %r9 movl $_Z15MulMatrixKernelPdS_S_i, %edi pushq 144(%rsp) .cfi_adjust_cfa_offset 8 pushq 160(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB4_12: movq 32(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movq 24(%rsp), %rsi movq %r12, %rdi movq %rbx, %rdx movl $2, %ecx callq hipMemcpy movq 32(%rsp), %rdi callq hipEventSynchronize movl %r14d, %ebx imull %ebx, %ebx imull %r14d, %ebx movslq %ebx, %r14 movl $0, 40(%rsp) movq 88(%rsp), %rsi movq 32(%rsp), %rdx leaq 40(%rsp), %rdi callq hipEventElapsedTime movss 40(%rsp), %xmm1 # xmm1 = mem[0],zero,zero,zero divss .LCPI4_0(%rip), %xmm1 cvtss2sd %xmm1, %xmm0 testq %r14, %r14 js .LBB4_13 # %bb.14: cvtsi2ss %ebx, %xmm2 jmp .LBB4_15 .LBB4_13: movq %r14, %rax shrq %rax andl $1, %ebx orq %rax, %rbx cvtsi2ss %rbx, %xmm2 addss %xmm2, %xmm2 .LBB4_15: divss %xmm1, %xmm2 xorps %xmm1, %xmm1 cvtss2sd %xmm2, %xmm1 mulsd .LCPI4_1(%rip), %xmm1 movl $.L.str.1, %edi movq 8(%rsp), %rdx # 8-byte Reload movl %edx, %esi # kill: def $edx killed $edx killed $rdx movb $2, %al callq printf leaq 96(%rsp), %rsi movl $4, %edi callq clock_gettime xorps %xmm0, %xmm0 movsd %xmm0, 8(%rsp) # 8-byte Spill xorpd %xmm1, %xmm1 testl %eax, %eax jne .LBB4_17 # %bb.16: xorps %xmm0, %xmm0 cvtsi2sdq 96(%rsp), %xmm0 xorps %xmm1, %xmm1 cvtsi2sdq 104(%rsp), %xmm1 mulsd .LCPI4_2(%rip), %xmm1 addsd %xmm0, %xmm1 .LBB4_17: # %_Z15get_elapsedtimev.exit movsd %xmm1, 16(%rsp) # 8-byte Spill leaq 96(%rsp), %rsi movl $4, %edi callq clock_gettime testl %eax, %eax jne .LBB4_19 # %bb.18: xorps %xmm0, %xmm0 cvtsi2sdq 96(%rsp), %xmm0 xorps %xmm1, %xmm1 cvtsi2sdq 104(%rsp), %xmm1 mulsd .LCPI4_2(%rip), %xmm1 addsd %xmm0, %xmm1 movsd %xmm1, 8(%rsp) # 8-byte Spill .LBB4_19: # %_Z15get_elapsedtimev.exit80 movsd 8(%rsp), %xmm0 # 8-byte Reload # xmm0 = mem[0],zero subsd 16(%rsp), %xmm0 # 8-byte Folded Reload movsd %xmm0, 8(%rsp) # 8-byte Spill movq stdout(%rip), %rcx movl $.L.str.2, %edi movl $22, %esi movl $1, %edx callq fwrite@PLT movq stdout(%rip), %rdi movl $.L.str.3, %esi movsd 8(%rsp), %xmm0 # 8-byte Reload # xmm0 = mem[0],zero movb $1, %al callq fprintf movq stdout(%rip), %rdi movq %r14, %xmm1 punpckldq .LCPI4_3(%rip), %xmm1 # xmm1 = xmm1[0],mem[0],xmm1[1],mem[1] subpd .LCPI4_4(%rip), %xmm1 movapd %xmm1, %xmm0 unpckhpd %xmm1, %xmm0 # xmm0 = xmm0[1],xmm1[1] addsd %xmm1, %xmm0 divsd 8(%rsp), %xmm0 # 8-byte Folded Reload mulsd .LCPI4_1(%rip), %xmm0 movl $.L.str.4, %esi movb $1, %al callq fprintf xorl %eax, %eax addq $200, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end4: .size main, .Lfunc_end4-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB5_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB5_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z15MulMatrixKernelPdS_S_i, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end5: .size __hip_module_ctor, .Lfunc_end5-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB6_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB6_2: retq .Lfunc_end6: .size __hip_module_dtor, .Lfunc_end6-__hip_module_dtor .cfi_endproc # -- End function .type _Z15MulMatrixKernelPdS_S_i,@object # @_Z15MulMatrixKernelPdS_S_i .section .rodata,"a",@progbits .globl _Z15MulMatrixKernelPdS_S_i .p2align 3, 0x0 _Z15MulMatrixKernelPdS_S_i: .quad _Z30__device_stub__MulMatrixKernelPdS_S_i .size _Z15MulMatrixKernelPdS_S_i, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "Matrix Multiplication\n Size: %dx%d\n" .size .L.str, 37 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "Matrice %dx%d\n\tTemps: %f s\n\tMFlops: %.2f\n" .size .L.str.1, 42 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "Performance results: \n" .size .L.str.2, 23 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz " Time: %lf s\n" .size .L.str.3, 15 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz " MFlops: %.2f\n" .size .L.str.4, 16 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z15MulMatrixKernelPdS_S_i" .size .L__unnamed_1, 27 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z30__device_stub__MulMatrixKernelPdS_S_i .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z15MulMatrixKernelPdS_S_i .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA device assembly to AMD device assembly.
code for sm_80 Function : _Z15MulMatrixKernelPdS_S_i .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R2, SR_CTAID.X ; /* 0x0000000000027919 */ /* 0x000e220000002500 */ /*0020*/ ISETP.NE.AND P0, PT, RZ, c[0x0][0xc], PT ; /* 0x00000300ff007a0c */ /* 0x000fe20003f05270 */ /*0030*/ HFMA2.MMA R17, -RZ, RZ, 0, 4.76837158203125e-07 ; /* 0x00000008ff117435 */ /* 0x000fe200000001ff */ /*0040*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*0050*/ S2R R9, SR_TID.X ; /* 0x0000000000097919 */ /* 0x000e220000002100 */ /*0060*/ CS2R R22, SRZ ; /* 0x0000000000167805 */ /* 0x000fc6000001ff00 */ /*0070*/ S2R R0, SR_CTAID.Y ; /* 0x0000000000007919 */ /* 0x000e680000002600 */ /*0080*/ S2R R3, SR_TID.Y ; /* 0x0000000000037919 */ /* 0x000e620000002200 */ /*0090*/ IMAD R2, R2, c[0x0][0x0], R9 ; /* 0x0000000002027a24 */ /* 0x001fe400078e0209 */ /*00a0*/ IMAD R5, R0, c[0x0][0x4], R3 ; /* 0x0000010000057a24 */ /* 0x002fc800078e0203 */ /*00b0*/ IMAD R16, R5, c[0x0][0x178], R2 ; /* 0x00005e0005107a24 */ /* 0x000fc800078e0202 */ /*00c0*/ IMAD.WIDE R16, R16, R17, c[0x0][0x170] ; /* 0x00005c0010107625 */ /* 0x000fe200078e0211 */ /*00d0*/ @!P0 BRA 0x730 ; /* 0x0000065000008947 */ /* 0x000fea0003800000 */ /*00e0*/ SHF.L.U32 R6, R3, 0x3, RZ ; /* 0x0000000303067819 */ /* 0x000fe200000006ff */ /*00f0*/ IMAD R4, R5, c[0x0][0x178], R9 ; /* 0x00005e0005047a24 */ /* 0x000fe200078e0209 */ /*0100*/ SHF.L.U32 R7, R9, 0x8, RZ ; /* 0x0000000809077819 */ /* 0x000fe200000006ff */ /*0110*/ IMAD R5, R3, c[0x0][0x178], R2 ; /* 0x00005e0003057a24 */ /* 0x000fe200078e0202 */ /*0120*/ LEA R6, R9, R6, 0x8 ; /* 0x0000000609067211 */ /* 0x000fe200078e40ff */ /*0130*/ CS2R R22, SRZ ; /* 0x0000000000167805 */ /* 0x000fe2000001ff00 */ /*0140*/ MOV R2, RZ ; /* 0x000000ff00027202 */ /* 0x000fe40000000f00 */ /*0150*/ HFMA2.MMA R28, -RZ, RZ, 0, 4.76837158203125e-07 ; /* 0x00000008ff1c7435 */ /* 0x000fd400000001ff */ /*0160*/ IMAD.WIDE.U32 R26, R4, R28, c[0x0][0x160] ; /* 0x00005800041a7625 */ /* 0x000fc800078e001c */ /*0170*/ IMAD.WIDE.U32 R28, R5, R28, c[0x0][0x168] ; /* 0x00005a00051c7625 */ /* 0x000fe400078e001c */ /*0180*/ LDG.E.64 R26, [R26.64] ; /* 0x000000041a1a7981 */ /* 0x000ea8000c1e1b00 */ /*0190*/ LDG.E.64 R28, [R28.64] ; /* 0x000000041c1c7981 */ /* 0x000ee2000c1e1b00 */ /*01a0*/ IADD3 R2, R2, 0x1, RZ ; /* 0x0000000102027810 */ /* 0x000fe20007ffe0ff */ /*01b0*/ IMAD R5, R0, c[0x0][0x178], R5 ; /* 0x00005e0000057a24 */ /* 0x000fe200078e0205 */ /*01c0*/ IADD3 R4, R4, c[0x0][0x0], RZ ; /* 0x0000000004047a10 */ /* 0x000fe40007ffe0ff */ /*01d0*/ ISETP.GE.U32.AND P0, PT, R2, c[0x0][0xc], PT ; /* 0x0000030002007a0c */ /* 0x000fe20003f06070 */ /*01e0*/ STS.64 [R6], R26 ; /* 0x0000001a06007388 */ /* 0x004fe80000000a00 */ /*01f0*/ STS.64 [R6+0x2000], R28 ; /* 0x0020001c06007388 */ /* 0x008fe80000000a00 */ /*0200*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0210*/ LDS.64 R18, [R3.X8] ; /* 0x0000000003127984 */ /* 0x000fe80000008a00 */ /*0220*/ LDS.128 R12, [R7+0x2000] ; /* 0x00200000070c7984 */ /* 0x000e280000000c00 */ /*0230*/ LDS.64 R24, [R3.X8+0x100] ; /* 0x0001000003187984 */ /* 0x000e680000008a00 */ /*0240*/ LDS.64 R20, [R3.X8+0x200] ; /* 0x0002000003147984 */ /* 0x000fe80000008a00 */ /*0250*/ LDS.128 R8, [R7+0x2010] ; /* 0x0020100007087984 */ /* 0x000ea20000000c00 */ /*0260*/ DFMA R12, R12, R18, R22 ; /* 0x000000120c0c722b */ /* 0x0010460000000016 */ /*0270*/ LDS.64 R22, [R3.X8+0x300] ; /* 0x0003000003167984 */ /* 0x001e260000008a00 */ /*0280*/ DFMA R24, R24, R14, R12 ; /* 0x0000000e1818722b */ /* 0x0022a2000000000c */ /*0290*/ LDS.64 R18, [R3.X8+0x400] ; /* 0x0004000003127984 */ /* 0x000fe80000008a00 */ /*02a0*/ LDS.128 R12, [R7+0x2020] ; /* 0x00202000070c7984 */ /* 0x002e620000000c00 */ /*02b0*/ DFMA R24, R8, R20, R24 ; /* 0x000000140818722b */ /* 0x0044060000000018 */ /*02c0*/ LDS.64 R20, [R3.X8+0x500] ; /* 0x0005000003147984 */ /* 0x004ea60000008a00 */ /*02d0*/ DFMA R24, R22, R10, R24 ; /* 0x0000000a1618722b */ /* 0x0010640000000018 */ /*02e0*/ LDS.64 R22, [R3.X8+0x600] ; /* 0x0006000003167984 */ /* 0x001fe80000008a00 */ /*02f0*/ LDS.128 R8, [R7+0x2030] ; /* 0x0020300007087984 */ /* 0x000e220000000c00 */ /*0300*/ DFMA R24, R12, R18, R24 ; /* 0x000000120c18722b */ /* 0x0022860000000018 */ /*0310*/ LDS.64 R18, [R3.X8+0x700] ; /* 0x0007000003127984 */ /* 0x002e660000008a00 */ /*0320*/ DFMA R24, R20, R14, R24 ; /* 0x0000000e1418722b */ /* 0x0044240000000018 */ /*0330*/ LDS.64 R20, [R3.X8+0x800] ; /* 0x0008000003147984 */ /* 0x004fe80000008a00 */ /*0340*/ LDS.128 R12, [R7+0x2040] ; /* 0x00204000070c7984 */ /* 0x000ea20000000c00 */ /*0350*/ DFMA R24, R8, R22, R24 ; /* 0x000000160818722b */ /* 0x0010460000000018 */ /*0360*/ LDS.64 R22, [R3.X8+0x900] ; /* 0x0009000003167984 */ /* 0x001e260000008a00 */ /*0370*/ DFMA R24, R18, R10, R24 ; /* 0x0000000a1218722b */ /* 0x0022a40000000018 */ /*0380*/ LDS.64 R18, [R3.X8+0xa00] ; /* 0x000a000003127984 */ /* 0x002fe80000008a00 */ /*0390*/ LDS.128 R8, [R7+0x2050] ; /* 0x0020500007087984 */ /* 0x000e620000000c00 */ /*03a0*/ DFMA R24, R12, R20, R24 ; /* 0x000000140c18722b */ /* 0x0044060000000018 */ /*03b0*/ LDS.64 R20, [R3.X8+0xb00] ; /* 0x000b000003147984 */ /* 0x004ea60000008a00 */ /*03c0*/ DFMA R24, R22, R14, R24 ; /* 0x0000000e1618722b */ /* 0x0010640000000018 */ /*03d0*/ LDS.64 R22, [R3.X8+0xc00] ; /* 0x000c000003167984 */ /* 0x001fe80000008a00 */ /*03e0*/ LDS.128 R12, [R7+0x2060] ; /* 0x00206000070c7984 */ /* 0x000e220000000c00 */ /*03f0*/ DFMA R24, R8, R18, R24 ; /* 0x000000120818722b */ /* 0x0022860000000018 */ /*0400*/ LDS.64 R18, [R3.X8+0xd00] ; /* 0x000d000003127984 */ /* 0x002e660000008a00 */ /*0410*/ DFMA R24, R20, R10, R24 ; /* 0x0000000a1418722b */ /* 0x0044240000000018 */ /*0420*/ LDS.64 R20, [R3.X8+0xe00] ; /* 0x000e000003147984 */ /* 0x004fe80000008a00 */ /*0430*/ LDS.128 R8, [R7+0x2070] ; /* 0x0020700007087984 */ /* 0x000ea20000000c00 */ /*0440*/ DFMA R24, R12, R22, R24 ; /* 0x000000160c18722b */ /* 0x0010460000000018 */ /*0450*/ LDS.64 R22, [R3.X8+0xf00] ; /* 0x000f000003167984 */ /* 0x001e260000008a00 */ /*0460*/ DFMA R24, R18, R14, R24 ; /* 0x0000000e1218722b */ /* 0x0022a40000000018 */ /*0470*/ LDS.64 R18, [R3.X8+0x1000] ; /* 0x0010000003127984 */ /* 0x002fe80000008a00 */ /*0480*/ LDS.128 R12, [R7+0x2080] ; /* 0x00208000070c7984 */ /* 0x000e620000000c00 */ /*0490*/ DFMA R24, R8, R20, R24 ; /* 0x000000140818722b */ /* 0x0044060000000018 */ /*04a0*/ LDS.64 R20, [R3.X8+0x1100] ; /* 0x0011000003147984 */ /* 0x004ea60000008a00 */ /*04b0*/ DFMA R24, R22, R10, R24 ; /* 0x0000000a1618722b */ /* 0x0010640000000018 */ /*04c0*/ LDS.64 R22, [R3.X8+0x1200] ; /* 0x0012000003167984 */ /* 0x001fe80000008a00 */ /*04d0*/ LDS.128 R8, [R7+0x2090] ; /* 0x0020900007087984 */ /* 0x000e220000000c00 */ /*04e0*/ DFMA R24, R12, R18, R24 ; /* 0x000000120c18722b */ /* 0x0022860000000018 */ /*04f0*/ LDS.64 R18, [R3.X8+0x1300] ; /* 0x0013000003127984 */ /* 0x002e660000008a00 */ /*0500*/ DFMA R24, R20, R14, R24 ; /* 0x0000000e1418722b */ /* 0x0044240000000018 */ /*0510*/ LDS.64 R20, [R3.X8+0x1400] ; /* 0x0014000003147984 */ /* 0x004fe80000008a00 */ /*0520*/ LDS.128 R12, [R7+0x20a0] ; /* 0x0020a000070c7984 */ /* 0x000ea20000000c00 */ /*0530*/ DFMA R24, R8, R22, R24 ; /* 0x000000160818722b */ /* 0x0010460000000018 */ /*0540*/ LDS.64 R22, [R3.X8+0x1500] ; /* 0x0015000003167984 */ /* 0x001e260000008a00 */ /*0550*/ DFMA R24, R18, R10, R24 ; /* 0x0000000a1218722b */ /* 0x0022a40000000018 */ /*0560*/ LDS.64 R18, [R3.X8+0x1600] ; /* 0x0016000003127984 */ /* 0x002fe80000008a00 */ /*0570*/ LDS.128 R8, [R7+0x20b0] ; /* 0x0020b00007087984 */ /* 0x000e620000000c00 */ /*0580*/ DFMA R24, R12, R20, R24 ; /* 0x000000140c18722b */ /* 0x0044060000000018 */ /*0590*/ LDS.64 R20, [R3.X8+0x1700] ; /* 0x0017000003147984 */ /* 0x004ea60000008a00 */ /*05a0*/ DFMA R24, R22, R14, R24 ; /* 0x0000000e1618722b */ /* 0x0010640000000018 */ /*05b0*/ LDS.64 R22, [R3.X8+0x1800] ; /* 0x0018000003167984 */ /* 0x001fe80000008a00 */ /*05c0*/ LDS.128 R12, [R7+0x20c0] ; /* 0x0020c000070c7984 */ /* 0x000e220000000c00 */ /*05d0*/ DFMA R24, R8, R18, R24 ; /* 0x000000120818722b */ /* 0x0022860000000018 */ /*05e0*/ LDS.64 R18, [R3.X8+0x1900] ; /* 0x0019000003127984 */ /* 0x002e660000008a00 */ /*05f0*/ DFMA R24, R20, R10, R24 ; /* 0x0000000a1418722b */ /* 0x0044240000000018 */ /*0600*/ LDS.64 R20, [R3.X8+0x1a00] ; /* 0x001a000003147984 */ /* 0x004fe80000008a00 */ /*0610*/ LDS.128 R8, [R7+0x20d0] ; /* 0x0020d00007087984 */ /* 0x000ea20000000c00 */ /*0620*/ DFMA R24, R12, R22, R24 ; /* 0x000000160c18722b */ /* 0x0010460000000018 */ /*0630*/ LDS.64 R22, [R3.X8+0x1b00] ; /* 0x001b000003167984 */ /* 0x001e260000008a00 */ /*0640*/ DFMA R24, R18, R14, R24 ; /* 0x0000000e1218722b */ /* 0x0022a40000000018 */ /*0650*/ LDS.64 R18, [R3.X8+0x1c00] ; /* 0x001c000003127984 */ /* 0x002fe80000008a00 */ /*0660*/ LDS.128 R12, [R7+0x20e0] ; /* 0x0020e000070c7984 */ /* 0x000e620000000c00 */ /*0670*/ DFMA R24, R8, R20, R24 ; /* 0x000000140818722b */ /* 0x0044060000000018 */ /*0680*/ LDS.64 R20, [R3.X8+0x1d00] ; /* 0x001d000003147984 */ /* 0x004ea60000008a00 */ /*0690*/ DFMA R24, R22, R10, R24 ; /* 0x0000000a1618722b */ /* 0x0010640000000018 */ /*06a0*/ LDS.64 R22, [R3.X8+0x1e00] ; /* 0x001e000003167984 */ /* 0x001fe80000008a00 */ /*06b0*/ LDS.128 R8, [R7+0x20f0] ; /* 0x0020f00007087984 */ /* 0x000e220000000c00 */ /*06c0*/ DFMA R18, R12, R18, R24 ; /* 0x000000120c12722b */ /* 0x0022860000000018 */ /*06d0*/ LDS.64 R12, [R3.X8+0x1f00] ; /* 0x001f0000030c7984 */ /* 0x002e660000008a00 */ /*06e0*/ DFMA R14, R20, R14, R18 ; /* 0x0000000e140e722b */ /* 0x004e0c0000000012 */ /*06f0*/ DFMA R22, R8, R22, R14 ; /* 0x000000160816722b */ /* 0x001e4c000000000e */ /*0700*/ DFMA R22, R12, R10, R22 ; /* 0x0000000a0c16722b */ /* 0x0020620000000016 */ /*0710*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0720*/ @!P0 BRA 0x150 ; /* 0xfffffa2000008947 */ /* 0x003fea000383ffff */ /*0730*/ STG.E.64 [R16.64], R22 ; /* 0x0000001610007986 */ /* 0x000fe2000c101b04 */ /*0740*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0750*/ BRA 0x750; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0760*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0770*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0780*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0790*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*07a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*07b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*07c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*07d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*07e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*07f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z15MulMatrixKernelPdS_S_i .globl _Z15MulMatrixKernelPdS_S_i .p2align 8 .type _Z15MulMatrixKernelPdS_S_i,@function _Z15MulMatrixKernelPdS_S_i: s_clause 0x2 s_load_b32 s4, s[0:1], 0x2c s_load_b32 s3, s[0:1], 0x20 s_load_b32 s2, s[0:1], 0x18 v_and_b32_e32 v4, 0x3ff, v0 v_bfe_u32 v3, v0, 10, 10 s_mov_b32 s9, 0 s_waitcnt lgkmcnt(0) s_and_b32 s8, s4, 0xffff s_lshr_b32 s4, s4, 16 v_mad_u64_u32 v[0:1], null, s14, s8, v[4:5] v_mad_u64_u32 v[1:2], null, s15, s4, v[3:4] s_cmp_eq_u32 s3, 0 s_cbranch_scc1 .LBB0_5 s_load_b128 s[4:7], s[0:1], 0x0 v_dual_mov_b32 v10, 0 :: v_dual_lshlrev_b32 v5, 8, v4 v_lshlrev_b32_e32 v2, 3, v3 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_nc_u32_e32 v8, 0x2000, v5 v_mad_u64_u32 v[6:7], null, v1, s2, v[4:5] s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_4) v_dual_mov_b32 v4, 0 :: v_dual_add_nc_u32 v7, v5, v2 v_mov_b32_e32 v5, 0 v_add_nc_u32_e32 v9, v8, v2 s_set_inst_prefetch_distance 0x1 .p2align 6 .LBB0_2: s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_3) | instid1(VALU_DEP_3) v_mad_u64_u32 v[11:12], null, s9, s15, v[3:4] v_mad_u64_u32 v[12:13], null, s9, s8, v[6:7] v_mov_b32_e32 v13, v10 s_mov_b32 s10, 0 v_mad_u64_u32 v[14:15], null, v11, s2, v[0:1] v_mov_b32_e32 v15, v10 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_lshlrev_b64 v[11:12], 3, v[12:13] v_lshlrev_b64 v[13:14], 3, v[14:15] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_co_u32 v11, vcc_lo, s4, v11 v_add_co_ci_u32_e32 v12, vcc_lo, s5, v12, vcc_lo s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_u32 v13, vcc_lo, s6, v13 v_add_co_ci_u32_e32 v14, vcc_lo, s7, v14, vcc_lo global_load_b64 v[15:16], v[11:12], off global_load_b64 v[12:13], v[13:14], off v_mov_b32_e32 v11, v2 s_waitcnt vmcnt(1) ds_store_b64 v7, v[15:16] s_waitcnt vmcnt(0) ds_store_b64 v9, v[12:13] s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv .LBB0_3: v_add_nc_u32_e32 v14, s10, v8 s_add_i32 s10, s10, 8 ds_load_b64 v[12:13], v11 ds_load_b64 v[14:15], v14 v_add_nc_u32_e32 v11, 0x100, v11 s_cmpk_eq_i32 s10, 0x100 s_waitcnt lgkmcnt(0) v_fma_f64 v[4:5], v[12:13], v[14:15], v[4:5] s_cbranch_scc0 .LBB0_3 s_add_i32 s9, s9, 1 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_eq_u32 s9, s3 s_barrier buffer_gl0_inv s_cbranch_scc0 .LBB0_2 s_branch .LBB0_6 .LBB0_5: v_mov_b32_e32 v4, 0 v_mov_b32_e32 v5, 0 .LBB0_6: s_set_inst_prefetch_distance 0x2 s_load_b64 s[0:1], s[0:1], 0x10 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_mad_u64_u32 v[2:3], null, v1, s2, v[0:1] v_ashrrev_i32_e32 v3, 31, v2 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[0:1], 3, v[2:3] s_waitcnt lgkmcnt(0) v_add_co_u32 v0, vcc_lo, s0, v0 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo global_store_b64 v[0:1], v[4:5], off s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z15MulMatrixKernelPdS_S_i .amdhsa_group_segment_fixed_size 16384 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 288 .amdhsa_user_sgpr_count 14 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 1 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 1 .amdhsa_next_free_vgpr 17 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z15MulMatrixKernelPdS_S_i, .Lfunc_end0-_Z15MulMatrixKernelPdS_S_i .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 8 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .offset: 24 .size: 4 .value_kind: by_value - .offset: 32 .size: 4 .value_kind: hidden_block_count_x - .offset: 36 .size: 4 .value_kind: hidden_block_count_y - .offset: 40 .size: 4 .value_kind: hidden_block_count_z - .offset: 44 .size: 2 .value_kind: hidden_group_size_x - .offset: 46 .size: 2 .value_kind: hidden_group_size_y - .offset: 48 .size: 2 .value_kind: hidden_group_size_z - .offset: 50 .size: 2 .value_kind: hidden_remainder_x - .offset: 52 .size: 2 .value_kind: hidden_remainder_y - .offset: 54 .size: 2 .value_kind: hidden_remainder_z - .offset: 72 .size: 8 .value_kind: hidden_global_offset_x - .offset: 80 .size: 8 .value_kind: hidden_global_offset_y - .offset: 88 .size: 8 .value_kind: hidden_global_offset_z - .offset: 96 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 16384 .kernarg_segment_align: 8 .kernarg_segment_size: 288 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z15MulMatrixKernelPdS_S_i .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z15MulMatrixKernelPdS_S_i.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 17 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_001155b8_00000000-6_dgemm_8.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2063: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2063: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z15get_elapsedtimev .type _Z15get_elapsedtimev, @function _Z15get_elapsedtimev: .LFB2057: .cfi_startproc endbr64 subq $40, %rsp .cfi_def_cfa_offset 48 movq %fs:40, %rax movq %rax, 24(%rsp) xorl %eax, %eax movq %rsp, %rsi movl $4, %edi call clock_gettime@PLT pxor %xmm0, %xmm0 testl %eax, %eax jne .L3 pxor %xmm0, %xmm0 cvtsi2sdq 8(%rsp), %xmm0 mulsd .LC1(%rip), %xmm0 pxor %xmm1, %xmm1 cvtsi2sdq (%rsp), %xmm1 addsd %xmm1, %xmm0 .L3: movq 24(%rsp), %rax subq %fs:40, %rax jne .L8 addq $40, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L8: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size _Z15get_elapsedtimev, .-_Z15get_elapsedtimev .globl _Z4initPPdS0_S0_i .type _Z4initPPdS0_S0_i, @function _Z4initPPdS0_S0_i: .LFB2058: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $8, %rsp .cfi_def_cfa_offset 64 movq %rdi, %r15 movq %rsi, %r13 movq %rdx, %r12 movl %ecx, %ebx movl $2019, %edi call srand@PLT testl %ebx, %ebx jle .L9 movq %r15, %rbp movslq %ebx, %rbx leaq 0(,%rbx,8), %r14 addq %r14, %r15 .L11: movl $0, %ebx .L12: call rand@PLT movq 0(%rbp), %rdx pxor %xmm0, %xmm0 cvtsi2sdl %eax, %xmm0 movsd %xmm0, (%rdx,%rbx) call rand@PLT movq 0(%r13), %rdx pxor %xmm0, %xmm0 cvtsi2sdl %eax, %xmm0 movsd %xmm0, (%rdx,%rbx) movq (%r12), %rax movq $0x000000000, (%rax,%rbx) addq $8, %rbx cmpq %r14, %rbx jne .L12 addq $8, %rbp addq $8, %r13 addq $8, %r12 cmpq %r15, %rbp jne .L11 .L9: addq $8, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2058: .size _Z4initPPdS0_S0_i, .-_Z4initPPdS0_S0_i .globl _Z4multPPdS0_S0_i .type _Z4multPPdS0_S0_i, @function _Z4multPPdS0_S0_i: .LFB2059: .cfi_startproc endbr64 movq %rdi, %r10 movq %rsi, %r8 movq %rdx, %r11 movl %ecx, %eax testl %ecx, %ecx jle .L15 cltq leaq 0(,%rax,8), %rsi movl $0, %r9d .L17: movl $0, %ecx .L20: movq (%r10,%r9), %rdi movl $0, %eax pxor %xmm1, %xmm1 .L18: movq (%r8,%rax), %rdx movsd (%rdx,%rcx), %xmm0 mulsd (%rdi,%rax), %xmm0 addsd %xmm0, %xmm1 addq $8, %rax cmpq %rsi, %rax jne .L18 movq (%r11,%r9), %rax movsd %xmm1, (%rax,%rcx) addq $8, %rcx cmpq %rsi, %rcx jne .L20 addq $8, %r9 cmpq %rsi, %r9 jne .L17 .L15: ret .cfi_endproc .LFE2059: .size _Z4multPPdS0_S0_i, .-_Z4multPPdS0_S0_i .globl _Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i .type _Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i, @function _Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i: .LFB2085: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 24(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movl %ecx, 4(%rsp) movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax leaq 24(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) leaq 4(%rsp), %rax movq %rax, 120(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 56(%rsp) movl $1, 60(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) leaq 40(%rsp), %rcx leaq 32(%rsp), %rdx leaq 60(%rsp), %rsi leaq 48(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L26 .L22: movq 136(%rsp), %rax subq %fs:40, %rax jne .L27 addq $152, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L26: .cfi_restore_state pushq 40(%rsp) .cfi_def_cfa_offset 168 pushq 40(%rsp) .cfi_def_cfa_offset 176 leaq 112(%rsp), %r9 movq 76(%rsp), %rcx movl 84(%rsp), %r8d movq 64(%rsp), %rsi movl 72(%rsp), %edx leaq _Z15MulMatrixKernelPdS_S_i(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L22 .L27: call __stack_chk_fail@PLT .cfi_endproc .LFE2085: .size _Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i, .-_Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i .globl _Z15MulMatrixKernelPdS_S_i .type _Z15MulMatrixKernelPdS_S_i, @function _Z15MulMatrixKernelPdS_S_i: .LFB2086: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2086: .size _Z15MulMatrixKernelPdS_S_i, .-_Z15MulMatrixKernelPdS_S_i .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC2: .string "Matrix Multiplication\n Size: %dx%d\n" .align 8 .LC6: .string "Matrice %dx%d\n\tTemps: %f s\n\tMFlops: %.2f\n" .section .rodata.str1.1,"aMS",@progbits,1 .LC7: .string "Performance results: \n" .LC8: .string " Time: %lf s\n" .LC9: .string " MFlops: %.2f\n" .text .globl main .type main, @function main: .LFB2060: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $120, %rsp .cfi_def_cfa_offset 176 movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax cmpl $1, %edi jg .L45 movl $2048, %r8d movl $2048, %ecx leaq .LC2(%rip), %rdx movl $2, %esi movq stdout(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movl $33554432, %edi call malloc@PLT movq %rax, 8(%rsp) movl $33554432, %edi call malloc@PLT movq %rax, 16(%rsp) movl $33554432, %edi call malloc@PLT movq %rax, %r15 movl $16384, %edi call malloc@PLT movq %rax, %r12 movl $16384, %edi call malloc@PLT movq %rax, %rbp movl $16384, %edi call malloc@PLT movq %rax, %rdx movl $2048, %r13d movl $16384, %ebx movl $33554432, %r14d .L40: movq 8(%rsp), %r8 movq 16(%rsp), %rdi movq %r15, %rsi movl $0, %ecx .L33: movq %r8, (%r12,%rcx) movq %rdi, 0(%rbp,%rcx) movq %rsi, (%rdx,%rcx) addq %rbx, %r8 addq %rbx, %rdi addq %rbx, %rsi addq $8, %rcx cmpq %rbx, %rcx jne .L33 .L32: movl %r13d, %ecx movq %rbp, %rsi movq %r12, %rdi call _Z4initPPdS0_S0_i leaq 40(%rsp), %rdi call cudaEventCreate@PLT leaq 48(%rsp), %rdi call cudaEventCreate@PLT leaq 56(%rsp), %rdi movq %r14, %rsi call cudaMalloc@PLT leaq 64(%rsp), %rdi movq %r14, %rsi call cudaMalloc@PLT leaq 72(%rsp), %rdi movq %r14, %rsi call cudaMalloc@PLT movl $1, %ecx movq %r14, %rdx movq 8(%rsp), %rsi movq 56(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movq %r14, %rdx movq 16(%rsp), %rsi movq 64(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movq %r14, %rdx movq %r15, %rsi movq 72(%rsp), %rdi call cudaMemcpy@PLT leal 31(%r13), %eax testl %r13d, %r13d cmovns %r13d, %eax sarl $5, %eax movl %r13d, %edx andl $31, %edx cmpl $1, %edx sbbl $-1, %eax movl %eax, 80(%rsp) movl %eax, 84(%rsp) movl $1, 88(%rsp) movl $32, 92(%rsp) movl $32, 96(%rsp) movl $1, 100(%rsp) movl $0, %esi movq 40(%rsp), %rdi call cudaEventRecord@PLT movl 100(%rsp), %ecx movl $0, %r9d movl $0, %r8d movq 92(%rsp), %rdx movq 80(%rsp), %rdi movl 88(%rsp), %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L46 .L35: movl $0, %esi movq 48(%rsp), %rdi call cudaEventRecord@PLT movl $2, %ecx movq %r14, %rdx movq 72(%rsp), %rsi movq %r15, %rdi call cudaMemcpy@PLT movq 48(%rsp), %rdi call cudaEventSynchronize@PLT movl %r13d, %ebx imull %r13d, %ebx imull %r13d, %ebx movslq %ebx, %rbx movl $0x00000000, 36(%rsp) leaq 36(%rsp), %rdi movq 48(%rsp), %rdx movq 40(%rsp), %rsi call cudaEventElapsedTime@PLT movss 36(%rsp), %xmm0 divss .LC4(%rip), %xmm0 testq %rbx, %rbx js .L36 pxor %xmm1, %xmm1 cvtsi2ssq %rbx, %xmm1 .L37: divss %xmm0, %xmm1 cvtss2sd %xmm1, %xmm1 cvtss2sd %xmm0, %xmm0 mulsd .LC5(%rip), %xmm1 movl %r13d, %ecx movl %r13d, %edx leaq .LC6(%rip), %rsi movl $2, %edi movl $2, %eax call __printf_chk@PLT call _Z15get_elapsedtimev movsd %xmm0, 8(%rsp) call _Z15get_elapsedtimev movapd %xmm0, %xmm2 subsd 8(%rsp), %xmm2 movsd %xmm2, 8(%rsp) leaq .LC7(%rip), %rdx movl $2, %esi movq stdout(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movsd 8(%rsp), %xmm0 leaq .LC8(%rip), %rdx movl $2, %esi movq stdout(%rip), %rdi movl $1, %eax call __fprintf_chk@PLT testq %rbx, %rbx js .L38 pxor %xmm0, %xmm0 cvtsi2sdq %rbx, %xmm0 .L39: divsd 8(%rsp), %xmm0 mulsd .LC5(%rip), %xmm0 leaq .LC9(%rip), %rdx movl $2, %esi movq stdout(%rip), %rdi movl $1, %eax call __fprintf_chk@PLT movq 104(%rsp), %rax subq %fs:40, %rax jne .L47 movl $0, %eax addq $120, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L45: .cfi_restore_state movq 8(%rsi), %rdi movl $10, %edx movl $0, %esi call __isoc23_strtol@PLT movq %rax, %rbx movl %eax, %r13d movl %eax, %r8d movl %eax, %ecx leaq .LC2(%rip), %rdx movl $2, %esi movq stdout(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movq %rbx, 24(%rsp) movslq 24(%rsp), %rbx movq %rbx, %r14 imulq %rbx, %r14 salq $3, %r14 movq %r14, %rdi call malloc@PLT movq %rax, 8(%rsp) movq %r14, %rdi call malloc@PLT movq %rax, 16(%rsp) movq %r14, %rdi call malloc@PLT movq %rax, %r15 salq $3, %rbx movq %rbx, %rdi call malloc@PLT movq %rax, %r12 movq %rbx, %rdi call malloc@PLT movq %rax, %rbp movq %rbx, %rdi call malloc@PLT movq %rax, %rdx cmpl $0, 24(%rsp) jg .L40 jmp .L32 .L46: movl %r13d, %ecx movq 72(%rsp), %rdx movq 64(%rsp), %rsi movq 56(%rsp), %rdi call _Z40__device_stub__Z15MulMatrixKernelPdS_S_iPdS_S_i jmp .L35 .L36: movq %rbx, %rax shrq %rax movq %rbx, %rdx andl $1, %edx orq %rdx, %rax pxor %xmm1, %xmm1 cvtsi2ssq %rax, %xmm1 addss %xmm1, %xmm1 jmp .L37 .L38: movq %rbx, %rax shrq %rax andl $1, %ebx orq %rbx, %rax pxor %xmm0, %xmm0 cvtsi2sdq %rax, %xmm0 addsd %xmm0, %xmm0 jmp .L39 .L47: call __stack_chk_fail@PLT .cfi_endproc .LFE2060: .size main, .-main .section .rodata.str1.1 .LC10: .string "_Z15MulMatrixKernelPdS_S_i" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2088: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC10(%rip), %rdx movq %rdx, %rcx leaq _Z15MulMatrixKernelPdS_S_i(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2088: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .section .rodata.cst8,"aM",@progbits,8 .align 8 .LC1: .long -400107883 .long 1041313291 .section .rodata.cst4,"aM",@progbits,4 .align 4 .LC4: .long 1148846080 .section .rodata.cst8 .align 8 .LC5: .long -1598689907 .long 1051772663 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "dgemm_8.hip" .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 # -- Begin function _Z15get_elapsedtimev .LCPI0_0: .quad 0x3e112e0be826d695 # double 1.0000000000000001E-9 .text .globl _Z15get_elapsedtimev .p2align 4, 0x90 .type _Z15get_elapsedtimev,@function _Z15get_elapsedtimev: # @_Z15get_elapsedtimev .cfi_startproc # %bb.0: subq $24, %rsp .cfi_def_cfa_offset 32 leaq 8(%rsp), %rsi movl $4, %edi callq clock_gettime xorps %xmm0, %xmm0 testl %eax, %eax je .LBB0_1 # %bb.2: addq $24, %rsp .cfi_def_cfa_offset 8 retq .LBB0_1: .cfi_def_cfa_offset 32 cvtsi2sdq 8(%rsp), %xmm1 xorps %xmm0, %xmm0 cvtsi2sdq 16(%rsp), %xmm0 mulsd .LCPI0_0(%rip), %xmm0 addsd %xmm1, %xmm0 addq $24, %rsp .cfi_def_cfa_offset 8 retq .Lfunc_end0: .size _Z15get_elapsedtimev, .Lfunc_end0-_Z15get_elapsedtimev .cfi_endproc # -- End function .globl _Z4initPPdS0_S0_i # -- Begin function _Z4initPPdS0_S0_i .p2align 4, 0x90 .type _Z4initPPdS0_S0_i,@function _Z4initPPdS0_S0_i: # @_Z4initPPdS0_S0_i .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 pushq %rax .cfi_def_cfa_offset 64 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl %ecx, %ebp movq %rdx, %rbx movq %rsi, %r14 movq %rdi, %r15 movl $2019, %edi # imm = 0x7E3 callq srand testl %ebp, %ebp jle .LBB1_5 # %bb.1: # %.preheader.lr.ph movl %ebp, %r12d xorl %r13d, %r13d .p2align 4, 0x90 .LBB1_2: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB1_3 Depth 2 xorl %ebp, %ebp .p2align 4, 0x90 .LBB1_3: # Parent Loop BB1_2 Depth=1 # => This Inner Loop Header: Depth=2 callq rand xorps %xmm0, %xmm0 cvtsi2sd %eax, %xmm0 movq (%r15,%r13,8), %rax movsd %xmm0, (%rax,%rbp,8) callq rand xorps %xmm0, %xmm0 cvtsi2sd %eax, %xmm0 movq (%r14,%r13,8), %rax movsd %xmm0, (%rax,%rbp,8) movq (%rbx,%r13,8), %rax movq $0, (%rax,%rbp,8) incq %rbp cmpq %rbp, %r12 jne .LBB1_3 # %bb.4: # %._crit_edge # in Loop: Header=BB1_2 Depth=1 incq %r13 cmpq %r12, %r13 jne .LBB1_2 .LBB1_5: # %._crit_edge18 addq $8, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end1: .size _Z4initPPdS0_S0_i, .Lfunc_end1-_Z4initPPdS0_S0_i .cfi_endproc # -- End function .globl _Z4multPPdS0_S0_i # -- Begin function _Z4multPPdS0_S0_i .p2align 4, 0x90 .type _Z4multPPdS0_S0_i,@function _Z4multPPdS0_S0_i: # @_Z4multPPdS0_S0_i .cfi_startproc # %bb.0: testl %ecx, %ecx jle .LBB2_8 # %bb.1: # %.preheader23.lr.ph pushq %rbx .cfi_def_cfa_offset 16 .cfi_offset %rbx, -16 movl %ecx, %eax xorl %ecx, %ecx .p2align 4, 0x90 .LBB2_2: # %.preheader23 # =>This Loop Header: Depth=1 # Child Loop BB2_3 Depth 2 # Child Loop BB2_4 Depth 3 movq (%rdx,%rcx,8), %r8 movq (%rdi,%rcx,8), %r9 xorl %r10d, %r10d .p2align 4, 0x90 .LBB2_3: # %.preheader # Parent Loop BB2_2 Depth=1 # => This Loop Header: Depth=2 # Child Loop BB2_4 Depth 3 xorpd %xmm0, %xmm0 xorl %r11d, %r11d .p2align 4, 0x90 .LBB2_4: # Parent Loop BB2_2 Depth=1 # Parent Loop BB2_3 Depth=2 # => This Inner Loop Header: Depth=3 movsd (%r9,%r11,8), %xmm1 # xmm1 = mem[0],zero movq (%rsi,%r11,8), %rbx mulsd (%rbx,%r10,8), %xmm1 addsd %xmm1, %xmm0 incq %r11 cmpq %r11, %rax jne .LBB2_4 # %bb.5: # %._crit_edge # in Loop: Header=BB2_3 Depth=2 movsd %xmm0, (%r8,%r10,8) incq %r10 cmpq %rax, %r10 jne .LBB2_3 # %bb.6: # %._crit_edge27 # in Loop: Header=BB2_2 Depth=1 incq %rcx cmpq %rax, %rcx jne .LBB2_2 # %bb.7: popq %rbx .cfi_def_cfa_offset 8 .cfi_restore %rbx .LBB2_8: # %._crit_edge29 retq .Lfunc_end2: .size _Z4multPPdS0_S0_i, .Lfunc_end2-_Z4multPPdS0_S0_i .cfi_endproc # -- End function .globl _Z30__device_stub__MulMatrixKernelPdS_S_i # -- Begin function _Z30__device_stub__MulMatrixKernelPdS_S_i .p2align 4, 0x90 .type _Z30__device_stub__MulMatrixKernelPdS_S_i,@function _Z30__device_stub__MulMatrixKernelPdS_S_i: # @_Z30__device_stub__MulMatrixKernelPdS_S_i .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movq %rdx, 56(%rsp) movl %ecx, 4(%rsp) leaq 72(%rsp), %rax movq %rax, 80(%rsp) leaq 64(%rsp), %rax movq %rax, 88(%rsp) leaq 56(%rsp), %rax movq %rax, 96(%rsp) leaq 4(%rsp), %rax movq %rax, 104(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 80(%rsp), %r9 movl $_Z15MulMatrixKernelPdS_S_i, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $136, %rsp .cfi_adjust_cfa_offset -136 retq .Lfunc_end3: .size _Z30__device_stub__MulMatrixKernelPdS_S_i, .Lfunc_end3-_Z30__device_stub__MulMatrixKernelPdS_S_i .cfi_endproc # -- End function .section .rodata.cst4,"aM",@progbits,4 .p2align 2, 0x0 # -- Begin function main .LCPI4_0: .long 0x447a0000 # float 1000 .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 .LCPI4_1: .quad 0x3eb0c6f7a0b5ed8d # double 9.9999999999999995E-7 .LCPI4_2: .quad 0x3e112e0be826d695 # double 1.0000000000000001E-9 .section .rodata.cst16,"aM",@progbits,16 .p2align 4, 0x0 .LCPI4_3: .long 1127219200 # 0x43300000 .long 1160773632 # 0x45300000 .long 0 # 0x0 .long 0 # 0x0 .LCPI4_4: .quad 0x4330000000000000 # double 4503599627370496 .quad 0x4530000000000000 # double 1.9342813113834067E+25 .text .globl main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 subq $200, %rsp .cfi_def_cfa_offset 256 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl $2048, %r14d # imm = 0x800 cmpl $2, %edi jl .LBB4_2 # %bb.1: movq 8(%rsi), %rdi xorl %esi, %esi movl $10, %edx callq __isoc23_strtol movq %rax, %r14 .LBB4_2: movq stdout(%rip), %rdi movl $.L.str, %esi movl %r14d, %edx movl %r14d, %ecx xorl %eax, %eax callq fprintf movslq %r14d, %rbx leaq (,%rbx,8), %r12 movq %r12, %r15 imulq %rbx, %r15 movq %r15, %rdi callq malloc movq %rax, 56(%rsp) # 8-byte Spill movq %r15, %rdi callq malloc movq %rax, 64(%rsp) # 8-byte Spill movq %r15, 136(%rsp) # 8-byte Spill movq %r15, %rdi callq malloc movq %rax, 16(%rsp) # 8-byte Spill movq %r12, %rdi callq malloc movq %rax, %rbp movq %r12, %rdi callq malloc movq %rax, %r15 movq %r12, %rdi callq malloc movq %rax, %r12 movl %r14d, %r13d testl %ebx, %ebx jle .LBB4_5 # %bb.3: # %.lr.ph.preheader leaq (,%r13,8), %rax xorl %ecx, %ecx movq 56(%rsp), %rdx # 8-byte Reload movq 64(%rsp), %rsi # 8-byte Reload movq 16(%rsp), %rdi # 8-byte Reload .p2align 4, 0x90 .LBB4_4: # %.lr.ph # =>This Inner Loop Header: Depth=1 movq %rdx, (%rbp,%rcx,8) movq %rsi, (%r15,%rcx,8) movq %rdi, (%r12,%rcx,8) incq %rcx addq %rax, %rdi addq %rax, %rsi addq %rax, %rdx cmpq %rcx, %r13 jne .LBB4_4 .LBB4_5: # %._crit_edge movl $2019, %edi # imm = 0x7E3 callq srand movq %r14, 8(%rsp) # 8-byte Spill testl %r14d, %r14d jle .LBB4_10 # %bb.6: # %.preheader.lr.ph.i xorl %ebx, %ebx .p2align 4, 0x90 .LBB4_7: # %.preheader.i # =>This Loop Header: Depth=1 # Child Loop BB4_8 Depth 2 xorl %r14d, %r14d .p2align 4, 0x90 .LBB4_8: # Parent Loop BB4_7 Depth=1 # => This Inner Loop Header: Depth=2 callq rand xorps %xmm0, %xmm0 cvtsi2sd %eax, %xmm0 movq (%rbp,%rbx,8), %rax movsd %xmm0, (%rax,%r14,8) callq rand xorps %xmm0, %xmm0 cvtsi2sd %eax, %xmm0 movq (%r15,%rbx,8), %rax movsd %xmm0, (%rax,%r14,8) movq (%r12,%rbx,8), %rax movq $0, (%rax,%r14,8) incq %r14 cmpq %r14, %r13 jne .LBB4_8 # %bb.9: # %._crit_edge.i # in Loop: Header=BB4_7 Depth=1 incq %rbx cmpq %r13, %rbx jne .LBB4_7 .LBB4_10: # %_Z4initPPdS0_S0_i.exit leaq 88(%rsp), %rdi callq hipEventCreate leaq 32(%rsp), %rdi callq hipEventCreate leaq 80(%rsp), %rdi movq 136(%rsp), %rbx # 8-byte Reload movq %rbx, %rsi callq hipMalloc leaq 72(%rsp), %rdi movq %rbx, %rsi callq hipMalloc leaq 24(%rsp), %rdi movq %rbx, %rsi callq hipMalloc movq 80(%rsp), %rdi movq 56(%rsp), %rsi # 8-byte Reload movq %rbx, %rdx movl $1, %ecx callq hipMemcpy movq 72(%rsp), %rdi movq 64(%rsp), %rsi # 8-byte Reload movq %rbx, %rdx movl $1, %ecx callq hipMemcpy movq 24(%rsp), %rdi movq 16(%rsp), %r12 # 8-byte Reload movq %r12, %rsi movq %rbx, %rdx movl $1, %ecx callq hipMemcpy movq 8(%rsp), %r14 # 8-byte Reload leal 31(%r14), %eax testl %r14d, %r14d cmovnsl %r14d, %eax sarl $5, %eax movl %r14d, %ecx andl $31, %ecx cmpl $1, %ecx sbbl $-1, %eax movq %rax, %r15 shlq $32, %r15 orq %rax, %r15 movq 88(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movabsq $137438953504, %rdx # imm = 0x2000000020 movq %r15, %rdi movl $1, %esi movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB4_12 # %bb.11: movq 80(%rsp), %rax movq 72(%rsp), %rcx movq 24(%rsp), %rdx movq %rax, 192(%rsp) movq %rcx, 184(%rsp) movq %rdx, 176(%rsp) movl %r14d, 52(%rsp) leaq 192(%rsp), %rax movq %rax, 96(%rsp) leaq 184(%rsp), %rax movq %rax, 104(%rsp) leaq 176(%rsp), %rax movq %rax, 112(%rsp) leaq 52(%rsp), %rax movq %rax, 120(%rsp) leaq 40(%rsp), %rdi leaq 160(%rsp), %rsi leaq 152(%rsp), %rdx leaq 144(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 160(%rsp), %rcx movl 168(%rsp), %r8d leaq 96(%rsp), %r9 movl $_Z15MulMatrixKernelPdS_S_i, %edi pushq 144(%rsp) .cfi_adjust_cfa_offset 8 pushq 160(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB4_12: movq 32(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movq 24(%rsp), %rsi movq %r12, %rdi movq %rbx, %rdx movl $2, %ecx callq hipMemcpy movq 32(%rsp), %rdi callq hipEventSynchronize movl %r14d, %ebx imull %ebx, %ebx imull %r14d, %ebx movslq %ebx, %r14 movl $0, 40(%rsp) movq 88(%rsp), %rsi movq 32(%rsp), %rdx leaq 40(%rsp), %rdi callq hipEventElapsedTime movss 40(%rsp), %xmm1 # xmm1 = mem[0],zero,zero,zero divss .LCPI4_0(%rip), %xmm1 cvtss2sd %xmm1, %xmm0 testq %r14, %r14 js .LBB4_13 # %bb.14: cvtsi2ss %ebx, %xmm2 jmp .LBB4_15 .LBB4_13: movq %r14, %rax shrq %rax andl $1, %ebx orq %rax, %rbx cvtsi2ss %rbx, %xmm2 addss %xmm2, %xmm2 .LBB4_15: divss %xmm1, %xmm2 xorps %xmm1, %xmm1 cvtss2sd %xmm2, %xmm1 mulsd .LCPI4_1(%rip), %xmm1 movl $.L.str.1, %edi movq 8(%rsp), %rdx # 8-byte Reload movl %edx, %esi # kill: def $edx killed $edx killed $rdx movb $2, %al callq printf leaq 96(%rsp), %rsi movl $4, %edi callq clock_gettime xorps %xmm0, %xmm0 movsd %xmm0, 8(%rsp) # 8-byte Spill xorpd %xmm1, %xmm1 testl %eax, %eax jne .LBB4_17 # %bb.16: xorps %xmm0, %xmm0 cvtsi2sdq 96(%rsp), %xmm0 xorps %xmm1, %xmm1 cvtsi2sdq 104(%rsp), %xmm1 mulsd .LCPI4_2(%rip), %xmm1 addsd %xmm0, %xmm1 .LBB4_17: # %_Z15get_elapsedtimev.exit movsd %xmm1, 16(%rsp) # 8-byte Spill leaq 96(%rsp), %rsi movl $4, %edi callq clock_gettime testl %eax, %eax jne .LBB4_19 # %bb.18: xorps %xmm0, %xmm0 cvtsi2sdq 96(%rsp), %xmm0 xorps %xmm1, %xmm1 cvtsi2sdq 104(%rsp), %xmm1 mulsd .LCPI4_2(%rip), %xmm1 addsd %xmm0, %xmm1 movsd %xmm1, 8(%rsp) # 8-byte Spill .LBB4_19: # %_Z15get_elapsedtimev.exit80 movsd 8(%rsp), %xmm0 # 8-byte Reload # xmm0 = mem[0],zero subsd 16(%rsp), %xmm0 # 8-byte Folded Reload movsd %xmm0, 8(%rsp) # 8-byte Spill movq stdout(%rip), %rcx movl $.L.str.2, %edi movl $22, %esi movl $1, %edx callq fwrite@PLT movq stdout(%rip), %rdi movl $.L.str.3, %esi movsd 8(%rsp), %xmm0 # 8-byte Reload # xmm0 = mem[0],zero movb $1, %al callq fprintf movq stdout(%rip), %rdi movq %r14, %xmm1 punpckldq .LCPI4_3(%rip), %xmm1 # xmm1 = xmm1[0],mem[0],xmm1[1],mem[1] subpd .LCPI4_4(%rip), %xmm1 movapd %xmm1, %xmm0 unpckhpd %xmm1, %xmm0 # xmm0 = xmm0[1],xmm1[1] addsd %xmm1, %xmm0 divsd 8(%rsp), %xmm0 # 8-byte Folded Reload mulsd .LCPI4_1(%rip), %xmm0 movl $.L.str.4, %esi movb $1, %al callq fprintf xorl %eax, %eax addq $200, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end4: .size main, .Lfunc_end4-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB5_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB5_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z15MulMatrixKernelPdS_S_i, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end5: .size __hip_module_ctor, .Lfunc_end5-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB6_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB6_2: retq .Lfunc_end6: .size __hip_module_dtor, .Lfunc_end6-__hip_module_dtor .cfi_endproc # -- End function .type _Z15MulMatrixKernelPdS_S_i,@object # @_Z15MulMatrixKernelPdS_S_i .section .rodata,"a",@progbits .globl _Z15MulMatrixKernelPdS_S_i .p2align 3, 0x0 _Z15MulMatrixKernelPdS_S_i: .quad _Z30__device_stub__MulMatrixKernelPdS_S_i .size _Z15MulMatrixKernelPdS_S_i, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "Matrix Multiplication\n Size: %dx%d\n" .size .L.str, 37 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "Matrice %dx%d\n\tTemps: %f s\n\tMFlops: %.2f\n" .size .L.str.1, 42 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "Performance results: \n" .size .L.str.2, 23 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz " Time: %lf s\n" .size .L.str.3, 15 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz " MFlops: %.2f\n" .size .L.str.4, 16 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z15MulMatrixKernelPdS_S_i" .size .L__unnamed_1, 27 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z30__device_stub__MulMatrixKernelPdS_S_i .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z15MulMatrixKernelPdS_S_i .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include <iostream> #include <algorithm> #include <vector> #include <iterator> #include <fstream> #include <math.h> #include <unistd.h> #include <stdlib.h> #include <string> #include <stdio.h> #include <cmath> #include<sys/stat.h> #include<ctime> #include <cuda_runtime.h> #include<thrust/reduce.h> #include<cuda_runtime.h> #include<thrust/sort.h> #include<thrust/device_ptr.h> #include<thrust/device_vector.h> #include<thrust/host_vector.h> #include<thrust/copy.h> #include<thrust/execution_policy.h> #include<thrust/scan.h> using namespace std; #define thrustSortBlockSize 4000000000 #define bucketNum 10 struct edge{ int src; int dst; }; struct cmpStruc{ __device__ bool operator () (const edge &a, const edge &b){ return (a.src < b.src) || (a.src == b.src && a.dst < b.dst) ; } }cmp; class edgeVector{ public: unsigned int capcity; unsigned int esize; edge *Edges; edgeVector(){esize = 0; capcity = 0;} void init(unsigned int s) { Edges = new edge [s]; capcity = s; return ;} void addEdge(edge * E){ if(esize >= capcity) { capcity *= 2; edge* tmpEdges = new edge [capcity]; memcpy(tmpEdges,Edges,sizeof(edge)*esize); delete [] Edges; Edges = tmpEdges; } memcpy(Edges+esize,E,sizeof(edge)); esize ++; } void clear() {delete [] Edges; return ;} }; unsigned int *edgeOffset; int *edgeRow; int *adjLength; edge *Edges; clock_t start_, end_; bool preProcess(const char *fileName, unsigned int &_edgeNum, unsigned &_nodeNum) { //get file size ifstream fin1(fileName,ios::in|ios::binary); fin1.seekg(0,ios::end); streampos Size = fin1.tellg(); fin1.close(); long int size = Size; cout << "the size of input file is " << size << " Byte. " << endl; unsigned int edgeNum = size/(sizeof(int)*2); Edges = new edge [edgeNum]; //read data ifstream fin(fileName, std::ios::binary); if (fin.bad()) { cout << "File not fould!" << endl; return false; } cout << "start read data... ..." << endl; fin.read((char *)Edges,sizeof(edge)*edgeNum); fin.close(); cout << "end read data" << endl; //pre work //fine node number int divideNum = 100; unsigned int *maxNodeIDs = new unsigned int [divideNum]; memset(maxNodeIDs,0,sizeof(unsigned int)*divideNum); #pragma omp parallel for for (int d = 0; d < divideNum; d++) { unsigned int step = edgeNum/divideNum; unsigned int s = d*step; unsigned int e = (d+1)*step; if (d == divideNum - 1) e = edgeNum; for(unsigned int i = s; i < e; i ++) { if (Edges[i].src > maxNodeIDs[d]) maxNodeIDs[d] = Edges[i].src; if (Edges[i].dst > maxNodeIDs[d]) maxNodeIDs[d] = Edges[i].dst; } } unsigned int maxNodeID = maxNodeIDs[0]; for (int i = 1; i < divideNum; i ++) if (maxNodeIDs[i] > maxNodeID) maxNodeID = maxNodeIDs[i]; cout << "get max nodeid" << endl; unsigned nodeNum = maxNodeID + 1; delete [] maxNodeIDs; //cal degrees int * degreeRecord = new int[nodeNum]; memset(degreeRecord,0,sizeof(int)*nodeNum); //############################################# for (unsigned int i = 0; i < edgeNum; i++) { degreeRecord[Edges[i].src]++; degreeRecord[Edges[i].dst]++; } #pragma omp parallel for for (unsigned int i = 0; i < edgeNum; i ++) { unsigned int src = Edges[i].src; unsigned int dst = Edges[i].dst; if (degreeRecord[src] > degreeRecord[dst] || (degreeRecord[src] == degreeRecord[dst] && src < dst)) { Edges[i].src = dst; Edges[i].dst = src; } } int * toBeMini = new int[nodeNum]; int wipedEdges = 0; memset(toBeMini,0,sizeof(int)*nodeNum); int totalMinied = 0; for (unsigned int i = 0; i < nodeNum; i ++) { if (degreeRecord[i] <= 1) { totalMinied ++; } toBeMini[i] = totalMinied; } #pragma omp parallen for for (unsigned int i = 0; i < edgeNum; i++) { unsigned int src = Edges[i].src; unsigned int dst = Edges[i].dst; if (degreeRecord[src] <= 1) { Edges[i].src = -1; wipedEdges ++; continue; } if (degreeRecord[dst] <= 1) { Edges[i].dst = -1; wipedEdges ++; continue; } if (src > 0) { Edges[i].src = src - toBeMini[src-1]; } if (dst > 0) Edges[i].dst = dst - toBeMini[dst-1]; } nodeNum = nodeNum - totalMinied; delete [] toBeMini; delete [] degreeRecord; cout << "end rearrange dst and src" << endl; //###################################### /*#pragma omp parallel for for (unsigned int i = 0; i < edgeNum; i++) { unsigned int src = Edges[i].src; unsigned int dst = Edges[i].dst; if (src < dst) { Edges[i].src = dst; Edges[i].dst = src; } }*/ //######################################### //sort edges //************sort edges && get nodeNum******** edgeVector * edgeBucket = new edgeVector [bucketNum]; for (int i = 0; i < bucketNum; i ++) edgeBucket[i].init(edgeNum/bucketNum); unsigned bucketStep = (nodeNum + bucketNum - 1)/bucketNum; for (int i = 0; i < edgeNum; i ++) { if (Edges[i].src == -1) continue; int bucketID = Edges[i].src/bucketStep; edgeBucket[bucketID].addEdge(Edges+i); } cout << "end pust edges in bucket" << endl; unsigned int *bucketEdgeOffset = new unsigned int [bucketNum]; bucketEdgeOffset[0] = 0; for (int i = 0; i < bucketNum-1; i ++) { unsigned int bucketSize = edgeBucket[i].esize; if (bucketSize > thrustSortBlockSize/sizeof(edge)) { cout << "bucket " << i << "size is " << bucketSize << ", it's too large!" << endl; return false; } bucketEdgeOffset[i+1] = bucketEdgeOffset[i] + bucketSize; } for (int i = 0; i < bucketNum; i++) { thrust::device_vector<edge> D (edgeBucket[i].Edges, edgeBucket[i].Edges+edgeBucket[i].esize); thrust::sort(D.begin(),D.begin()+edgeBucket[i].esize,cmp); thrust::copy(D.begin(),D.begin()+edgeBucket[i].esize,edgeBucket[i].Edges); } cout << "end sort edges in GPU " << endl; for(int i = 0; i < bucketNum; i ++) { memcpy(Edges+bucketEdgeOffset[i],edgeBucket[i].Edges,sizeof(edge)*edgeBucket[i].esize); } cout << "end copy result to Edges" << endl; delete [] bucketEdgeOffset; for (int i = 0; i < bucketNum; i ++) edgeBucket[i].clear(); delete [] edgeBucket; //************end sort edges && get nodeNum******** edgeNum = edgeNum - wipedEdges;//************************************************ //unsigned int nodeNum = Edges[edgeNum-1].src + 1; edgeOffset = new unsigned int [nodeNum+2]; edgeOffset[0] = 0; edgeRow = new int [edgeNum+1]; adjLength = new int[nodeNum+1]; memset(adjLength,0,sizeof(int)*(nodeNum+1)); unsigned int nodePos = 0; unsigned int edgePos = 0; edge * edgePtr; int formerSrc = -1,formerDst = -1; start_ = clock(); // for (int i = 0; i < edgeNum; i++) // printf("%d %d\n",Edges[i].src,Edges[i].dst); for (unsigned int i = 0; i < edgeNum; i++) { edgePtr = Edges + i; if (edgePtr->src == -1 || edgePtr->dst == -1) continue; if (edgePtr->src == edgePtr->dst) { formerSrc = edgePtr->src; formerDst = edgePtr->dst; int curSrc = edgePtr->src; for (unsigned j = nodePos + 1; j <= curSrc; j++) { edgeOffset[j] = edgePos; adjLength[j-1] = edgeOffset[j]-edgeOffset[j-1]; } nodePos = curSrc; continue; } if ((i > 0) && (edgePtr->src == formerSrc)) { //TODO find a more efficienty way if(edgePtr->dst == formerDst){ continue; } edgeRow[edgePos++] = edgePtr->dst; formerDst = edgePtr->dst; continue; } int curSrc = edgePtr->src; for (unsigned j = nodePos + 1; j <= curSrc; j++) { edgeOffset[j] = edgePos; adjLength[j-1] = edgeOffset[j]-edgeOffset[j-1]; } nodePos = curSrc; edgeRow[edgePos++] = edgePtr->dst; formerSrc = edgePtr->src; formerDst = edgePtr->dst; // cout << " end an edge in a loop " << endl; } for (unsigned i = nodePos + 1; i < nodeNum; i ++) { edgeOffset[i] = edgePos; adjLength[i-1] = edgeOffset[i] - edgeOffset[i-1]; } end_ = clock(); cout << "merge and make csr use " << (double)1000*(end_-start_)/CLOCKS_PER_SEC << " ms." << endl; edgeOffset[nodeNum] = edgePos; edgeOffset[nodeNum+1] = edgePos + 1; adjLength[nodeNum-1] = edgeOffset[nodeNum] - edgeOffset[nodeNum-1]; adjLength[nodeNum] = 1024; edgeRow[edgePos] = nodeNum; cout << "csr built, edgeNum is "<< edgePos<< ", the node num is " << nodeNum << ", origin egde num is " << edgeNum << endl; //TODO remove empty node in edgeOffset int maxDegreeStored = 0; for (int i = 0; i < nodeNum; i ++) if (adjLength[i] > maxDegreeStored) maxDegreeStored = adjLength[i]; cout << "The max stored degree is " << maxDegreeStored << endl; _edgeNum = edgeOffset[nodeNum]; _nodeNum = nodeNum; delete [] Edges; return true; }
#include <iostream> #include <algorithm> #include <vector> #include <iterator> #include <fstream> #include <math.h> #include <unistd.h> #include <stdlib.h> #include <string> #include <stdio.h> #include <cmath> #include<sys/stat.h> #include<ctime> #include <hip/hip_runtime.h> #include<thrust/reduce.h> #include<thrust/sort.h> #include<thrust/device_ptr.h> #include<thrust/device_vector.h> #include<thrust/host_vector.h> #include<thrust/copy.h> #include<thrust/execution_policy.h> #include<thrust/scan.h> using namespace std; #define thrustSortBlockSize 4000000000 #define bucketNum 10 struct edge{ int src; int dst; }; struct cmpStruc{ __device__ bool operator () (const edge &a, const edge &b){ return (a.src < b.src) || (a.src == b.src && a.dst < b.dst) ; } }cmp; class edgeVector{ public: unsigned int capcity; unsigned int esize; edge *Edges; edgeVector(){esize = 0; capcity = 0;} void init(unsigned int s) { Edges = new edge [s]; capcity = s; return ;} void addEdge(edge * E){ if(esize >= capcity) { capcity *= 2; edge* tmpEdges = new edge [capcity]; memcpy(tmpEdges,Edges,sizeof(edge)*esize); delete [] Edges; Edges = tmpEdges; } memcpy(Edges+esize,E,sizeof(edge)); esize ++; } void clear() {delete [] Edges; return ;} }; unsigned int *edgeOffset; int *edgeRow; int *adjLength; edge *Edges; clock_t start_, end_; bool preProcess(const char *fileName, unsigned int &_edgeNum, unsigned &_nodeNum) { //get file size ifstream fin1(fileName,ios::in|ios::binary); fin1.seekg(0,ios::end); streampos Size = fin1.tellg(); fin1.close(); long int size = Size; cout << "the size of input file is " << size << " Byte. " << endl; unsigned int edgeNum = size/(sizeof(int)*2); Edges = new edge [edgeNum]; //read data ifstream fin(fileName, std::ios::binary); if (fin.bad()) { cout << "File not fould!" << endl; return false; } cout << "start read data... ..." << endl; fin.read((char *)Edges,sizeof(edge)*edgeNum); fin.close(); cout << "end read data" << endl; //pre work //fine node number int divideNum = 100; unsigned int *maxNodeIDs = new unsigned int [divideNum]; memset(maxNodeIDs,0,sizeof(unsigned int)*divideNum); #pragma omp parallel for for (int d = 0; d < divideNum; d++) { unsigned int step = edgeNum/divideNum; unsigned int s = d*step; unsigned int e = (d+1)*step; if (d == divideNum - 1) e = edgeNum; for(unsigned int i = s; i < e; i ++) { if (Edges[i].src > maxNodeIDs[d]) maxNodeIDs[d] = Edges[i].src; if (Edges[i].dst > maxNodeIDs[d]) maxNodeIDs[d] = Edges[i].dst; } } unsigned int maxNodeID = maxNodeIDs[0]; for (int i = 1; i < divideNum; i ++) if (maxNodeIDs[i] > maxNodeID) maxNodeID = maxNodeIDs[i]; cout << "get max nodeid" << endl; unsigned nodeNum = maxNodeID + 1; delete [] maxNodeIDs; //cal degrees int * degreeRecord = new int[nodeNum]; memset(degreeRecord,0,sizeof(int)*nodeNum); //############################################# for (unsigned int i = 0; i < edgeNum; i++) { degreeRecord[Edges[i].src]++; degreeRecord[Edges[i].dst]++; } #pragma omp parallel for for (unsigned int i = 0; i < edgeNum; i ++) { unsigned int src = Edges[i].src; unsigned int dst = Edges[i].dst; if (degreeRecord[src] > degreeRecord[dst] || (degreeRecord[src] == degreeRecord[dst] && src < dst)) { Edges[i].src = dst; Edges[i].dst = src; } } int * toBeMini = new int[nodeNum]; int wipedEdges = 0; memset(toBeMini,0,sizeof(int)*nodeNum); int totalMinied = 0; for (unsigned int i = 0; i < nodeNum; i ++) { if (degreeRecord[i] <= 1) { totalMinied ++; } toBeMini[i] = totalMinied; } #pragma omp parallen for for (unsigned int i = 0; i < edgeNum; i++) { unsigned int src = Edges[i].src; unsigned int dst = Edges[i].dst; if (degreeRecord[src] <= 1) { Edges[i].src = -1; wipedEdges ++; continue; } if (degreeRecord[dst] <= 1) { Edges[i].dst = -1; wipedEdges ++; continue; } if (src > 0) { Edges[i].src = src - toBeMini[src-1]; } if (dst > 0) Edges[i].dst = dst - toBeMini[dst-1]; } nodeNum = nodeNum - totalMinied; delete [] toBeMini; delete [] degreeRecord; cout << "end rearrange dst and src" << endl; //###################################### /*#pragma omp parallel for for (unsigned int i = 0; i < edgeNum; i++) { unsigned int src = Edges[i].src; unsigned int dst = Edges[i].dst; if (src < dst) { Edges[i].src = dst; Edges[i].dst = src; } }*/ //######################################### //sort edges //************sort edges && get nodeNum******** edgeVector * edgeBucket = new edgeVector [bucketNum]; for (int i = 0; i < bucketNum; i ++) edgeBucket[i].init(edgeNum/bucketNum); unsigned bucketStep = (nodeNum + bucketNum - 1)/bucketNum; for (int i = 0; i < edgeNum; i ++) { if (Edges[i].src == -1) continue; int bucketID = Edges[i].src/bucketStep; edgeBucket[bucketID].addEdge(Edges+i); } cout << "end pust edges in bucket" << endl; unsigned int *bucketEdgeOffset = new unsigned int [bucketNum]; bucketEdgeOffset[0] = 0; for (int i = 0; i < bucketNum-1; i ++) { unsigned int bucketSize = edgeBucket[i].esize; if (bucketSize > thrustSortBlockSize/sizeof(edge)) { cout << "bucket " << i << "size is " << bucketSize << ", it's too large!" << endl; return false; } bucketEdgeOffset[i+1] = bucketEdgeOffset[i] + bucketSize; } for (int i = 0; i < bucketNum; i++) { thrust::device_vector<edge> D (edgeBucket[i].Edges, edgeBucket[i].Edges+edgeBucket[i].esize); thrust::sort(D.begin(),D.begin()+edgeBucket[i].esize,cmp); thrust::copy(D.begin(),D.begin()+edgeBucket[i].esize,edgeBucket[i].Edges); } cout << "end sort edges in GPU " << endl; for(int i = 0; i < bucketNum; i ++) { memcpy(Edges+bucketEdgeOffset[i],edgeBucket[i].Edges,sizeof(edge)*edgeBucket[i].esize); } cout << "end copy result to Edges" << endl; delete [] bucketEdgeOffset; for (int i = 0; i < bucketNum; i ++) edgeBucket[i].clear(); delete [] edgeBucket; //************end sort edges && get nodeNum******** edgeNum = edgeNum - wipedEdges;//************************************************ //unsigned int nodeNum = Edges[edgeNum-1].src + 1; edgeOffset = new unsigned int [nodeNum+2]; edgeOffset[0] = 0; edgeRow = new int [edgeNum+1]; adjLength = new int[nodeNum+1]; memset(adjLength,0,sizeof(int)*(nodeNum+1)); unsigned int nodePos = 0; unsigned int edgePos = 0; edge * edgePtr; int formerSrc = -1,formerDst = -1; start_ = clock(); // for (int i = 0; i < edgeNum; i++) // printf("%d %d\n",Edges[i].src,Edges[i].dst); for (unsigned int i = 0; i < edgeNum; i++) { edgePtr = Edges + i; if (edgePtr->src == -1 || edgePtr->dst == -1) continue; if (edgePtr->src == edgePtr->dst) { formerSrc = edgePtr->src; formerDst = edgePtr->dst; int curSrc = edgePtr->src; for (unsigned j = nodePos + 1; j <= curSrc; j++) { edgeOffset[j] = edgePos; adjLength[j-1] = edgeOffset[j]-edgeOffset[j-1]; } nodePos = curSrc; continue; } if ((i > 0) && (edgePtr->src == formerSrc)) { //TODO find a more efficienty way if(edgePtr->dst == formerDst){ continue; } edgeRow[edgePos++] = edgePtr->dst; formerDst = edgePtr->dst; continue; } int curSrc = edgePtr->src; for (unsigned j = nodePos + 1; j <= curSrc; j++) { edgeOffset[j] = edgePos; adjLength[j-1] = edgeOffset[j]-edgeOffset[j-1]; } nodePos = curSrc; edgeRow[edgePos++] = edgePtr->dst; formerSrc = edgePtr->src; formerDst = edgePtr->dst; // cout << " end an edge in a loop " << endl; } for (unsigned i = nodePos + 1; i < nodeNum; i ++) { edgeOffset[i] = edgePos; adjLength[i-1] = edgeOffset[i] - edgeOffset[i-1]; } end_ = clock(); cout << "merge and make csr use " << (double)1000*(end_-start_)/CLOCKS_PER_SEC << " ms." << endl; edgeOffset[nodeNum] = edgePos; edgeOffset[nodeNum+1] = edgePos + 1; adjLength[nodeNum-1] = edgeOffset[nodeNum] - edgeOffset[nodeNum-1]; adjLength[nodeNum] = 1024; edgeRow[edgePos] = nodeNum; cout << "csr built, edgeNum is "<< edgePos<< ", the node num is " << nodeNum << ", origin egde num is " << edgeNum << endl; //TODO remove empty node in edgeOffset int maxDegreeStored = 0; for (int i = 0; i < nodeNum; i ++) if (adjLength[i] > maxDegreeStored) maxDegreeStored = adjLength[i]; cout << "The max stored degree is " << maxDegreeStored << endl; _edgeNum = edgeOffset[nodeNum]; _nodeNum = nodeNum; delete [] Edges; return true; }
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
/* ============================================================================ Name : sem1.cu Author : maminov Version : Copyright : copyleft Description : CUDA compute reciprocals ============================================================================ */ #include <cuda_runtime.h> #include <device_launch_parameters.h> #include <iostream> #include <numeric> #include <stdlib.h> static void CheckCudaErrorAux(const char *, unsigned, const char *, cudaError_t); #define CUDA_CHECK_RETURN(value) CheckCudaErrorAux(__FILE__,__LINE__, #value, value) /** * CUDA kernel that computes reciprocal values for a given vector */ __global__ void heatKernel(float *data, unsigned size, unsigned iterations, float tempPerIteration, float dx, float dt) { unsigned idx = blockIdx.x*blockDim.x + threadIdx.x; if (idx < size) for (int i = 0; i < iterations; ++i) { if (idx == 0) data[size + 1] += tempPerIteration; data[idx + 1] += (data[idx - 1 + 1] - 2 * data[idx + 1] + data[idx + 1 + 1]) * dt / (dx*dx); __syncthreads(); } } void initialize(float *data, unsigned size) { for (unsigned i = 0; i < size; ++i) data[i] = 0.; } int main(void) { static const int SIM_TIME = 5; static const int STEPS_PER_SECOND = 10; static const int CYL_LENGTH = 10; static const int STEPS_PER_METER = 10; static const int WORK_SIZE = CYL_LENGTH * STEPS_PER_METER; static const int BLOCK_SIZE = 256; static const float TEMP_PER_SEC = 5.; float *hostData = new float[WORK_SIZE + 2]; float *gpuData; initialize(hostData, WORK_SIZE + 2); CUDA_CHECK_RETURN(cudaMalloc((void **)&gpuData, sizeof(float)*(WORK_SIZE + 2))); CUDA_CHECK_RETURN(cudaMemcpy(gpuData, hostData, sizeof(float)*(WORK_SIZE + 2), cudaMemcpyHostToDevice)); float *result = new float[WORK_SIZE + 2]; const int blockCount = (WORK_SIZE + BLOCK_SIZE - 1) / BLOCK_SIZE; heatKernel <<< blockCount, BLOCK_SIZE >>> (gpuData, WORK_SIZE, SIM_TIME * STEPS_PER_SECOND, TEMP_PER_SEC / (STEPS_PER_SECOND*1.0f), CYL_LENGTH / (STEPS_PER_METER*1.0f), SIM_TIME / (STEPS_PER_SECOND*1.0f)); CUDA_CHECK_RETURN(cudaMemcpy(result, gpuData, sizeof(float)*(WORK_SIZE + 2), cudaMemcpyDeviceToHost)); /* Verify the results */ for (int i = 0; i < WORK_SIZE; ++i) std::cout << result[i + 1] << std::endl; /* Free memory */ CUDA_CHECK_RETURN(cudaFree(gpuData)); delete[] hostData; delete[] result; return 0; } /** * Check the return value of the CUDA runtime API call and exit * the application if the call has failed. */ static void CheckCudaErrorAux(const char *file, unsigned line, const char *statement, cudaError_t err) { if (err == cudaSuccess) return; std::cerr << statement << " returned " << cudaGetErrorString(err) << "(" << err << ") at " << file << ":" << line << std::endl; exit(1); }
code for sm_80 Function : _Z10heatKernelPfjjfff .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R0, SR_CTAID.X ; /* 0x0000000000007919 */ /* 0x000e220000002500 */ /*0020*/ ISETP.NE.AND P0, PT, RZ, c[0x0][0x16c], PT ; /* 0x00005b00ff007a0c */ /* 0x000fc60003f05270 */ /*0030*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e240000002100 */ /*0040*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x001fca00078e0203 */ /*0050*/ ISETP.GE.U32.OR P0, PT, R0, c[0x0][0x168], !P0 ; /* 0x00005a0000007a0c */ /* 0x000fda0004706470 */ /*0060*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0070*/ MOV R10, c[0x0][0x16c] ; /* 0x00005b00000a7a02 */ /* 0x000fe20000000f00 */ /*0080*/ IMAD.MOV.U32 R2, RZ, RZ, 0x1 ; /* 0x00000001ff027424 */ /* 0x000fe200078e00ff */ /*0090*/ IADD3 R8, R0, 0x1, RZ ; /* 0x0000000100087810 */ /* 0x000fe20007ffe0ff */ /*00a0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*00b0*/ IADD3 R3, R10, -0x1, RZ ; /* 0xffffffff0a037810 */ /* 0x000fe40007ffe0ff */ /*00c0*/ IADD3 R6, R0, 0x2, RZ ; /* 0x0000000200067810 */ /* 0x000fe40007ffe0ff */ /*00d0*/ ISETP.GE.U32.AND P0, PT, R3, 0x3, PT ; /* 0x000000030300780c */ /* 0x000fe20003f06070 */ /*00e0*/ IMAD.MOV.U32 R3, RZ, RZ, 0x4 ; /* 0x00000004ff037424 */ /* 0x000fe200078e00ff */ /*00f0*/ MOV R11, c[0x0][0x174] ; /* 0x00005d00000b7a02 */ /* 0x000fc40000000f00 */ /*0100*/ IADD3 R2, R2, c[0x0][0x168], RZ ; /* 0x00005a0002027a10 */ /* 0x000fe20007ffe0ff */ /*0110*/ IMAD.WIDE.U32 R8, R8, R3, c[0x0][0x160] ; /* 0x0000580008087625 */ /* 0x000fe200078e0003 */ /*0120*/ LOP3.LUT R10, R10, 0x3, RZ, 0xc0, !PT ; /* 0x000000030a0a7812 */ /* 0x000fc600078ec0ff */ /*0130*/ IMAD.WIDE.U32 R6, R6, R3, c[0x0][0x160] ; /* 0x0000580006067625 */ /* 0x000fc800078e0003 */ /*0140*/ IMAD.WIDE.U32 R4, R0, R3, c[0x0][0x160] ; /* 0x0000580000047625 */ /* 0x000fc800078e0003 */ /*0150*/ FMUL R11, R11, c[0x0][0x174] ; /* 0x00005d000b0b7a20 */ /* 0x000fe40000400000 */ /*0160*/ IMAD.WIDE.U32 R2, R2, R3, c[0x0][0x160] ; /* 0x0000580002027625 */ /* 0x000fe200078e0003 */ /*0170*/ @!P0 BRA 0x8e0 ; /* 0x0000076000008947 */ /* 0x000fea0003800000 */ /*0180*/ IADD3 R12, -R10, c[0x0][0x16c], RZ ; /* 0x00005b000a0c7a10 */ /* 0x000fe40007ffe1ff */ /*0190*/ ISETP.NE.AND P2, PT, R0, RZ, PT ; /* 0x000000ff0000720c */ /* 0x000fda0003f45270 */ /*01a0*/ @!P2 LDG.E R13, [R2.64] ; /* 0x00000004020da981 */ /* 0x001ea4000c1e1900 */ /*01b0*/ @!P2 FADD R15, R13, c[0x0][0x170] ; /* 0x00005c000d0fa621 */ /* 0x004fca0000000000 */ /*01c0*/ @!P2 STG.E [R2.64], R15 ; /* 0x0000000f0200a986 */ /* 0x0001e8000c101904 */ /*01d0*/ LDG.E R13, [R8.64] ; /* 0x00000004080d7981 */ /* 0x000ea8000c1e1900 */ /*01e0*/ LDG.E R14, [R4.64] ; /* 0x00000004040e7981 */ /* 0x000ee8000c1e1900 */ /*01f0*/ LDG.E R19, [R6.64] ; /* 0x0000000406137981 */ /* 0x000f22000c1e1900 */ /*0200*/ MUFU.RCP R16, R11 ; /* 0x0000000b00107308 */ /* 0x000e620000001000 */ /*0210*/ IADD3 R12, R12, -0x4, RZ ; /* 0xfffffffc0c0c7810 */ /* 0x000fe20007ffe0ff */ /*0220*/ BSSY B0, 0x330 ; /* 0x0000010000007945 */ /* 0x000fe60003800000 */ /*0230*/ ISETP.NE.AND P3, PT, R12, RZ, PT ; /* 0x000000ff0c00720c */ /* 0x000fe20003f65270 */ /*0240*/ FADD R17, R13, R13 ; /* 0x0000000d0d117221 */ /* 0x004fc80000000000 */ /*0250*/ FADD R14, R14, -R17 ; /* 0x800000110e0e7221 */ /* 0x008fe40000000000 */ /*0260*/ FFMA R17, -R11, R16, 1 ; /* 0x3f8000000b117423 */ /* 0x002fe40000000110 */ /*0270*/ FADD R14, R14, R19 ; /* 0x000000130e0e7221 */ /* 0x010fe40000000000 */ /*0280*/ FFMA R17, R16, R17, R16 ; /* 0x0000001110117223 */ /* 0x000fe40000000010 */ /*0290*/ FMUL R14, R14, c[0x0][0x178] ; /* 0x00005e000e0e7a20 */ /* 0x000fc80000400000 */ /*02a0*/ FCHK P0, R14, R11 ; /* 0x0000000b0e007302 */ /* 0x000e620000000000 */ /*02b0*/ FFMA R16, R14, R17, RZ ; /* 0x000000110e107223 */ /* 0x000fc800000000ff */ /*02c0*/ FFMA R15, -R11, R16, R14 ; /* 0x000000100b0f7223 */ /* 0x001fc8000000010e */ /*02d0*/ FFMA R16, R17, R15, R16 ; /* 0x0000000f11107223 */ /* 0x000fe20000000010 */ /*02e0*/ @!P0 BRA 0x320 ; /* 0x0000003000008947 */ /* 0x002fea0003800000 */ /*02f0*/ MOV R17, R14 ; /* 0x0000000e00117202 */ /* 0x000fe40000000f00 */ /*0300*/ MOV R14, 0x320 ; /* 0x00000320000e7802 */ /* 0x000fe40000000f00 */ /*0310*/ CALL.REL.NOINC 0xb00 ; /* 0x000007e000007944 */ /* 0x000fea0003c00000 */ /*0320*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0330*/ FADD R13, R13, R16 ; /* 0x000000100d0d7221 */ /* 0x000fe20000000000 */ /*0340*/ BSSY B0, 0x3c0 ; /* 0x0000007000007945 */ /* 0x000fe80003800000 */ /*0350*/ STG.E [R8.64], R13 ; /* 0x0000000d08007986 */ /* 0x0001e8000c101904 */ /*0360*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0370*/ @P2 BRA 0x3b0 ; /* 0x0000003000002947 */ /* 0x000fea0003800000 */ /*0380*/ LDG.E R13, [R2.64] ; /* 0x00000004020d7981 */ /* 0x001ea4000c1e1900 */ /*0390*/ FADD R13, R13, c[0x0][0x170] ; /* 0x00005c000d0d7621 */ /* 0x004fca0000000000 */ /*03a0*/ STG.E [R2.64], R13 ; /* 0x0000000d02007986 */ /* 0x0001e4000c101904 */ /*03b0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*03c0*/ LDG.E R13, [R8.64] ; /* 0x00000004080d7981 */ /* 0x001ea8000c1e1900 */ /*03d0*/ LDG.E R14, [R4.64] ; /* 0x00000004040e7981 */ /* 0x000ee8000c1e1900 */ /*03e0*/ LDG.E R17, [R6.64] ; /* 0x0000000406117981 */ /* 0x000f22000c1e1900 */ /*03f0*/ MUFU.RCP R16, R11 ; /* 0x0000000b00107308 */ /* 0x000e220000001000 */ /*0400*/ BSSY B0, 0x500 ; /* 0x000000f000007945 */ /* 0x000fe20003800000 */ /*0410*/ FADD R15, R13, R13 ; /* 0x0000000d0d0f7221 */ /* 0x004fc80000000000 */ /*0420*/ FADD R14, R14, -R15 ; /* 0x8000000f0e0e7221 */ /* 0x008fe40000000000 */ /*0430*/ FFMA R15, -R11, R16, 1 ; /* 0x3f8000000b0f7423 */ /* 0x001fe40000000110 */ /*0440*/ FADD R14, R14, R17 ; /* 0x000000110e0e7221 */ /* 0x010fe40000000000 */ /*0450*/ FFMA R17, R16, R15, R16 ; /* 0x0000000f10117223 */ /* 0x000fe40000000010 */ /*0460*/ FMUL R14, R14, c[0x0][0x178] ; /* 0x00005e000e0e7a20 */ /* 0x000fc80000400000 */ /*0470*/ FCHK P0, R14, R11 ; /* 0x0000000b0e007302 */ /* 0x000e220000000000 */ /*0480*/ FFMA R15, R17, R14, RZ ; /* 0x0000000e110f7223 */ /* 0x000fc800000000ff */ /*0490*/ FFMA R16, -R11, R15, R14 ; /* 0x0000000f0b107223 */ /* 0x000fc8000000010e */ /*04a0*/ FFMA R16, R17, R16, R15 ; /* 0x0000001011107223 */ /* 0x000fe2000000000f */ /*04b0*/ @!P0 BRA 0x4f0 ; /* 0x0000003000008947 */ /* 0x001fea0003800000 */ /*04c0*/ IMAD.MOV.U32 R17, RZ, RZ, R14 ; /* 0x000000ffff117224 */ /* 0x000fe200078e000e */ /*04d0*/ MOV R14, 0x4f0 ; /* 0x000004f0000e7802 */ /* 0x000fe40000000f00 */ /*04e0*/ CALL.REL.NOINC 0xb00 ; /* 0x0000061000007944 */ /* 0x000fea0003c00000 */ /*04f0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0500*/ FADD R13, R13, R16 ; /* 0x000000100d0d7221 */ /* 0x000fe20000000000 */ /*0510*/ BSSY B0, 0x590 ; /* 0x0000007000007945 */ /* 0x000fe80003800000 */ /*0520*/ STG.E [R8.64], R13 ; /* 0x0000000d08007986 */ /* 0x0001e8000c101904 */ /*0530*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0540*/ @P2 BRA 0x580 ; /* 0x0000003000002947 */ /* 0x000fea0003800000 */ /*0550*/ LDG.E R13, [R2.64] ; /* 0x00000004020d7981 */ /* 0x001ea4000c1e1900 */ /*0560*/ FADD R13, R13, c[0x0][0x170] ; /* 0x00005c000d0d7621 */ /* 0x004fca0000000000 */ /*0570*/ STG.E [R2.64], R13 ; /* 0x0000000d02007986 */ /* 0x0001e4000c101904 */ /*0580*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0590*/ LDG.E R13, [R8.64] ; /* 0x00000004080d7981 */ /* 0x001ea8000c1e1900 */ /*05a0*/ LDG.E R14, [R4.64] ; /* 0x00000004040e7981 */ /* 0x000ee8000c1e1900 */ /*05b0*/ LDG.E R17, [R6.64] ; /* 0x0000000406117981 */ /* 0x000f22000c1e1900 */ /*05c0*/ MUFU.RCP R16, R11 ; /* 0x0000000b00107308 */ /* 0x000e220000001000 */ /*05d0*/ BSSY B0, 0x6d0 ; /* 0x000000f000007945 */ /* 0x000fe20003800000 */ /*05e0*/ FADD R15, R13, R13 ; /* 0x0000000d0d0f7221 */ /* 0x004fc80000000000 */ /*05f0*/ FADD R14, R14, -R15 ; /* 0x8000000f0e0e7221 */ /* 0x008fe40000000000 */ /*0600*/ FFMA R15, -R11, R16, 1 ; /* 0x3f8000000b0f7423 */ /* 0x001fe40000000110 */ /*0610*/ FADD R14, R14, R17 ; /* 0x000000110e0e7221 */ /* 0x010fe40000000000 */ /*0620*/ FFMA R17, R16, R15, R16 ; /* 0x0000000f10117223 */ /* 0x000fe40000000010 */ /*0630*/ FMUL R14, R14, c[0x0][0x178] ; /* 0x00005e000e0e7a20 */ /* 0x000fc80000400000 */ /*0640*/ FCHK P0, R14, R11 ; /* 0x0000000b0e007302 */ /* 0x000e220000000000 */ /*0650*/ FFMA R15, R17, R14, RZ ; /* 0x0000000e110f7223 */ /* 0x000fc800000000ff */ /*0660*/ FFMA R16, -R11, R15, R14 ; /* 0x0000000f0b107223 */ /* 0x000fc8000000010e */ /*0670*/ FFMA R16, R17, R16, R15 ; /* 0x0000001011107223 */ /* 0x000fe2000000000f */ /*0680*/ @!P0 BRA 0x6c0 ; /* 0x0000003000008947 */ /* 0x001fea0003800000 */ /*0690*/ MOV R17, R14 ; /* 0x0000000e00117202 */ /* 0x000fe40000000f00 */ /*06a0*/ MOV R14, 0x6c0 ; /* 0x000006c0000e7802 */ /* 0x000fe40000000f00 */ /*06b0*/ CALL.REL.NOINC 0xb00 ; /* 0x0000044000007944 */ /* 0x000fea0003c00000 */ /*06c0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*06d0*/ FADD R13, R13, R16 ; /* 0x000000100d0d7221 */ /* 0x000fe20000000000 */ /*06e0*/ BSSY B0, 0x760 ; /* 0x0000007000007945 */ /* 0x000fe80003800000 */ /*06f0*/ STG.E [R8.64], R13 ; /* 0x0000000d08007986 */ /* 0x0001e8000c101904 */ /*0700*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0710*/ @P2 BRA 0x750 ; /* 0x0000003000002947 */ /* 0x000fea0003800000 */ /*0720*/ LDG.E R13, [R2.64] ; /* 0x00000004020d7981 */ /* 0x001ea4000c1e1900 */ /*0730*/ FADD R13, R13, c[0x0][0x170] ; /* 0x00005c000d0d7621 */ /* 0x004fca0000000000 */ /*0740*/ STG.E [R2.64], R13 ; /* 0x0000000d02007986 */ /* 0x0001e4000c101904 */ /*0750*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0760*/ LDG.E R13, [R8.64] ; /* 0x00000004080d7981 */ /* 0x001ea8000c1e1900 */ /*0770*/ LDG.E R14, [R4.64] ; /* 0x00000004040e7981 */ /* 0x000ee8000c1e1900 */ /*0780*/ LDG.E R17, [R6.64] ; /* 0x0000000406117981 */ /* 0x000f22000c1e1900 */ /*0790*/ MUFU.RCP R16, R11 ; /* 0x0000000b00107308 */ /* 0x000e220000001000 */ /*07a0*/ BSSY B0, 0x8a0 ; /* 0x000000f000007945 */ /* 0x000fe20003800000 */ /*07b0*/ FADD R15, R13, R13 ; /* 0x0000000d0d0f7221 */ /* 0x004fc80000000000 */ /*07c0*/ FADD R14, R14, -R15 ; /* 0x8000000f0e0e7221 */ /* 0x008fe40000000000 */ /*07d0*/ FFMA R15, -R11, R16, 1 ; /* 0x3f8000000b0f7423 */ /* 0x001fe40000000110 */ /*07e0*/ FADD R14, R14, R17 ; /* 0x000000110e0e7221 */ /* 0x010fe40000000000 */ /*07f0*/ FFMA R17, R16, R15, R16 ; /* 0x0000000f10117223 */ /* 0x000fe40000000010 */ /*0800*/ FMUL R14, R14, c[0x0][0x178] ; /* 0x00005e000e0e7a20 */ /* 0x000fc80000400000 */ /*0810*/ FCHK P0, R14, R11 ; /* 0x0000000b0e007302 */ /* 0x000e220000000000 */ /*0820*/ FFMA R15, R17, R14, RZ ; /* 0x0000000e110f7223 */ /* 0x000fc800000000ff */ /*0830*/ FFMA R16, -R11, R15, R14 ; /* 0x0000000f0b107223 */ /* 0x000fc8000000010e */ /*0840*/ FFMA R16, R17, R16, R15 ; /* 0x0000001011107223 */ /* 0x000fe2000000000f */ /*0850*/ @!P0 BRA 0x890 ; /* 0x0000003000008947 */ /* 0x001fea0003800000 */ /*0860*/ IMAD.MOV.U32 R17, RZ, RZ, R14 ; /* 0x000000ffff117224 */ /* 0x000fe200078e000e */ /*0870*/ MOV R14, 0x890 ; /* 0x00000890000e7802 */ /* 0x000fe40000000f00 */ /*0880*/ CALL.REL.NOINC 0xb00 ; /* 0x0000027000007944 */ /* 0x000fea0003c00000 */ /*0890*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*08a0*/ FADD R13, R13, R16 ; /* 0x000000100d0d7221 */ /* 0x000fca0000000000 */ /*08b0*/ STG.E [R8.64], R13 ; /* 0x0000000d08007986 */ /* 0x0001e8000c101904 */ /*08c0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*08d0*/ @P3 BRA 0x190 ; /* 0xfffff8b000003947 */ /* 0x000fea000383ffff */ /*08e0*/ ISETP.NE.AND P0, PT, R10, RZ, PT ; /* 0x000000ff0a00720c */ /* 0x000fda0003f05270 */ /*08f0*/ @!P0 EXIT ; /* 0x000000000000894d */ /* 0x000fea0003800000 */ /*0900*/ ISETP.NE.AND P0, PT, R0, RZ, PT ; /* 0x000000ff0000720c */ /* 0x000fda0003f05270 */ /*0910*/ @!P0 LDG.E R12, [R2.64] ; /* 0x00000004020c8981 */ /* 0x000ea4000c1e1900 */ /*0920*/ @!P0 FADD R13, R12, c[0x0][0x170] ; /* 0x00005c000c0d8621 */ /* 0x005fca0000000000 */ /*0930*/ @!P0 STG.E [R2.64], R13 ; /* 0x0000000d02008986 */ /* 0x0001e8000c101904 */ /*0940*/ LDG.E R12, [R8.64] ; /* 0x00000004080c7981 */ /* 0x000ea8000c1e1900 */ /*0950*/ LDG.E R14, [R4.64] ; /* 0x00000004040e7981 */ /* 0x000ee8000c1e1900 */ /*0960*/ LDG.E R17, [R6.64] ; /* 0x0000000406117981 */ /* 0x000f22000c1e1900 */ /*0970*/ MUFU.RCP R16, R11 ; /* 0x0000000b00107308 */ /* 0x000e620000001000 */ /*0980*/ BSSY B0, 0xa90 ; /* 0x0000010000007945 */ /* 0x000fe20003800000 */ /*0990*/ FADD R15, R12, R12 ; /* 0x0000000c0c0f7221 */ /* 0x004fc80000000000 */ /*09a0*/ FADD R14, R14, -R15 ; /* 0x8000000f0e0e7221 */ /* 0x008fe40000000000 */ /*09b0*/ FFMA R15, -R11, R16, 1 ; /* 0x3f8000000b0f7423 */ /* 0x002fe40000000110 */ /*09c0*/ FADD R14, R14, R17 ; /* 0x000000110e0e7221 */ /* 0x010fe40000000000 */ /*09d0*/ FFMA R15, R16, R15, R16 ; /* 0x0000000f100f7223 */ /* 0x000fe40000000010 */ /*09e0*/ FMUL R14, R14, c[0x0][0x178] ; /* 0x00005e000e0e7a20 */ /* 0x000fc80000400000 */ /*09f0*/ FCHK P0, R14, R11 ; /* 0x0000000b0e007302 */ /* 0x000e620000000000 */ /*0a00*/ FFMA R13, R15, R14, RZ ; /* 0x0000000e0f0d7223 */ /* 0x001fc800000000ff */ /*0a10*/ FFMA R16, -R11, R13, R14 ; /* 0x0000000d0b107223 */ /* 0x000fc8000000010e */ /*0a20*/ FFMA R13, R15, R16, R13 ; /* 0x000000100f0d7223 */ /* 0x000fe2000000000d */ /*0a30*/ @!P0 BRA 0xa80 ; /* 0x0000004000008947 */ /* 0x002fea0003800000 */ /*0a40*/ MOV R17, R14 ; /* 0x0000000e00117202 */ /* 0x000fe40000000f00 */ /*0a50*/ MOV R14, 0xa70 ; /* 0x00000a70000e7802 */ /* 0x000fe40000000f00 */ /*0a60*/ CALL.REL.NOINC 0xb00 ; /* 0x0000009000007944 */ /* 0x000fea0003c00000 */ /*0a70*/ IMAD.MOV.U32 R13, RZ, RZ, R16 ; /* 0x000000ffff0d7224 */ /* 0x000fe400078e0010 */ /*0a80*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0a90*/ IADD3 R10, R10, -0x1, RZ ; /* 0xffffffff0a0a7810 */ /* 0x000fe20007ffe0ff */ /*0aa0*/ FADD R13, R12, R13 ; /* 0x0000000d0c0d7221 */ /* 0x000fc60000000000 */ /*0ab0*/ ISETP.NE.AND P0, PT, R10, RZ, PT ; /* 0x000000ff0a00720c */ /* 0x000fe40003f05270 */ /*0ac0*/ STG.E [R8.64], R13 ; /* 0x0000000d08007986 */ /* 0x0001e8000c101904 */ /*0ad0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fee0000010000 */ /*0ae0*/ @P0 BRA 0x900 ; /* 0xfffffe1000000947 */ /* 0x000fea000383ffff */ /*0af0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0b00*/ SHF.R.U32.HI R15, RZ, 0x17, R11 ; /* 0x00000017ff0f7819 */ /* 0x000fe2000001160b */ /*0b10*/ BSSY B1, 0x1150 ; /* 0x0000063000017945 */ /* 0x000fe20003800000 */ /*0b20*/ SHF.R.U32.HI R18, RZ, 0x17, R17 ; /* 0x00000017ff127819 */ /* 0x000fc40000011611 */ /*0b30*/ LOP3.LUT R15, R15, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff0f0f7812 */ /* 0x000fe400078ec0ff */ /*0b40*/ LOP3.LUT R18, R18, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff12127812 */ /* 0x000fe400078ec0ff */ /*0b50*/ IADD3 R21, R15, -0x1, RZ ; /* 0xffffffff0f157810 */ /* 0x000fe40007ffe0ff */ /*0b60*/ IADD3 R19, R18, -0x1, RZ ; /* 0xffffffff12137810 */ /* 0x000fe40007ffe0ff */ /*0b70*/ ISETP.GT.U32.AND P0, PT, R21, 0xfd, PT ; /* 0x000000fd1500780c */ /* 0x000fe40003f04070 */ /*0b80*/ MOV R20, R11 ; /* 0x0000000b00147202 */ /* 0x000fc40000000f00 */ /*0b90*/ ISETP.GT.U32.OR P0, PT, R19, 0xfd, P0 ; /* 0x000000fd1300780c */ /* 0x000fda0000704470 */ /*0ba0*/ @!P0 IMAD.MOV.U32 R16, RZ, RZ, RZ ; /* 0x000000ffff108224 */ /* 0x000fe200078e00ff */ /*0bb0*/ @!P0 BRA 0xd30 ; /* 0x0000017000008947 */ /* 0x000fea0003800000 */ /*0bc0*/ FSETP.GTU.FTZ.AND P0, PT, |R17|, +INF , PT ; /* 0x7f8000001100780b */ /* 0x000fe40003f1c200 */ /*0bd0*/ FSETP.GTU.FTZ.AND P1, PT, |R11|, +INF , PT ; /* 0x7f8000000b00780b */ /* 0x000fc80003f3c200 */ /*0be0*/ PLOP3.LUT P0, PT, P0, P1, PT, 0xa8, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000703570 */ /*0bf0*/ @P0 BRA 0x1130 ; /* 0x0000053000000947 */ /* 0x000fea0003800000 */ /*0c00*/ LOP3.LUT P0, RZ, R20, 0x7fffffff, R17, 0xc8, !PT ; /* 0x7fffffff14ff7812 */ /* 0x000fda000780c811 */ /*0c10*/ @!P0 BRA 0x1110 ; /* 0x000004f000008947 */ /* 0x000fea0003800000 */ /*0c20*/ FSETP.NEU.FTZ.AND P4, PT, |R17|, +INF , PT ; /* 0x7f8000001100780b */ /* 0x000fe40003f9d200 */ /*0c30*/ FSETP.NEU.FTZ.AND P1, PT, |R11|, +INF , PT ; /* 0x7f8000000b00780b */ /* 0x000fe40003f3d200 */ /*0c40*/ FSETP.NEU.FTZ.AND P0, PT, |R17|, +INF , PT ; /* 0x7f8000001100780b */ /* 0x000fd60003f1d200 */ /*0c50*/ @!P1 BRA !P4, 0x1110 ; /* 0x000004b000009947 */ /* 0x000fea0006000000 */ /*0c60*/ LOP3.LUT P4, RZ, R17, 0x7fffffff, RZ, 0xc0, !PT ; /* 0x7fffffff11ff7812 */ /* 0x000fc8000788c0ff */ /*0c70*/ PLOP3.LUT P1, PT, P1, P4, PT, 0x2a, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000f28572 */ /*0c80*/ @P1 BRA 0x10f0 ; /* 0x0000046000001947 */ /* 0x000fea0003800000 */ /*0c90*/ LOP3.LUT P1, RZ, R20, 0x7fffffff, RZ, 0xc0, !PT ; /* 0x7fffffff14ff7812 */ /* 0x000fc8000782c0ff */ /*0ca0*/ PLOP3.LUT P0, PT, P0, P1, PT, 0x2a, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000702572 */ /*0cb0*/ @P0 BRA 0x10c0 ; /* 0x0000040000000947 */ /* 0x000fea0003800000 */ /*0cc0*/ ISETP.GE.AND P0, PT, R19, RZ, PT ; /* 0x000000ff1300720c */ /* 0x000fe40003f06270 */ /*0cd0*/ ISETP.GE.AND P1, PT, R21, RZ, PT ; /* 0x000000ff1500720c */ /* 0x000fd60003f26270 */ /*0ce0*/ @P0 MOV R16, RZ ; /* 0x000000ff00100202 */ /* 0x000fe20000000f00 */ /*0cf0*/ @!P0 IMAD.MOV.U32 R16, RZ, RZ, -0x40 ; /* 0xffffffc0ff108424 */ /* 0x000fe400078e00ff */ /*0d00*/ @!P0 FFMA R17, R17, 1.84467440737095516160e+19, RZ ; /* 0x5f80000011118823 */ /* 0x000fe400000000ff */ /*0d10*/ @!P1 FFMA R20, R11, 1.84467440737095516160e+19, RZ ; /* 0x5f8000000b149823 */ /* 0x000fe200000000ff */ /*0d20*/ @!P1 IADD3 R16, R16, 0x40, RZ ; /* 0x0000004010109810 */ /* 0x000fe40007ffe0ff */ /*0d30*/ LEA R19, R15, 0xc0800000, 0x17 ; /* 0xc08000000f137811 */ /* 0x000fe200078eb8ff */ /*0d40*/ BSSY B2, 0x10b0 ; /* 0x0000036000027945 */ /* 0x000fe20003800000 */ /*0d50*/ IADD3 R18, R18, -0x7f, RZ ; /* 0xffffff8112127810 */ /* 0x000fe40007ffe0ff */ /*0d60*/ IADD3 R23, -R19, R20, RZ ; /* 0x0000001413177210 */ /* 0x000fc60007ffe1ff */ /*0d70*/ IMAD R22, R18.reuse, -0x800000, R17 ; /* 0xff80000012167824 */ /* 0x040fe200078e0211 */ /*0d80*/ MUFU.RCP R20, R23 ; /* 0x0000001700147308 */ /* 0x0000620000001000 */ /*0d90*/ FADD.FTZ R19, -R23, -RZ ; /* 0x800000ff17137221 */ /* 0x000fe20000010100 */ /*0da0*/ IADD3 R23, R18, 0x7f, -R15 ; /* 0x0000007f12177810 */ /* 0x001fca0007ffe80f */ /*0db0*/ IMAD.IADD R18, R23, 0x1, R16 ; /* 0x0000000117127824 */ /* 0x000fe400078e0210 */ /*0dc0*/ FFMA R21, R20, R19, 1 ; /* 0x3f80000014157423 */ /* 0x002fc80000000013 */ /*0dd0*/ FFMA R17, R20, R21, R20 ; /* 0x0000001514117223 */ /* 0x000fc80000000014 */ /*0de0*/ FFMA R20, R22, R17, RZ ; /* 0x0000001116147223 */ /* 0x000fc800000000ff */ /*0df0*/ FFMA R21, R19, R20, R22 ; /* 0x0000001413157223 */ /* 0x000fc80000000016 */ /*0e00*/ FFMA R20, R17, R21, R20 ; /* 0x0000001511147223 */ /* 0x000fc80000000014 */ /*0e10*/ FFMA R19, R19, R20, R22 ; /* 0x0000001413137223 */ /* 0x000fc80000000016 */ /*0e20*/ FFMA R21, R17, R19, R20 ; /* 0x0000001311157223 */ /* 0x000fca0000000014 */ /*0e30*/ SHF.R.U32.HI R15, RZ, 0x17, R21 ; /* 0x00000017ff0f7819 */ /* 0x000fc80000011615 */ /*0e40*/ LOP3.LUT R15, R15, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff0f0f7812 */ /* 0x000fc800078ec0ff */ /*0e50*/ IADD3 R16, R15, R18, RZ ; /* 0x000000120f107210 */ /* 0x000fc80007ffe0ff */ /*0e60*/ IADD3 R15, R16, -0x1, RZ ; /* 0xffffffff100f7810 */ /* 0x000fc80007ffe0ff */ /*0e70*/ ISETP.GE.U32.AND P0, PT, R15, 0xfe, PT ; /* 0x000000fe0f00780c */ /* 0x000fda0003f06070 */ /*0e80*/ @!P0 BRA 0x1090 ; /* 0x0000020000008947 */ /* 0x000fea0003800000 */ /*0e90*/ ISETP.GT.AND P0, PT, R16, 0xfe, PT ; /* 0x000000fe1000780c */ /* 0x000fda0003f04270 */ /*0ea0*/ @P0 BRA 0x1060 ; /* 0x000001b000000947 */ /* 0x000fea0003800000 */ /*0eb0*/ ISETP.GE.AND P0, PT, R16, 0x1, PT ; /* 0x000000011000780c */ /* 0x000fda0003f06270 */ /*0ec0*/ @P0 BRA 0x10a0 ; /* 0x000001d000000947 */ /* 0x000fea0003800000 */ /*0ed0*/ ISETP.GE.AND P0, PT, R16, -0x18, PT ; /* 0xffffffe81000780c */ /* 0x000fe40003f06270 */ /*0ee0*/ LOP3.LUT R21, R21, 0x80000000, RZ, 0xc0, !PT ; /* 0x8000000015157812 */ /* 0x000fd600078ec0ff */ /*0ef0*/ @!P0 BRA 0x10a0 ; /* 0x000001a000008947 */ /* 0x000fea0003800000 */ /*0f00*/ FFMA.RZ R15, R17, R19.reuse, R20.reuse ; /* 0x00000013110f7223 */ /* 0x180fe2000000c014 */ /*0f10*/ ISETP.NE.AND P4, PT, R16.reuse, RZ, PT ; /* 0x000000ff1000720c */ /* 0x040fe40003f85270 */ /*0f20*/ ISETP.NE.AND P1, PT, R16.reuse, RZ, PT ; /* 0x000000ff1000720c */ /* 0x040fe40003f25270 */ /*0f30*/ LOP3.LUT R18, R15, 0x7fffff, RZ, 0xc0, !PT ; /* 0x007fffff0f127812 */ /* 0x000fe200078ec0ff */ /*0f40*/ FFMA.RP R15, R17.reuse, R19.reuse, R20.reuse ; /* 0x00000013110f7223 */ /* 0x1c0fe40000008014 */ /*0f50*/ FFMA.RM R20, R17, R19, R20 ; /* 0x0000001311147223 */ /* 0x000fe20000004014 */ /*0f60*/ IADD3 R17, R16, 0x20, RZ ; /* 0x0000002010117810 */ /* 0x000fe20007ffe0ff */ /*0f70*/ IMAD.MOV R16, RZ, RZ, -R16 ; /* 0x000000ffff107224 */ /* 0x000fe200078e0a10 */ /*0f80*/ LOP3.LUT R18, R18, 0x800000, RZ, 0xfc, !PT ; /* 0x0080000012127812 */ /* 0x000fc400078efcff */ /*0f90*/ FSETP.NEU.FTZ.AND P0, PT, R15, R20, PT ; /* 0x000000140f00720b */ /* 0x000fe40003f1d000 */ /*0fa0*/ SHF.L.U32 R17, R18, R17, RZ ; /* 0x0000001112117219 */ /* 0x000fe400000006ff */ /*0fb0*/ SEL R15, R16, RZ, P4 ; /* 0x000000ff100f7207 */ /* 0x000fe40002000000 */ /*0fc0*/ ISETP.NE.AND P1, PT, R17, RZ, P1 ; /* 0x000000ff1100720c */ /* 0x000fe40000f25270 */ /*0fd0*/ SHF.R.U32.HI R15, RZ, R15, R18 ; /* 0x0000000fff0f7219 */ /* 0x000fe40000011612 */ /*0fe0*/ PLOP3.LUT P0, PT, P0, P1, PT, 0xa8, 0x0 ; /* 0x000000000000781c */ /* 0x000fc40000703570 */ /*0ff0*/ SHF.R.U32.HI R17, RZ, 0x1, R15 ; /* 0x00000001ff117819 */ /* 0x000fe4000001160f */ /*1000*/ SEL R16, RZ, 0x1, !P0 ; /* 0x00000001ff107807 */ /* 0x000fc80004000000 */ /*1010*/ LOP3.LUT R16, R16, 0x1, R17, 0xf8, !PT ; /* 0x0000000110107812 */ /* 0x000fc800078ef811 */ /*1020*/ LOP3.LUT R16, R16, R15, RZ, 0xc0, !PT ; /* 0x0000000f10107212 */ /* 0x000fc800078ec0ff */ /*1030*/ IADD3 R16, R17, R16, RZ ; /* 0x0000001011107210 */ /* 0x000fc80007ffe0ff */ /*1040*/ LOP3.LUT R21, R16, R21, RZ, 0xfc, !PT ; /* 0x0000001510157212 */ /* 0x000fe200078efcff */ /*1050*/ BRA 0x10a0 ; /* 0x0000004000007947 */ /* 0x000fea0003800000 */ /*1060*/ LOP3.LUT R21, R21, 0x80000000, RZ, 0xc0, !PT ; /* 0x8000000015157812 */ /* 0x000fc800078ec0ff */ /*1070*/ LOP3.LUT R21, R21, 0x7f800000, RZ, 0xfc, !PT ; /* 0x7f80000015157812 */ /* 0x000fe200078efcff */ /*1080*/ BRA 0x10a0 ; /* 0x0000001000007947 */ /* 0x000fea0003800000 */ /*1090*/ IMAD R21, R18, 0x800000, R21 ; /* 0x0080000012157824 */ /* 0x000fe400078e0215 */ /*10a0*/ BSYNC B2 ; /* 0x0000000000027941 */ /* 0x000fea0003800000 */ /*10b0*/ BRA 0x1140 ; /* 0x0000008000007947 */ /* 0x000fea0003800000 */ /*10c0*/ LOP3.LUT R17, R20, 0x80000000, R17, 0x48, !PT ; /* 0x8000000014117812 */ /* 0x000fc800078e4811 */ /*10d0*/ LOP3.LUT R21, R17, 0x7f800000, RZ, 0xfc, !PT ; /* 0x7f80000011157812 */ /* 0x000fe200078efcff */ /*10e0*/ BRA 0x1140 ; /* 0x0000005000007947 */ /* 0x000fea0003800000 */ /*10f0*/ LOP3.LUT R21, R20, 0x80000000, R17, 0x48, !PT ; /* 0x8000000014157812 */ /* 0x000fe200078e4811 */ /*1100*/ BRA 0x1140 ; /* 0x0000003000007947 */ /* 0x000fea0003800000 */ /*1110*/ MUFU.RSQ R21, -QNAN ; /* 0xffc0000000157908 */ /* 0x000e220000001400 */ /*1120*/ BRA 0x1140 ; /* 0x0000001000007947 */ /* 0x000fea0003800000 */ /*1130*/ FADD.FTZ R21, R17, R11 ; /* 0x0000000b11157221 */ /* 0x000fe40000010000 */ /*1140*/ BSYNC B1 ; /* 0x0000000000017941 */ /* 0x000fea0003800000 */ /*1150*/ IMAD.MOV.U32 R15, RZ, RZ, 0x0 ; /* 0x00000000ff0f7424 */ /* 0x000fe200078e00ff */ /*1160*/ MOV R16, R21 ; /* 0x0000001500107202 */ /* 0x001fc60000000f00 */ /*1170*/ RET.REL.NODEC R14 0x0 ; /* 0xffffee800e007950 */ /* 0x000fea0003c3ffff */ /*1180*/ BRA 0x1180; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*1190*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*11a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*11b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*11c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*11d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*11e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*11f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1200*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1210*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1220*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1230*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1240*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1250*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1260*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1270*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
/* ============================================================================ Name : sem1.cu Author : maminov Version : Copyright : copyleft Description : CUDA compute reciprocals ============================================================================ */ #include <cuda_runtime.h> #include <device_launch_parameters.h> #include <iostream> #include <numeric> #include <stdlib.h> static void CheckCudaErrorAux(const char *, unsigned, const char *, cudaError_t); #define CUDA_CHECK_RETURN(value) CheckCudaErrorAux(__FILE__,__LINE__, #value, value) /** * CUDA kernel that computes reciprocal values for a given vector */ __global__ void heatKernel(float *data, unsigned size, unsigned iterations, float tempPerIteration, float dx, float dt) { unsigned idx = blockIdx.x*blockDim.x + threadIdx.x; if (idx < size) for (int i = 0; i < iterations; ++i) { if (idx == 0) data[size + 1] += tempPerIteration; data[idx + 1] += (data[idx - 1 + 1] - 2 * data[idx + 1] + data[idx + 1 + 1]) * dt / (dx*dx); __syncthreads(); } } void initialize(float *data, unsigned size) { for (unsigned i = 0; i < size; ++i) data[i] = 0.; } int main(void) { static const int SIM_TIME = 5; static const int STEPS_PER_SECOND = 10; static const int CYL_LENGTH = 10; static const int STEPS_PER_METER = 10; static const int WORK_SIZE = CYL_LENGTH * STEPS_PER_METER; static const int BLOCK_SIZE = 256; static const float TEMP_PER_SEC = 5.; float *hostData = new float[WORK_SIZE + 2]; float *gpuData; initialize(hostData, WORK_SIZE + 2); CUDA_CHECK_RETURN(cudaMalloc((void **)&gpuData, sizeof(float)*(WORK_SIZE + 2))); CUDA_CHECK_RETURN(cudaMemcpy(gpuData, hostData, sizeof(float)*(WORK_SIZE + 2), cudaMemcpyHostToDevice)); float *result = new float[WORK_SIZE + 2]; const int blockCount = (WORK_SIZE + BLOCK_SIZE - 1) / BLOCK_SIZE; heatKernel <<< blockCount, BLOCK_SIZE >>> (gpuData, WORK_SIZE, SIM_TIME * STEPS_PER_SECOND, TEMP_PER_SEC / (STEPS_PER_SECOND*1.0f), CYL_LENGTH / (STEPS_PER_METER*1.0f), SIM_TIME / (STEPS_PER_SECOND*1.0f)); CUDA_CHECK_RETURN(cudaMemcpy(result, gpuData, sizeof(float)*(WORK_SIZE + 2), cudaMemcpyDeviceToHost)); /* Verify the results */ for (int i = 0; i < WORK_SIZE; ++i) std::cout << result[i + 1] << std::endl; /* Free memory */ CUDA_CHECK_RETURN(cudaFree(gpuData)); delete[] hostData; delete[] result; return 0; } /** * Check the return value of the CUDA runtime API call and exit * the application if the call has failed. */ static void CheckCudaErrorAux(const char *file, unsigned line, const char *statement, cudaError_t err) { if (err == cudaSuccess) return; std::cerr << statement << " returned " << cudaGetErrorString(err) << "(" << err << ") at " << file << ":" << line << std::endl; exit(1); }
.file "tmpxft_00059100_00000000-6_kernel.cudafe1.cpp" .text #APP .globl _ZSt21ios_base_library_initv .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string " returned " .LC1: .string "(" .LC2: .string ") at " .LC3: .string ":" #NO_APP .text .type _ZL17CheckCudaErrorAuxPKcjS0_9cudaError, @function _ZL17CheckCudaErrorAuxPKcjS0_9cudaError: .LFB3710: .cfi_startproc testl %ecx, %ecx jne .L7 ret .L7: pushq %r13 .cfi_def_cfa_offset 16 .cfi_offset 13, -16 pushq %r12 .cfi_def_cfa_offset 24 .cfi_offset 12, -24 pushq %rbp .cfi_def_cfa_offset 32 .cfi_offset 6, -32 pushq %rbx .cfi_def_cfa_offset 40 .cfi_offset 3, -40 subq $8, %rsp .cfi_def_cfa_offset 48 movq %rdi, %rbp movl %esi, %r12d movl %ecx, %ebx movq %rdx, %rsi leaq _ZSt4cerr(%rip), %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi leaq .LC0(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %r13 movl %ebx, %edi call cudaGetErrorString@PLT movq %rax, %rsi movq %r13, %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi leaq .LC1(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi movl %ebx, %esi call _ZNSolsEi@PLT movq %rax, %rdi leaq .LC2(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi movq %rbp, %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi leaq .LC3(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi movl %r12d, %esi call _ZNSo9_M_insertImEERSoT_@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movl $1, %edi call exit@PLT .cfi_endproc .LFE3710: .size _ZL17CheckCudaErrorAuxPKcjS0_9cudaError, .-_ZL17CheckCudaErrorAuxPKcjS0_9cudaError .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB3713: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE3713: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z10initializePfj .type _Z10initializePfj, @function _Z10initializePfj: .LFB3708: .cfi_startproc endbr64 testl %esi, %esi je .L10 movq %rdi, %rax movl %esi, %esi leaq (%rdi,%rsi,4), %rdx .L12: movl $0x00000000, (%rax) addq $4, %rax cmpq %rdx, %rax jne .L12 .L10: ret .cfi_endproc .LFE3708: .size _Z10initializePfj, .-_Z10initializePfj .globl _Z35__device_stub__Z10heatKernelPfjjfffPfjjfff .type _Z35__device_stub__Z10heatKernelPfjjfffPfjjfff, @function _Z35__device_stub__Z10heatKernelPfjjfffPfjjfff: .LFB3735: .cfi_startproc endbr64 subq $168, %rsp .cfi_def_cfa_offset 176 movq %rdi, 24(%rsp) movl %esi, 20(%rsp) movl %edx, 16(%rsp) movss %xmm0, 12(%rsp) movss %xmm1, 8(%rsp) movss %xmm2, 4(%rsp) movq %fs:40, %rax movq %rax, 152(%rsp) xorl %eax, %eax leaq 24(%rsp), %rax movq %rax, 96(%rsp) leaq 20(%rsp), %rax movq %rax, 104(%rsp) leaq 16(%rsp), %rax movq %rax, 112(%rsp) leaq 12(%rsp), %rax movq %rax, 120(%rsp) leaq 8(%rsp), %rax movq %rax, 128(%rsp) leaq 4(%rsp), %rax movq %rax, 136(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 56(%rsp) movl $1, 60(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) leaq 40(%rsp), %rcx leaq 32(%rsp), %rdx leaq 60(%rsp), %rsi leaq 48(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L18 .L14: movq 152(%rsp), %rax subq %fs:40, %rax jne .L19 addq $168, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L18: .cfi_restore_state pushq 40(%rsp) .cfi_def_cfa_offset 184 pushq 40(%rsp) .cfi_def_cfa_offset 192 leaq 112(%rsp), %r9 movq 76(%rsp), %rcx movl 84(%rsp), %r8d movq 64(%rsp), %rsi movl 72(%rsp), %edx leaq _Z10heatKernelPfjjfff(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 176 jmp .L14 .L19: call __stack_chk_fail@PLT .cfi_endproc .LFE3735: .size _Z35__device_stub__Z10heatKernelPfjjfffPfjjfff, .-_Z35__device_stub__Z10heatKernelPfjjfffPfjjfff .globl _Z10heatKernelPfjjfff .type _Z10heatKernelPfjjfff, @function _Z10heatKernelPfjjfff: .LFB3736: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z35__device_stub__Z10heatKernelPfjjfffPfjjfff addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE3736: .size _Z10heatKernelPfjjfff, .-_Z10heatKernelPfjjfff .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC5: .string "cudaMalloc((void **)&gpuData, sizeof(float)*(WORK_SIZE + 2))" .align 8 .LC6: .string "/home/ubuntu/Datasets/stackv2/train-structured/xobotun/TP_CUDA/master/Sem1/Sem1/kernel.cu" .align 8 .LC7: .string "cudaMemcpy(gpuData, hostData, sizeof(float)*(WORK_SIZE + 2), cudaMemcpyHostToDevice)" .align 8 .LC10: .string "cudaMemcpy(result, gpuData, sizeof(float)*(WORK_SIZE + 2), cudaMemcpyDeviceToHost)" .section .rodata.str1.1 .LC11: .string "cudaFree(gpuData)" .text .globl main .type main, @function main: .LFB3709: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $72, %rsp .cfi_def_cfa_offset 128 movq %fs:40, %rax movq %rax, 56(%rsp) xorl %eax, %eax movl $408, %edi call _Znam@PLT movq %rax, %r15 movq %rax, 8(%rsp) movl $102, %esi movq %rax, %rdi call _Z10initializePfj leaq 24(%rsp), %rdi movl $408, %esi call cudaMalloc@PLT movl %eax, %ecx leaq .LC5(%rip), %rdx movl $58, %esi leaq .LC6(%rip), %rbx movq %rbx, %rdi call _ZL17CheckCudaErrorAuxPKcjS0_9cudaError movl $1, %ecx movl $408, %edx movq %r15, %rsi movq 24(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %ecx leaq .LC7(%rip), %rdx movl $59, %esi movq %rbx, %rdi call _ZL17CheckCudaErrorAuxPKcjS0_9cudaError movl $408, %edi call _Znam@PLT movq %rax, %r15 movl $256, 44(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) movl $1, 40(%rsp) movl $0, %r9d movl $0, %r8d movq 44(%rsp), %rdx movl $1, %ecx movq 32(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L32 .L23: movl $2, %ecx movl $408, %edx movq 24(%rsp), %rsi movq %r15, %rdi call cudaMemcpy@PLT movl %eax, %ecx leaq .LC10(%rip), %rdx movl $66, %esi leaq .LC6(%rip), %rdi call _ZL17CheckCudaErrorAuxPKcjS0_9cudaError leaq 4(%r15), %r12 leaq 404(%r15), %r14 leaq _ZSt4cout(%rip), %r13 jmp .L28 .L32: movss .LC8(%rip), %xmm2 movss .LC9(%rip), %xmm1 movaps %xmm2, %xmm0 movl $50, %edx movl $100, %esi movq 24(%rsp), %rdi call _Z35__device_stub__Z10heatKernelPfjjfffPfjjfff jmp .L23 .L35: movq 56(%rsp), %rax subq %fs:40, %rax jne .L33 call _ZSt16__throw_bad_castv@PLT .L33: call __stack_chk_fail@PLT .L36: movzbl 67(%rbp), %esi .L27: movsbl %sil, %esi movq %rbx, %rdi call _ZNSo3putEc@PLT movq %rax, %rdi call _ZNSo5flushEv@PLT addq $4, %r12 cmpq %r14, %r12 je .L34 .L28: pxor %xmm0, %xmm0 cvtss2sd (%r12), %xmm0 movq %r13, %rdi call _ZNSo9_M_insertIdEERSoT_@PLT movq %rax, %rbx movq (%rax), %rax movq -24(%rax), %rax movq 240(%rbx,%rax), %rbp testq %rbp, %rbp je .L35 cmpb $0, 56(%rbp) jne .L36 movq %rbp, %rdi call _ZNKSt5ctypeIcE13_M_widen_initEv@PLT movq 0(%rbp), %rax movl $10, %esi movq %rbp, %rdi call *48(%rax) movl %eax, %esi jmp .L27 .L34: movq 24(%rsp), %rdi call cudaFree@PLT movl %eax, %ecx leaq .LC11(%rip), %rdx movl $74, %esi leaq .LC6(%rip), %rdi call _ZL17CheckCudaErrorAuxPKcjS0_9cudaError movq 8(%rsp), %rdi call _ZdaPv@PLT movq %r15, %rdi call _ZdaPv@PLT movq 56(%rsp), %rax subq %fs:40, %rax jne .L37 movl $0, %eax addq $72, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L37: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE3709: .size main, .-main .section .rodata.str1.1 .LC12: .string "_Z10heatKernelPfjjfff" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB3738: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC12(%rip), %rdx movq %rdx, %rcx leaq _Z10heatKernelPfjjfff(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE3738: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .section .rodata.cst4,"aM",@progbits,4 .align 4 .LC8: .long 1056964608 .align 4 .LC9: .long 1065353216 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
/* ============================================================================ Name : sem1.cu Author : maminov Version : Copyright : copyleft Description : CUDA compute reciprocals ============================================================================ */ #include <cuda_runtime.h> #include <device_launch_parameters.h> #include <iostream> #include <numeric> #include <stdlib.h> static void CheckCudaErrorAux(const char *, unsigned, const char *, cudaError_t); #define CUDA_CHECK_RETURN(value) CheckCudaErrorAux(__FILE__,__LINE__, #value, value) /** * CUDA kernel that computes reciprocal values for a given vector */ __global__ void heatKernel(float *data, unsigned size, unsigned iterations, float tempPerIteration, float dx, float dt) { unsigned idx = blockIdx.x*blockDim.x + threadIdx.x; if (idx < size) for (int i = 0; i < iterations; ++i) { if (idx == 0) data[size + 1] += tempPerIteration; data[idx + 1] += (data[idx - 1 + 1] - 2 * data[idx + 1] + data[idx + 1 + 1]) * dt / (dx*dx); __syncthreads(); } } void initialize(float *data, unsigned size) { for (unsigned i = 0; i < size; ++i) data[i] = 0.; } int main(void) { static const int SIM_TIME = 5; static const int STEPS_PER_SECOND = 10; static const int CYL_LENGTH = 10; static const int STEPS_PER_METER = 10; static const int WORK_SIZE = CYL_LENGTH * STEPS_PER_METER; static const int BLOCK_SIZE = 256; static const float TEMP_PER_SEC = 5.; float *hostData = new float[WORK_SIZE + 2]; float *gpuData; initialize(hostData, WORK_SIZE + 2); CUDA_CHECK_RETURN(cudaMalloc((void **)&gpuData, sizeof(float)*(WORK_SIZE + 2))); CUDA_CHECK_RETURN(cudaMemcpy(gpuData, hostData, sizeof(float)*(WORK_SIZE + 2), cudaMemcpyHostToDevice)); float *result = new float[WORK_SIZE + 2]; const int blockCount = (WORK_SIZE + BLOCK_SIZE - 1) / BLOCK_SIZE; heatKernel <<< blockCount, BLOCK_SIZE >>> (gpuData, WORK_SIZE, SIM_TIME * STEPS_PER_SECOND, TEMP_PER_SEC / (STEPS_PER_SECOND*1.0f), CYL_LENGTH / (STEPS_PER_METER*1.0f), SIM_TIME / (STEPS_PER_SECOND*1.0f)); CUDA_CHECK_RETURN(cudaMemcpy(result, gpuData, sizeof(float)*(WORK_SIZE + 2), cudaMemcpyDeviceToHost)); /* Verify the results */ for (int i = 0; i < WORK_SIZE; ++i) std::cout << result[i + 1] << std::endl; /* Free memory */ CUDA_CHECK_RETURN(cudaFree(gpuData)); delete[] hostData; delete[] result; return 0; } /** * Check the return value of the CUDA runtime API call and exit * the application if the call has failed. */ static void CheckCudaErrorAux(const char *file, unsigned line, const char *statement, cudaError_t err) { if (err == cudaSuccess) return; std::cerr << statement << " returned " << cudaGetErrorString(err) << "(" << err << ") at " << file << ":" << line << std::endl; exit(1); }
/* ============================================================================ Name : sem1.cu Author : maminov Version : Copyright : copyleft Description : CUDA compute reciprocals ============================================================================ */ #include <hip/hip_runtime.h> #include <iostream> #include <numeric> #include <stdlib.h> static void CheckCudaErrorAux(const char *, unsigned, const char *, hipError_t); #define CUDA_CHECK_RETURN(value) CheckCudaErrorAux(__FILE__,__LINE__, #value, value) /** * CUDA kernel that computes reciprocal values for a given vector */ __global__ void heatKernel(float *data, unsigned size, unsigned iterations, float tempPerIteration, float dx, float dt) { unsigned idx = blockIdx.x*blockDim.x + threadIdx.x; if (idx < size) for (int i = 0; i < iterations; ++i) { if (idx == 0) data[size + 1] += tempPerIteration; data[idx + 1] += (data[idx - 1 + 1] - 2 * data[idx + 1] + data[idx + 1 + 1]) * dt / (dx*dx); __syncthreads(); } } void initialize(float *data, unsigned size) { for (unsigned i = 0; i < size; ++i) data[i] = 0.; } int main(void) { static const int SIM_TIME = 5; static const int STEPS_PER_SECOND = 10; static const int CYL_LENGTH = 10; static const int STEPS_PER_METER = 10; static const int WORK_SIZE = CYL_LENGTH * STEPS_PER_METER; static const int BLOCK_SIZE = 256; static const float TEMP_PER_SEC = 5.; float *hostData = new float[WORK_SIZE + 2]; float *gpuData; initialize(hostData, WORK_SIZE + 2); CUDA_CHECK_RETURN(hipMalloc((void **)&gpuData, sizeof(float)*(WORK_SIZE + 2))); CUDA_CHECK_RETURN(hipMemcpy(gpuData, hostData, sizeof(float)*(WORK_SIZE + 2), hipMemcpyHostToDevice)); float *result = new float[WORK_SIZE + 2]; const int blockCount = (WORK_SIZE + BLOCK_SIZE - 1) / BLOCK_SIZE; heatKernel <<< blockCount, BLOCK_SIZE >>> (gpuData, WORK_SIZE, SIM_TIME * STEPS_PER_SECOND, TEMP_PER_SEC / (STEPS_PER_SECOND*1.0f), CYL_LENGTH / (STEPS_PER_METER*1.0f), SIM_TIME / (STEPS_PER_SECOND*1.0f)); CUDA_CHECK_RETURN(hipMemcpy(result, gpuData, sizeof(float)*(WORK_SIZE + 2), hipMemcpyDeviceToHost)); /* Verify the results */ for (int i = 0; i < WORK_SIZE; ++i) std::cout << result[i + 1] << std::endl; /* Free memory */ CUDA_CHECK_RETURN(hipFree(gpuData)); delete[] hostData; delete[] result; return 0; } /** * Check the return value of the CUDA runtime API call and exit * the application if the call has failed. */ static void CheckCudaErrorAux(const char *file, unsigned line, const char *statement, hipError_t err) { if (err == hipSuccess) return; std::cerr << statement << " returned " << hipGetErrorString(err) << "(" << err << ") at " << file << ":" << line << std::endl; exit(1); }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
/* ============================================================================ Name : sem1.cu Author : maminov Version : Copyright : copyleft Description : CUDA compute reciprocals ============================================================================ */ #include <hip/hip_runtime.h> #include <iostream> #include <numeric> #include <stdlib.h> static void CheckCudaErrorAux(const char *, unsigned, const char *, hipError_t); #define CUDA_CHECK_RETURN(value) CheckCudaErrorAux(__FILE__,__LINE__, #value, value) /** * CUDA kernel that computes reciprocal values for a given vector */ __global__ void heatKernel(float *data, unsigned size, unsigned iterations, float tempPerIteration, float dx, float dt) { unsigned idx = blockIdx.x*blockDim.x + threadIdx.x; if (idx < size) for (int i = 0; i < iterations; ++i) { if (idx == 0) data[size + 1] += tempPerIteration; data[idx + 1] += (data[idx - 1 + 1] - 2 * data[idx + 1] + data[idx + 1 + 1]) * dt / (dx*dx); __syncthreads(); } } void initialize(float *data, unsigned size) { for (unsigned i = 0; i < size; ++i) data[i] = 0.; } int main(void) { static const int SIM_TIME = 5; static const int STEPS_PER_SECOND = 10; static const int CYL_LENGTH = 10; static const int STEPS_PER_METER = 10; static const int WORK_SIZE = CYL_LENGTH * STEPS_PER_METER; static const int BLOCK_SIZE = 256; static const float TEMP_PER_SEC = 5.; float *hostData = new float[WORK_SIZE + 2]; float *gpuData; initialize(hostData, WORK_SIZE + 2); CUDA_CHECK_RETURN(hipMalloc((void **)&gpuData, sizeof(float)*(WORK_SIZE + 2))); CUDA_CHECK_RETURN(hipMemcpy(gpuData, hostData, sizeof(float)*(WORK_SIZE + 2), hipMemcpyHostToDevice)); float *result = new float[WORK_SIZE + 2]; const int blockCount = (WORK_SIZE + BLOCK_SIZE - 1) / BLOCK_SIZE; heatKernel <<< blockCount, BLOCK_SIZE >>> (gpuData, WORK_SIZE, SIM_TIME * STEPS_PER_SECOND, TEMP_PER_SEC / (STEPS_PER_SECOND*1.0f), CYL_LENGTH / (STEPS_PER_METER*1.0f), SIM_TIME / (STEPS_PER_SECOND*1.0f)); CUDA_CHECK_RETURN(hipMemcpy(result, gpuData, sizeof(float)*(WORK_SIZE + 2), hipMemcpyDeviceToHost)); /* Verify the results */ for (int i = 0; i < WORK_SIZE; ++i) std::cout << result[i + 1] << std::endl; /* Free memory */ CUDA_CHECK_RETURN(hipFree(gpuData)); delete[] hostData; delete[] result; return 0; } /** * Check the return value of the CUDA runtime API call and exit * the application if the call has failed. */ static void CheckCudaErrorAux(const char *file, unsigned line, const char *statement, hipError_t err) { if (err == hipSuccess) return; std::cerr << statement << " returned " << hipGetErrorString(err) << "(" << err << ") at " << file << ":" << line << std::endl; exit(1); }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z10heatKernelPfjjfff .globl _Z10heatKernelPfjjfff .p2align 8 .type _Z10heatKernelPfjjfff,@function _Z10heatKernelPfjjfff: s_clause 0x1 s_load_b32 s2, s[0:1], 0x2c s_load_b64 s[4:5], s[0:1], 0x8 s_mov_b32 s9, 0 s_waitcnt lgkmcnt(0) s_and_b32 s2, s2, 0xffff s_cmp_lg_u32 s5, 0 v_mad_u64_u32 v[1:2], null, s15, s2, v[0:1] s_cselect_b32 s2, -1, 0 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(SALU_CYCLE_1) v_cmp_gt_u32_e32 vcc_lo, s4, v1 s_and_b32 s2, vcc_lo, s2 s_and_saveexec_b32 s3, s2 s_cbranch_execz .LBB0_5 s_clause 0x2 s_load_b64 s[10:11], s[0:1], 0x0 s_load_b64 s[6:7], s[0:1], 0x10 s_load_b32 s3, s[0:1], 0x18 v_dual_mov_b32 v2, 0 :: v_dual_add_nc_u32 v3, 1, v1 v_cmp_eq_u32_e64 s2, 0, v1 s_add_i32 s8, s4, 1 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_3) | instid1(VALU_DEP_3) v_mov_b32_e32 v4, v2 v_lshlrev_b64 v[5:6], 2, v[1:2] v_add_nc_u32_e32 v1, 2, v1 s_lshl_b64 s[0:1], s[8:9], 2 v_lshlrev_b64 v[3:4], 2, v[3:4] s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_3) | instid1(VALU_DEP_4) v_lshlrev_b64 v[7:8], 2, v[1:2] s_waitcnt lgkmcnt(0) v_add_co_u32 v0, vcc_lo, s10, v5 v_add_co_ci_u32_e32 v1, vcc_lo, s11, v6, vcc_lo v_add_co_u32 v2, vcc_lo, s10, v3 v_add_co_ci_u32_e32 v3, vcc_lo, s11, v4, vcc_lo v_add_co_u32 v4, vcc_lo, s10, v7 v_add_co_ci_u32_e32 v5, vcc_lo, s11, v8, vcc_lo v_mul_f32_e64 v6, s7, s7 v_mov_b32_e32 v7, 0 s_add_u32 s0, s10, s0 s_addc_u32 s1, s11, s1 s_set_inst_prefetch_distance 0x1 s_branch .LBB0_3 .p2align 6 .LBB0_2: s_or_b32 exec_lo, exec_lo, s4 s_clause 0x2 global_load_b32 v8, v[0:1], off global_load_b32 v9, v[2:3], off global_load_b32 v10, v[4:5], off s_add_i32 s5, s5, -1 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_3) | instid1(VALU_DEP_1) s_cmp_lg_u32 s5, 0 s_waitcnt vmcnt(1) v_fmac_f32_e32 v8, -2.0, v9 s_waitcnt vmcnt(0) v_add_f32_e32 v8, v10, v8 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_f32_e32 v8, s3, v8 v_div_scale_f32 v10, null, v6, v6, v8 v_div_scale_f32 v13, vcc_lo, v8, v6, v8 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_2) | instid1(VALU_DEP_1) v_rcp_f32_e32 v11, v10 s_waitcnt_depctr 0xfff v_fma_f32 v12, -v10, v11, 1.0 v_fmac_f32_e32 v11, v12, v11 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_f32_e32 v12, v13, v11 v_fma_f32 v14, -v10, v12, v13 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmac_f32_e32 v12, v14, v11 v_fma_f32 v10, -v10, v12, v13 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_div_fmas_f32 v10, v10, v11, v12 v_div_fixup_f32 v8, v10, v6, v8 s_delay_alu instid0(VALU_DEP_1) v_add_f32_e32 v8, v9, v8 global_store_b32 v[2:3], v8, off s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv s_cbranch_scc0 .LBB0_5 .LBB0_3: s_and_saveexec_b32 s4, s2 s_cbranch_execz .LBB0_2 global_load_b32 v8, v7, s[0:1] s_waitcnt vmcnt(0) v_add_f32_e32 v8, s6, v8 global_store_b32 v7, v8, s[0:1] s_branch .LBB0_2 .LBB0_5: s_set_inst_prefetch_distance 0x2 s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z10heatKernelPfjjfff .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 288 .amdhsa_user_sgpr_count 15 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 0 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 0 .amdhsa_next_free_vgpr 15 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z10heatKernelPfjjfff, .Lfunc_end0-_Z10heatKernelPfjjfff .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .offset: 8 .size: 4 .value_kind: by_value - .offset: 12 .size: 4 .value_kind: by_value - .offset: 16 .size: 4 .value_kind: by_value - .offset: 20 .size: 4 .value_kind: by_value - .offset: 24 .size: 4 .value_kind: by_value - .offset: 32 .size: 4 .value_kind: hidden_block_count_x - .offset: 36 .size: 4 .value_kind: hidden_block_count_y - .offset: 40 .size: 4 .value_kind: hidden_block_count_z - .offset: 44 .size: 2 .value_kind: hidden_group_size_x - .offset: 46 .size: 2 .value_kind: hidden_group_size_y - .offset: 48 .size: 2 .value_kind: hidden_group_size_z - .offset: 50 .size: 2 .value_kind: hidden_remainder_x - .offset: 52 .size: 2 .value_kind: hidden_remainder_y - .offset: 54 .size: 2 .value_kind: hidden_remainder_z - .offset: 72 .size: 8 .value_kind: hidden_global_offset_x - .offset: 80 .size: 8 .value_kind: hidden_global_offset_y - .offset: 88 .size: 8 .value_kind: hidden_global_offset_z - .offset: 96 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 288 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z10heatKernelPfjjfff .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z10heatKernelPfjjfff.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 15 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
/* ============================================================================ Name : sem1.cu Author : maminov Version : Copyright : copyleft Description : CUDA compute reciprocals ============================================================================ */ #include <hip/hip_runtime.h> #include <iostream> #include <numeric> #include <stdlib.h> static void CheckCudaErrorAux(const char *, unsigned, const char *, hipError_t); #define CUDA_CHECK_RETURN(value) CheckCudaErrorAux(__FILE__,__LINE__, #value, value) /** * CUDA kernel that computes reciprocal values for a given vector */ __global__ void heatKernel(float *data, unsigned size, unsigned iterations, float tempPerIteration, float dx, float dt) { unsigned idx = blockIdx.x*blockDim.x + threadIdx.x; if (idx < size) for (int i = 0; i < iterations; ++i) { if (idx == 0) data[size + 1] += tempPerIteration; data[idx + 1] += (data[idx - 1 + 1] - 2 * data[idx + 1] + data[idx + 1 + 1]) * dt / (dx*dx); __syncthreads(); } } void initialize(float *data, unsigned size) { for (unsigned i = 0; i < size; ++i) data[i] = 0.; } int main(void) { static const int SIM_TIME = 5; static const int STEPS_PER_SECOND = 10; static const int CYL_LENGTH = 10; static const int STEPS_PER_METER = 10; static const int WORK_SIZE = CYL_LENGTH * STEPS_PER_METER; static const int BLOCK_SIZE = 256; static const float TEMP_PER_SEC = 5.; float *hostData = new float[WORK_SIZE + 2]; float *gpuData; initialize(hostData, WORK_SIZE + 2); CUDA_CHECK_RETURN(hipMalloc((void **)&gpuData, sizeof(float)*(WORK_SIZE + 2))); CUDA_CHECK_RETURN(hipMemcpy(gpuData, hostData, sizeof(float)*(WORK_SIZE + 2), hipMemcpyHostToDevice)); float *result = new float[WORK_SIZE + 2]; const int blockCount = (WORK_SIZE + BLOCK_SIZE - 1) / BLOCK_SIZE; heatKernel <<< blockCount, BLOCK_SIZE >>> (gpuData, WORK_SIZE, SIM_TIME * STEPS_PER_SECOND, TEMP_PER_SEC / (STEPS_PER_SECOND*1.0f), CYL_LENGTH / (STEPS_PER_METER*1.0f), SIM_TIME / (STEPS_PER_SECOND*1.0f)); CUDA_CHECK_RETURN(hipMemcpy(result, gpuData, sizeof(float)*(WORK_SIZE + 2), hipMemcpyDeviceToHost)); /* Verify the results */ for (int i = 0; i < WORK_SIZE; ++i) std::cout << result[i + 1] << std::endl; /* Free memory */ CUDA_CHECK_RETURN(hipFree(gpuData)); delete[] hostData; delete[] result; return 0; } /** * Check the return value of the CUDA runtime API call and exit * the application if the call has failed. */ static void CheckCudaErrorAux(const char *file, unsigned line, const char *statement, hipError_t err) { if (err == hipSuccess) return; std::cerr << statement << " returned " << hipGetErrorString(err) << "(" << err << ") at " << file << ":" << line << std::endl; exit(1); }
.text .file "kernel.hip" # Start of file scope inline assembly .globl _ZSt21ios_base_library_initv # End of file scope inline assembly .globl _Z25__device_stub__heatKernelPfjjfff # -- Begin function _Z25__device_stub__heatKernelPfjjfff .p2align 4, 0x90 .type _Z25__device_stub__heatKernelPfjjfff,@function _Z25__device_stub__heatKernelPfjjfff: # @_Z25__device_stub__heatKernelPfjjfff .cfi_startproc # %bb.0: subq $136, %rsp .cfi_def_cfa_offset 144 movq %rdi, 72(%rsp) movl %esi, 20(%rsp) movl %edx, 16(%rsp) movss %xmm0, 12(%rsp) movss %xmm1, 8(%rsp) movss %xmm2, 4(%rsp) leaq 72(%rsp), %rax movq %rax, 80(%rsp) leaq 20(%rsp), %rax movq %rax, 88(%rsp) leaq 16(%rsp), %rax movq %rax, 96(%rsp) leaq 12(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) leaq 4(%rsp), %rax movq %rax, 120(%rsp) leaq 56(%rsp), %rdi leaq 40(%rsp), %rsi leaq 32(%rsp), %rdx leaq 24(%rsp), %rcx callq __hipPopCallConfiguration movq 56(%rsp), %rsi movl 64(%rsp), %edx movq 40(%rsp), %rcx movl 48(%rsp), %r8d leaq 80(%rsp), %r9 movl $_Z10heatKernelPfjjfff, %edi pushq 24(%rsp) .cfi_adjust_cfa_offset 8 pushq 40(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $152, %rsp .cfi_adjust_cfa_offset -152 retq .Lfunc_end0: .size _Z25__device_stub__heatKernelPfjjfff, .Lfunc_end0-_Z25__device_stub__heatKernelPfjjfff .cfi_endproc # -- End function .globl _Z10initializePfj # -- Begin function _Z10initializePfj .p2align 4, 0x90 .type _Z10initializePfj,@function _Z10initializePfj: # @_Z10initializePfj .cfi_startproc # %bb.0: testl %esi, %esi je .LBB1_1 # %bb.2: # %.lr.ph.preheader movl %esi, %edx shlq $2, %rdx xorl %esi, %esi jmp memset@PLT # TAILCALL .LBB1_1: # %._crit_edge retq .Lfunc_end1: .size _Z10initializePfj, .Lfunc_end1-_Z10initializePfj .cfi_endproc # -- End function .globl main # -- Begin function main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %r15 .cfi_def_cfa_offset 16 pushq %r14 .cfi_def_cfa_offset 24 pushq %r13 .cfi_def_cfa_offset 32 pushq %r12 .cfi_def_cfa_offset 40 pushq %rbx .cfi_def_cfa_offset 48 subq $144, %rsp .cfi_def_cfa_offset 192 .cfi_offset %rbx, -48 .cfi_offset %r12, -40 .cfi_offset %r13, -32 .cfi_offset %r14, -24 .cfi_offset %r15, -16 movl $408, %edi # imm = 0x198 callq _Znam movq %rax, %rbx movl $408, %edx # imm = 0x198 movq %rax, %rdi xorl %esi, %esi callq memset@PLT leaq 8(%rsp), %rdi movl $408, %esi # imm = 0x198 callq hipMalloc movl $.L.str.1, %esi movl $58, %edi movl %eax, %edx callq _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t movq 8(%rsp), %rdi movl $408, %edx # imm = 0x198 movq %rbx, %rsi movl $1, %ecx callq hipMemcpy movl $.L.str.2, %esi movl $59, %edi movl %eax, %edx callq _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t movl $408, %edi # imm = 0x198 callq _Znam movq %rax, %r14 movabsq $4294967297, %rdi # imm = 0x100000001 leaq 255(%rdi), %rdx movl $1, %esi movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB2_2 # %bb.1: movq 8(%rsp), %rax movq %rax, 88(%rsp) movl $100, 36(%rsp) movl $50, 32(%rsp) movl $1056964608, 28(%rsp) # imm = 0x3F000000 movl $1065353216, 24(%rsp) # imm = 0x3F800000 movl $1056964608, 20(%rsp) # imm = 0x3F000000 leaq 88(%rsp), %rax movq %rax, 96(%rsp) leaq 36(%rsp), %rax movq %rax, 104(%rsp) leaq 32(%rsp), %rax movq %rax, 112(%rsp) leaq 28(%rsp), %rax movq %rax, 120(%rsp) leaq 24(%rsp), %rax movq %rax, 128(%rsp) leaq 20(%rsp), %rax movq %rax, 136(%rsp) leaq 72(%rsp), %rdi leaq 56(%rsp), %rsi leaq 48(%rsp), %rdx leaq 40(%rsp), %rcx callq __hipPopCallConfiguration movq 72(%rsp), %rsi movl 80(%rsp), %edx movq 56(%rsp), %rcx movl 64(%rsp), %r8d leaq 96(%rsp), %r9 movl $_Z10heatKernelPfjjfff, %edi pushq 40(%rsp) .cfi_adjust_cfa_offset 8 pushq 56(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB2_2: movq 8(%rsp), %rsi movl $408, %edx # imm = 0x198 movq %r14, %rdi movl $2, %ecx callq hipMemcpy movl $.L.str.3, %esi movl $66, %edi movl %eax, %edx callq _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t movl $1, %r12d jmp .LBB2_3 .p2align 4, 0x90 .LBB2_6: # in Loop: Header=BB2_3 Depth=1 movq %r15, %rdi movq %rax, %r13 callq _ZNKSt5ctypeIcE13_M_widen_initEv movq (%r15), %rax movq %r15, %rdi movl $10, %esi callq *48(%rax) movl %eax, %ecx movq %r13, %rax .LBB2_7: # %_ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_.exit # in Loop: Header=BB2_3 Depth=1 movsbl %cl, %esi movq %rax, %rdi callq _ZNSo3putEc movq %rax, %rdi callq _ZNSo5flushEv incq %r12 cmpq $101, %r12 je .LBB2_8 .LBB2_3: # =>This Inner Loop Header: Depth=1 movss (%r14,%r12,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $_ZSt4cout, %edi callq _ZNSo9_M_insertIdEERSoT_ movq (%rax), %rcx movq -24(%rcx), %rcx movq 240(%rax,%rcx), %r15 testq %r15, %r15 je .LBB2_9 # %bb.4: # %_ZSt13__check_facetISt5ctypeIcEERKT_PS3_.exit.i.i # in Loop: Header=BB2_3 Depth=1 cmpb $0, 56(%r15) je .LBB2_6 # %bb.5: # in Loop: Header=BB2_3 Depth=1 movzbl 67(%r15), %ecx jmp .LBB2_7 .LBB2_8: movq 8(%rsp), %rdi callq hipFree movl $.L.str.4, %esi movl $74, %edi movl %eax, %edx callq _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t movq %rbx, %rdi callq _ZdaPv movq %r14, %rdi callq _ZdaPv xorl %eax, %eax addq $144, %rsp .cfi_def_cfa_offset 48 popq %rbx .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 retq .LBB2_9: .cfi_def_cfa_offset 192 callq _ZSt16__throw_bad_castv .Lfunc_end2: .size main, .Lfunc_end2-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t .type _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t,@function _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t: # @_ZL17CheckCudaErrorAuxPKcjS0_10hipError_t .cfi_startproc # %bb.0: testl %edx, %edx jne .LBB3_2 # %bb.1: retq .LBB3_2: pushq %rbp .cfi_def_cfa_offset 16 pushq %r14 .cfi_def_cfa_offset 24 pushq %rbx .cfi_def_cfa_offset 32 .cfi_offset %rbx, -32 .cfi_offset %r14, -24 .cfi_offset %rbp, -16 movl %edi, %ebx movl $_ZSt4cerr, %edi movl %edx, %ebp callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movl $.L.str.5, %esi movq %rax, %rdi callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movq %rax, %r14 movl %ebp, %edi callq hipGetErrorString movq %r14, %rdi movq %rax, %rsi callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movl $.L.str.6, %esi movq %rax, %rdi callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movq %rax, %rdi movl %ebp, %esi callq _ZNSolsEi movl $.L.str.7, %esi movq %rax, %rdi callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movl $.L.str, %esi movq %rax, %rdi callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movl $.L.str.8, %esi movq %rax, %rdi callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movq %rax, %rdi movl %ebx, %esi callq _ZNSolsEj movq %rax, %rdi callq _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_ movl $1, %edi callq exit .Lfunc_end3: .size _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t, .Lfunc_end3-_ZL17CheckCudaErrorAuxPKcjS0_10hipError_t .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB4_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB4_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z10heatKernelPfjjfff, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end4: .size __hip_module_ctor, .Lfunc_end4-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB5_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB5_2: retq .Lfunc_end5: .size __hip_module_dtor, .Lfunc_end5-__hip_module_dtor .cfi_endproc # -- End function .type _Z10heatKernelPfjjfff,@object # @_Z10heatKernelPfjjfff .section .rodata,"a",@progbits .globl _Z10heatKernelPfjjfff .p2align 3, 0x0 _Z10heatKernelPfjjfff: .quad _Z25__device_stub__heatKernelPfjjfff .size _Z10heatKernelPfjjfff, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "/home/ubuntu/Datasets/stackv2/train-structured-repos-hip/xobotun/TP_CUDA/master/Sem1/Sem1/kernel.hip" .size .L.str, 101 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "hipMalloc((void **)&gpuData, sizeof(float)*(WORK_SIZE + 2))" .size .L.str.1, 60 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "hipMemcpy(gpuData, hostData, sizeof(float)*(WORK_SIZE + 2), hipMemcpyHostToDevice)" .size .L.str.2, 83 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "hipMemcpy(result, gpuData, sizeof(float)*(WORK_SIZE + 2), hipMemcpyDeviceToHost)" .size .L.str.3, 81 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "hipFree(gpuData)" .size .L.str.4, 17 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz " returned " .size .L.str.5, 11 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz "(" .size .L.str.6, 2 .type .L.str.7,@object # @.str.7 .L.str.7: .asciz ") at " .size .L.str.7, 6 .type .L.str.8,@object # @.str.8 .L.str.8: .asciz ":" .size .L.str.8, 2 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z10heatKernelPfjjfff" .size .L__unnamed_1, 22 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z25__device_stub__heatKernelPfjjfff .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z10heatKernelPfjjfff .addrsig_sym _ZSt4cout .addrsig_sym _ZSt4cerr .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA device assembly to AMD device assembly.
code for sm_80 Function : _Z10heatKernelPfjjfff .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R0, SR_CTAID.X ; /* 0x0000000000007919 */ /* 0x000e220000002500 */ /*0020*/ ISETP.NE.AND P0, PT, RZ, c[0x0][0x16c], PT ; /* 0x00005b00ff007a0c */ /* 0x000fc60003f05270 */ /*0030*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e240000002100 */ /*0040*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x001fca00078e0203 */ /*0050*/ ISETP.GE.U32.OR P0, PT, R0, c[0x0][0x168], !P0 ; /* 0x00005a0000007a0c */ /* 0x000fda0004706470 */ /*0060*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0070*/ MOV R10, c[0x0][0x16c] ; /* 0x00005b00000a7a02 */ /* 0x000fe20000000f00 */ /*0080*/ IMAD.MOV.U32 R2, RZ, RZ, 0x1 ; /* 0x00000001ff027424 */ /* 0x000fe200078e00ff */ /*0090*/ IADD3 R8, R0, 0x1, RZ ; /* 0x0000000100087810 */ /* 0x000fe20007ffe0ff */ /*00a0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*00b0*/ IADD3 R3, R10, -0x1, RZ ; /* 0xffffffff0a037810 */ /* 0x000fe40007ffe0ff */ /*00c0*/ IADD3 R6, R0, 0x2, RZ ; /* 0x0000000200067810 */ /* 0x000fe40007ffe0ff */ /*00d0*/ ISETP.GE.U32.AND P0, PT, R3, 0x3, PT ; /* 0x000000030300780c */ /* 0x000fe20003f06070 */ /*00e0*/ IMAD.MOV.U32 R3, RZ, RZ, 0x4 ; /* 0x00000004ff037424 */ /* 0x000fe200078e00ff */ /*00f0*/ MOV R11, c[0x0][0x174] ; /* 0x00005d00000b7a02 */ /* 0x000fc40000000f00 */ /*0100*/ IADD3 R2, R2, c[0x0][0x168], RZ ; /* 0x00005a0002027a10 */ /* 0x000fe20007ffe0ff */ /*0110*/ IMAD.WIDE.U32 R8, R8, R3, c[0x0][0x160] ; /* 0x0000580008087625 */ /* 0x000fe200078e0003 */ /*0120*/ LOP3.LUT R10, R10, 0x3, RZ, 0xc0, !PT ; /* 0x000000030a0a7812 */ /* 0x000fc600078ec0ff */ /*0130*/ IMAD.WIDE.U32 R6, R6, R3, c[0x0][0x160] ; /* 0x0000580006067625 */ /* 0x000fc800078e0003 */ /*0140*/ IMAD.WIDE.U32 R4, R0, R3, c[0x0][0x160] ; /* 0x0000580000047625 */ /* 0x000fc800078e0003 */ /*0150*/ FMUL R11, R11, c[0x0][0x174] ; /* 0x00005d000b0b7a20 */ /* 0x000fe40000400000 */ /*0160*/ IMAD.WIDE.U32 R2, R2, R3, c[0x0][0x160] ; /* 0x0000580002027625 */ /* 0x000fe200078e0003 */ /*0170*/ @!P0 BRA 0x8e0 ; /* 0x0000076000008947 */ /* 0x000fea0003800000 */ /*0180*/ IADD3 R12, -R10, c[0x0][0x16c], RZ ; /* 0x00005b000a0c7a10 */ /* 0x000fe40007ffe1ff */ /*0190*/ ISETP.NE.AND P2, PT, R0, RZ, PT ; /* 0x000000ff0000720c */ /* 0x000fda0003f45270 */ /*01a0*/ @!P2 LDG.E R13, [R2.64] ; /* 0x00000004020da981 */ /* 0x001ea4000c1e1900 */ /*01b0*/ @!P2 FADD R15, R13, c[0x0][0x170] ; /* 0x00005c000d0fa621 */ /* 0x004fca0000000000 */ /*01c0*/ @!P2 STG.E [R2.64], R15 ; /* 0x0000000f0200a986 */ /* 0x0001e8000c101904 */ /*01d0*/ LDG.E R13, [R8.64] ; /* 0x00000004080d7981 */ /* 0x000ea8000c1e1900 */ /*01e0*/ LDG.E R14, [R4.64] ; /* 0x00000004040e7981 */ /* 0x000ee8000c1e1900 */ /*01f0*/ LDG.E R19, [R6.64] ; /* 0x0000000406137981 */ /* 0x000f22000c1e1900 */ /*0200*/ MUFU.RCP R16, R11 ; /* 0x0000000b00107308 */ /* 0x000e620000001000 */ /*0210*/ IADD3 R12, R12, -0x4, RZ ; /* 0xfffffffc0c0c7810 */ /* 0x000fe20007ffe0ff */ /*0220*/ BSSY B0, 0x330 ; /* 0x0000010000007945 */ /* 0x000fe60003800000 */ /*0230*/ ISETP.NE.AND P3, PT, R12, RZ, PT ; /* 0x000000ff0c00720c */ /* 0x000fe20003f65270 */ /*0240*/ FADD R17, R13, R13 ; /* 0x0000000d0d117221 */ /* 0x004fc80000000000 */ /*0250*/ FADD R14, R14, -R17 ; /* 0x800000110e0e7221 */ /* 0x008fe40000000000 */ /*0260*/ FFMA R17, -R11, R16, 1 ; /* 0x3f8000000b117423 */ /* 0x002fe40000000110 */ /*0270*/ FADD R14, R14, R19 ; /* 0x000000130e0e7221 */ /* 0x010fe40000000000 */ /*0280*/ FFMA R17, R16, R17, R16 ; /* 0x0000001110117223 */ /* 0x000fe40000000010 */ /*0290*/ FMUL R14, R14, c[0x0][0x178] ; /* 0x00005e000e0e7a20 */ /* 0x000fc80000400000 */ /*02a0*/ FCHK P0, R14, R11 ; /* 0x0000000b0e007302 */ /* 0x000e620000000000 */ /*02b0*/ FFMA R16, R14, R17, RZ ; /* 0x000000110e107223 */ /* 0x000fc800000000ff */ /*02c0*/ FFMA R15, -R11, R16, R14 ; /* 0x000000100b0f7223 */ /* 0x001fc8000000010e */ /*02d0*/ FFMA R16, R17, R15, R16 ; /* 0x0000000f11107223 */ /* 0x000fe20000000010 */ /*02e0*/ @!P0 BRA 0x320 ; /* 0x0000003000008947 */ /* 0x002fea0003800000 */ /*02f0*/ MOV R17, R14 ; /* 0x0000000e00117202 */ /* 0x000fe40000000f00 */ /*0300*/ MOV R14, 0x320 ; /* 0x00000320000e7802 */ /* 0x000fe40000000f00 */ /*0310*/ CALL.REL.NOINC 0xb00 ; /* 0x000007e000007944 */ /* 0x000fea0003c00000 */ /*0320*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0330*/ FADD R13, R13, R16 ; /* 0x000000100d0d7221 */ /* 0x000fe20000000000 */ /*0340*/ BSSY B0, 0x3c0 ; /* 0x0000007000007945 */ /* 0x000fe80003800000 */ /*0350*/ STG.E [R8.64], R13 ; /* 0x0000000d08007986 */ /* 0x0001e8000c101904 */ /*0360*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0370*/ @P2 BRA 0x3b0 ; /* 0x0000003000002947 */ /* 0x000fea0003800000 */ /*0380*/ LDG.E R13, [R2.64] ; /* 0x00000004020d7981 */ /* 0x001ea4000c1e1900 */ /*0390*/ FADD R13, R13, c[0x0][0x170] ; /* 0x00005c000d0d7621 */ /* 0x004fca0000000000 */ /*03a0*/ STG.E [R2.64], R13 ; /* 0x0000000d02007986 */ /* 0x0001e4000c101904 */ /*03b0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*03c0*/ LDG.E R13, [R8.64] ; /* 0x00000004080d7981 */ /* 0x001ea8000c1e1900 */ /*03d0*/ LDG.E R14, [R4.64] ; /* 0x00000004040e7981 */ /* 0x000ee8000c1e1900 */ /*03e0*/ LDG.E R17, [R6.64] ; /* 0x0000000406117981 */ /* 0x000f22000c1e1900 */ /*03f0*/ MUFU.RCP R16, R11 ; /* 0x0000000b00107308 */ /* 0x000e220000001000 */ /*0400*/ BSSY B0, 0x500 ; /* 0x000000f000007945 */ /* 0x000fe20003800000 */ /*0410*/ FADD R15, R13, R13 ; /* 0x0000000d0d0f7221 */ /* 0x004fc80000000000 */ /*0420*/ FADD R14, R14, -R15 ; /* 0x8000000f0e0e7221 */ /* 0x008fe40000000000 */ /*0430*/ FFMA R15, -R11, R16, 1 ; /* 0x3f8000000b0f7423 */ /* 0x001fe40000000110 */ /*0440*/ FADD R14, R14, R17 ; /* 0x000000110e0e7221 */ /* 0x010fe40000000000 */ /*0450*/ FFMA R17, R16, R15, R16 ; /* 0x0000000f10117223 */ /* 0x000fe40000000010 */ /*0460*/ FMUL R14, R14, c[0x0][0x178] ; /* 0x00005e000e0e7a20 */ /* 0x000fc80000400000 */ /*0470*/ FCHK P0, R14, R11 ; /* 0x0000000b0e007302 */ /* 0x000e220000000000 */ /*0480*/ FFMA R15, R17, R14, RZ ; /* 0x0000000e110f7223 */ /* 0x000fc800000000ff */ /*0490*/ FFMA R16, -R11, R15, R14 ; /* 0x0000000f0b107223 */ /* 0x000fc8000000010e */ /*04a0*/ FFMA R16, R17, R16, R15 ; /* 0x0000001011107223 */ /* 0x000fe2000000000f */ /*04b0*/ @!P0 BRA 0x4f0 ; /* 0x0000003000008947 */ /* 0x001fea0003800000 */ /*04c0*/ IMAD.MOV.U32 R17, RZ, RZ, R14 ; /* 0x000000ffff117224 */ /* 0x000fe200078e000e */ /*04d0*/ MOV R14, 0x4f0 ; /* 0x000004f0000e7802 */ /* 0x000fe40000000f00 */ /*04e0*/ CALL.REL.NOINC 0xb00 ; /* 0x0000061000007944 */ /* 0x000fea0003c00000 */ /*04f0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0500*/ FADD R13, R13, R16 ; /* 0x000000100d0d7221 */ /* 0x000fe20000000000 */ /*0510*/ BSSY B0, 0x590 ; /* 0x0000007000007945 */ /* 0x000fe80003800000 */ /*0520*/ STG.E [R8.64], R13 ; /* 0x0000000d08007986 */ /* 0x0001e8000c101904 */ /*0530*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0540*/ @P2 BRA 0x580 ; /* 0x0000003000002947 */ /* 0x000fea0003800000 */ /*0550*/ LDG.E R13, [R2.64] ; /* 0x00000004020d7981 */ /* 0x001ea4000c1e1900 */ /*0560*/ FADD R13, R13, c[0x0][0x170] ; /* 0x00005c000d0d7621 */ /* 0x004fca0000000000 */ /*0570*/ STG.E [R2.64], R13 ; /* 0x0000000d02007986 */ /* 0x0001e4000c101904 */ /*0580*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0590*/ LDG.E R13, [R8.64] ; /* 0x00000004080d7981 */ /* 0x001ea8000c1e1900 */ /*05a0*/ LDG.E R14, [R4.64] ; /* 0x00000004040e7981 */ /* 0x000ee8000c1e1900 */ /*05b0*/ LDG.E R17, [R6.64] ; /* 0x0000000406117981 */ /* 0x000f22000c1e1900 */ /*05c0*/ MUFU.RCP R16, R11 ; /* 0x0000000b00107308 */ /* 0x000e220000001000 */ /*05d0*/ BSSY B0, 0x6d0 ; /* 0x000000f000007945 */ /* 0x000fe20003800000 */ /*05e0*/ FADD R15, R13, R13 ; /* 0x0000000d0d0f7221 */ /* 0x004fc80000000000 */ /*05f0*/ FADD R14, R14, -R15 ; /* 0x8000000f0e0e7221 */ /* 0x008fe40000000000 */ /*0600*/ FFMA R15, -R11, R16, 1 ; /* 0x3f8000000b0f7423 */ /* 0x001fe40000000110 */ /*0610*/ FADD R14, R14, R17 ; /* 0x000000110e0e7221 */ /* 0x010fe40000000000 */ /*0620*/ FFMA R17, R16, R15, R16 ; /* 0x0000000f10117223 */ /* 0x000fe40000000010 */ /*0630*/ FMUL R14, R14, c[0x0][0x178] ; /* 0x00005e000e0e7a20 */ /* 0x000fc80000400000 */ /*0640*/ FCHK P0, R14, R11 ; /* 0x0000000b0e007302 */ /* 0x000e220000000000 */ /*0650*/ FFMA R15, R17, R14, RZ ; /* 0x0000000e110f7223 */ /* 0x000fc800000000ff */ /*0660*/ FFMA R16, -R11, R15, R14 ; /* 0x0000000f0b107223 */ /* 0x000fc8000000010e */ /*0670*/ FFMA R16, R17, R16, R15 ; /* 0x0000001011107223 */ /* 0x000fe2000000000f */ /*0680*/ @!P0 BRA 0x6c0 ; /* 0x0000003000008947 */ /* 0x001fea0003800000 */ /*0690*/ MOV R17, R14 ; /* 0x0000000e00117202 */ /* 0x000fe40000000f00 */ /*06a0*/ MOV R14, 0x6c0 ; /* 0x000006c0000e7802 */ /* 0x000fe40000000f00 */ /*06b0*/ CALL.REL.NOINC 0xb00 ; /* 0x0000044000007944 */ /* 0x000fea0003c00000 */ /*06c0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*06d0*/ FADD R13, R13, R16 ; /* 0x000000100d0d7221 */ /* 0x000fe20000000000 */ /*06e0*/ BSSY B0, 0x760 ; /* 0x0000007000007945 */ /* 0x000fe80003800000 */ /*06f0*/ STG.E [R8.64], R13 ; /* 0x0000000d08007986 */ /* 0x0001e8000c101904 */ /*0700*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0710*/ @P2 BRA 0x750 ; /* 0x0000003000002947 */ /* 0x000fea0003800000 */ /*0720*/ LDG.E R13, [R2.64] ; /* 0x00000004020d7981 */ /* 0x001ea4000c1e1900 */ /*0730*/ FADD R13, R13, c[0x0][0x170] ; /* 0x00005c000d0d7621 */ /* 0x004fca0000000000 */ /*0740*/ STG.E [R2.64], R13 ; /* 0x0000000d02007986 */ /* 0x0001e4000c101904 */ /*0750*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0760*/ LDG.E R13, [R8.64] ; /* 0x00000004080d7981 */ /* 0x001ea8000c1e1900 */ /*0770*/ LDG.E R14, [R4.64] ; /* 0x00000004040e7981 */ /* 0x000ee8000c1e1900 */ /*0780*/ LDG.E R17, [R6.64] ; /* 0x0000000406117981 */ /* 0x000f22000c1e1900 */ /*0790*/ MUFU.RCP R16, R11 ; /* 0x0000000b00107308 */ /* 0x000e220000001000 */ /*07a0*/ BSSY B0, 0x8a0 ; /* 0x000000f000007945 */ /* 0x000fe20003800000 */ /*07b0*/ FADD R15, R13, R13 ; /* 0x0000000d0d0f7221 */ /* 0x004fc80000000000 */ /*07c0*/ FADD R14, R14, -R15 ; /* 0x8000000f0e0e7221 */ /* 0x008fe40000000000 */ /*07d0*/ FFMA R15, -R11, R16, 1 ; /* 0x3f8000000b0f7423 */ /* 0x001fe40000000110 */ /*07e0*/ FADD R14, R14, R17 ; /* 0x000000110e0e7221 */ /* 0x010fe40000000000 */ /*07f0*/ FFMA R17, R16, R15, R16 ; /* 0x0000000f10117223 */ /* 0x000fe40000000010 */ /*0800*/ FMUL R14, R14, c[0x0][0x178] ; /* 0x00005e000e0e7a20 */ /* 0x000fc80000400000 */ /*0810*/ FCHK P0, R14, R11 ; /* 0x0000000b0e007302 */ /* 0x000e220000000000 */ /*0820*/ FFMA R15, R17, R14, RZ ; /* 0x0000000e110f7223 */ /* 0x000fc800000000ff */ /*0830*/ FFMA R16, -R11, R15, R14 ; /* 0x0000000f0b107223 */ /* 0x000fc8000000010e */ /*0840*/ FFMA R16, R17, R16, R15 ; /* 0x0000001011107223 */ /* 0x000fe2000000000f */ /*0850*/ @!P0 BRA 0x890 ; /* 0x0000003000008947 */ /* 0x001fea0003800000 */ /*0860*/ IMAD.MOV.U32 R17, RZ, RZ, R14 ; /* 0x000000ffff117224 */ /* 0x000fe200078e000e */ /*0870*/ MOV R14, 0x890 ; /* 0x00000890000e7802 */ /* 0x000fe40000000f00 */ /*0880*/ CALL.REL.NOINC 0xb00 ; /* 0x0000027000007944 */ /* 0x000fea0003c00000 */ /*0890*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*08a0*/ FADD R13, R13, R16 ; /* 0x000000100d0d7221 */ /* 0x000fca0000000000 */ /*08b0*/ STG.E [R8.64], R13 ; /* 0x0000000d08007986 */ /* 0x0001e8000c101904 */ /*08c0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*08d0*/ @P3 BRA 0x190 ; /* 0xfffff8b000003947 */ /* 0x000fea000383ffff */ /*08e0*/ ISETP.NE.AND P0, PT, R10, RZ, PT ; /* 0x000000ff0a00720c */ /* 0x000fda0003f05270 */ /*08f0*/ @!P0 EXIT ; /* 0x000000000000894d */ /* 0x000fea0003800000 */ /*0900*/ ISETP.NE.AND P0, PT, R0, RZ, PT ; /* 0x000000ff0000720c */ /* 0x000fda0003f05270 */ /*0910*/ @!P0 LDG.E R12, [R2.64] ; /* 0x00000004020c8981 */ /* 0x000ea4000c1e1900 */ /*0920*/ @!P0 FADD R13, R12, c[0x0][0x170] ; /* 0x00005c000c0d8621 */ /* 0x005fca0000000000 */ /*0930*/ @!P0 STG.E [R2.64], R13 ; /* 0x0000000d02008986 */ /* 0x0001e8000c101904 */ /*0940*/ LDG.E R12, [R8.64] ; /* 0x00000004080c7981 */ /* 0x000ea8000c1e1900 */ /*0950*/ LDG.E R14, [R4.64] ; /* 0x00000004040e7981 */ /* 0x000ee8000c1e1900 */ /*0960*/ LDG.E R17, [R6.64] ; /* 0x0000000406117981 */ /* 0x000f22000c1e1900 */ /*0970*/ MUFU.RCP R16, R11 ; /* 0x0000000b00107308 */ /* 0x000e620000001000 */ /*0980*/ BSSY B0, 0xa90 ; /* 0x0000010000007945 */ /* 0x000fe20003800000 */ /*0990*/ FADD R15, R12, R12 ; /* 0x0000000c0c0f7221 */ /* 0x004fc80000000000 */ /*09a0*/ FADD R14, R14, -R15 ; /* 0x8000000f0e0e7221 */ /* 0x008fe40000000000 */ /*09b0*/ FFMA R15, -R11, R16, 1 ; /* 0x3f8000000b0f7423 */ /* 0x002fe40000000110 */ /*09c0*/ FADD R14, R14, R17 ; /* 0x000000110e0e7221 */ /* 0x010fe40000000000 */ /*09d0*/ FFMA R15, R16, R15, R16 ; /* 0x0000000f100f7223 */ /* 0x000fe40000000010 */ /*09e0*/ FMUL R14, R14, c[0x0][0x178] ; /* 0x00005e000e0e7a20 */ /* 0x000fc80000400000 */ /*09f0*/ FCHK P0, R14, R11 ; /* 0x0000000b0e007302 */ /* 0x000e620000000000 */ /*0a00*/ FFMA R13, R15, R14, RZ ; /* 0x0000000e0f0d7223 */ /* 0x001fc800000000ff */ /*0a10*/ FFMA R16, -R11, R13, R14 ; /* 0x0000000d0b107223 */ /* 0x000fc8000000010e */ /*0a20*/ FFMA R13, R15, R16, R13 ; /* 0x000000100f0d7223 */ /* 0x000fe2000000000d */ /*0a30*/ @!P0 BRA 0xa80 ; /* 0x0000004000008947 */ /* 0x002fea0003800000 */ /*0a40*/ MOV R17, R14 ; /* 0x0000000e00117202 */ /* 0x000fe40000000f00 */ /*0a50*/ MOV R14, 0xa70 ; /* 0x00000a70000e7802 */ /* 0x000fe40000000f00 */ /*0a60*/ CALL.REL.NOINC 0xb00 ; /* 0x0000009000007944 */ /* 0x000fea0003c00000 */ /*0a70*/ IMAD.MOV.U32 R13, RZ, RZ, R16 ; /* 0x000000ffff0d7224 */ /* 0x000fe400078e0010 */ /*0a80*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0a90*/ IADD3 R10, R10, -0x1, RZ ; /* 0xffffffff0a0a7810 */ /* 0x000fe20007ffe0ff */ /*0aa0*/ FADD R13, R12, R13 ; /* 0x0000000d0c0d7221 */ /* 0x000fc60000000000 */ /*0ab0*/ ISETP.NE.AND P0, PT, R10, RZ, PT ; /* 0x000000ff0a00720c */ /* 0x000fe40003f05270 */ /*0ac0*/ STG.E [R8.64], R13 ; /* 0x0000000d08007986 */ /* 0x0001e8000c101904 */ /*0ad0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fee0000010000 */ /*0ae0*/ @P0 BRA 0x900 ; /* 0xfffffe1000000947 */ /* 0x000fea000383ffff */ /*0af0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0b00*/ SHF.R.U32.HI R15, RZ, 0x17, R11 ; /* 0x00000017ff0f7819 */ /* 0x000fe2000001160b */ /*0b10*/ BSSY B1, 0x1150 ; /* 0x0000063000017945 */ /* 0x000fe20003800000 */ /*0b20*/ SHF.R.U32.HI R18, RZ, 0x17, R17 ; /* 0x00000017ff127819 */ /* 0x000fc40000011611 */ /*0b30*/ LOP3.LUT R15, R15, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff0f0f7812 */ /* 0x000fe400078ec0ff */ /*0b40*/ LOP3.LUT R18, R18, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff12127812 */ /* 0x000fe400078ec0ff */ /*0b50*/ IADD3 R21, R15, -0x1, RZ ; /* 0xffffffff0f157810 */ /* 0x000fe40007ffe0ff */ /*0b60*/ IADD3 R19, R18, -0x1, RZ ; /* 0xffffffff12137810 */ /* 0x000fe40007ffe0ff */ /*0b70*/ ISETP.GT.U32.AND P0, PT, R21, 0xfd, PT ; /* 0x000000fd1500780c */ /* 0x000fe40003f04070 */ /*0b80*/ MOV R20, R11 ; /* 0x0000000b00147202 */ /* 0x000fc40000000f00 */ /*0b90*/ ISETP.GT.U32.OR P0, PT, R19, 0xfd, P0 ; /* 0x000000fd1300780c */ /* 0x000fda0000704470 */ /*0ba0*/ @!P0 IMAD.MOV.U32 R16, RZ, RZ, RZ ; /* 0x000000ffff108224 */ /* 0x000fe200078e00ff */ /*0bb0*/ @!P0 BRA 0xd30 ; /* 0x0000017000008947 */ /* 0x000fea0003800000 */ /*0bc0*/ FSETP.GTU.FTZ.AND P0, PT, |R17|, +INF , PT ; /* 0x7f8000001100780b */ /* 0x000fe40003f1c200 */ /*0bd0*/ FSETP.GTU.FTZ.AND P1, PT, |R11|, +INF , PT ; /* 0x7f8000000b00780b */ /* 0x000fc80003f3c200 */ /*0be0*/ PLOP3.LUT P0, PT, P0, P1, PT, 0xa8, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000703570 */ /*0bf0*/ @P0 BRA 0x1130 ; /* 0x0000053000000947 */ /* 0x000fea0003800000 */ /*0c00*/ LOP3.LUT P0, RZ, R20, 0x7fffffff, R17, 0xc8, !PT ; /* 0x7fffffff14ff7812 */ /* 0x000fda000780c811 */ /*0c10*/ @!P0 BRA 0x1110 ; /* 0x000004f000008947 */ /* 0x000fea0003800000 */ /*0c20*/ FSETP.NEU.FTZ.AND P4, PT, |R17|, +INF , PT ; /* 0x7f8000001100780b */ /* 0x000fe40003f9d200 */ /*0c30*/ FSETP.NEU.FTZ.AND P1, PT, |R11|, +INF , PT ; /* 0x7f8000000b00780b */ /* 0x000fe40003f3d200 */ /*0c40*/ FSETP.NEU.FTZ.AND P0, PT, |R17|, +INF , PT ; /* 0x7f8000001100780b */ /* 0x000fd60003f1d200 */ /*0c50*/ @!P1 BRA !P4, 0x1110 ; /* 0x000004b000009947 */ /* 0x000fea0006000000 */ /*0c60*/ LOP3.LUT P4, RZ, R17, 0x7fffffff, RZ, 0xc0, !PT ; /* 0x7fffffff11ff7812 */ /* 0x000fc8000788c0ff */ /*0c70*/ PLOP3.LUT P1, PT, P1, P4, PT, 0x2a, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000f28572 */ /*0c80*/ @P1 BRA 0x10f0 ; /* 0x0000046000001947 */ /* 0x000fea0003800000 */ /*0c90*/ LOP3.LUT P1, RZ, R20, 0x7fffffff, RZ, 0xc0, !PT ; /* 0x7fffffff14ff7812 */ /* 0x000fc8000782c0ff */ /*0ca0*/ PLOP3.LUT P0, PT, P0, P1, PT, 0x2a, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000702572 */ /*0cb0*/ @P0 BRA 0x10c0 ; /* 0x0000040000000947 */ /* 0x000fea0003800000 */ /*0cc0*/ ISETP.GE.AND P0, PT, R19, RZ, PT ; /* 0x000000ff1300720c */ /* 0x000fe40003f06270 */ /*0cd0*/ ISETP.GE.AND P1, PT, R21, RZ, PT ; /* 0x000000ff1500720c */ /* 0x000fd60003f26270 */ /*0ce0*/ @P0 MOV R16, RZ ; /* 0x000000ff00100202 */ /* 0x000fe20000000f00 */ /*0cf0*/ @!P0 IMAD.MOV.U32 R16, RZ, RZ, -0x40 ; /* 0xffffffc0ff108424 */ /* 0x000fe400078e00ff */ /*0d00*/ @!P0 FFMA R17, R17, 1.84467440737095516160e+19, RZ ; /* 0x5f80000011118823 */ /* 0x000fe400000000ff */ /*0d10*/ @!P1 FFMA R20, R11, 1.84467440737095516160e+19, RZ ; /* 0x5f8000000b149823 */ /* 0x000fe200000000ff */ /*0d20*/ @!P1 IADD3 R16, R16, 0x40, RZ ; /* 0x0000004010109810 */ /* 0x000fe40007ffe0ff */ /*0d30*/ LEA R19, R15, 0xc0800000, 0x17 ; /* 0xc08000000f137811 */ /* 0x000fe200078eb8ff */ /*0d40*/ BSSY B2, 0x10b0 ; /* 0x0000036000027945 */ /* 0x000fe20003800000 */ /*0d50*/ IADD3 R18, R18, -0x7f, RZ ; /* 0xffffff8112127810 */ /* 0x000fe40007ffe0ff */ /*0d60*/ IADD3 R23, -R19, R20, RZ ; /* 0x0000001413177210 */ /* 0x000fc60007ffe1ff */ /*0d70*/ IMAD R22, R18.reuse, -0x800000, R17 ; /* 0xff80000012167824 */ /* 0x040fe200078e0211 */ /*0d80*/ MUFU.RCP R20, R23 ; /* 0x0000001700147308 */ /* 0x0000620000001000 */ /*0d90*/ FADD.FTZ R19, -R23, -RZ ; /* 0x800000ff17137221 */ /* 0x000fe20000010100 */ /*0da0*/ IADD3 R23, R18, 0x7f, -R15 ; /* 0x0000007f12177810 */ /* 0x001fca0007ffe80f */ /*0db0*/ IMAD.IADD R18, R23, 0x1, R16 ; /* 0x0000000117127824 */ /* 0x000fe400078e0210 */ /*0dc0*/ FFMA R21, R20, R19, 1 ; /* 0x3f80000014157423 */ /* 0x002fc80000000013 */ /*0dd0*/ FFMA R17, R20, R21, R20 ; /* 0x0000001514117223 */ /* 0x000fc80000000014 */ /*0de0*/ FFMA R20, R22, R17, RZ ; /* 0x0000001116147223 */ /* 0x000fc800000000ff */ /*0df0*/ FFMA R21, R19, R20, R22 ; /* 0x0000001413157223 */ /* 0x000fc80000000016 */ /*0e00*/ FFMA R20, R17, R21, R20 ; /* 0x0000001511147223 */ /* 0x000fc80000000014 */ /*0e10*/ FFMA R19, R19, R20, R22 ; /* 0x0000001413137223 */ /* 0x000fc80000000016 */ /*0e20*/ FFMA R21, R17, R19, R20 ; /* 0x0000001311157223 */ /* 0x000fca0000000014 */ /*0e30*/ SHF.R.U32.HI R15, RZ, 0x17, R21 ; /* 0x00000017ff0f7819 */ /* 0x000fc80000011615 */ /*0e40*/ LOP3.LUT R15, R15, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff0f0f7812 */ /* 0x000fc800078ec0ff */ /*0e50*/ IADD3 R16, R15, R18, RZ ; /* 0x000000120f107210 */ /* 0x000fc80007ffe0ff */ /*0e60*/ IADD3 R15, R16, -0x1, RZ ; /* 0xffffffff100f7810 */ /* 0x000fc80007ffe0ff */ /*0e70*/ ISETP.GE.U32.AND P0, PT, R15, 0xfe, PT ; /* 0x000000fe0f00780c */ /* 0x000fda0003f06070 */ /*0e80*/ @!P0 BRA 0x1090 ; /* 0x0000020000008947 */ /* 0x000fea0003800000 */ /*0e90*/ ISETP.GT.AND P0, PT, R16, 0xfe, PT ; /* 0x000000fe1000780c */ /* 0x000fda0003f04270 */ /*0ea0*/ @P0 BRA 0x1060 ; /* 0x000001b000000947 */ /* 0x000fea0003800000 */ /*0eb0*/ ISETP.GE.AND P0, PT, R16, 0x1, PT ; /* 0x000000011000780c */ /* 0x000fda0003f06270 */ /*0ec0*/ @P0 BRA 0x10a0 ; /* 0x000001d000000947 */ /* 0x000fea0003800000 */ /*0ed0*/ ISETP.GE.AND P0, PT, R16, -0x18, PT ; /* 0xffffffe81000780c */ /* 0x000fe40003f06270 */ /*0ee0*/ LOP3.LUT R21, R21, 0x80000000, RZ, 0xc0, !PT ; /* 0x8000000015157812 */ /* 0x000fd600078ec0ff */ /*0ef0*/ @!P0 BRA 0x10a0 ; /* 0x000001a000008947 */ /* 0x000fea0003800000 */ /*0f00*/ FFMA.RZ R15, R17, R19.reuse, R20.reuse ; /* 0x00000013110f7223 */ /* 0x180fe2000000c014 */ /*0f10*/ ISETP.NE.AND P4, PT, R16.reuse, RZ, PT ; /* 0x000000ff1000720c */ /* 0x040fe40003f85270 */ /*0f20*/ ISETP.NE.AND P1, PT, R16.reuse, RZ, PT ; /* 0x000000ff1000720c */ /* 0x040fe40003f25270 */ /*0f30*/ LOP3.LUT R18, R15, 0x7fffff, RZ, 0xc0, !PT ; /* 0x007fffff0f127812 */ /* 0x000fe200078ec0ff */ /*0f40*/ FFMA.RP R15, R17.reuse, R19.reuse, R20.reuse ; /* 0x00000013110f7223 */ /* 0x1c0fe40000008014 */ /*0f50*/ FFMA.RM R20, R17, R19, R20 ; /* 0x0000001311147223 */ /* 0x000fe20000004014 */ /*0f60*/ IADD3 R17, R16, 0x20, RZ ; /* 0x0000002010117810 */ /* 0x000fe20007ffe0ff */ /*0f70*/ IMAD.MOV R16, RZ, RZ, -R16 ; /* 0x000000ffff107224 */ /* 0x000fe200078e0a10 */ /*0f80*/ LOP3.LUT R18, R18, 0x800000, RZ, 0xfc, !PT ; /* 0x0080000012127812 */ /* 0x000fc400078efcff */ /*0f90*/ FSETP.NEU.FTZ.AND P0, PT, R15, R20, PT ; /* 0x000000140f00720b */ /* 0x000fe40003f1d000 */ /*0fa0*/ SHF.L.U32 R17, R18, R17, RZ ; /* 0x0000001112117219 */ /* 0x000fe400000006ff */ /*0fb0*/ SEL R15, R16, RZ, P4 ; /* 0x000000ff100f7207 */ /* 0x000fe40002000000 */ /*0fc0*/ ISETP.NE.AND P1, PT, R17, RZ, P1 ; /* 0x000000ff1100720c */ /* 0x000fe40000f25270 */ /*0fd0*/ SHF.R.U32.HI R15, RZ, R15, R18 ; /* 0x0000000fff0f7219 */ /* 0x000fe40000011612 */ /*0fe0*/ PLOP3.LUT P0, PT, P0, P1, PT, 0xa8, 0x0 ; /* 0x000000000000781c */ /* 0x000fc40000703570 */ /*0ff0*/ SHF.R.U32.HI R17, RZ, 0x1, R15 ; /* 0x00000001ff117819 */ /* 0x000fe4000001160f */ /*1000*/ SEL R16, RZ, 0x1, !P0 ; /* 0x00000001ff107807 */ /* 0x000fc80004000000 */ /*1010*/ LOP3.LUT R16, R16, 0x1, R17, 0xf8, !PT ; /* 0x0000000110107812 */ /* 0x000fc800078ef811 */ /*1020*/ LOP3.LUT R16, R16, R15, RZ, 0xc0, !PT ; /* 0x0000000f10107212 */ /* 0x000fc800078ec0ff */ /*1030*/ IADD3 R16, R17, R16, RZ ; /* 0x0000001011107210 */ /* 0x000fc80007ffe0ff */ /*1040*/ LOP3.LUT R21, R16, R21, RZ, 0xfc, !PT ; /* 0x0000001510157212 */ /* 0x000fe200078efcff */ /*1050*/ BRA 0x10a0 ; /* 0x0000004000007947 */ /* 0x000fea0003800000 */ /*1060*/ LOP3.LUT R21, R21, 0x80000000, RZ, 0xc0, !PT ; /* 0x8000000015157812 */ /* 0x000fc800078ec0ff */ /*1070*/ LOP3.LUT R21, R21, 0x7f800000, RZ, 0xfc, !PT ; /* 0x7f80000015157812 */ /* 0x000fe200078efcff */ /*1080*/ BRA 0x10a0 ; /* 0x0000001000007947 */ /* 0x000fea0003800000 */ /*1090*/ IMAD R21, R18, 0x800000, R21 ; /* 0x0080000012157824 */ /* 0x000fe400078e0215 */ /*10a0*/ BSYNC B2 ; /* 0x0000000000027941 */ /* 0x000fea0003800000 */ /*10b0*/ BRA 0x1140 ; /* 0x0000008000007947 */ /* 0x000fea0003800000 */ /*10c0*/ LOP3.LUT R17, R20, 0x80000000, R17, 0x48, !PT ; /* 0x8000000014117812 */ /* 0x000fc800078e4811 */ /*10d0*/ LOP3.LUT R21, R17, 0x7f800000, RZ, 0xfc, !PT ; /* 0x7f80000011157812 */ /* 0x000fe200078efcff */ /*10e0*/ BRA 0x1140 ; /* 0x0000005000007947 */ /* 0x000fea0003800000 */ /*10f0*/ LOP3.LUT R21, R20, 0x80000000, R17, 0x48, !PT ; /* 0x8000000014157812 */ /* 0x000fe200078e4811 */ /*1100*/ BRA 0x1140 ; /* 0x0000003000007947 */ /* 0x000fea0003800000 */ /*1110*/ MUFU.RSQ R21, -QNAN ; /* 0xffc0000000157908 */ /* 0x000e220000001400 */ /*1120*/ BRA 0x1140 ; /* 0x0000001000007947 */ /* 0x000fea0003800000 */ /*1130*/ FADD.FTZ R21, R17, R11 ; /* 0x0000000b11157221 */ /* 0x000fe40000010000 */ /*1140*/ BSYNC B1 ; /* 0x0000000000017941 */ /* 0x000fea0003800000 */ /*1150*/ IMAD.MOV.U32 R15, RZ, RZ, 0x0 ; /* 0x00000000ff0f7424 */ /* 0x000fe200078e00ff */ /*1160*/ MOV R16, R21 ; /* 0x0000001500107202 */ /* 0x001fc60000000f00 */ /*1170*/ RET.REL.NODEC R14 0x0 ; /* 0xffffee800e007950 */ /* 0x000fea0003c3ffff */ /*1180*/ BRA 0x1180; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*1190*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*11a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*11b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*11c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*11d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*11e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*11f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1200*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1210*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1220*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1230*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1240*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1250*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1260*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1270*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z10heatKernelPfjjfff .globl _Z10heatKernelPfjjfff .p2align 8 .type _Z10heatKernelPfjjfff,@function _Z10heatKernelPfjjfff: s_clause 0x1 s_load_b32 s2, s[0:1], 0x2c s_load_b64 s[4:5], s[0:1], 0x8 s_mov_b32 s9, 0 s_waitcnt lgkmcnt(0) s_and_b32 s2, s2, 0xffff s_cmp_lg_u32 s5, 0 v_mad_u64_u32 v[1:2], null, s15, s2, v[0:1] s_cselect_b32 s2, -1, 0 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(SALU_CYCLE_1) v_cmp_gt_u32_e32 vcc_lo, s4, v1 s_and_b32 s2, vcc_lo, s2 s_and_saveexec_b32 s3, s2 s_cbranch_execz .LBB0_5 s_clause 0x2 s_load_b64 s[10:11], s[0:1], 0x0 s_load_b64 s[6:7], s[0:1], 0x10 s_load_b32 s3, s[0:1], 0x18 v_dual_mov_b32 v2, 0 :: v_dual_add_nc_u32 v3, 1, v1 v_cmp_eq_u32_e64 s2, 0, v1 s_add_i32 s8, s4, 1 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_3) | instid1(VALU_DEP_3) v_mov_b32_e32 v4, v2 v_lshlrev_b64 v[5:6], 2, v[1:2] v_add_nc_u32_e32 v1, 2, v1 s_lshl_b64 s[0:1], s[8:9], 2 v_lshlrev_b64 v[3:4], 2, v[3:4] s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_3) | instid1(VALU_DEP_4) v_lshlrev_b64 v[7:8], 2, v[1:2] s_waitcnt lgkmcnt(0) v_add_co_u32 v0, vcc_lo, s10, v5 v_add_co_ci_u32_e32 v1, vcc_lo, s11, v6, vcc_lo v_add_co_u32 v2, vcc_lo, s10, v3 v_add_co_ci_u32_e32 v3, vcc_lo, s11, v4, vcc_lo v_add_co_u32 v4, vcc_lo, s10, v7 v_add_co_ci_u32_e32 v5, vcc_lo, s11, v8, vcc_lo v_mul_f32_e64 v6, s7, s7 v_mov_b32_e32 v7, 0 s_add_u32 s0, s10, s0 s_addc_u32 s1, s11, s1 s_set_inst_prefetch_distance 0x1 s_branch .LBB0_3 .p2align 6 .LBB0_2: s_or_b32 exec_lo, exec_lo, s4 s_clause 0x2 global_load_b32 v8, v[0:1], off global_load_b32 v9, v[2:3], off global_load_b32 v10, v[4:5], off s_add_i32 s5, s5, -1 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_3) | instid1(VALU_DEP_1) s_cmp_lg_u32 s5, 0 s_waitcnt vmcnt(1) v_fmac_f32_e32 v8, -2.0, v9 s_waitcnt vmcnt(0) v_add_f32_e32 v8, v10, v8 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_f32_e32 v8, s3, v8 v_div_scale_f32 v10, null, v6, v6, v8 v_div_scale_f32 v13, vcc_lo, v8, v6, v8 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_2) | instid1(VALU_DEP_1) v_rcp_f32_e32 v11, v10 s_waitcnt_depctr 0xfff v_fma_f32 v12, -v10, v11, 1.0 v_fmac_f32_e32 v11, v12, v11 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_f32_e32 v12, v13, v11 v_fma_f32 v14, -v10, v12, v13 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmac_f32_e32 v12, v14, v11 v_fma_f32 v10, -v10, v12, v13 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_div_fmas_f32 v10, v10, v11, v12 v_div_fixup_f32 v8, v10, v6, v8 s_delay_alu instid0(VALU_DEP_1) v_add_f32_e32 v8, v9, v8 global_store_b32 v[2:3], v8, off s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv s_cbranch_scc0 .LBB0_5 .LBB0_3: s_and_saveexec_b32 s4, s2 s_cbranch_execz .LBB0_2 global_load_b32 v8, v7, s[0:1] s_waitcnt vmcnt(0) v_add_f32_e32 v8, s6, v8 global_store_b32 v7, v8, s[0:1] s_branch .LBB0_2 .LBB0_5: s_set_inst_prefetch_distance 0x2 s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z10heatKernelPfjjfff .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 288 .amdhsa_user_sgpr_count 15 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 0 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 0 .amdhsa_next_free_vgpr 15 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z10heatKernelPfjjfff, .Lfunc_end0-_Z10heatKernelPfjjfff .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .offset: 8 .size: 4 .value_kind: by_value - .offset: 12 .size: 4 .value_kind: by_value - .offset: 16 .size: 4 .value_kind: by_value - .offset: 20 .size: 4 .value_kind: by_value - .offset: 24 .size: 4 .value_kind: by_value - .offset: 32 .size: 4 .value_kind: hidden_block_count_x - .offset: 36 .size: 4 .value_kind: hidden_block_count_y - .offset: 40 .size: 4 .value_kind: hidden_block_count_z - .offset: 44 .size: 2 .value_kind: hidden_group_size_x - .offset: 46 .size: 2 .value_kind: hidden_group_size_y - .offset: 48 .size: 2 .value_kind: hidden_group_size_z - .offset: 50 .size: 2 .value_kind: hidden_remainder_x - .offset: 52 .size: 2 .value_kind: hidden_remainder_y - .offset: 54 .size: 2 .value_kind: hidden_remainder_z - .offset: 72 .size: 8 .value_kind: hidden_global_offset_x - .offset: 80 .size: 8 .value_kind: hidden_global_offset_y - .offset: 88 .size: 8 .value_kind: hidden_global_offset_z - .offset: 96 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 288 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z10heatKernelPfjjfff .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z10heatKernelPfjjfff.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 15 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_00059100_00000000-6_kernel.cudafe1.cpp" .text #APP .globl _ZSt21ios_base_library_initv .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string " returned " .LC1: .string "(" .LC2: .string ") at " .LC3: .string ":" #NO_APP .text .type _ZL17CheckCudaErrorAuxPKcjS0_9cudaError, @function _ZL17CheckCudaErrorAuxPKcjS0_9cudaError: .LFB3710: .cfi_startproc testl %ecx, %ecx jne .L7 ret .L7: pushq %r13 .cfi_def_cfa_offset 16 .cfi_offset 13, -16 pushq %r12 .cfi_def_cfa_offset 24 .cfi_offset 12, -24 pushq %rbp .cfi_def_cfa_offset 32 .cfi_offset 6, -32 pushq %rbx .cfi_def_cfa_offset 40 .cfi_offset 3, -40 subq $8, %rsp .cfi_def_cfa_offset 48 movq %rdi, %rbp movl %esi, %r12d movl %ecx, %ebx movq %rdx, %rsi leaq _ZSt4cerr(%rip), %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi leaq .LC0(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %r13 movl %ebx, %edi call cudaGetErrorString@PLT movq %rax, %rsi movq %r13, %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi leaq .LC1(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi movl %ebx, %esi call _ZNSolsEi@PLT movq %rax, %rdi leaq .LC2(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi movq %rbp, %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi leaq .LC3(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi movl %r12d, %esi call _ZNSo9_M_insertImEERSoT_@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movl $1, %edi call exit@PLT .cfi_endproc .LFE3710: .size _ZL17CheckCudaErrorAuxPKcjS0_9cudaError, .-_ZL17CheckCudaErrorAuxPKcjS0_9cudaError .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB3713: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE3713: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z10initializePfj .type _Z10initializePfj, @function _Z10initializePfj: .LFB3708: .cfi_startproc endbr64 testl %esi, %esi je .L10 movq %rdi, %rax movl %esi, %esi leaq (%rdi,%rsi,4), %rdx .L12: movl $0x00000000, (%rax) addq $4, %rax cmpq %rdx, %rax jne .L12 .L10: ret .cfi_endproc .LFE3708: .size _Z10initializePfj, .-_Z10initializePfj .globl _Z35__device_stub__Z10heatKernelPfjjfffPfjjfff .type _Z35__device_stub__Z10heatKernelPfjjfffPfjjfff, @function _Z35__device_stub__Z10heatKernelPfjjfffPfjjfff: .LFB3735: .cfi_startproc endbr64 subq $168, %rsp .cfi_def_cfa_offset 176 movq %rdi, 24(%rsp) movl %esi, 20(%rsp) movl %edx, 16(%rsp) movss %xmm0, 12(%rsp) movss %xmm1, 8(%rsp) movss %xmm2, 4(%rsp) movq %fs:40, %rax movq %rax, 152(%rsp) xorl %eax, %eax leaq 24(%rsp), %rax movq %rax, 96(%rsp) leaq 20(%rsp), %rax movq %rax, 104(%rsp) leaq 16(%rsp), %rax movq %rax, 112(%rsp) leaq 12(%rsp), %rax movq %rax, 120(%rsp) leaq 8(%rsp), %rax movq %rax, 128(%rsp) leaq 4(%rsp), %rax movq %rax, 136(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 56(%rsp) movl $1, 60(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) leaq 40(%rsp), %rcx leaq 32(%rsp), %rdx leaq 60(%rsp), %rsi leaq 48(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L18 .L14: movq 152(%rsp), %rax subq %fs:40, %rax jne .L19 addq $168, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L18: .cfi_restore_state pushq 40(%rsp) .cfi_def_cfa_offset 184 pushq 40(%rsp) .cfi_def_cfa_offset 192 leaq 112(%rsp), %r9 movq 76(%rsp), %rcx movl 84(%rsp), %r8d movq 64(%rsp), %rsi movl 72(%rsp), %edx leaq _Z10heatKernelPfjjfff(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 176 jmp .L14 .L19: call __stack_chk_fail@PLT .cfi_endproc .LFE3735: .size _Z35__device_stub__Z10heatKernelPfjjfffPfjjfff, .-_Z35__device_stub__Z10heatKernelPfjjfffPfjjfff .globl _Z10heatKernelPfjjfff .type _Z10heatKernelPfjjfff, @function _Z10heatKernelPfjjfff: .LFB3736: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z35__device_stub__Z10heatKernelPfjjfffPfjjfff addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE3736: .size _Z10heatKernelPfjjfff, .-_Z10heatKernelPfjjfff .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC5: .string "cudaMalloc((void **)&gpuData, sizeof(float)*(WORK_SIZE + 2))" .align 8 .LC6: .string "/home/ubuntu/Datasets/stackv2/train-structured/xobotun/TP_CUDA/master/Sem1/Sem1/kernel.cu" .align 8 .LC7: .string "cudaMemcpy(gpuData, hostData, sizeof(float)*(WORK_SIZE + 2), cudaMemcpyHostToDevice)" .align 8 .LC10: .string "cudaMemcpy(result, gpuData, sizeof(float)*(WORK_SIZE + 2), cudaMemcpyDeviceToHost)" .section .rodata.str1.1 .LC11: .string "cudaFree(gpuData)" .text .globl main .type main, @function main: .LFB3709: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $72, %rsp .cfi_def_cfa_offset 128 movq %fs:40, %rax movq %rax, 56(%rsp) xorl %eax, %eax movl $408, %edi call _Znam@PLT movq %rax, %r15 movq %rax, 8(%rsp) movl $102, %esi movq %rax, %rdi call _Z10initializePfj leaq 24(%rsp), %rdi movl $408, %esi call cudaMalloc@PLT movl %eax, %ecx leaq .LC5(%rip), %rdx movl $58, %esi leaq .LC6(%rip), %rbx movq %rbx, %rdi call _ZL17CheckCudaErrorAuxPKcjS0_9cudaError movl $1, %ecx movl $408, %edx movq %r15, %rsi movq 24(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %ecx leaq .LC7(%rip), %rdx movl $59, %esi movq %rbx, %rdi call _ZL17CheckCudaErrorAuxPKcjS0_9cudaError movl $408, %edi call _Znam@PLT movq %rax, %r15 movl $256, 44(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) movl $1, 40(%rsp) movl $0, %r9d movl $0, %r8d movq 44(%rsp), %rdx movl $1, %ecx movq 32(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L32 .L23: movl $2, %ecx movl $408, %edx movq 24(%rsp), %rsi movq %r15, %rdi call cudaMemcpy@PLT movl %eax, %ecx leaq .LC10(%rip), %rdx movl $66, %esi leaq .LC6(%rip), %rdi call _ZL17CheckCudaErrorAuxPKcjS0_9cudaError leaq 4(%r15), %r12 leaq 404(%r15), %r14 leaq _ZSt4cout(%rip), %r13 jmp .L28 .L32: movss .LC8(%rip), %xmm2 movss .LC9(%rip), %xmm1 movaps %xmm2, %xmm0 movl $50, %edx movl $100, %esi movq 24(%rsp), %rdi call _Z35__device_stub__Z10heatKernelPfjjfffPfjjfff jmp .L23 .L35: movq 56(%rsp), %rax subq %fs:40, %rax jne .L33 call _ZSt16__throw_bad_castv@PLT .L33: call __stack_chk_fail@PLT .L36: movzbl 67(%rbp), %esi .L27: movsbl %sil, %esi movq %rbx, %rdi call _ZNSo3putEc@PLT movq %rax, %rdi call _ZNSo5flushEv@PLT addq $4, %r12 cmpq %r14, %r12 je .L34 .L28: pxor %xmm0, %xmm0 cvtss2sd (%r12), %xmm0 movq %r13, %rdi call _ZNSo9_M_insertIdEERSoT_@PLT movq %rax, %rbx movq (%rax), %rax movq -24(%rax), %rax movq 240(%rbx,%rax), %rbp testq %rbp, %rbp je .L35 cmpb $0, 56(%rbp) jne .L36 movq %rbp, %rdi call _ZNKSt5ctypeIcE13_M_widen_initEv@PLT movq 0(%rbp), %rax movl $10, %esi movq %rbp, %rdi call *48(%rax) movl %eax, %esi jmp .L27 .L34: movq 24(%rsp), %rdi call cudaFree@PLT movl %eax, %ecx leaq .LC11(%rip), %rdx movl $74, %esi leaq .LC6(%rip), %rdi call _ZL17CheckCudaErrorAuxPKcjS0_9cudaError movq 8(%rsp), %rdi call _ZdaPv@PLT movq %r15, %rdi call _ZdaPv@PLT movq 56(%rsp), %rax subq %fs:40, %rax jne .L37 movl $0, %eax addq $72, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L37: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE3709: .size main, .-main .section .rodata.str1.1 .LC12: .string "_Z10heatKernelPfjjfff" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB3738: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC12(%rip), %rdx movq %rdx, %rcx leaq _Z10heatKernelPfjjfff(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE3738: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .section .rodata.cst4,"aM",@progbits,4 .align 4 .LC8: .long 1056964608 .align 4 .LC9: .long 1065353216 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "kernel.hip" # Start of file scope inline assembly .globl _ZSt21ios_base_library_initv # End of file scope inline assembly .globl _Z25__device_stub__heatKernelPfjjfff # -- Begin function _Z25__device_stub__heatKernelPfjjfff .p2align 4, 0x90 .type _Z25__device_stub__heatKernelPfjjfff,@function _Z25__device_stub__heatKernelPfjjfff: # @_Z25__device_stub__heatKernelPfjjfff .cfi_startproc # %bb.0: subq $136, %rsp .cfi_def_cfa_offset 144 movq %rdi, 72(%rsp) movl %esi, 20(%rsp) movl %edx, 16(%rsp) movss %xmm0, 12(%rsp) movss %xmm1, 8(%rsp) movss %xmm2, 4(%rsp) leaq 72(%rsp), %rax movq %rax, 80(%rsp) leaq 20(%rsp), %rax movq %rax, 88(%rsp) leaq 16(%rsp), %rax movq %rax, 96(%rsp) leaq 12(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) leaq 4(%rsp), %rax movq %rax, 120(%rsp) leaq 56(%rsp), %rdi leaq 40(%rsp), %rsi leaq 32(%rsp), %rdx leaq 24(%rsp), %rcx callq __hipPopCallConfiguration movq 56(%rsp), %rsi movl 64(%rsp), %edx movq 40(%rsp), %rcx movl 48(%rsp), %r8d leaq 80(%rsp), %r9 movl $_Z10heatKernelPfjjfff, %edi pushq 24(%rsp) .cfi_adjust_cfa_offset 8 pushq 40(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $152, %rsp .cfi_adjust_cfa_offset -152 retq .Lfunc_end0: .size _Z25__device_stub__heatKernelPfjjfff, .Lfunc_end0-_Z25__device_stub__heatKernelPfjjfff .cfi_endproc # -- End function .globl _Z10initializePfj # -- Begin function _Z10initializePfj .p2align 4, 0x90 .type _Z10initializePfj,@function _Z10initializePfj: # @_Z10initializePfj .cfi_startproc # %bb.0: testl %esi, %esi je .LBB1_1 # %bb.2: # %.lr.ph.preheader movl %esi, %edx shlq $2, %rdx xorl %esi, %esi jmp memset@PLT # TAILCALL .LBB1_1: # %._crit_edge retq .Lfunc_end1: .size _Z10initializePfj, .Lfunc_end1-_Z10initializePfj .cfi_endproc # -- End function .globl main # -- Begin function main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %r15 .cfi_def_cfa_offset 16 pushq %r14 .cfi_def_cfa_offset 24 pushq %r13 .cfi_def_cfa_offset 32 pushq %r12 .cfi_def_cfa_offset 40 pushq %rbx .cfi_def_cfa_offset 48 subq $144, %rsp .cfi_def_cfa_offset 192 .cfi_offset %rbx, -48 .cfi_offset %r12, -40 .cfi_offset %r13, -32 .cfi_offset %r14, -24 .cfi_offset %r15, -16 movl $408, %edi # imm = 0x198 callq _Znam movq %rax, %rbx movl $408, %edx # imm = 0x198 movq %rax, %rdi xorl %esi, %esi callq memset@PLT leaq 8(%rsp), %rdi movl $408, %esi # imm = 0x198 callq hipMalloc movl $.L.str.1, %esi movl $58, %edi movl %eax, %edx callq _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t movq 8(%rsp), %rdi movl $408, %edx # imm = 0x198 movq %rbx, %rsi movl $1, %ecx callq hipMemcpy movl $.L.str.2, %esi movl $59, %edi movl %eax, %edx callq _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t movl $408, %edi # imm = 0x198 callq _Znam movq %rax, %r14 movabsq $4294967297, %rdi # imm = 0x100000001 leaq 255(%rdi), %rdx movl $1, %esi movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB2_2 # %bb.1: movq 8(%rsp), %rax movq %rax, 88(%rsp) movl $100, 36(%rsp) movl $50, 32(%rsp) movl $1056964608, 28(%rsp) # imm = 0x3F000000 movl $1065353216, 24(%rsp) # imm = 0x3F800000 movl $1056964608, 20(%rsp) # imm = 0x3F000000 leaq 88(%rsp), %rax movq %rax, 96(%rsp) leaq 36(%rsp), %rax movq %rax, 104(%rsp) leaq 32(%rsp), %rax movq %rax, 112(%rsp) leaq 28(%rsp), %rax movq %rax, 120(%rsp) leaq 24(%rsp), %rax movq %rax, 128(%rsp) leaq 20(%rsp), %rax movq %rax, 136(%rsp) leaq 72(%rsp), %rdi leaq 56(%rsp), %rsi leaq 48(%rsp), %rdx leaq 40(%rsp), %rcx callq __hipPopCallConfiguration movq 72(%rsp), %rsi movl 80(%rsp), %edx movq 56(%rsp), %rcx movl 64(%rsp), %r8d leaq 96(%rsp), %r9 movl $_Z10heatKernelPfjjfff, %edi pushq 40(%rsp) .cfi_adjust_cfa_offset 8 pushq 56(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB2_2: movq 8(%rsp), %rsi movl $408, %edx # imm = 0x198 movq %r14, %rdi movl $2, %ecx callq hipMemcpy movl $.L.str.3, %esi movl $66, %edi movl %eax, %edx callq _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t movl $1, %r12d jmp .LBB2_3 .p2align 4, 0x90 .LBB2_6: # in Loop: Header=BB2_3 Depth=1 movq %r15, %rdi movq %rax, %r13 callq _ZNKSt5ctypeIcE13_M_widen_initEv movq (%r15), %rax movq %r15, %rdi movl $10, %esi callq *48(%rax) movl %eax, %ecx movq %r13, %rax .LBB2_7: # %_ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_.exit # in Loop: Header=BB2_3 Depth=1 movsbl %cl, %esi movq %rax, %rdi callq _ZNSo3putEc movq %rax, %rdi callq _ZNSo5flushEv incq %r12 cmpq $101, %r12 je .LBB2_8 .LBB2_3: # =>This Inner Loop Header: Depth=1 movss (%r14,%r12,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $_ZSt4cout, %edi callq _ZNSo9_M_insertIdEERSoT_ movq (%rax), %rcx movq -24(%rcx), %rcx movq 240(%rax,%rcx), %r15 testq %r15, %r15 je .LBB2_9 # %bb.4: # %_ZSt13__check_facetISt5ctypeIcEERKT_PS3_.exit.i.i # in Loop: Header=BB2_3 Depth=1 cmpb $0, 56(%r15) je .LBB2_6 # %bb.5: # in Loop: Header=BB2_3 Depth=1 movzbl 67(%r15), %ecx jmp .LBB2_7 .LBB2_8: movq 8(%rsp), %rdi callq hipFree movl $.L.str.4, %esi movl $74, %edi movl %eax, %edx callq _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t movq %rbx, %rdi callq _ZdaPv movq %r14, %rdi callq _ZdaPv xorl %eax, %eax addq $144, %rsp .cfi_def_cfa_offset 48 popq %rbx .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 retq .LBB2_9: .cfi_def_cfa_offset 192 callq _ZSt16__throw_bad_castv .Lfunc_end2: .size main, .Lfunc_end2-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t .type _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t,@function _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t: # @_ZL17CheckCudaErrorAuxPKcjS0_10hipError_t .cfi_startproc # %bb.0: testl %edx, %edx jne .LBB3_2 # %bb.1: retq .LBB3_2: pushq %rbp .cfi_def_cfa_offset 16 pushq %r14 .cfi_def_cfa_offset 24 pushq %rbx .cfi_def_cfa_offset 32 .cfi_offset %rbx, -32 .cfi_offset %r14, -24 .cfi_offset %rbp, -16 movl %edi, %ebx movl $_ZSt4cerr, %edi movl %edx, %ebp callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movl $.L.str.5, %esi movq %rax, %rdi callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movq %rax, %r14 movl %ebp, %edi callq hipGetErrorString movq %r14, %rdi movq %rax, %rsi callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movl $.L.str.6, %esi movq %rax, %rdi callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movq %rax, %rdi movl %ebp, %esi callq _ZNSolsEi movl $.L.str.7, %esi movq %rax, %rdi callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movl $.L.str, %esi movq %rax, %rdi callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movl $.L.str.8, %esi movq %rax, %rdi callq _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc movq %rax, %rdi movl %ebx, %esi callq _ZNSolsEj movq %rax, %rdi callq _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_ movl $1, %edi callq exit .Lfunc_end3: .size _ZL17CheckCudaErrorAuxPKcjS0_10hipError_t, .Lfunc_end3-_ZL17CheckCudaErrorAuxPKcjS0_10hipError_t .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB4_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB4_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z10heatKernelPfjjfff, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end4: .size __hip_module_ctor, .Lfunc_end4-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB5_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB5_2: retq .Lfunc_end5: .size __hip_module_dtor, .Lfunc_end5-__hip_module_dtor .cfi_endproc # -- End function .type _Z10heatKernelPfjjfff,@object # @_Z10heatKernelPfjjfff .section .rodata,"a",@progbits .globl _Z10heatKernelPfjjfff .p2align 3, 0x0 _Z10heatKernelPfjjfff: .quad _Z25__device_stub__heatKernelPfjjfff .size _Z10heatKernelPfjjfff, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "/home/ubuntu/Datasets/stackv2/train-structured-repos-hip/xobotun/TP_CUDA/master/Sem1/Sem1/kernel.hip" .size .L.str, 101 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "hipMalloc((void **)&gpuData, sizeof(float)*(WORK_SIZE + 2))" .size .L.str.1, 60 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "hipMemcpy(gpuData, hostData, sizeof(float)*(WORK_SIZE + 2), hipMemcpyHostToDevice)" .size .L.str.2, 83 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "hipMemcpy(result, gpuData, sizeof(float)*(WORK_SIZE + 2), hipMemcpyDeviceToHost)" .size .L.str.3, 81 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "hipFree(gpuData)" .size .L.str.4, 17 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz " returned " .size .L.str.5, 11 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz "(" .size .L.str.6, 2 .type .L.str.7,@object # @.str.7 .L.str.7: .asciz ") at " .size .L.str.7, 6 .type .L.str.8,@object # @.str.8 .L.str.8: .asciz ":" .size .L.str.8, 2 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z10heatKernelPfjjfff" .size .L__unnamed_1, 22 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z25__device_stub__heatKernelPfjjfff .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z10heatKernelPfjjfff .addrsig_sym _ZSt4cout .addrsig_sym _ZSt4cerr .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
#include "includes.h" __global__ void gLogSoftmaxGrad(float* grad, const float* adj, const float* val, const int rows, const int cols) { for(int bid = 0; bid < rows; bid += gridDim.x) { int j = bid + blockIdx.x; if(j < rows) { extern __shared__ float _share[]; float* _sum = _share + blockDim.x; float* gradRow = grad + j * cols; const float* adjRow = adj + j * cols; const float* valRow = val + j * cols; _sum[threadIdx.x] = 0.0; for(int tid = 0; tid < cols; tid += blockDim.x) { int id = tid + threadIdx.x; if(id < cols) { _sum[threadIdx.x] += adjRow[id]; } } __syncthreads(); int len = blockDim.x; while(len != 1) { __syncthreads(); int skip = (len + 1) >> 1; if(threadIdx.x < (len >> 1)) _sum[threadIdx.x] += _sum[threadIdx.x + skip]; len = (len + 1) >> 1; } __syncthreads(); for(int tid = 0; tid < cols; tid += blockDim.x) { int id = tid + threadIdx.x; if(id < cols) gradRow[id] += adjRow[id] - (expf(valRow[id]) * _sum[0]); } } } }
code for sm_80 Function : _Z15gLogSoftmaxGradPfPKfS1_ii .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ IMAD.MOV.U32 R1, RZ, RZ, c[0x0][0x28] ; /* 0x00000a00ff017624 */ /* 0x000fe400078e00ff */ /*0010*/ IMAD.MOV.U32 R0, RZ, RZ, c[0x0][0x178] ; /* 0x00005e00ff007624 */ /* 0x000fca00078e00ff */ /*0020*/ ISETP.GE.AND P0, PT, R0, 0x1, PT ; /* 0x000000010000780c */ /* 0x000fda0003f06270 */ /*0030*/ @!P0 EXIT ; /* 0x000000000000894d */ /* 0x000fea0003800000 */ /*0040*/ S2R R0, SR_TID.X ; /* 0x0000000000007919 */ /* 0x000e220000002100 */ /*0050*/ S2UR UR5, SR_CTAID.X ; /* 0x00000000000579c3 */ /* 0x000e620000002500 */ /*0060*/ ULDC UR4, c[0x0][0x0] ; /* 0x0000000000047ab9 */ /* 0x000fe20000000800 */ /*0070*/ IMAD.MOV.U32 R13, RZ, RZ, RZ ; /* 0x000000ffff0d7224 */ /* 0x000fe200078e00ff */ /*0080*/ USHF.L.U32 UR4, UR4, 0x2, URZ ; /* 0x0000000204047899 */ /* 0x000fe4000800063f */ /*0090*/ ULDC.64 UR6, c[0x0][0x118] ; /* 0x0000460000067ab9 */ /* 0x000fc80000000a00 */ /*00a0*/ LEA R2, R0, UR4, 0x2 ; /* 0x0000000400027c11 */ /* 0x001fe4000f8e10ff */ /*00b0*/ IADD3 R4, R13.reuse, UR5, RZ ; /* 0x000000050d047c10 */ /* 0x042fe4000fffe0ff */ /*00c0*/ IADD3 R13, R13, c[0x0][0xc], RZ ; /* 0x000003000d0d7a10 */ /* 0x000fe40007ffe0ff */ /*00d0*/ ISETP.GE.AND P1, PT, R4, c[0x0][0x178], PT ; /* 0x00005e0004007a0c */ /* 0x000fe40003f26270 */ /*00e0*/ ISETP.GE.AND P0, PT, R13, c[0x0][0x178], PT ; /* 0x00005e000d007a0c */ /* 0x000fd60003f06270 */ /*00f0*/ @P1 BRA 0x600 ; /* 0x0000050000001947 */ /* 0x001fea0003800000 */ /*0100*/ STS [R0.X4+UR4], RZ ; /* 0x000000ff00007988 */ /* 0x0001e20008004804 */ /*0110*/ IMAD.MOV.U32 R3, RZ, RZ, c[0x0][0x17c] ; /* 0x00005f00ff037624 */ /* 0x000fca00078e00ff */ /*0120*/ ISETP.GE.AND P1, PT, R3, 0x1, PT ; /* 0x000000010300780c */ /* 0x000fe20003f26270 */ /*0130*/ IMAD R3, R4, c[0x0][0x17c], RZ ; /* 0x00005f0004037a24 */ /* 0x000fca00078e02ff */ /*0140*/ SHF.R.S32.HI R4, RZ, 0x1f, R3 ; /* 0x0000001fff047819 */ /* 0x000fce0000011403 */ /*0150*/ @!P1 BRA 0x270 ; /* 0x0000011000009947 */ /* 0x000fea0003800000 */ /*0160*/ IMAD.MOV.U32 R5, RZ, RZ, RZ ; /* 0x000000ffff057224 */ /* 0x001fe400078e00ff */ /*0170*/ IMAD.MOV.U32 R9, RZ, RZ, RZ ; /* 0x000000ffff097224 */ /* 0x000fe400078e00ff */ /*0180*/ IMAD.IADD R8, R0, 0x1, R5 ; /* 0x0000000100087824 */ /* 0x000fe200078e0205 */ /*0190*/ IADD3 R5, R5, c[0x0][0x0], RZ ; /* 0x0000000005057a10 */ /* 0x000fe20007ffe0ff */ /*01a0*/ BSSY B0, 0x260 ; /* 0x000000b000007945 */ /* 0x000fe60003800000 */ /*01b0*/ ISETP.GE.AND P3, PT, R8, c[0x0][0x17c], PT ; /* 0x00005f0008007a0c */ /* 0x000fe40003f66270 */ /*01c0*/ ISETP.GE.AND P2, PT, R5, c[0x0][0x17c], PT ; /* 0x00005f0005007a0c */ /* 0x000fd60003f46270 */ /*01d0*/ @P3 BRA 0x250 ; /* 0x0000007000003947 */ /* 0x001fea0003800000 */ /*01e0*/ IADD3 R7, P3, R3, R8, RZ ; /* 0x0000000803077210 */ /* 0x000fc80007f7e0ff */ /*01f0*/ LEA.HI.X.SX32 R8, R8, R4, 0x1, P3 ; /* 0x0000000408087211 */ /* 0x000fe400018f0eff */ /*0200*/ LEA R6, P3, R7, c[0x0][0x168], 0x2 ; /* 0x00005a0007067a11 */ /* 0x000fc800078610ff */ /*0210*/ LEA.HI.X R7, R7, c[0x0][0x16c], R8, 0x2, P3 ; /* 0x00005b0007077a11 */ /* 0x000fca00018f1408 */ /*0220*/ LDG.E R6, [R6.64] ; /* 0x0000000606067981 */ /* 0x000ea4000c1e1900 */ /*0230*/ FADD R9, R9, R6 ; /* 0x0000000609097221 */ /* 0x004fca0000000000 */ /*0240*/ STS [R0.X4+UR4], R9 ; /* 0x0000000900007988 */ /* 0x0001e40008004804 */ /*0250*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0260*/ @!P2 BRA 0x180 ; /* 0xffffff100000a947 */ /* 0x000fea000383ffff */ /*0270*/ IMAD.MOV.U32 R5, RZ, RZ, c[0x0][0x0] ; /* 0x00000000ff057624 */ /* 0x001fe200078e00ff */ /*0280*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fe80000010000 */ /*0290*/ ISETP.NE.AND P2, PT, R5, 0x1, PT ; /* 0x000000010500780c */ /* 0x000fda0003f45270 */ /*02a0*/ @!P2 BRA 0x3a0 ; /* 0x000000f00000a947 */ /* 0x000fea0003800000 */ /*02b0*/ IMAD.MOV.U32 R5, RZ, RZ, c[0x0][0x0] ; /* 0x00000000ff057624 */ /* 0x000fca00078e00ff */ /*02c0*/ SHF.R.S32.HI R7, RZ, 0x1, R5 ; /* 0x00000001ff077819 */ /* 0x000fe20000011405 */ /*02d0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fe60000010000 */ /*02e0*/ ISETP.GE.U32.AND P2, PT, R0, R7, PT ; /* 0x000000070000720c */ /* 0x000fda0003f46070 */ /*02f0*/ @!P2 LEA R6, R5.reuse, 0x2, 0x1 ; /* 0x000000020506a811 */ /* 0x040fe400078e08ff */ /*0300*/ IADD3 R5, R5, 0x1, RZ ; /* 0x0000000105057810 */ /* 0x000fe40007ffe0ff */ /*0310*/ @!P2 LOP3.LUT R7, R6, 0xfffffffc, RZ, 0xc0, !PT ; /* 0xfffffffc0607a812 */ /* 0x000fe400078ec0ff */ /*0320*/ SHF.R.S32.HI R5, RZ, 0x1, R5 ; /* 0x00000001ff057819 */ /* 0x000fe20000011405 */ /*0330*/ @!P2 LDS R6, [R0.X4+UR4] ; /* 0x000000040006a984 */ /* 0x000fe40008004800 */ /*0340*/ @!P2 IMAD.IADD R7, R2, 0x1, R7 ; /* 0x000000010207a824 */ /* 0x000fcc00078e0207 */ /*0350*/ @!P2 LDS R7, [R7] ; /* 0x000000000707a984 */ /* 0x000e240000000800 */ /*0360*/ @!P2 FADD R9, R6, R7 ; /* 0x000000070609a221 */ /* 0x001fca0000000000 */ /*0370*/ @!P2 STS [R0.X4+UR4], R9 ; /* 0x000000090000a988 */ /* 0x0001e20008004804 */ /*0380*/ ISETP.NE.AND P2, PT, R5, 0x1, PT ; /* 0x000000010500780c */ /* 0x000fda0003f45270 */ /*0390*/ @P2 BRA 0x2c0 ; /* 0xffffff2000002947 */ /* 0x001fea000383ffff */ /*03a0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*03b0*/ @!P1 BRA 0x600 ; /* 0x0000024000009947 */ /* 0x000fea0003800000 */ /*03c0*/ IMAD.MOV.U32 R5, RZ, RZ, RZ ; /* 0x000000ffff057224 */ /* 0x000fc800078e00ff */ /*03d0*/ IMAD.IADD R8, R0, 0x1, R5 ; /* 0x0000000100087824 */ /* 0x000fe200078e0205 */ /*03e0*/ IADD3 R5, R5, c[0x0][0x0], RZ ; /* 0x0000000005057a10 */ /* 0x000fe20007ffe0ff */ /*03f0*/ BSSY B0, 0x5f0 ; /* 0x000001f000007945 */ /* 0x000fe60003800000 */ /*0400*/ ISETP.GE.AND P2, PT, R8, c[0x0][0x17c], PT ; /* 0x00005f0008007a0c */ /* 0x000fc40003f46270 */ /*0410*/ ISETP.GE.AND P1, PT, R5, c[0x0][0x17c], PT ; /* 0x00005f0005007a0c */ /* 0x000fd60003f26270 */ /*0420*/ @P2 BRA 0x5e0 ; /* 0x000001b000002947 */ /* 0x001fea0003800000 */ /*0430*/ IADD3 R7, P2, R3, R8, RZ ; /* 0x0000000803077210 */ /* 0x000fe20007f5e0ff */ /*0440*/ IMAD.MOV.U32 R19, RZ, RZ, 0x3bbb989d ; /* 0x3bbb989dff137424 */ /* 0x000fe200078e00ff */ /*0450*/ LDS R15, [UR4] ; /* 0x00000004ff0f7984 */ /* 0x000e240008000800 */ /*0460*/ LEA.HI.X.SX32 R8, R8, R4, 0x1, P2 ; /* 0x0000000408087211 */ /* 0x000fe200010f0eff */ /*0470*/ IMAD.SHL.U32 R6, R7, 0x4, RZ ; /* 0x0000000407067824 */ /* 0x000fc600078e00ff */ /*0480*/ SHF.L.U64.HI R7, R7, 0x2, R8 ; /* 0x0000000207077819 */ /* 0x000fe40000010208 */ /*0490*/ IADD3 R8, P2, R6, c[0x0][0x170], RZ ; /* 0x00005c0006087a10 */ /* 0x000fc80007f5e0ff */ /*04a0*/ IADD3.X R9, R7, c[0x0][0x174], RZ, P2, !PT ; /* 0x00005d0007097a10 */ /* 0x000fca00017fe4ff */ /*04b0*/ LDG.E R8, [R8.64] ; /* 0x0000000608087981 */ /* 0x000ea2000c1e1900 */ /*04c0*/ IADD3 R10, P2, R6.reuse, c[0x0][0x168], RZ ; /* 0x00005a00060a7a10 */ /* 0x040fe40007f5e0ff */ /*04d0*/ IADD3 R6, P3, R6, c[0x0][0x160], RZ ; /* 0x0000580006067a10 */ /* 0x000fe40007f7e0ff */ /*04e0*/ IADD3.X R11, R7.reuse, c[0x0][0x16c], RZ, P2, !PT ; /* 0x00005b00070b7a10 */ /* 0x040fe400017fe4ff */ /*04f0*/ IADD3.X R7, R7, c[0x0][0x164], RZ, P3, !PT ; /* 0x0000590007077a10 */ /* 0x000fc60001ffe4ff */ /*0500*/ LDG.E R10, [R10.64] ; /* 0x000000060a0a7981 */ /* 0x000e28000c1e1900 */ /*0510*/ LDG.E R17, [R6.64] ; /* 0x0000000606117981 */ /* 0x000ee2000c1e1900 */ /*0520*/ IMAD.MOV.U32 R21, RZ, RZ, 0x437c0000 ; /* 0x437c0000ff157424 */ /* 0x000fe400078e00ff */ /*0530*/ FFMA.SAT R12, R8, R19, 0.5 ; /* 0x3f000000080c7423 */ /* 0x004fc80000002013 */ /*0540*/ FFMA.RM R12, R12, R21, 12582913 ; /* 0x4b4000010c0c7423 */ /* 0x000fc80000004015 */ /*0550*/ FADD R9, R12.reuse, -12583039 ; /* 0xcb40007f0c097421 */ /* 0x040fe40000000000 */ /*0560*/ IMAD.SHL.U32 R12, R12, 0x800000, RZ ; /* 0x008000000c0c7824 */ /* 0x000fe400078e00ff */ /*0570*/ FFMA R9, R8, 1.4426950216293334961, -R9 ; /* 0x3fb8aa3b08097823 */ /* 0x000fc80000000809 */ /*0580*/ FFMA R9, R8, 1.925963033500011079e-08, R9 ; /* 0x32a5706008097823 */ /* 0x000fcc0000000009 */ /*0590*/ MUFU.EX2 R9, R9 ; /* 0x0000000900097308 */ /* 0x000e640000000800 */ /*05a0*/ FMUL R12, R12, R9 ; /* 0x000000090c0c7220 */ /* 0x002fc80000400000 */ /*05b0*/ FFMA R10, -R12, R15, R10 ; /* 0x0000000f0c0a7223 */ /* 0x001fc8000000010a */ /*05c0*/ FADD R17, R10, R17 ; /* 0x000000110a117221 */ /* 0x008fca0000000000 */ /*05d0*/ STG.E [R6.64], R17 ; /* 0x0000001106007986 */ /* 0x0001e4000c101906 */ /*05e0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*05f0*/ @!P1 BRA 0x3d0 ; /* 0xfffffdd000009947 */ /* 0x000fea000383ffff */ /*0600*/ @!P0 BRA 0xb0 ; /* 0xfffffaa000008947 */ /* 0x000fea000383ffff */ /*0610*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0620*/ BRA 0x620; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0630*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0640*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0650*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0660*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0670*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0680*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0690*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*06a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*06b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*06c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*06d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*06e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*06f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include "includes.h" __global__ void gLogSoftmaxGrad(float* grad, const float* adj, const float* val, const int rows, const int cols) { for(int bid = 0; bid < rows; bid += gridDim.x) { int j = bid + blockIdx.x; if(j < rows) { extern __shared__ float _share[]; float* _sum = _share + blockDim.x; float* gradRow = grad + j * cols; const float* adjRow = adj + j * cols; const float* valRow = val + j * cols; _sum[threadIdx.x] = 0.0; for(int tid = 0; tid < cols; tid += blockDim.x) { int id = tid + threadIdx.x; if(id < cols) { _sum[threadIdx.x] += adjRow[id]; } } __syncthreads(); int len = blockDim.x; while(len != 1) { __syncthreads(); int skip = (len + 1) >> 1; if(threadIdx.x < (len >> 1)) _sum[threadIdx.x] += _sum[threadIdx.x + skip]; len = (len + 1) >> 1; } __syncthreads(); for(int tid = 0; tid < cols; tid += blockDim.x) { int id = tid + threadIdx.x; if(id < cols) gradRow[id] += adjRow[id] - (expf(valRow[id]) * _sum[0]); } } } }
.file "tmpxft_0012e87e_00000000-6_gLogSoftmaxGrad.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2029: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2029: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z43__device_stub__Z15gLogSoftmaxGradPfPKfS1_iiPfPKfS1_ii .type _Z43__device_stub__Z15gLogSoftmaxGradPfPKfS1_iiPfPKfS1_ii, @function _Z43__device_stub__Z15gLogSoftmaxGradPfPKfS1_iiPfPKfS1_ii: .LFB2051: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 24(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movl %ecx, 4(%rsp) movl %r8d, (%rsp) movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax leaq 24(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) leaq 4(%rsp), %rax movq %rax, 120(%rsp) movq %rsp, %rax movq %rax, 128(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 56(%rsp) movl $1, 60(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) leaq 40(%rsp), %rcx leaq 32(%rsp), %rdx leaq 60(%rsp), %rsi leaq 48(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 136(%rsp), %rax subq %fs:40, %rax jne .L8 addq $152, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 40(%rsp) .cfi_def_cfa_offset 168 pushq 40(%rsp) .cfi_def_cfa_offset 176 leaq 112(%rsp), %r9 movq 76(%rsp), %rcx movl 84(%rsp), %r8d movq 64(%rsp), %rsi movl 72(%rsp), %edx leaq _Z15gLogSoftmaxGradPfPKfS1_ii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z43__device_stub__Z15gLogSoftmaxGradPfPKfS1_iiPfPKfS1_ii, .-_Z43__device_stub__Z15gLogSoftmaxGradPfPKfS1_iiPfPKfS1_ii .globl _Z15gLogSoftmaxGradPfPKfS1_ii .type _Z15gLogSoftmaxGradPfPKfS1_ii, @function _Z15gLogSoftmaxGradPfPKfS1_ii: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z43__device_stub__Z15gLogSoftmaxGradPfPKfS1_iiPfPKfS1_ii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z15gLogSoftmaxGradPfPKfS1_ii, .-_Z15gLogSoftmaxGradPfPKfS1_ii .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "_Z15gLogSoftmaxGradPfPKfS1_ii" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2054: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC0(%rip), %rdx movq %rdx, %rcx leaq _Z15gLogSoftmaxGradPfPKfS1_ii(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2054: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include "includes.h" __global__ void gLogSoftmaxGrad(float* grad, const float* adj, const float* val, const int rows, const int cols) { for(int bid = 0; bid < rows; bid += gridDim.x) { int j = bid + blockIdx.x; if(j < rows) { extern __shared__ float _share[]; float* _sum = _share + blockDim.x; float* gradRow = grad + j * cols; const float* adjRow = adj + j * cols; const float* valRow = val + j * cols; _sum[threadIdx.x] = 0.0; for(int tid = 0; tid < cols; tid += blockDim.x) { int id = tid + threadIdx.x; if(id < cols) { _sum[threadIdx.x] += adjRow[id]; } } __syncthreads(); int len = blockDim.x; while(len != 1) { __syncthreads(); int skip = (len + 1) >> 1; if(threadIdx.x < (len >> 1)) _sum[threadIdx.x] += _sum[threadIdx.x + skip]; len = (len + 1) >> 1; } __syncthreads(); for(int tid = 0; tid < cols; tid += blockDim.x) { int id = tid + threadIdx.x; if(id < cols) gradRow[id] += adjRow[id] - (expf(valRow[id]) * _sum[0]); } } } }
#include <hip/hip_runtime.h> #include "includes.h" __global__ void gLogSoftmaxGrad(float* grad, const float* adj, const float* val, const int rows, const int cols) { for(int bid = 0; bid < rows; bid += gridDim.x) { int j = bid + blockIdx.x; if(j < rows) { extern __shared__ float _share[]; float* _sum = _share + blockDim.x; float* gradRow = grad + j * cols; const float* adjRow = adj + j * cols; const float* valRow = val + j * cols; _sum[threadIdx.x] = 0.0; for(int tid = 0; tid < cols; tid += blockDim.x) { int id = tid + threadIdx.x; if(id < cols) { _sum[threadIdx.x] += adjRow[id]; } } __syncthreads(); int len = blockDim.x; while(len != 1) { __syncthreads(); int skip = (len + 1) >> 1; if(threadIdx.x < (len >> 1)) _sum[threadIdx.x] += _sum[threadIdx.x + skip]; len = (len + 1) >> 1; } __syncthreads(); for(int tid = 0; tid < cols; tid += blockDim.x) { int id = tid + threadIdx.x; if(id < cols) gradRow[id] += adjRow[id] - (expf(valRow[id]) * _sum[0]); } } } }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include "includes.h" __global__ void gLogSoftmaxGrad(float* grad, const float* adj, const float* val, const int rows, const int cols) { for(int bid = 0; bid < rows; bid += gridDim.x) { int j = bid + blockIdx.x; if(j < rows) { extern __shared__ float _share[]; float* _sum = _share + blockDim.x; float* gradRow = grad + j * cols; const float* adjRow = adj + j * cols; const float* valRow = val + j * cols; _sum[threadIdx.x] = 0.0; for(int tid = 0; tid < cols; tid += blockDim.x) { int id = tid + threadIdx.x; if(id < cols) { _sum[threadIdx.x] += adjRow[id]; } } __syncthreads(); int len = blockDim.x; while(len != 1) { __syncthreads(); int skip = (len + 1) >> 1; if(threadIdx.x < (len >> 1)) _sum[threadIdx.x] += _sum[threadIdx.x + skip]; len = (len + 1) >> 1; } __syncthreads(); for(int tid = 0; tid < cols; tid += blockDim.x) { int id = tid + threadIdx.x; if(id < cols) gradRow[id] += adjRow[id] - (expf(valRow[id]) * _sum[0]); } } } }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z15gLogSoftmaxGradPfPKfS1_ii .globl _Z15gLogSoftmaxGradPfPKfS1_ii .p2align 8 .type _Z15gLogSoftmaxGradPfPKfS1_ii,@function _Z15gLogSoftmaxGradPfPKfS1_ii: s_load_b32 s12, s[0:1], 0x18 s_waitcnt lgkmcnt(0) s_cmp_lt_i32 s12, 1 s_cbranch_scc1 .LBB0_19 s_clause 0x2 s_load_b64 s[2:3], s[0:1], 0x1c s_load_b128 s[4:7], s[0:1], 0x0 s_load_b64 s[8:9], s[0:1], 0x10 s_add_u32 s0, s0, 32 s_addc_u32 s1, s1, 0 v_mov_b32_e32 v3, 0 s_mov_b32 s14, 0 s_waitcnt lgkmcnt(0) s_cmp_gt_i32 s2, 0 s_cselect_b32 s13, -1, 0 s_branch .LBB0_3 .LBB0_2: s_add_i32 s14, s3, s14 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_ge_i32 s14, s12 s_cbranch_scc1 .LBB0_19 .LBB0_3: s_add_i32 s10, s14, s15 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_ge_i32 s10, s12 s_cbranch_scc1 .LBB0_2 s_load_b32 s11, s[0:1], 0xc s_mul_i32 s10, s10, s2 s_waitcnt lgkmcnt(0) s_and_b32 s16, s11, 0xffff s_ashr_i32 s11, s10, 31 s_lshl_b32 s17, s16, 2 s_lshl_b64 s[10:11], s[10:11], 2 s_add_i32 s17, s17, 0 s_add_u32 s18, s6, s10 v_lshl_add_u32 v4, v0, 2, s17 s_addc_u32 s19, s7, s11 s_and_not1_b32 vcc_lo, exec_lo, s13 ds_store_b32 v4, v3 s_cbranch_vccnz .LBB0_9 s_mov_b32 s20, 0 s_branch .LBB0_7 .p2align 6 .LBB0_6: s_or_b32 exec_lo, exec_lo, s21 s_add_i32 s20, s20, s16 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_ge_i32 s20, s2 s_cbranch_scc1 .LBB0_9 .LBB0_7: v_add_nc_u32_e32 v1, s20, v0 s_mov_b32 s21, exec_lo s_delay_alu instid0(VALU_DEP_1) v_cmpx_gt_i32_e64 s2, v1 s_cbranch_execz .LBB0_6 v_ashrrev_i32_e32 v2, 31, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[1:2], 2, v[1:2] v_add_co_u32 v1, vcc_lo, s18, v1 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v2, vcc_lo, s19, v2, vcc_lo global_load_b32 v1, v[1:2], off ds_load_b32 v2, v4 s_waitcnt vmcnt(0) lgkmcnt(0) v_add_f32_e32 v1, v1, v2 ds_store_b32 v4, v1 s_branch .LBB0_6 .LBB0_9: s_cmp_eq_u32 s16, 1 s_waitcnt lgkmcnt(0) s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv s_cbranch_scc1 .LBB0_14 s_mov_b32 s20, s16 s_branch .LBB0_12 .p2align 6 .LBB0_11: s_or_b32 exec_lo, exec_lo, s21 s_cmp_lg_u32 s20, 1 s_cbranch_scc0 .LBB0_14 .LBB0_12: s_lshr_b32 s21, s20, 1 s_add_i32 s20, s20, 1 v_cmp_gt_u32_e32 vcc_lo, s21, v0 s_lshr_b32 s20, s20, 1 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_and_saveexec_b32 s21, vcc_lo s_cbranch_execz .LBB0_11 v_lshl_add_u32 v1, s20, 2, v4 ds_load_b32 v1, v1 ds_load_b32 v2, v4 s_waitcnt lgkmcnt(0) v_add_f32_e32 v1, v1, v2 ds_store_b32 v4, v1 s_branch .LBB0_11 .LBB0_14: s_and_not1_b32 vcc_lo, exec_lo, s13 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_cbranch_vccnz .LBB0_2 s_add_u32 s20, s4, s10 s_addc_u32 s21, s5, s11 s_add_u32 s10, s8, s10 s_addc_u32 s11, s9, s11 s_mov_b32 s22, 0 s_branch .LBB0_17 .LBB0_16: s_or_b32 exec_lo, exec_lo, s23 s_add_i32 s22, s22, s16 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_lt_i32 s22, s2 s_cbranch_scc0 .LBB0_2 .LBB0_17: v_add_nc_u32_e32 v1, s22, v0 s_mov_b32 s23, exec_lo s_delay_alu instid0(VALU_DEP_1) v_cmpx_gt_i32_e64 s2, v1 s_cbranch_execz .LBB0_16 v_ashrrev_i32_e32 v2, 31, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[1:2], 2, v[1:2] v_add_co_u32 v4, vcc_lo, s10, v1 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v5, vcc_lo, s11, v2, vcc_lo global_load_b32 v6, v[4:5], off v_add_co_u32 v4, vcc_lo, s18, v1 v_add_co_ci_u32_e32 v5, vcc_lo, s19, v2, vcc_lo v_add_co_u32 v1, vcc_lo, s20, v1 v_add_co_ci_u32_e32 v2, vcc_lo, s21, v2, vcc_lo global_load_b32 v4, v[4:5], off global_load_b32 v5, v[1:2], off s_waitcnt vmcnt(2) v_mul_f32_e32 v7, 0x3fb8aa3b, v6 v_cmp_ngt_f32_e32 vcc_lo, 0xc2ce8ed0, v6 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_fma_f32 v8, v6, 0x3fb8aa3b, -v7 v_rndne_f32_e32 v9, v7 v_dual_fmac_f32 v8, 0x32a5705f, v6 :: v_dual_sub_f32 v7, v7, v9 v_cvt_i32_f32_e32 v9, v9 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_dual_add_f32 v7, v7, v8 :: v_dual_mov_b32 v8, s17 v_exp_f32_e32 v7, v7 ds_load_b32 v8, v8 v_ldexp_f32 v7, v7, v9 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_cndmask_b32_e32 v7, 0, v7, vcc_lo v_cmp_nlt_f32_e32 vcc_lo, 0x42b17218, v6 v_cndmask_b32_e32 v6, 0x7f800000, v7, vcc_lo s_waitcnt vmcnt(1) lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_fma_f32 v4, -v6, v8, v4 s_waitcnt vmcnt(0) v_add_f32_e32 v4, v5, v4 global_store_b32 v[1:2], v4, off s_branch .LBB0_16 .LBB0_19: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z15gLogSoftmaxGradPfPKfS1_ii .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 288 .amdhsa_user_sgpr_count 15 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 0 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 0 .amdhsa_next_free_vgpr 10 .amdhsa_next_free_sgpr 24 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z15gLogSoftmaxGradPfPKfS1_ii, .Lfunc_end0-_Z15gLogSoftmaxGradPfPKfS1_ii .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 8 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .offset: 24 .size: 4 .value_kind: by_value - .offset: 28 .size: 4 .value_kind: by_value - .offset: 32 .size: 4 .value_kind: hidden_block_count_x - .offset: 36 .size: 4 .value_kind: hidden_block_count_y - .offset: 40 .size: 4 .value_kind: hidden_block_count_z - .offset: 44 .size: 2 .value_kind: hidden_group_size_x - .offset: 46 .size: 2 .value_kind: hidden_group_size_y - .offset: 48 .size: 2 .value_kind: hidden_group_size_z - .offset: 50 .size: 2 .value_kind: hidden_remainder_x - .offset: 52 .size: 2 .value_kind: hidden_remainder_y - .offset: 54 .size: 2 .value_kind: hidden_remainder_z - .offset: 72 .size: 8 .value_kind: hidden_global_offset_x - .offset: 80 .size: 8 .value_kind: hidden_global_offset_y - .offset: 88 .size: 8 .value_kind: hidden_global_offset_z - .offset: 96 .size: 2 .value_kind: hidden_grid_dims - .offset: 152 .size: 4 .value_kind: hidden_dynamic_lds_size .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 288 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z15gLogSoftmaxGradPfPKfS1_ii .private_segment_fixed_size: 0 .sgpr_count: 26 .sgpr_spill_count: 0 .symbol: _Z15gLogSoftmaxGradPfPKfS1_ii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 10 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <hip/hip_runtime.h> #include "includes.h" __global__ void gLogSoftmaxGrad(float* grad, const float* adj, const float* val, const int rows, const int cols) { for(int bid = 0; bid < rows; bid += gridDim.x) { int j = bid + blockIdx.x; if(j < rows) { extern __shared__ float _share[]; float* _sum = _share + blockDim.x; float* gradRow = grad + j * cols; const float* adjRow = adj + j * cols; const float* valRow = val + j * cols; _sum[threadIdx.x] = 0.0; for(int tid = 0; tid < cols; tid += blockDim.x) { int id = tid + threadIdx.x; if(id < cols) { _sum[threadIdx.x] += adjRow[id]; } } __syncthreads(); int len = blockDim.x; while(len != 1) { __syncthreads(); int skip = (len + 1) >> 1; if(threadIdx.x < (len >> 1)) _sum[threadIdx.x] += _sum[threadIdx.x + skip]; len = (len + 1) >> 1; } __syncthreads(); for(int tid = 0; tid < cols; tid += blockDim.x) { int id = tid + threadIdx.x; if(id < cols) gradRow[id] += adjRow[id] - (expf(valRow[id]) * _sum[0]); } } } }
.text .file "gLogSoftmaxGrad.hip" .globl _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii # -- Begin function _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii .p2align 4, 0x90 .type _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii,@function _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii: # @_Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movq %rdx, 56(%rsp) movl %ecx, 4(%rsp) movl %r8d, (%rsp) leaq 72(%rsp), %rax movq %rax, 80(%rsp) leaq 64(%rsp), %rax movq %rax, 88(%rsp) leaq 56(%rsp), %rax movq %rax, 96(%rsp) leaq 4(%rsp), %rax movq %rax, 104(%rsp) movq %rsp, %rax movq %rax, 112(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 80(%rsp), %r9 movl $_Z15gLogSoftmaxGradPfPKfS1_ii, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $136, %rsp .cfi_adjust_cfa_offset -136 retq .Lfunc_end0: .size _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii, .Lfunc_end0-_Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB1_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB1_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z15gLogSoftmaxGradPfPKfS1_ii, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end1: .size __hip_module_ctor, .Lfunc_end1-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB2_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB2_2: retq .Lfunc_end2: .size __hip_module_dtor, .Lfunc_end2-__hip_module_dtor .cfi_endproc # -- End function .type _Z15gLogSoftmaxGradPfPKfS1_ii,@object # @_Z15gLogSoftmaxGradPfPKfS1_ii .section .rodata,"a",@progbits .globl _Z15gLogSoftmaxGradPfPKfS1_ii .p2align 3, 0x0 _Z15gLogSoftmaxGradPfPKfS1_ii: .quad _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii .size _Z15gLogSoftmaxGradPfPKfS1_ii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z15gLogSoftmaxGradPfPKfS1_ii" .size .L__unnamed_1, 30 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z15gLogSoftmaxGradPfPKfS1_ii .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA device assembly to AMD device assembly.
code for sm_80 Function : _Z15gLogSoftmaxGradPfPKfS1_ii .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ IMAD.MOV.U32 R1, RZ, RZ, c[0x0][0x28] ; /* 0x00000a00ff017624 */ /* 0x000fe400078e00ff */ /*0010*/ IMAD.MOV.U32 R0, RZ, RZ, c[0x0][0x178] ; /* 0x00005e00ff007624 */ /* 0x000fca00078e00ff */ /*0020*/ ISETP.GE.AND P0, PT, R0, 0x1, PT ; /* 0x000000010000780c */ /* 0x000fda0003f06270 */ /*0030*/ @!P0 EXIT ; /* 0x000000000000894d */ /* 0x000fea0003800000 */ /*0040*/ S2R R0, SR_TID.X ; /* 0x0000000000007919 */ /* 0x000e220000002100 */ /*0050*/ S2UR UR5, SR_CTAID.X ; /* 0x00000000000579c3 */ /* 0x000e620000002500 */ /*0060*/ ULDC UR4, c[0x0][0x0] ; /* 0x0000000000047ab9 */ /* 0x000fe20000000800 */ /*0070*/ IMAD.MOV.U32 R13, RZ, RZ, RZ ; /* 0x000000ffff0d7224 */ /* 0x000fe200078e00ff */ /*0080*/ USHF.L.U32 UR4, UR4, 0x2, URZ ; /* 0x0000000204047899 */ /* 0x000fe4000800063f */ /*0090*/ ULDC.64 UR6, c[0x0][0x118] ; /* 0x0000460000067ab9 */ /* 0x000fc80000000a00 */ /*00a0*/ LEA R2, R0, UR4, 0x2 ; /* 0x0000000400027c11 */ /* 0x001fe4000f8e10ff */ /*00b0*/ IADD3 R4, R13.reuse, UR5, RZ ; /* 0x000000050d047c10 */ /* 0x042fe4000fffe0ff */ /*00c0*/ IADD3 R13, R13, c[0x0][0xc], RZ ; /* 0x000003000d0d7a10 */ /* 0x000fe40007ffe0ff */ /*00d0*/ ISETP.GE.AND P1, PT, R4, c[0x0][0x178], PT ; /* 0x00005e0004007a0c */ /* 0x000fe40003f26270 */ /*00e0*/ ISETP.GE.AND P0, PT, R13, c[0x0][0x178], PT ; /* 0x00005e000d007a0c */ /* 0x000fd60003f06270 */ /*00f0*/ @P1 BRA 0x600 ; /* 0x0000050000001947 */ /* 0x001fea0003800000 */ /*0100*/ STS [R0.X4+UR4], RZ ; /* 0x000000ff00007988 */ /* 0x0001e20008004804 */ /*0110*/ IMAD.MOV.U32 R3, RZ, RZ, c[0x0][0x17c] ; /* 0x00005f00ff037624 */ /* 0x000fca00078e00ff */ /*0120*/ ISETP.GE.AND P1, PT, R3, 0x1, PT ; /* 0x000000010300780c */ /* 0x000fe20003f26270 */ /*0130*/ IMAD R3, R4, c[0x0][0x17c], RZ ; /* 0x00005f0004037a24 */ /* 0x000fca00078e02ff */ /*0140*/ SHF.R.S32.HI R4, RZ, 0x1f, R3 ; /* 0x0000001fff047819 */ /* 0x000fce0000011403 */ /*0150*/ @!P1 BRA 0x270 ; /* 0x0000011000009947 */ /* 0x000fea0003800000 */ /*0160*/ IMAD.MOV.U32 R5, RZ, RZ, RZ ; /* 0x000000ffff057224 */ /* 0x001fe400078e00ff */ /*0170*/ IMAD.MOV.U32 R9, RZ, RZ, RZ ; /* 0x000000ffff097224 */ /* 0x000fe400078e00ff */ /*0180*/ IMAD.IADD R8, R0, 0x1, R5 ; /* 0x0000000100087824 */ /* 0x000fe200078e0205 */ /*0190*/ IADD3 R5, R5, c[0x0][0x0], RZ ; /* 0x0000000005057a10 */ /* 0x000fe20007ffe0ff */ /*01a0*/ BSSY B0, 0x260 ; /* 0x000000b000007945 */ /* 0x000fe60003800000 */ /*01b0*/ ISETP.GE.AND P3, PT, R8, c[0x0][0x17c], PT ; /* 0x00005f0008007a0c */ /* 0x000fe40003f66270 */ /*01c0*/ ISETP.GE.AND P2, PT, R5, c[0x0][0x17c], PT ; /* 0x00005f0005007a0c */ /* 0x000fd60003f46270 */ /*01d0*/ @P3 BRA 0x250 ; /* 0x0000007000003947 */ /* 0x001fea0003800000 */ /*01e0*/ IADD3 R7, P3, R3, R8, RZ ; /* 0x0000000803077210 */ /* 0x000fc80007f7e0ff */ /*01f0*/ LEA.HI.X.SX32 R8, R8, R4, 0x1, P3 ; /* 0x0000000408087211 */ /* 0x000fe400018f0eff */ /*0200*/ LEA R6, P3, R7, c[0x0][0x168], 0x2 ; /* 0x00005a0007067a11 */ /* 0x000fc800078610ff */ /*0210*/ LEA.HI.X R7, R7, c[0x0][0x16c], R8, 0x2, P3 ; /* 0x00005b0007077a11 */ /* 0x000fca00018f1408 */ /*0220*/ LDG.E R6, [R6.64] ; /* 0x0000000606067981 */ /* 0x000ea4000c1e1900 */ /*0230*/ FADD R9, R9, R6 ; /* 0x0000000609097221 */ /* 0x004fca0000000000 */ /*0240*/ STS [R0.X4+UR4], R9 ; /* 0x0000000900007988 */ /* 0x0001e40008004804 */ /*0250*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0260*/ @!P2 BRA 0x180 ; /* 0xffffff100000a947 */ /* 0x000fea000383ffff */ /*0270*/ IMAD.MOV.U32 R5, RZ, RZ, c[0x0][0x0] ; /* 0x00000000ff057624 */ /* 0x001fe200078e00ff */ /*0280*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fe80000010000 */ /*0290*/ ISETP.NE.AND P2, PT, R5, 0x1, PT ; /* 0x000000010500780c */ /* 0x000fda0003f45270 */ /*02a0*/ @!P2 BRA 0x3a0 ; /* 0x000000f00000a947 */ /* 0x000fea0003800000 */ /*02b0*/ IMAD.MOV.U32 R5, RZ, RZ, c[0x0][0x0] ; /* 0x00000000ff057624 */ /* 0x000fca00078e00ff */ /*02c0*/ SHF.R.S32.HI R7, RZ, 0x1, R5 ; /* 0x00000001ff077819 */ /* 0x000fe20000011405 */ /*02d0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fe60000010000 */ /*02e0*/ ISETP.GE.U32.AND P2, PT, R0, R7, PT ; /* 0x000000070000720c */ /* 0x000fda0003f46070 */ /*02f0*/ @!P2 LEA R6, R5.reuse, 0x2, 0x1 ; /* 0x000000020506a811 */ /* 0x040fe400078e08ff */ /*0300*/ IADD3 R5, R5, 0x1, RZ ; /* 0x0000000105057810 */ /* 0x000fe40007ffe0ff */ /*0310*/ @!P2 LOP3.LUT R7, R6, 0xfffffffc, RZ, 0xc0, !PT ; /* 0xfffffffc0607a812 */ /* 0x000fe400078ec0ff */ /*0320*/ SHF.R.S32.HI R5, RZ, 0x1, R5 ; /* 0x00000001ff057819 */ /* 0x000fe20000011405 */ /*0330*/ @!P2 LDS R6, [R0.X4+UR4] ; /* 0x000000040006a984 */ /* 0x000fe40008004800 */ /*0340*/ @!P2 IMAD.IADD R7, R2, 0x1, R7 ; /* 0x000000010207a824 */ /* 0x000fcc00078e0207 */ /*0350*/ @!P2 LDS R7, [R7] ; /* 0x000000000707a984 */ /* 0x000e240000000800 */ /*0360*/ @!P2 FADD R9, R6, R7 ; /* 0x000000070609a221 */ /* 0x001fca0000000000 */ /*0370*/ @!P2 STS [R0.X4+UR4], R9 ; /* 0x000000090000a988 */ /* 0x0001e20008004804 */ /*0380*/ ISETP.NE.AND P2, PT, R5, 0x1, PT ; /* 0x000000010500780c */ /* 0x000fda0003f45270 */ /*0390*/ @P2 BRA 0x2c0 ; /* 0xffffff2000002947 */ /* 0x001fea000383ffff */ /*03a0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*03b0*/ @!P1 BRA 0x600 ; /* 0x0000024000009947 */ /* 0x000fea0003800000 */ /*03c0*/ IMAD.MOV.U32 R5, RZ, RZ, RZ ; /* 0x000000ffff057224 */ /* 0x000fc800078e00ff */ /*03d0*/ IMAD.IADD R8, R0, 0x1, R5 ; /* 0x0000000100087824 */ /* 0x000fe200078e0205 */ /*03e0*/ IADD3 R5, R5, c[0x0][0x0], RZ ; /* 0x0000000005057a10 */ /* 0x000fe20007ffe0ff */ /*03f0*/ BSSY B0, 0x5f0 ; /* 0x000001f000007945 */ /* 0x000fe60003800000 */ /*0400*/ ISETP.GE.AND P2, PT, R8, c[0x0][0x17c], PT ; /* 0x00005f0008007a0c */ /* 0x000fc40003f46270 */ /*0410*/ ISETP.GE.AND P1, PT, R5, c[0x0][0x17c], PT ; /* 0x00005f0005007a0c */ /* 0x000fd60003f26270 */ /*0420*/ @P2 BRA 0x5e0 ; /* 0x000001b000002947 */ /* 0x001fea0003800000 */ /*0430*/ IADD3 R7, P2, R3, R8, RZ ; /* 0x0000000803077210 */ /* 0x000fe20007f5e0ff */ /*0440*/ IMAD.MOV.U32 R19, RZ, RZ, 0x3bbb989d ; /* 0x3bbb989dff137424 */ /* 0x000fe200078e00ff */ /*0450*/ LDS R15, [UR4] ; /* 0x00000004ff0f7984 */ /* 0x000e240008000800 */ /*0460*/ LEA.HI.X.SX32 R8, R8, R4, 0x1, P2 ; /* 0x0000000408087211 */ /* 0x000fe200010f0eff */ /*0470*/ IMAD.SHL.U32 R6, R7, 0x4, RZ ; /* 0x0000000407067824 */ /* 0x000fc600078e00ff */ /*0480*/ SHF.L.U64.HI R7, R7, 0x2, R8 ; /* 0x0000000207077819 */ /* 0x000fe40000010208 */ /*0490*/ IADD3 R8, P2, R6, c[0x0][0x170], RZ ; /* 0x00005c0006087a10 */ /* 0x000fc80007f5e0ff */ /*04a0*/ IADD3.X R9, R7, c[0x0][0x174], RZ, P2, !PT ; /* 0x00005d0007097a10 */ /* 0x000fca00017fe4ff */ /*04b0*/ LDG.E R8, [R8.64] ; /* 0x0000000608087981 */ /* 0x000ea2000c1e1900 */ /*04c0*/ IADD3 R10, P2, R6.reuse, c[0x0][0x168], RZ ; /* 0x00005a00060a7a10 */ /* 0x040fe40007f5e0ff */ /*04d0*/ IADD3 R6, P3, R6, c[0x0][0x160], RZ ; /* 0x0000580006067a10 */ /* 0x000fe40007f7e0ff */ /*04e0*/ IADD3.X R11, R7.reuse, c[0x0][0x16c], RZ, P2, !PT ; /* 0x00005b00070b7a10 */ /* 0x040fe400017fe4ff */ /*04f0*/ IADD3.X R7, R7, c[0x0][0x164], RZ, P3, !PT ; /* 0x0000590007077a10 */ /* 0x000fc60001ffe4ff */ /*0500*/ LDG.E R10, [R10.64] ; /* 0x000000060a0a7981 */ /* 0x000e28000c1e1900 */ /*0510*/ LDG.E R17, [R6.64] ; /* 0x0000000606117981 */ /* 0x000ee2000c1e1900 */ /*0520*/ IMAD.MOV.U32 R21, RZ, RZ, 0x437c0000 ; /* 0x437c0000ff157424 */ /* 0x000fe400078e00ff */ /*0530*/ FFMA.SAT R12, R8, R19, 0.5 ; /* 0x3f000000080c7423 */ /* 0x004fc80000002013 */ /*0540*/ FFMA.RM R12, R12, R21, 12582913 ; /* 0x4b4000010c0c7423 */ /* 0x000fc80000004015 */ /*0550*/ FADD R9, R12.reuse, -12583039 ; /* 0xcb40007f0c097421 */ /* 0x040fe40000000000 */ /*0560*/ IMAD.SHL.U32 R12, R12, 0x800000, RZ ; /* 0x008000000c0c7824 */ /* 0x000fe400078e00ff */ /*0570*/ FFMA R9, R8, 1.4426950216293334961, -R9 ; /* 0x3fb8aa3b08097823 */ /* 0x000fc80000000809 */ /*0580*/ FFMA R9, R8, 1.925963033500011079e-08, R9 ; /* 0x32a5706008097823 */ /* 0x000fcc0000000009 */ /*0590*/ MUFU.EX2 R9, R9 ; /* 0x0000000900097308 */ /* 0x000e640000000800 */ /*05a0*/ FMUL R12, R12, R9 ; /* 0x000000090c0c7220 */ /* 0x002fc80000400000 */ /*05b0*/ FFMA R10, -R12, R15, R10 ; /* 0x0000000f0c0a7223 */ /* 0x001fc8000000010a */ /*05c0*/ FADD R17, R10, R17 ; /* 0x000000110a117221 */ /* 0x008fca0000000000 */ /*05d0*/ STG.E [R6.64], R17 ; /* 0x0000001106007986 */ /* 0x0001e4000c101906 */ /*05e0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*05f0*/ @!P1 BRA 0x3d0 ; /* 0xfffffdd000009947 */ /* 0x000fea000383ffff */ /*0600*/ @!P0 BRA 0xb0 ; /* 0xfffffaa000008947 */ /* 0x000fea000383ffff */ /*0610*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0620*/ BRA 0x620; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0630*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0640*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0650*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0660*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0670*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0680*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0690*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*06a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*06b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*06c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*06d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*06e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*06f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z15gLogSoftmaxGradPfPKfS1_ii .globl _Z15gLogSoftmaxGradPfPKfS1_ii .p2align 8 .type _Z15gLogSoftmaxGradPfPKfS1_ii,@function _Z15gLogSoftmaxGradPfPKfS1_ii: s_load_b32 s12, s[0:1], 0x18 s_waitcnt lgkmcnt(0) s_cmp_lt_i32 s12, 1 s_cbranch_scc1 .LBB0_19 s_clause 0x2 s_load_b64 s[2:3], s[0:1], 0x1c s_load_b128 s[4:7], s[0:1], 0x0 s_load_b64 s[8:9], s[0:1], 0x10 s_add_u32 s0, s0, 32 s_addc_u32 s1, s1, 0 v_mov_b32_e32 v3, 0 s_mov_b32 s14, 0 s_waitcnt lgkmcnt(0) s_cmp_gt_i32 s2, 0 s_cselect_b32 s13, -1, 0 s_branch .LBB0_3 .LBB0_2: s_add_i32 s14, s3, s14 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_ge_i32 s14, s12 s_cbranch_scc1 .LBB0_19 .LBB0_3: s_add_i32 s10, s14, s15 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_ge_i32 s10, s12 s_cbranch_scc1 .LBB0_2 s_load_b32 s11, s[0:1], 0xc s_mul_i32 s10, s10, s2 s_waitcnt lgkmcnt(0) s_and_b32 s16, s11, 0xffff s_ashr_i32 s11, s10, 31 s_lshl_b32 s17, s16, 2 s_lshl_b64 s[10:11], s[10:11], 2 s_add_i32 s17, s17, 0 s_add_u32 s18, s6, s10 v_lshl_add_u32 v4, v0, 2, s17 s_addc_u32 s19, s7, s11 s_and_not1_b32 vcc_lo, exec_lo, s13 ds_store_b32 v4, v3 s_cbranch_vccnz .LBB0_9 s_mov_b32 s20, 0 s_branch .LBB0_7 .p2align 6 .LBB0_6: s_or_b32 exec_lo, exec_lo, s21 s_add_i32 s20, s20, s16 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_ge_i32 s20, s2 s_cbranch_scc1 .LBB0_9 .LBB0_7: v_add_nc_u32_e32 v1, s20, v0 s_mov_b32 s21, exec_lo s_delay_alu instid0(VALU_DEP_1) v_cmpx_gt_i32_e64 s2, v1 s_cbranch_execz .LBB0_6 v_ashrrev_i32_e32 v2, 31, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[1:2], 2, v[1:2] v_add_co_u32 v1, vcc_lo, s18, v1 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v2, vcc_lo, s19, v2, vcc_lo global_load_b32 v1, v[1:2], off ds_load_b32 v2, v4 s_waitcnt vmcnt(0) lgkmcnt(0) v_add_f32_e32 v1, v1, v2 ds_store_b32 v4, v1 s_branch .LBB0_6 .LBB0_9: s_cmp_eq_u32 s16, 1 s_waitcnt lgkmcnt(0) s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv s_cbranch_scc1 .LBB0_14 s_mov_b32 s20, s16 s_branch .LBB0_12 .p2align 6 .LBB0_11: s_or_b32 exec_lo, exec_lo, s21 s_cmp_lg_u32 s20, 1 s_cbranch_scc0 .LBB0_14 .LBB0_12: s_lshr_b32 s21, s20, 1 s_add_i32 s20, s20, 1 v_cmp_gt_u32_e32 vcc_lo, s21, v0 s_lshr_b32 s20, s20, 1 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_and_saveexec_b32 s21, vcc_lo s_cbranch_execz .LBB0_11 v_lshl_add_u32 v1, s20, 2, v4 ds_load_b32 v1, v1 ds_load_b32 v2, v4 s_waitcnt lgkmcnt(0) v_add_f32_e32 v1, v1, v2 ds_store_b32 v4, v1 s_branch .LBB0_11 .LBB0_14: s_and_not1_b32 vcc_lo, exec_lo, s13 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_cbranch_vccnz .LBB0_2 s_add_u32 s20, s4, s10 s_addc_u32 s21, s5, s11 s_add_u32 s10, s8, s10 s_addc_u32 s11, s9, s11 s_mov_b32 s22, 0 s_branch .LBB0_17 .LBB0_16: s_or_b32 exec_lo, exec_lo, s23 s_add_i32 s22, s22, s16 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_lt_i32 s22, s2 s_cbranch_scc0 .LBB0_2 .LBB0_17: v_add_nc_u32_e32 v1, s22, v0 s_mov_b32 s23, exec_lo s_delay_alu instid0(VALU_DEP_1) v_cmpx_gt_i32_e64 s2, v1 s_cbranch_execz .LBB0_16 v_ashrrev_i32_e32 v2, 31, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[1:2], 2, v[1:2] v_add_co_u32 v4, vcc_lo, s10, v1 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v5, vcc_lo, s11, v2, vcc_lo global_load_b32 v6, v[4:5], off v_add_co_u32 v4, vcc_lo, s18, v1 v_add_co_ci_u32_e32 v5, vcc_lo, s19, v2, vcc_lo v_add_co_u32 v1, vcc_lo, s20, v1 v_add_co_ci_u32_e32 v2, vcc_lo, s21, v2, vcc_lo global_load_b32 v4, v[4:5], off global_load_b32 v5, v[1:2], off s_waitcnt vmcnt(2) v_mul_f32_e32 v7, 0x3fb8aa3b, v6 v_cmp_ngt_f32_e32 vcc_lo, 0xc2ce8ed0, v6 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_fma_f32 v8, v6, 0x3fb8aa3b, -v7 v_rndne_f32_e32 v9, v7 v_dual_fmac_f32 v8, 0x32a5705f, v6 :: v_dual_sub_f32 v7, v7, v9 v_cvt_i32_f32_e32 v9, v9 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_dual_add_f32 v7, v7, v8 :: v_dual_mov_b32 v8, s17 v_exp_f32_e32 v7, v7 ds_load_b32 v8, v8 v_ldexp_f32 v7, v7, v9 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_cndmask_b32_e32 v7, 0, v7, vcc_lo v_cmp_nlt_f32_e32 vcc_lo, 0x42b17218, v6 v_cndmask_b32_e32 v6, 0x7f800000, v7, vcc_lo s_waitcnt vmcnt(1) lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_fma_f32 v4, -v6, v8, v4 s_waitcnt vmcnt(0) v_add_f32_e32 v4, v5, v4 global_store_b32 v[1:2], v4, off s_branch .LBB0_16 .LBB0_19: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z15gLogSoftmaxGradPfPKfS1_ii .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 288 .amdhsa_user_sgpr_count 15 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 0 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 0 .amdhsa_next_free_vgpr 10 .amdhsa_next_free_sgpr 24 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z15gLogSoftmaxGradPfPKfS1_ii, .Lfunc_end0-_Z15gLogSoftmaxGradPfPKfS1_ii .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 8 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .offset: 24 .size: 4 .value_kind: by_value - .offset: 28 .size: 4 .value_kind: by_value - .offset: 32 .size: 4 .value_kind: hidden_block_count_x - .offset: 36 .size: 4 .value_kind: hidden_block_count_y - .offset: 40 .size: 4 .value_kind: hidden_block_count_z - .offset: 44 .size: 2 .value_kind: hidden_group_size_x - .offset: 46 .size: 2 .value_kind: hidden_group_size_y - .offset: 48 .size: 2 .value_kind: hidden_group_size_z - .offset: 50 .size: 2 .value_kind: hidden_remainder_x - .offset: 52 .size: 2 .value_kind: hidden_remainder_y - .offset: 54 .size: 2 .value_kind: hidden_remainder_z - .offset: 72 .size: 8 .value_kind: hidden_global_offset_x - .offset: 80 .size: 8 .value_kind: hidden_global_offset_y - .offset: 88 .size: 8 .value_kind: hidden_global_offset_z - .offset: 96 .size: 2 .value_kind: hidden_grid_dims - .offset: 152 .size: 4 .value_kind: hidden_dynamic_lds_size .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 288 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z15gLogSoftmaxGradPfPKfS1_ii .private_segment_fixed_size: 0 .sgpr_count: 26 .sgpr_spill_count: 0 .symbol: _Z15gLogSoftmaxGradPfPKfS1_ii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 10 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_0012e87e_00000000-6_gLogSoftmaxGrad.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2029: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2029: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z43__device_stub__Z15gLogSoftmaxGradPfPKfS1_iiPfPKfS1_ii .type _Z43__device_stub__Z15gLogSoftmaxGradPfPKfS1_iiPfPKfS1_ii, @function _Z43__device_stub__Z15gLogSoftmaxGradPfPKfS1_iiPfPKfS1_ii: .LFB2051: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 24(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movl %ecx, 4(%rsp) movl %r8d, (%rsp) movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax leaq 24(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) leaq 4(%rsp), %rax movq %rax, 120(%rsp) movq %rsp, %rax movq %rax, 128(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 56(%rsp) movl $1, 60(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) leaq 40(%rsp), %rcx leaq 32(%rsp), %rdx leaq 60(%rsp), %rsi leaq 48(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 136(%rsp), %rax subq %fs:40, %rax jne .L8 addq $152, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 40(%rsp) .cfi_def_cfa_offset 168 pushq 40(%rsp) .cfi_def_cfa_offset 176 leaq 112(%rsp), %r9 movq 76(%rsp), %rcx movl 84(%rsp), %r8d movq 64(%rsp), %rsi movl 72(%rsp), %edx leaq _Z15gLogSoftmaxGradPfPKfS1_ii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z43__device_stub__Z15gLogSoftmaxGradPfPKfS1_iiPfPKfS1_ii, .-_Z43__device_stub__Z15gLogSoftmaxGradPfPKfS1_iiPfPKfS1_ii .globl _Z15gLogSoftmaxGradPfPKfS1_ii .type _Z15gLogSoftmaxGradPfPKfS1_ii, @function _Z15gLogSoftmaxGradPfPKfS1_ii: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z43__device_stub__Z15gLogSoftmaxGradPfPKfS1_iiPfPKfS1_ii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z15gLogSoftmaxGradPfPKfS1_ii, .-_Z15gLogSoftmaxGradPfPKfS1_ii .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "_Z15gLogSoftmaxGradPfPKfS1_ii" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2054: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC0(%rip), %rdx movq %rdx, %rcx leaq _Z15gLogSoftmaxGradPfPKfS1_ii(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2054: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "gLogSoftmaxGrad.hip" .globl _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii # -- Begin function _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii .p2align 4, 0x90 .type _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii,@function _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii: # @_Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movq %rdx, 56(%rsp) movl %ecx, 4(%rsp) movl %r8d, (%rsp) leaq 72(%rsp), %rax movq %rax, 80(%rsp) leaq 64(%rsp), %rax movq %rax, 88(%rsp) leaq 56(%rsp), %rax movq %rax, 96(%rsp) leaq 4(%rsp), %rax movq %rax, 104(%rsp) movq %rsp, %rax movq %rax, 112(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 80(%rsp), %r9 movl $_Z15gLogSoftmaxGradPfPKfS1_ii, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $136, %rsp .cfi_adjust_cfa_offset -136 retq .Lfunc_end0: .size _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii, .Lfunc_end0-_Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB1_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB1_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z15gLogSoftmaxGradPfPKfS1_ii, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end1: .size __hip_module_ctor, .Lfunc_end1-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB2_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB2_2: retq .Lfunc_end2: .size __hip_module_dtor, .Lfunc_end2-__hip_module_dtor .cfi_endproc # -- End function .type _Z15gLogSoftmaxGradPfPKfS1_ii,@object # @_Z15gLogSoftmaxGradPfPKfS1_ii .section .rodata,"a",@progbits .globl _Z15gLogSoftmaxGradPfPKfS1_ii .p2align 3, 0x0 _Z15gLogSoftmaxGradPfPKfS1_ii: .quad _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii .size _Z15gLogSoftmaxGradPfPKfS1_ii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z15gLogSoftmaxGradPfPKfS1_ii" .size .L__unnamed_1, 30 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z30__device_stub__gLogSoftmaxGradPfPKfS1_ii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z15gLogSoftmaxGradPfPKfS1_ii .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include "includes.h" #define INF 2147483647 extern "C" { } __global__ void oneReduction(int * tab, int len, int mod) { __shared__ int begin, end; __shared__ int tmp_T[1024]; if(threadIdx.x == 0) { begin = blockIdx.x*len; end = blockIdx.x*len + len; } __syncthreads(); if(blockIdx.x % mod < mod/2) { for(int k = len/2; k >= 1024; k /= 2) { for(int g = begin; g < end; g += 2*k) { for(int j = g; j < g + k; j += 512) { __syncthreads(); if(threadIdx.x < 512) tmp_T[threadIdx.x] = tab[j + threadIdx.x]; else tmp_T[threadIdx.x] = tab[j + threadIdx.x - 512 + k]; __syncthreads(); if(threadIdx.x < 512 && tmp_T[threadIdx.x] > tmp_T[threadIdx.x + 512]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; tmp_T[threadIdx.x + 512] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; } __syncthreads(); if(threadIdx.x < 512) tab[j + threadIdx.x] = tmp_T[threadIdx.x]; else tab[j + threadIdx.x - 512 + k] = tmp_T[threadIdx.x]; } } } for(int i = begin; i < begin+len; i += 1024) { __syncthreads(); tmp_T[threadIdx.x] = tab[i + threadIdx.x]; __syncthreads(); for(int jump = 512; jump >= 1; jump /= 2) { if(threadIdx.x % (jump*2) < jump && threadIdx.x + jump < 1024 && tmp_T[threadIdx.x] > tmp_T[threadIdx.x + jump]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; tmp_T[threadIdx.x + jump] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; } __syncthreads(); } tab[i + threadIdx.x] = tmp_T[threadIdx.x]; } } else { for(int k = len/2; k >= 1024; k /= 2) { for(int g = begin; g < end; g += 2*k) { for(int j = g; j < g + k; j += 512) { __syncthreads(); if(threadIdx.x < 512) tmp_T[threadIdx.x] = tab[j + threadIdx.x]; else tmp_T[threadIdx.x] = tab[j + threadIdx.x - 512 + k]; __syncthreads(); if(threadIdx.x < 512 && tmp_T[threadIdx.x] < tmp_T[threadIdx.x + 512]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; tmp_T[threadIdx.x + 512] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; } __syncthreads(); if(threadIdx.x < 512) tab[j + threadIdx.x] = tmp_T[threadIdx.x]; else tab[j + threadIdx.x - 512 + k] = tmp_T[threadIdx.x]; } } } for(int i = begin; i < begin + len; i += 1024) { __syncthreads(); tmp_T[threadIdx.x] = tab[i + threadIdx.x]; __syncthreads(); for(int jump = 512; jump >= 1; jump /= 2) { if(threadIdx.x % (jump*2) < jump && threadIdx.x + jump < 1024 && tmp_T[threadIdx.x] < tmp_T[threadIdx.x + jump]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; tmp_T[threadIdx.x + jump] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; } __syncthreads(); } tab[i + threadIdx.x] = tmp_T[threadIdx.x]; } } }
.file "tmpxft_00002763_00000000-6_oneReduction.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2029: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2029: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z34__device_stub__Z12oneReductionPiiiPiii .type _Z34__device_stub__Z12oneReductionPiiiPiii, @function _Z34__device_stub__Z12oneReductionPiiiPiii: .LFB2051: .cfi_startproc endbr64 subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 8(%rsp) movl %esi, 4(%rsp) movl %edx, (%rsp) movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax leaq 8(%rsp), %rax movq %rax, 80(%rsp) leaq 4(%rsp), %rax movq %rax, 88(%rsp) movq %rsp, %rax movq %rax, 96(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) movl $1, 40(%rsp) movl $1, 44(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) leaq 24(%rsp), %rcx leaq 16(%rsp), %rdx leaq 44(%rsp), %rsi leaq 32(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 104(%rsp), %rax subq %fs:40, %rax jne .L8 addq $120, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 24(%rsp) .cfi_def_cfa_offset 136 pushq 24(%rsp) .cfi_def_cfa_offset 144 leaq 96(%rsp), %r9 movq 60(%rsp), %rcx movl 68(%rsp), %r8d movq 48(%rsp), %rsi movl 56(%rsp), %edx leaq _Z12oneReductionPiii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 128 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z34__device_stub__Z12oneReductionPiiiPiii, .-_Z34__device_stub__Z12oneReductionPiiiPiii .globl _Z12oneReductionPiii .type _Z12oneReductionPiii, @function _Z12oneReductionPiii: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z34__device_stub__Z12oneReductionPiiiPiii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z12oneReductionPiii, .-_Z12oneReductionPiii .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "_Z12oneReductionPiii" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2054: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC0(%rip), %rdx movq %rdx, %rcx leaq _Z12oneReductionPiii(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2054: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include "includes.h" #define INF 2147483647 extern "C" { } __global__ void oneReduction(int * tab, int len, int mod) { __shared__ int begin, end; __shared__ int tmp_T[1024]; if(threadIdx.x == 0) { begin = blockIdx.x*len; end = blockIdx.x*len + len; } __syncthreads(); if(blockIdx.x % mod < mod/2) { for(int k = len/2; k >= 1024; k /= 2) { for(int g = begin; g < end; g += 2*k) { for(int j = g; j < g + k; j += 512) { __syncthreads(); if(threadIdx.x < 512) tmp_T[threadIdx.x] = tab[j + threadIdx.x]; else tmp_T[threadIdx.x] = tab[j + threadIdx.x - 512 + k]; __syncthreads(); if(threadIdx.x < 512 && tmp_T[threadIdx.x] > tmp_T[threadIdx.x + 512]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; tmp_T[threadIdx.x + 512] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; } __syncthreads(); if(threadIdx.x < 512) tab[j + threadIdx.x] = tmp_T[threadIdx.x]; else tab[j + threadIdx.x - 512 + k] = tmp_T[threadIdx.x]; } } } for(int i = begin; i < begin+len; i += 1024) { __syncthreads(); tmp_T[threadIdx.x] = tab[i + threadIdx.x]; __syncthreads(); for(int jump = 512; jump >= 1; jump /= 2) { if(threadIdx.x % (jump*2) < jump && threadIdx.x + jump < 1024 && tmp_T[threadIdx.x] > tmp_T[threadIdx.x + jump]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; tmp_T[threadIdx.x + jump] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; } __syncthreads(); } tab[i + threadIdx.x] = tmp_T[threadIdx.x]; } } else { for(int k = len/2; k >= 1024; k /= 2) { for(int g = begin; g < end; g += 2*k) { for(int j = g; j < g + k; j += 512) { __syncthreads(); if(threadIdx.x < 512) tmp_T[threadIdx.x] = tab[j + threadIdx.x]; else tmp_T[threadIdx.x] = tab[j + threadIdx.x - 512 + k]; __syncthreads(); if(threadIdx.x < 512 && tmp_T[threadIdx.x] < tmp_T[threadIdx.x + 512]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; tmp_T[threadIdx.x + 512] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; } __syncthreads(); if(threadIdx.x < 512) tab[j + threadIdx.x] = tmp_T[threadIdx.x]; else tab[j + threadIdx.x - 512 + k] = tmp_T[threadIdx.x]; } } } for(int i = begin; i < begin + len; i += 1024) { __syncthreads(); tmp_T[threadIdx.x] = tab[i + threadIdx.x]; __syncthreads(); for(int jump = 512; jump >= 1; jump /= 2) { if(threadIdx.x % (jump*2) < jump && threadIdx.x + jump < 1024 && tmp_T[threadIdx.x] < tmp_T[threadIdx.x + jump]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; tmp_T[threadIdx.x + jump] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; } __syncthreads(); } tab[i + threadIdx.x] = tmp_T[threadIdx.x]; } } }
#include <hip/hip_runtime.h> #include "includes.h" #define INF 2147483647 extern "C" { } __global__ void oneReduction(int * tab, int len, int mod) { __shared__ int begin, end; __shared__ int tmp_T[1024]; if(threadIdx.x == 0) { begin = blockIdx.x*len; end = blockIdx.x*len + len; } __syncthreads(); if(blockIdx.x % mod < mod/2) { for(int k = len/2; k >= 1024; k /= 2) { for(int g = begin; g < end; g += 2*k) { for(int j = g; j < g + k; j += 512) { __syncthreads(); if(threadIdx.x < 512) tmp_T[threadIdx.x] = tab[j + threadIdx.x]; else tmp_T[threadIdx.x] = tab[j + threadIdx.x - 512 + k]; __syncthreads(); if(threadIdx.x < 512 && tmp_T[threadIdx.x] > tmp_T[threadIdx.x + 512]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; tmp_T[threadIdx.x + 512] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; } __syncthreads(); if(threadIdx.x < 512) tab[j + threadIdx.x] = tmp_T[threadIdx.x]; else tab[j + threadIdx.x - 512 + k] = tmp_T[threadIdx.x]; } } } for(int i = begin; i < begin+len; i += 1024) { __syncthreads(); tmp_T[threadIdx.x] = tab[i + threadIdx.x]; __syncthreads(); for(int jump = 512; jump >= 1; jump /= 2) { if(threadIdx.x % (jump*2) < jump && threadIdx.x + jump < 1024 && tmp_T[threadIdx.x] > tmp_T[threadIdx.x + jump]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; tmp_T[threadIdx.x + jump] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; } __syncthreads(); } tab[i + threadIdx.x] = tmp_T[threadIdx.x]; } } else { for(int k = len/2; k >= 1024; k /= 2) { for(int g = begin; g < end; g += 2*k) { for(int j = g; j < g + k; j += 512) { __syncthreads(); if(threadIdx.x < 512) tmp_T[threadIdx.x] = tab[j + threadIdx.x]; else tmp_T[threadIdx.x] = tab[j + threadIdx.x - 512 + k]; __syncthreads(); if(threadIdx.x < 512 && tmp_T[threadIdx.x] < tmp_T[threadIdx.x + 512]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; tmp_T[threadIdx.x + 512] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; } __syncthreads(); if(threadIdx.x < 512) tab[j + threadIdx.x] = tmp_T[threadIdx.x]; else tab[j + threadIdx.x - 512 + k] = tmp_T[threadIdx.x]; } } } for(int i = begin; i < begin + len; i += 1024) { __syncthreads(); tmp_T[threadIdx.x] = tab[i + threadIdx.x]; __syncthreads(); for(int jump = 512; jump >= 1; jump /= 2) { if(threadIdx.x % (jump*2) < jump && threadIdx.x + jump < 1024 && tmp_T[threadIdx.x] < tmp_T[threadIdx.x + jump]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; tmp_T[threadIdx.x + jump] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; } __syncthreads(); } tab[i + threadIdx.x] = tmp_T[threadIdx.x]; } } }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include "includes.h" #define INF 2147483647 extern "C" { } __global__ void oneReduction(int * tab, int len, int mod) { __shared__ int begin, end; __shared__ int tmp_T[1024]; if(threadIdx.x == 0) { begin = blockIdx.x*len; end = blockIdx.x*len + len; } __syncthreads(); if(blockIdx.x % mod < mod/2) { for(int k = len/2; k >= 1024; k /= 2) { for(int g = begin; g < end; g += 2*k) { for(int j = g; j < g + k; j += 512) { __syncthreads(); if(threadIdx.x < 512) tmp_T[threadIdx.x] = tab[j + threadIdx.x]; else tmp_T[threadIdx.x] = tab[j + threadIdx.x - 512 + k]; __syncthreads(); if(threadIdx.x < 512 && tmp_T[threadIdx.x] > tmp_T[threadIdx.x + 512]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; tmp_T[threadIdx.x + 512] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; } __syncthreads(); if(threadIdx.x < 512) tab[j + threadIdx.x] = tmp_T[threadIdx.x]; else tab[j + threadIdx.x - 512 + k] = tmp_T[threadIdx.x]; } } } for(int i = begin; i < begin+len; i += 1024) { __syncthreads(); tmp_T[threadIdx.x] = tab[i + threadIdx.x]; __syncthreads(); for(int jump = 512; jump >= 1; jump /= 2) { if(threadIdx.x % (jump*2) < jump && threadIdx.x + jump < 1024 && tmp_T[threadIdx.x] > tmp_T[threadIdx.x + jump]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; tmp_T[threadIdx.x + jump] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; } __syncthreads(); } tab[i + threadIdx.x] = tmp_T[threadIdx.x]; } } else { for(int k = len/2; k >= 1024; k /= 2) { for(int g = begin; g < end; g += 2*k) { for(int j = g; j < g + k; j += 512) { __syncthreads(); if(threadIdx.x < 512) tmp_T[threadIdx.x] = tab[j + threadIdx.x]; else tmp_T[threadIdx.x] = tab[j + threadIdx.x - 512 + k]; __syncthreads(); if(threadIdx.x < 512 && tmp_T[threadIdx.x] < tmp_T[threadIdx.x + 512]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; tmp_T[threadIdx.x + 512] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; } __syncthreads(); if(threadIdx.x < 512) tab[j + threadIdx.x] = tmp_T[threadIdx.x]; else tab[j + threadIdx.x - 512 + k] = tmp_T[threadIdx.x]; } } } for(int i = begin; i < begin + len; i += 1024) { __syncthreads(); tmp_T[threadIdx.x] = tab[i + threadIdx.x]; __syncthreads(); for(int jump = 512; jump >= 1; jump /= 2) { if(threadIdx.x % (jump*2) < jump && threadIdx.x + jump < 1024 && tmp_T[threadIdx.x] < tmp_T[threadIdx.x + jump]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; tmp_T[threadIdx.x + jump] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; } __syncthreads(); } tab[i + threadIdx.x] = tmp_T[threadIdx.x]; } } }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z12oneReductionPiii .globl _Z12oneReductionPiii .p2align 8 .type _Z12oneReductionPiii,@function _Z12oneReductionPiii: s_load_b32 s4, s[0:1], 0x8 s_mov_b32 s2, exec_lo v_cmpx_eq_u32_e32 0, v0 s_cbranch_execz .LBB0_2 s_waitcnt lgkmcnt(0) s_mul_i32 s3, s15, s4 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_1) | instid1(SALU_CYCLE_1) v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, s3 s_add_i32 s5, s15, 1 s_mul_i32 s3, s5, s4 s_movk_i32 s5, 0x1000 v_mov_b32_e32 v3, s3 v_add_nc_u32_e32 v1, s5, v1 ds_store_2addr_b32 v1, v3, v2 offset1:1 .LBB0_2: s_or_b32 exec_lo, exec_lo, s2 s_load_b32 s5, s[0:1], 0xc s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv v_cvt_f32_u32_e32 v1, s5 s_sub_i32 s3, 0, s5 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_1) v_rcp_iflag_f32_e32 v1, v1 s_waitcnt_depctr 0xfff v_mul_f32_e32 v1, 0x4f7ffffe, v1 v_cvt_u32_f32_e32 v1, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_readfirstlane_b32 s2, v1 s_mul_i32 s3, s3, s2 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_mul_hi_u32 s3, s2, s3 s_add_i32 s2, s2, s3 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_mul_hi_u32 s2, s15, s2 s_mul_i32 s2, s2, s5 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_4) | instid1(SALU_CYCLE_1) s_sub_i32 s6, s15, s2 s_load_b64 s[2:3], s[0:1], 0x0 s_sub_i32 s7, s6, s5 s_cmp_ge_u32 s6, s5 s_cselect_b32 s0, s7, s6 s_sub_i32 s1, s0, s5 s_cmp_ge_u32 s0, s5 s_cselect_b32 s0, s1, s0 s_lshr_b32 s1, s5, 31 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_add_i32 s5, s5, s1 s_ashr_i32 s5, s5, 1 s_cmpk_gt_i32 s4, 0x7ff s_cselect_b32 s1, -1, 0 s_cmp_ge_u32 s0, s5 s_mov_b32 s0, -1 s_cbranch_scc0 .LBB0_25 s_and_not1_b32 vcc_lo, exec_lo, s1 s_cbranch_vccnz .LBB0_15 v_dual_mov_b32 v2, 0 :: v_dual_lshlrev_b32 v3, 2, v0 v_cmp_gt_u32_e64 s0, 0x200, v0 v_add_nc_u32_e32 v4, 0xfffffe00, v0 s_mov_b32 s5, s4 s_delay_alu instid0(VALU_DEP_3) v_add_nc_u32_e32 v5, 0x800, v3 s_branch .LBB0_6 .LBB0_5: s_cmpk_lt_u32 s5, 0x1000 s_mov_b32 s5, s6 s_cbranch_scc1 .LBB0_15 .LBB0_6: s_movk_i32 s6, 0x1000 s_delay_alu instid0(SALU_CYCLE_1) v_add_nc_u32_e32 v1, s6, v2 s_lshr_b32 s6, s5, 1 ds_load_2addr_b32 v[6:7], v1 offset1:1 s_waitcnt lgkmcnt(0) v_cmp_ge_i32_e32 vcc_lo, v7, v6 v_readfirstlane_b32 s7, v7 s_cbranch_vccnz .LBB0_5 v_add_nc_u32_e32 v1, s6, v4 s_and_b32 s8, s5, 0x7ffffffe v_add_nc_u32_e32 v6, s6, v0 s_cmp_gt_u32 s5, 1 s_cselect_b32 s9, -1, 0 v_cndmask_b32_e64 v7, v1, v0, s0 s_branch .LBB0_9 .LBB0_8: s_set_inst_prefetch_distance 0x2 ds_load_b32 v1, v2 offset:4096 s_add_i32 s7, s7, s8 s_waitcnt lgkmcnt(0) v_cmp_ge_i32_e32 vcc_lo, s7, v1 s_cbranch_vccnz .LBB0_5 .LBB0_9: s_and_not1_b32 vcc_lo, exec_lo, s9 s_cbranch_vccnz .LBB0_8 s_add_i32 s10, s7, s6 s_mov_b32 s11, s7 s_set_inst_prefetch_distance 0x1 s_branch .LBB0_12 .p2align 6 .LBB0_11: s_or_b32 exec_lo, exec_lo, s12 v_add_nc_u32_e32 v1, s11, v0 v_add3_u32 v8, v6, s11, 0xfffffe00 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv ds_load_b32 v10, v3 v_cndmask_b32_e64 v1, v8, v1, s0 s_addk_i32 s11, 0x200 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1) s_cmp_ge_i32 s11, s10 v_lshlrev_b64 v[8:9], 2, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v8, vcc_lo, s2, v8 v_add_co_ci_u32_e32 v9, vcc_lo, s3, v9, vcc_lo s_waitcnt lgkmcnt(0) global_store_b32 v[8:9], v10, off s_cbranch_scc1 .LBB0_8 .LBB0_12: v_add_nc_u32_e32 v1, s11, v7 s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv v_lshlrev_b64 v[8:9], 2, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v8, vcc_lo, s2, v8 v_add_co_ci_u32_e32 v9, vcc_lo, s3, v9, vcc_lo global_load_b32 v1, v[8:9], off s_waitcnt vmcnt(0) ds_store_b32 v3, v1 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_and_saveexec_b32 s12, s0 s_cbranch_execz .LBB0_11 ds_load_b32 v1, v3 ds_load_b32 v8, v5 s_waitcnt lgkmcnt(0) v_cmp_lt_i32_e32 vcc_lo, v1, v8 s_and_b32 exec_lo, exec_lo, vcc_lo s_cbranch_execz .LBB0_11 ds_store_b32 v5, v1 ds_store_b32 v3, v8 s_branch .LBB0_11 .LBB0_15: s_cmp_lt_i32 s4, 1 s_cbranch_scc1 .LBB0_24 v_mov_b32_e32 v2, 0 v_lshlrev_b32_e32 v6, 2, v0 ds_load_b32 v5, v2 offset:4100 s_branch .LBB0_18 .LBB0_17: s_set_inst_prefetch_distance 0x2 ds_load_b32 v1, v2 offset:4100 ds_load_b32 v7, v6 v_add_nc_u32_e32 v5, 0x400, v5 s_waitcnt lgkmcnt(1) v_add_nc_u32_e32 v1, s4, v1 s_waitcnt lgkmcnt(0) global_store_b32 v[3:4], v7, off v_cmp_ge_i32_e32 vcc_lo, v5, v1 s_cbranch_vccnz .LBB0_24 .LBB0_18: s_waitcnt lgkmcnt(0) v_add_nc_u32_e32 v1, v5, v0 s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv s_movk_i32 s0, 0x200 v_lshlrev_b64 v[3:4], 2, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v3, vcc_lo, s2, v3 v_add_co_ci_u32_e32 v4, vcc_lo, s3, v4, vcc_lo global_load_b32 v1, v[3:4], off s_waitcnt vmcnt(0) ds_store_b32 v6, v1 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_set_inst_prefetch_distance 0x1 s_branch .LBB0_20 .p2align 6 .LBB0_19: s_or_b32 exec_lo, exec_lo, s5 s_lshr_b32 s5, s0, 1 s_cmp_lt_u32 s0, 2 s_mov_b32 s0, s5 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_cbranch_scc1 .LBB0_17 .LBB0_20: s_lshl_b32 s5, s0, 1 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_add_i32 s5, s5, -1 v_and_b32_e32 v1, s5, v0 s_mov_b32 s5, exec_lo s_delay_alu instid0(VALU_DEP_1) v_cmpx_gt_u32_e64 s0, v1 s_cbranch_execz .LBB0_19 v_add_nc_u32_e32 v1, s0, v0 s_delay_alu instid0(VALU_DEP_1) v_cmp_gt_u32_e32 vcc_lo, 0x400, v1 s_and_b32 exec_lo, exec_lo, vcc_lo s_cbranch_execz .LBB0_19 v_lshlrev_b32_e32 v1, 2, v1 ds_load_b32 v7, v6 ds_load_b32 v8, v1 s_waitcnt lgkmcnt(0) v_cmp_lt_i32_e32 vcc_lo, v7, v8 s_and_b32 exec_lo, exec_lo, vcc_lo s_cbranch_execz .LBB0_19 v_xor_b32_e32 v9, v8, v7 ds_store_b32 v6, v9 ds_store_b32 v1, v7 ds_store_b32 v6, v8 s_branch .LBB0_19 .LBB0_24: s_mov_b32 s0, 0 .LBB0_25: s_delay_alu instid0(SALU_CYCLE_1) s_and_b32 vcc_lo, exec_lo, s0 s_cbranch_vccz .LBB0_47 s_and_not1_b32 vcc_lo, exec_lo, s1 s_cbranch_vccnz .LBB0_38 v_dual_mov_b32 v2, 0 :: v_dual_lshlrev_b32 v3, 2, v0 v_cmp_gt_u32_e64 s0, 0x200, v0 v_add_nc_u32_e32 v4, 0xfffffe00, v0 s_mov_b32 s1, s4 s_delay_alu instid0(VALU_DEP_3) v_add_nc_u32_e32 v5, 0x800, v3 s_branch .LBB0_29 .LBB0_28: s_cmpk_lt_u32 s1, 0x1000 s_mov_b32 s1, s5 s_cbranch_scc1 .LBB0_38 .LBB0_29: s_movk_i32 s5, 0x1000 s_delay_alu instid0(SALU_CYCLE_1) v_add_nc_u32_e32 v1, s5, v2 s_lshr_b32 s5, s1, 1 ds_load_2addr_b32 v[6:7], v1 offset1:1 s_waitcnt lgkmcnt(0) v_cmp_ge_i32_e32 vcc_lo, v7, v6 v_readfirstlane_b32 s6, v7 s_cbranch_vccnz .LBB0_28 v_add_nc_u32_e32 v1, s5, v4 s_and_b32 s7, s1, 0x7ffffffe v_add_nc_u32_e32 v6, s5, v0 s_cmp_gt_u32 s1, 1 s_cselect_b32 s8, -1, 0 v_cndmask_b32_e64 v7, v1, v0, s0 s_branch .LBB0_32 .LBB0_31: s_set_inst_prefetch_distance 0x2 ds_load_b32 v1, v2 offset:4096 s_add_i32 s6, s6, s7 s_waitcnt lgkmcnt(0) v_cmp_ge_i32_e32 vcc_lo, s6, v1 s_cbranch_vccnz .LBB0_28 .LBB0_32: s_and_not1_b32 vcc_lo, exec_lo, s8 s_cbranch_vccnz .LBB0_31 s_add_i32 s9, s6, s5 s_mov_b32 s10, s6 s_set_inst_prefetch_distance 0x1 s_branch .LBB0_35 .p2align 6 .LBB0_34: s_or_b32 exec_lo, exec_lo, s11 v_add_nc_u32_e32 v1, s10, v0 v_add3_u32 v8, v6, s10, 0xfffffe00 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv ds_load_b32 v10, v3 v_cndmask_b32_e64 v1, v8, v1, s0 s_addk_i32 s10, 0x200 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1) s_cmp_ge_i32 s10, s9 v_lshlrev_b64 v[8:9], 2, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v8, vcc_lo, s2, v8 v_add_co_ci_u32_e32 v9, vcc_lo, s3, v9, vcc_lo s_waitcnt lgkmcnt(0) global_store_b32 v[8:9], v10, off s_cbranch_scc1 .LBB0_31 .LBB0_35: v_add_nc_u32_e32 v1, s10, v7 s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv v_lshlrev_b64 v[8:9], 2, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v8, vcc_lo, s2, v8 v_add_co_ci_u32_e32 v9, vcc_lo, s3, v9, vcc_lo global_load_b32 v1, v[8:9], off s_waitcnt vmcnt(0) ds_store_b32 v3, v1 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_and_saveexec_b32 s11, s0 s_cbranch_execz .LBB0_34 ds_load_b32 v1, v3 ds_load_b32 v8, v5 s_waitcnt lgkmcnt(0) v_cmp_gt_i32_e32 vcc_lo, v1, v8 s_and_b32 exec_lo, exec_lo, vcc_lo s_cbranch_execz .LBB0_34 ds_store_b32 v5, v1 ds_store_b32 v3, v8 s_branch .LBB0_34 .LBB0_38: s_cmp_lt_i32 s4, 1 s_cbranch_scc1 .LBB0_47 v_mov_b32_e32 v2, 0 v_lshlrev_b32_e32 v6, 2, v0 ds_load_b32 v5, v2 offset:4100 s_branch .LBB0_41 .LBB0_40: s_set_inst_prefetch_distance 0x2 ds_load_b32 v1, v2 offset:4100 ds_load_b32 v7, v6 v_add_nc_u32_e32 v5, 0x400, v5 s_waitcnt lgkmcnt(1) v_add_nc_u32_e32 v1, s4, v1 s_waitcnt lgkmcnt(0) global_store_b32 v[3:4], v7, off v_cmp_ge_i32_e32 vcc_lo, v5, v1 s_cbranch_vccnz .LBB0_47 .LBB0_41: s_waitcnt lgkmcnt(0) v_add_nc_u32_e32 v1, v5, v0 s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv s_movk_i32 s0, 0x200 v_lshlrev_b64 v[3:4], 2, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v3, vcc_lo, s2, v3 v_add_co_ci_u32_e32 v4, vcc_lo, s3, v4, vcc_lo global_load_b32 v1, v[3:4], off s_waitcnt vmcnt(0) ds_store_b32 v6, v1 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_set_inst_prefetch_distance 0x1 s_branch .LBB0_43 .p2align 6 .LBB0_42: s_or_b32 exec_lo, exec_lo, s1 s_lshr_b32 s1, s0, 1 s_cmp_lt_u32 s0, 2 s_mov_b32 s0, s1 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_cbranch_scc1 .LBB0_40 .LBB0_43: s_lshl_b32 s1, s0, 1 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_add_i32 s1, s1, -1 v_and_b32_e32 v1, s1, v0 s_mov_b32 s1, exec_lo s_delay_alu instid0(VALU_DEP_1) v_cmpx_gt_u32_e64 s0, v1 s_cbranch_execz .LBB0_42 v_add_nc_u32_e32 v1, s0, v0 s_delay_alu instid0(VALU_DEP_1) v_cmp_gt_u32_e32 vcc_lo, 0x400, v1 s_and_b32 exec_lo, exec_lo, vcc_lo s_cbranch_execz .LBB0_42 v_lshlrev_b32_e32 v1, 2, v1 ds_load_b32 v7, v6 ds_load_b32 v8, v1 s_waitcnt lgkmcnt(0) v_cmp_gt_i32_e32 vcc_lo, v7, v8 s_and_b32 exec_lo, exec_lo, vcc_lo s_cbranch_execz .LBB0_42 v_xor_b32_e32 v9, v8, v7 ds_store_b32 v6, v9 ds_store_b32 v1, v7 ds_store_b32 v6, v8 s_branch .LBB0_42 .LBB0_47: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z12oneReductionPiii .amdhsa_group_segment_fixed_size 4104 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 16 .amdhsa_user_sgpr_count 15 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 0 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 0 .amdhsa_next_free_vgpr 11 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z12oneReductionPiii, .Lfunc_end0-_Z12oneReductionPiii .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .offset: 8 .size: 4 .value_kind: by_value - .offset: 12 .size: 4 .value_kind: by_value .group_segment_fixed_size: 4104 .kernarg_segment_align: 8 .kernarg_segment_size: 16 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z12oneReductionPiii .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z12oneReductionPiii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 11 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <hip/hip_runtime.h> #include "includes.h" #define INF 2147483647 extern "C" { } __global__ void oneReduction(int * tab, int len, int mod) { __shared__ int begin, end; __shared__ int tmp_T[1024]; if(threadIdx.x == 0) { begin = blockIdx.x*len; end = blockIdx.x*len + len; } __syncthreads(); if(blockIdx.x % mod < mod/2) { for(int k = len/2; k >= 1024; k /= 2) { for(int g = begin; g < end; g += 2*k) { for(int j = g; j < g + k; j += 512) { __syncthreads(); if(threadIdx.x < 512) tmp_T[threadIdx.x] = tab[j + threadIdx.x]; else tmp_T[threadIdx.x] = tab[j + threadIdx.x - 512 + k]; __syncthreads(); if(threadIdx.x < 512 && tmp_T[threadIdx.x] > tmp_T[threadIdx.x + 512]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; tmp_T[threadIdx.x + 512] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; } __syncthreads(); if(threadIdx.x < 512) tab[j + threadIdx.x] = tmp_T[threadIdx.x]; else tab[j + threadIdx.x - 512 + k] = tmp_T[threadIdx.x]; } } } for(int i = begin; i < begin+len; i += 1024) { __syncthreads(); tmp_T[threadIdx.x] = tab[i + threadIdx.x]; __syncthreads(); for(int jump = 512; jump >= 1; jump /= 2) { if(threadIdx.x % (jump*2) < jump && threadIdx.x + jump < 1024 && tmp_T[threadIdx.x] > tmp_T[threadIdx.x + jump]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; tmp_T[threadIdx.x + jump] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; } __syncthreads(); } tab[i + threadIdx.x] = tmp_T[threadIdx.x]; } } else { for(int k = len/2; k >= 1024; k /= 2) { for(int g = begin; g < end; g += 2*k) { for(int j = g; j < g + k; j += 512) { __syncthreads(); if(threadIdx.x < 512) tmp_T[threadIdx.x] = tab[j + threadIdx.x]; else tmp_T[threadIdx.x] = tab[j + threadIdx.x - 512 + k]; __syncthreads(); if(threadIdx.x < 512 && tmp_T[threadIdx.x] < tmp_T[threadIdx.x + 512]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; tmp_T[threadIdx.x + 512] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + 512]; } __syncthreads(); if(threadIdx.x < 512) tab[j + threadIdx.x] = tmp_T[threadIdx.x]; else tab[j + threadIdx.x - 512 + k] = tmp_T[threadIdx.x]; } } } for(int i = begin; i < begin + len; i += 1024) { __syncthreads(); tmp_T[threadIdx.x] = tab[i + threadIdx.x]; __syncthreads(); for(int jump = 512; jump >= 1; jump /= 2) { if(threadIdx.x % (jump*2) < jump && threadIdx.x + jump < 1024 && tmp_T[threadIdx.x] < tmp_T[threadIdx.x + jump]) { tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; tmp_T[threadIdx.x + jump] ^= tmp_T[threadIdx.x]; tmp_T[threadIdx.x] ^= tmp_T[threadIdx.x + jump]; } __syncthreads(); } tab[i + threadIdx.x] = tmp_T[threadIdx.x]; } } }
.text .file "oneReduction.hip" .globl _Z27__device_stub__oneReductionPiii # -- Begin function _Z27__device_stub__oneReductionPiii .p2align 4, 0x90 .type _Z27__device_stub__oneReductionPiii,@function _Z27__device_stub__oneReductionPiii: # @_Z27__device_stub__oneReductionPiii .cfi_startproc # %bb.0: subq $88, %rsp .cfi_def_cfa_offset 96 movq %rdi, 56(%rsp) movl %esi, 4(%rsp) movl %edx, (%rsp) leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 4(%rsp), %rax movq %rax, 72(%rsp) movq %rsp, %rax movq %rax, 80(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 64(%rsp), %r9 movl $_Z12oneReductionPiii, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $104, %rsp .cfi_adjust_cfa_offset -104 retq .Lfunc_end0: .size _Z27__device_stub__oneReductionPiii, .Lfunc_end0-_Z27__device_stub__oneReductionPiii .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB1_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB1_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z12oneReductionPiii, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end1: .size __hip_module_ctor, .Lfunc_end1-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB2_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB2_2: retq .Lfunc_end2: .size __hip_module_dtor, .Lfunc_end2-__hip_module_dtor .cfi_endproc # -- End function .type _Z12oneReductionPiii,@object # @_Z12oneReductionPiii .section .rodata,"a",@progbits .globl _Z12oneReductionPiii .p2align 3, 0x0 _Z12oneReductionPiii: .quad _Z27__device_stub__oneReductionPiii .size _Z12oneReductionPiii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z12oneReductionPiii" .size .L__unnamed_1, 21 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z27__device_stub__oneReductionPiii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z12oneReductionPiii .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_00002763_00000000-6_oneReduction.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2029: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2029: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z34__device_stub__Z12oneReductionPiiiPiii .type _Z34__device_stub__Z12oneReductionPiiiPiii, @function _Z34__device_stub__Z12oneReductionPiiiPiii: .LFB2051: .cfi_startproc endbr64 subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 8(%rsp) movl %esi, 4(%rsp) movl %edx, (%rsp) movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax leaq 8(%rsp), %rax movq %rax, 80(%rsp) leaq 4(%rsp), %rax movq %rax, 88(%rsp) movq %rsp, %rax movq %rax, 96(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) movl $1, 40(%rsp) movl $1, 44(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) leaq 24(%rsp), %rcx leaq 16(%rsp), %rdx leaq 44(%rsp), %rsi leaq 32(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 104(%rsp), %rax subq %fs:40, %rax jne .L8 addq $120, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 24(%rsp) .cfi_def_cfa_offset 136 pushq 24(%rsp) .cfi_def_cfa_offset 144 leaq 96(%rsp), %r9 movq 60(%rsp), %rcx movl 68(%rsp), %r8d movq 48(%rsp), %rsi movl 56(%rsp), %edx leaq _Z12oneReductionPiii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 128 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z34__device_stub__Z12oneReductionPiiiPiii, .-_Z34__device_stub__Z12oneReductionPiiiPiii .globl _Z12oneReductionPiii .type _Z12oneReductionPiii, @function _Z12oneReductionPiii: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z34__device_stub__Z12oneReductionPiiiPiii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z12oneReductionPiii, .-_Z12oneReductionPiii .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "_Z12oneReductionPiii" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2054: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC0(%rip), %rdx movq %rdx, %rcx leaq _Z12oneReductionPiii(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2054: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "oneReduction.hip" .globl _Z27__device_stub__oneReductionPiii # -- Begin function _Z27__device_stub__oneReductionPiii .p2align 4, 0x90 .type _Z27__device_stub__oneReductionPiii,@function _Z27__device_stub__oneReductionPiii: # @_Z27__device_stub__oneReductionPiii .cfi_startproc # %bb.0: subq $88, %rsp .cfi_def_cfa_offset 96 movq %rdi, 56(%rsp) movl %esi, 4(%rsp) movl %edx, (%rsp) leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 4(%rsp), %rax movq %rax, 72(%rsp) movq %rsp, %rax movq %rax, 80(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 64(%rsp), %r9 movl $_Z12oneReductionPiii, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $104, %rsp .cfi_adjust_cfa_offset -104 retq .Lfunc_end0: .size _Z27__device_stub__oneReductionPiii, .Lfunc_end0-_Z27__device_stub__oneReductionPiii .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB1_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB1_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z12oneReductionPiii, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end1: .size __hip_module_ctor, .Lfunc_end1-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB2_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB2_2: retq .Lfunc_end2: .size __hip_module_dtor, .Lfunc_end2-__hip_module_dtor .cfi_endproc # -- End function .type _Z12oneReductionPiii,@object # @_Z12oneReductionPiii .section .rodata,"a",@progbits .globl _Z12oneReductionPiii .p2align 3, 0x0 _Z12oneReductionPiii: .quad _Z27__device_stub__oneReductionPiii .size _Z12oneReductionPiii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z12oneReductionPiii" .size .L__unnamed_1, 21 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z27__device_stub__oneReductionPiii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z12oneReductionPiii .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
#include <stdio.h> #include <stdlib.h> #include <sys/time.h> #include <cuda_runtime.h> #define WIDTH 32 #define ROW_M 256 #define COL_M 256 #define ROW_N 256 #define COL_N 256 #define ROW_MxN ROW_M #define COL_MxN COL_N __global__ void MatMulKernel(float *Md, float *Nd, float *Pd, int width); void MatMul(float *M, float *N, float *P, int width); void printTwoDimDynamicArray(float *Array, const int col, const int row); int main(int argc, char *argv[]) { int i, j, k; int width = WIDTH; // Dynamic float *M = (float *)malloc(ROW_M * COL_M * sizeof(float)); float *N = (float *)malloc(ROW_N * COL_N * sizeof(float)); float *P = (float *)malloc(ROW_MxN * COL_MxN * sizeof(float)); float *MxN = (float *)malloc(ROW_MxN * COL_MxN * sizeof(float)); int pass = 1; // Initial for (i = 0; i < ROW_M; ++i) { for (j = 0; j < COL_M; ++j) { M[i*COL_M + j] = rand() % 5; } } for (i = 0; i < ROW_N; ++i) { for (j = 0; j < COL_N; ++j) { N[i*COL_N + j] = rand() % 5; } } struct timeval starttime, endtime; gettimeofday(&starttime, NULL); // CPU for (i = 0; i < ROW_M; ++i) { for (j = 0; j < COL_N; ++j) { for (k = 0; k < COL_M; ++k) { MxN[i*COL_N + j] += M[i*COL_M + k] * N[k*COL_N + j]; } } } gettimeofday(&endtime, NULL); double executime; executime = (endtime.tv_sec - starttime.tv_sec) * 1000.0; executime += (endtime.tv_usec - starttime.tv_usec) / 1000.0; printf("CPU time: %13lf msec\n", executime); /* printTwoDimDynamicArray(M, COL_M, ROW_M); printf("============================\n"); printTwoDimDynamicArray(N, COL_N, ROW_N); printf("============================\n"); printTwoDimDynamicArray(MxN, COL_MxN, ROW_MxN); */ // GPU MatMul((float *)M, (float *)N, (float *)P, width); // Compare for(i = 0; i < ROW_MxN; i++) { for(j = 0; j < COL_MxN; j++) { if(MxN[i*COL_MxN + j] != P[i*COL_MxN + j]) { printf("MxN[%d][%d] = %2.0f P[%d][%d] = %2.0f\n", i, j, MxN[i*COL_MxN + j], i, j, P[i*COL_MxN + j]); pass = 0; } } } free(M); free(N); free(P); free(MxN); printf("Test %s\n", (pass)?"PASSED":"FAILED"); return 0; } void printTwoDimDynamicArray(float *Array, const int col, const int row) { int x, y; for(y = 0; y != row; ++y) { for(x = 0; x != col; ++x) printf("%f ", Array[y*col + x]); printf("\n"); } } // Matrix multiplication kernel called by MatMul() __global__ void MatMulKernel(float *Md, float *Nd, float *Pd, int width){ int row = (blockIdx.y * blockDim.y) + threadIdx.y; int col = (blockIdx.x * blockDim.x) + threadIdx.x; float Pvalue = 0; // Multiply M and N if(row < ROW_MxN && col < COL_MxN){ for (int k = 0; k < COL_M; ++k) { float Melement = *(Md + row*COL_M + k); float Nelement = *(Nd + k*COL_N + col); Pvalue += Melement * Nelement; } *(Pd + row*COL_N + col) = Pvalue; } } // Matrix multiplication - Host code void MatMul(float *M, float *N, float *P, int width) { size_t size_M = ROW_M * COL_M * sizeof(float); size_t size_N = ROW_N * COL_N * sizeof(float); size_t size_P = ROW_MxN * COL_MxN * sizeof(float); float *Md, *Nd, *Pd; // Allocate and Load M, N to device memory cudaMalloc((void **)&Md, size_M); cudaMemcpy(Md, M, size_M, cudaMemcpyHostToDevice); cudaMalloc((void **)&Nd, size_N); cudaMemcpy(Nd, N, size_N, cudaMemcpyHostToDevice); // Allocate P on the device cudaMalloc((void **)&Pd, size_P); // Setup the execution configuration int gridDim_X = (ROW_MxN + 31)/32; int gridDim_Y = (COL_MxN + 31)/32; dim3 dimGrid(gridDim_X, gridDim_Y); dim3 dimBlock(width, width); printf("============================\n"); // Get start time event cudaEvent_t start, stop; cudaEventCreate(&start); cudaEventCreate(&stop); cudaEventRecord(start, 0); // Invoke kernel MatMulKernel<<<dimGrid, dimBlock>>>(Md, Nd, Pd, width); cudaError_t cuda_err = cudaGetLastError(); if ( cudaSuccess != cuda_err ){ printf("before kernel call: error = %s\n", cudaGetErrorString (cuda_err)); exit(1) ; } // Get stop time event cudaEventRecord(stop, 0); cudaEventSynchronize(stop); // Compute execution time float elapsedTime; cudaEventElapsedTime(&elapsedTime, start, stop); printf("GPU time: %13f msec\n", elapsedTime); cudaEventDestroy(start); cudaEventDestroy(stop); // Read P from device memory cudaMemcpy(P, Pd, size_P, cudaMemcpyDeviceToHost); // Free device memory cudaFree(Md); cudaFree(Nd); cudaFree(Pd); }
code for sm_80 Function : _Z12MatMulKernelPfS_S_i .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R0, SR_CTAID.X ; /* 0x0000000000007919 */ /* 0x000e280000002500 */ /*0020*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e280000002100 */ /*0030*/ S2R R7, SR_CTAID.Y ; /* 0x0000000000077919 */ /* 0x000e680000002600 */ /*0040*/ S2R R2, SR_TID.Y ; /* 0x0000000000027919 */ /* 0x000e620000002200 */ /*0050*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x001fca00078e0203 */ /*0060*/ ISETP.GT.AND P0, PT, R0, 0xff, PT ; /* 0x000000ff0000780c */ /* 0x000fe20003f04270 */ /*0070*/ IMAD R7, R7, c[0x0][0x4], R2 ; /* 0x0000010007077a24 */ /* 0x002fca00078e0202 */ /*0080*/ ISETP.GT.OR P0, PT, R7, 0xff, P0 ; /* 0x000000ff0700780c */ /* 0x000fda0000704670 */ /*0090*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*00a0*/ HFMA2.MMA R4, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff047435 */ /* 0x000fe200000001ff */ /*00b0*/ SHF.L.U32 R7, R7, 0x8, RZ ; /* 0x0000000807077819 */ /* 0x000fe200000006ff */ /*00c0*/ HFMA2.MMA R8, -RZ, RZ, 0, 0 ; /* 0x00000000ff087435 */ /* 0x000fe200000001ff */ /*00d0*/ SHF.R.S32.HI R6, RZ, 0x1f, R0 ; /* 0x0000001fff067819 */ /* 0x000fe20000011400 */ /*00e0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*00f0*/ MOV R15, RZ ; /* 0x000000ff000f7202 */ /* 0x000fca0000000f00 */ /*0100*/ IMAD.WIDE R2, R0, R4, c[0x0][0x168] ; /* 0x00005a0000027625 */ /* 0x000fc800078e0204 */ /*0110*/ IMAD.WIDE R4, R7, R4, c[0x0][0x160] ; /* 0x0000580007047625 */ /* 0x000fe200078e0204 */ /*0120*/ IADD3 R10, P0, R2, 0x2000, RZ ; /* 0x00002000020a7810 */ /* 0x000fc80007f1e0ff */ /*0130*/ IADD3 R2, P1, R4, 0x20, RZ ; /* 0x0000002004027810 */ /* 0x000fe40007f3e0ff */ /*0140*/ IADD3.X R11, RZ, R3, RZ, P0, !PT ; /* 0x00000003ff0b7210 */ /* 0x000fc600007fe4ff */ /*0150*/ IMAD.X R3, RZ, RZ, R5, P1 ; /* 0x000000ffff037224 */ /* 0x000fc600008e0605 */ /*0160*/ IMAD.MOV.U32 R4, RZ, RZ, R10 ; /* 0x000000ffff047224 */ /* 0x000fe200078e000a */ /*0170*/ MOV R5, R11 ; /* 0x0000000b00057202 */ /* 0x000fe20000000f00 */ /*0180*/ LDG.E R17, [R2.64+-0x20] ; /* 0xffffe00402117981 */ /* 0x000ea8000c1e1900 */ /*0190*/ LDG.E R16, [R4.64+-0x2000] ; /* 0xffe0000404107981 */ /* 0x000ea8000c1e1900 */ /*01a0*/ LDG.E R18, [R4.64+-0x1c00] ; /* 0xffe4000404127981 */ /* 0x000ee8000c1e1900 */ /*01b0*/ LDG.E R25, [R2.64+-0x1c] ; /* 0xffffe40402197981 */ /* 0x000ee8000c1e1900 */ /*01c0*/ LDG.E R19, [R4.64+-0x1800] ; /* 0xffe8000404137981 */ /* 0x000f28000c1e1900 */ /*01d0*/ LDG.E R20, [R2.64+-0x18] ; /* 0xffffe80402147981 */ /* 0x000f28000c1e1900 */ /*01e0*/ LDG.E R22, [R4.64+-0x1400] ; /* 0xffec000404167981 */ /* 0x000f68000c1e1900 */ /*01f0*/ LDG.E R21, [R2.64+-0x14] ; /* 0xffffec0402157981 */ /* 0x000f68000c1e1900 */ /*0200*/ LDG.E R23, [R4.64+-0x1000] ; /* 0xfff0000404177981 */ /* 0x000f68000c1e1900 */ /*0210*/ LDG.E R24, [R2.64+-0x10] ; /* 0xfffff00402187981 */ /* 0x000f68000c1e1900 */ /*0220*/ LDG.E R9, [R4.64+-0xc00] ; /* 0xfff4000404097981 */ /* 0x000f68000c1e1900 */ /*0230*/ LDG.E R10, [R2.64+-0xc] ; /* 0xfffff404020a7981 */ /* 0x000f68000c1e1900 */ /*0240*/ LDG.E R11, [R4.64+-0x800] ; /* 0xfff80004040b7981 */ /* 0x000f68000c1e1900 */ /*0250*/ LDG.E R12, [R2.64+-0x8] ; /* 0xfffff804020c7981 */ /* 0x000f68000c1e1900 */ /*0260*/ LDG.E R13, [R4.64+-0x400] ; /* 0xfffc0004040d7981 */ /* 0x000f68000c1e1900 */ /*0270*/ LDG.E R14, [R2.64+-0x4] ; /* 0xfffffc04020e7981 */ /* 0x000f62000c1e1900 */ /*0280*/ FFMA R17, R16, R17, R15 ; /* 0x0000001110117223 */ /* 0x004fc6000000000f */ /*0290*/ LDG.E R15, [R4.64] ; /* 0x00000004040f7981 */ /* 0x000ea8000c1e1900 */ /*02a0*/ LDG.E R16, [R2.64] ; /* 0x0000000402107981 */ /* 0x0000a2000c1e1900 */ /*02b0*/ FFMA R25, R18, R25, R17 ; /* 0x0000001912197223 */ /* 0x008fc60000000011 */ /*02c0*/ LDG.E R17, [R4.64+0x400] ; /* 0x0004000404117981 */ /* 0x000ee8000c1e1900 */ /*02d0*/ LDG.E R18, [R2.64+0x4] ; /* 0x0000040402127981 */ /* 0x0000e2000c1e1900 */ /*02e0*/ FFMA R25, R19, R20, R25 ; /* 0x0000001413197223 */ /* 0x010fc60000000019 */ /*02f0*/ LDG.E R19, [R4.64+0x800] ; /* 0x0008000404137981 */ /* 0x000f28000c1e1900 */ /*0300*/ LDG.E R20, [R2.64+0x8] ; /* 0x0000080402147981 */ /* 0x000122000c1e1900 */ /*0310*/ FFMA R25, R22, R21, R25 ; /* 0x0000001516197223 */ /* 0x020fc60000000019 */ /*0320*/ LDG.E R21, [R4.64+0xc00] ; /* 0x000c000404157981 */ /* 0x000f68000c1e1900 */ /*0330*/ LDG.E R22, [R2.64+0xc] ; /* 0x00000c0402167981 */ /* 0x000162000c1e1900 */ /*0340*/ FFMA R25, R23, R24, R25 ; /* 0x0000001817197223 */ /* 0x000fc60000000019 */ /*0350*/ LDG.E R23, [R4.64+0x1000] ; /* 0x0010000404177981 */ /* 0x000f68000c1e1900 */ /*0360*/ LDG.E R24, [R2.64+0x10] ; /* 0x0000100402187981 */ /* 0x000162000c1e1900 */ /*0370*/ FFMA R25, R9, R10, R25 ; /* 0x0000000a09197223 */ /* 0x000fc60000000019 */ /*0380*/ LDG.E R9, [R4.64+0x1400] ; /* 0x0014000404097981 */ /* 0x000f68000c1e1900 */ /*0390*/ LDG.E R10, [R2.64+0x14] ; /* 0x00001404020a7981 */ /* 0x000162000c1e1900 */ /*03a0*/ FFMA R25, R11, R12, R25 ; /* 0x0000000c0b197223 */ /* 0x000fc60000000019 */ /*03b0*/ LDG.E R12, [R4.64+0x1800] ; /* 0x00180004040c7981 */ /* 0x000f68000c1e1900 */ /*03c0*/ LDG.E R11, [R2.64+0x18] ; /* 0x00001804020b7981 */ /* 0x000162000c1e1900 */ /*03d0*/ FFMA R25, R13, R14, R25 ; /* 0x0000000e0d197223 */ /* 0x000fc60000000019 */ /*03e0*/ LDG.E R13, [R4.64+0x1c00] ; /* 0x001c0004040d7981 */ /* 0x000f68000c1e1900 */ /*03f0*/ LDG.E R14, [R2.64+0x1c] ; /* 0x00001c04020e7981 */ /* 0x000162000c1e1900 */ /*0400*/ IADD3 R8, R8, 0x10, RZ ; /* 0x0000001008087810 */ /* 0x000fc80007ffe0ff */ /*0410*/ ISETP.NE.AND P0, PT, R8, 0x100, PT ; /* 0x000001000800780c */ /* 0x000fe40003f05270 */ /*0420*/ IADD3 R2, P2, R2, 0x40, RZ ; /* 0x0000004002027810 */ /* 0x001fc80007f5e0ff */ /*0430*/ IADD3.X R3, RZ, R3, RZ, P2, !PT ; /* 0x00000003ff037210 */ /* 0x000fe200017fe4ff */ /*0440*/ FFMA R15, R15, R16, R25 ; /* 0x000000100f0f7223 */ /* 0x004fc80000000019 */ /*0450*/ FFMA R15, R17, R18, R15 ; /* 0x00000012110f7223 */ /* 0x008fc8000000000f */ /*0460*/ FFMA R15, R19, R20, R15 ; /* 0x00000014130f7223 */ /* 0x010fc8000000000f */ /*0470*/ FFMA R15, R21, R22, R15 ; /* 0x00000016150f7223 */ /* 0x020fc8000000000f */ /*0480*/ FFMA R15, R23, R24, R15 ; /* 0x00000018170f7223 */ /* 0x000fc8000000000f */ /*0490*/ FFMA R9, R9, R10, R15 ; /* 0x0000000a09097223 */ /* 0x000fe2000000000f */ /*04a0*/ IADD3 R10, P1, R4, 0x4000, RZ ; /* 0x00004000040a7810 */ /* 0x000fc60007f3e0ff */ /*04b0*/ FFMA R9, R12, R11, R9 ; /* 0x0000000b0c097223 */ /* 0x000fe20000000009 */ /*04c0*/ IADD3.X R11, RZ, R5, RZ, P1, !PT ; /* 0x00000005ff0b7210 */ /* 0x000fc60000ffe4ff */ /*04d0*/ FFMA R15, R13, R14, R9 ; /* 0x0000000e0d0f7223 */ /* 0x000fe20000000009 */ /*04e0*/ @P0 BRA 0x160 ; /* 0xfffffc7000000947 */ /* 0x000fea000383ffff */ /*04f0*/ IADD3 R0, P0, R0, R7, RZ ; /* 0x0000000700007210 */ /* 0x000fc80007f1e0ff */ /*0500*/ LEA.HI.X.SX32 R7, R7, R6, 0x1, P0 ; /* 0x0000000607077211 */ /* 0x000fe400000f0eff */ /*0510*/ LEA R2, P0, R0, c[0x0][0x170], 0x2 ; /* 0x00005c0000027a11 */ /* 0x000fc800078010ff */ /*0520*/ LEA.HI.X R3, R0, c[0x0][0x174], R7, 0x2, P0 ; /* 0x00005d0000037a11 */ /* 0x000fca00000f1407 */ /*0530*/ STG.E [R2.64], R15 ; /* 0x0000000f02007986 */ /* 0x000fe2000c101904 */ /*0540*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0550*/ BRA 0x550; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0560*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0570*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0580*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0590*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include <stdio.h> #include <stdlib.h> #include <sys/time.h> #include <cuda_runtime.h> #define WIDTH 32 #define ROW_M 256 #define COL_M 256 #define ROW_N 256 #define COL_N 256 #define ROW_MxN ROW_M #define COL_MxN COL_N __global__ void MatMulKernel(float *Md, float *Nd, float *Pd, int width); void MatMul(float *M, float *N, float *P, int width); void printTwoDimDynamicArray(float *Array, const int col, const int row); int main(int argc, char *argv[]) { int i, j, k; int width = WIDTH; // Dynamic float *M = (float *)malloc(ROW_M * COL_M * sizeof(float)); float *N = (float *)malloc(ROW_N * COL_N * sizeof(float)); float *P = (float *)malloc(ROW_MxN * COL_MxN * sizeof(float)); float *MxN = (float *)malloc(ROW_MxN * COL_MxN * sizeof(float)); int pass = 1; // Initial for (i = 0; i < ROW_M; ++i) { for (j = 0; j < COL_M; ++j) { M[i*COL_M + j] = rand() % 5; } } for (i = 0; i < ROW_N; ++i) { for (j = 0; j < COL_N; ++j) { N[i*COL_N + j] = rand() % 5; } } struct timeval starttime, endtime; gettimeofday(&starttime, NULL); // CPU for (i = 0; i < ROW_M; ++i) { for (j = 0; j < COL_N; ++j) { for (k = 0; k < COL_M; ++k) { MxN[i*COL_N + j] += M[i*COL_M + k] * N[k*COL_N + j]; } } } gettimeofday(&endtime, NULL); double executime; executime = (endtime.tv_sec - starttime.tv_sec) * 1000.0; executime += (endtime.tv_usec - starttime.tv_usec) / 1000.0; printf("CPU time: %13lf msec\n", executime); /* printTwoDimDynamicArray(M, COL_M, ROW_M); printf("============================\n"); printTwoDimDynamicArray(N, COL_N, ROW_N); printf("============================\n"); printTwoDimDynamicArray(MxN, COL_MxN, ROW_MxN); */ // GPU MatMul((float *)M, (float *)N, (float *)P, width); // Compare for(i = 0; i < ROW_MxN; i++) { for(j = 0; j < COL_MxN; j++) { if(MxN[i*COL_MxN + j] != P[i*COL_MxN + j]) { printf("MxN[%d][%d] = %2.0f P[%d][%d] = %2.0f\n", i, j, MxN[i*COL_MxN + j], i, j, P[i*COL_MxN + j]); pass = 0; } } } free(M); free(N); free(P); free(MxN); printf("Test %s\n", (pass)?"PASSED":"FAILED"); return 0; } void printTwoDimDynamicArray(float *Array, const int col, const int row) { int x, y; for(y = 0; y != row; ++y) { for(x = 0; x != col; ++x) printf("%f ", Array[y*col + x]); printf("\n"); } } // Matrix multiplication kernel called by MatMul() __global__ void MatMulKernel(float *Md, float *Nd, float *Pd, int width){ int row = (blockIdx.y * blockDim.y) + threadIdx.y; int col = (blockIdx.x * blockDim.x) + threadIdx.x; float Pvalue = 0; // Multiply M and N if(row < ROW_MxN && col < COL_MxN){ for (int k = 0; k < COL_M; ++k) { float Melement = *(Md + row*COL_M + k); float Nelement = *(Nd + k*COL_N + col); Pvalue += Melement * Nelement; } *(Pd + row*COL_N + col) = Pvalue; } } // Matrix multiplication - Host code void MatMul(float *M, float *N, float *P, int width) { size_t size_M = ROW_M * COL_M * sizeof(float); size_t size_N = ROW_N * COL_N * sizeof(float); size_t size_P = ROW_MxN * COL_MxN * sizeof(float); float *Md, *Nd, *Pd; // Allocate and Load M, N to device memory cudaMalloc((void **)&Md, size_M); cudaMemcpy(Md, M, size_M, cudaMemcpyHostToDevice); cudaMalloc((void **)&Nd, size_N); cudaMemcpy(Nd, N, size_N, cudaMemcpyHostToDevice); // Allocate P on the device cudaMalloc((void **)&Pd, size_P); // Setup the execution configuration int gridDim_X = (ROW_MxN + 31)/32; int gridDim_Y = (COL_MxN + 31)/32; dim3 dimGrid(gridDim_X, gridDim_Y); dim3 dimBlock(width, width); printf("============================\n"); // Get start time event cudaEvent_t start, stop; cudaEventCreate(&start); cudaEventCreate(&stop); cudaEventRecord(start, 0); // Invoke kernel MatMulKernel<<<dimGrid, dimBlock>>>(Md, Nd, Pd, width); cudaError_t cuda_err = cudaGetLastError(); if ( cudaSuccess != cuda_err ){ printf("before kernel call: error = %s\n", cudaGetErrorString (cuda_err)); exit(1) ; } // Get stop time event cudaEventRecord(stop, 0); cudaEventSynchronize(stop); // Compute execution time float elapsedTime; cudaEventElapsedTime(&elapsedTime, start, stop); printf("GPU time: %13f msec\n", elapsedTime); cudaEventDestroy(start); cudaEventDestroy(stop); // Read P from device memory cudaMemcpy(P, Pd, size_P, cudaMemcpyDeviceToHost); // Free device memory cudaFree(Md); cudaFree(Nd); cudaFree(Pd); }
.file "tmpxft_001bb9c0_00000000-6_matrix.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2062: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2062: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "%f " .LC1: .string "\n" .text .globl _Z23printTwoDimDynamicArrayPfii .type _Z23printTwoDimDynamicArrayPfii, @function _Z23printTwoDimDynamicArrayPfii: .LFB2058: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $40, %rsp .cfi_def_cfa_offset 96 movq %rdi, 16(%rsp) testl %edx, %edx je .L3 movl %esi, %r15d movslq %edx, %rax movq %rax, 8(%rsp) movl $0, %r14d movl $0, %r13d movslq %esi, %rax movq %rax, 24(%rsp) leaq .LC0(%rip), %r12 jmp .L5 .L7: movslq %r14d, %rax movq 16(%rsp), %rcx leaq (%rcx,%rax,4), %rbx movq 24(%rsp), %rdx addq %rdx, %rax leaq (%rcx,%rax,4), %rbp .L6: pxor %xmm0, %xmm0 cvtss2sd (%rbx), %xmm0 movq %r12, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT addq $4, %rbx cmpq %rbp, %rbx jne .L6 .L8: leaq .LC1(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT addq $1, %r13 addl %r15d, %r14d movq 8(%rsp), %rax cmpq %rax, %r13 je .L3 .L5: testl %r15d, %r15d jne .L7 jmp .L8 .L3: addq $40, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2058: .size _Z23printTwoDimDynamicArrayPfii, .-_Z23printTwoDimDynamicArrayPfii .globl _Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i .type _Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i, @function _Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i: .LFB2084: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 24(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movl %ecx, 4(%rsp) movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax leaq 24(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) leaq 4(%rsp), %rax movq %rax, 120(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 56(%rsp) movl $1, 60(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) leaq 40(%rsp), %rcx leaq 32(%rsp), %rdx leaq 60(%rsp), %rsi leaq 48(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L15 .L11: movq 136(%rsp), %rax subq %fs:40, %rax jne .L16 addq $152, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L15: .cfi_restore_state pushq 40(%rsp) .cfi_def_cfa_offset 168 pushq 40(%rsp) .cfi_def_cfa_offset 176 leaq 112(%rsp), %r9 movq 76(%rsp), %rcx movl 84(%rsp), %r8d movq 64(%rsp), %rsi movl 72(%rsp), %edx leaq _Z12MatMulKernelPfS_S_i(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L11 .L16: call __stack_chk_fail@PLT .cfi_endproc .LFE2084: .size _Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i, .-_Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i .globl _Z12MatMulKernelPfS_S_i .type _Z12MatMulKernelPfS_S_i, @function _Z12MatMulKernelPfS_S_i: .LFB2085: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2085: .size _Z12MatMulKernelPfS_S_i, .-_Z12MatMulKernelPfS_S_i .section .rodata.str1.1 .LC2: .string "============================\n" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC3: .string "before kernel call: error = %s\n" .section .rodata.str1.1 .LC4: .string "GPU time: %13f msec\n" .text .globl _Z6MatMulPfS_S_i .type _Z6MatMulPfS_S_i, @function _Z6MatMulPfS_S_i: .LFB2059: .cfi_startproc endbr64 pushq %r13 .cfi_def_cfa_offset 16 .cfi_offset 13, -16 pushq %r12 .cfi_def_cfa_offset 24 .cfi_offset 12, -24 pushq %rbp .cfi_def_cfa_offset 32 .cfi_offset 6, -32 pushq %rbx .cfi_def_cfa_offset 40 .cfi_offset 3, -40 subq $88, %rsp .cfi_def_cfa_offset 128 movq %rdi, %r12 movq %rsi, %rbp movq %rdx, %r13 movl %ecx, %ebx movq %fs:40, %rax movq %rax, 72(%rsp) xorl %eax, %eax leaq 8(%rsp), %rdi movl $262144, %esi call cudaMalloc@PLT movl $1, %ecx movl $262144, %edx movq %r12, %rsi movq 8(%rsp), %rdi call cudaMemcpy@PLT leaq 16(%rsp), %rdi movl $262144, %esi call cudaMalloc@PLT movl $1, %ecx movl $262144, %edx movq %rbp, %rsi movq 16(%rsp), %rdi call cudaMemcpy@PLT leaq 24(%rsp), %rdi movl $262144, %esi call cudaMalloc@PLT movl $8, 48(%rsp) movl $8, 52(%rsp) movl $1, 56(%rsp) movl %ebx, 60(%rsp) movl %ebx, 64(%rsp) movl $1, 68(%rsp) leaq .LC2(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT leaq 32(%rsp), %rdi call cudaEventCreate@PLT leaq 40(%rsp), %rdi call cudaEventCreate@PLT movl $0, %esi movq 32(%rsp), %rdi call cudaEventRecord@PLT movl 68(%rsp), %ecx movl $0, %r9d movl $0, %r8d movq 60(%rsp), %rdx movq 48(%rsp), %rdi movl 56(%rsp), %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L24 .L20: call cudaGetLastError@PLT testl %eax, %eax jne .L25 movl $0, %esi movq 40(%rsp), %rdi call cudaEventRecord@PLT movq 40(%rsp), %rdi call cudaEventSynchronize@PLT leaq 4(%rsp), %rdi movq 40(%rsp), %rdx movq 32(%rsp), %rsi call cudaEventElapsedTime@PLT pxor %xmm0, %xmm0 cvtss2sd 4(%rsp), %xmm0 leaq .LC4(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movq 32(%rsp), %rdi call cudaEventDestroy@PLT movq 40(%rsp), %rdi call cudaEventDestroy@PLT movl $2, %ecx movl $262144, %edx movq 24(%rsp), %rsi movq %r13, %rdi call cudaMemcpy@PLT movq 8(%rsp), %rdi call cudaFree@PLT movq 16(%rsp), %rdi call cudaFree@PLT movq 24(%rsp), %rdi call cudaFree@PLT movq 72(%rsp), %rax subq %fs:40, %rax jne .L26 addq $88, %rsp .cfi_remember_state .cfi_def_cfa_offset 40 popq %rbx .cfi_def_cfa_offset 32 popq %rbp .cfi_def_cfa_offset 24 popq %r12 .cfi_def_cfa_offset 16 popq %r13 .cfi_def_cfa_offset 8 ret .L24: .cfi_restore_state movl %ebx, %ecx movq 24(%rsp), %rdx movq 16(%rsp), %rsi movq 8(%rsp), %rdi call _Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i jmp .L20 .L25: movl %eax, %edi call cudaGetErrorString@PLT movq %rax, %rdx leaq .LC3(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $1, %edi call exit@PLT .L26: call __stack_chk_fail@PLT .cfi_endproc .LFE2059: .size _Z6MatMulPfS_S_i, .-_Z6MatMulPfS_S_i .section .rodata.str1.1 .LC5: .string "PASSED" .LC6: .string "FAILED" .LC8: .string "CPU time: %13lf msec\n" .section .rodata.str1.8 .align 8 .LC9: .string "MxN[%d][%d] = %2.0f P[%d][%d] = %2.0f\n" .section .rodata.str1.1 .LC10: .string "Test %s\n" .text .globl main .type main, @function main: .LFB2057: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $88, %rsp .cfi_def_cfa_offset 144 movq %fs:40, %rax movq %rax, 72(%rsp) xorl %eax, %eax movl $262144, %edi call malloc@PLT movq %rax, %r14 movl $262144, %edi call malloc@PLT movq %rax, %r15 movl $262144, %edi call malloc@PLT movq %rax, 16(%rsp) movl $262144, %edi call malloc@PLT movq %rax, %rbx leaq 1024(%r14), %r12 leaq 263168(%r14), %r13 .L28: leaq -1024(%r12), %rbp .L29: call rand@PLT movslq %eax, %rdx imulq $1717986919, %rdx, %rdx sarq $33, %rdx movl %eax, %ecx sarl $31, %ecx subl %ecx, %edx leal (%rdx,%rdx,4), %edx subl %edx, %eax pxor %xmm0, %xmm0 cvtsi2ssl %eax, %xmm0 movss %xmm0, 0(%rbp) addq $4, %rbp cmpq %r12, %rbp jne .L29 addq $1024, %r12 cmpq %r13, %r12 jne .L28 leaq 1024(%r15), %r12 leaq 263168(%r15), %r13 .L30: leaq -1024(%r12), %rbp .L31: call rand@PLT movslq %eax, %rdx imulq $1717986919, %rdx, %rdx sarq $33, %rdx movl %eax, %ecx sarl $31, %ecx subl %ecx, %edx leal (%rdx,%rdx,4), %edx subl %edx, %eax pxor %xmm0, %xmm0 cvtsi2ssl %eax, %xmm0 movss %xmm0, 0(%rbp) addq $4, %rbp cmpq %r12, %rbp jne .L31 addq $1024, %r12 cmpq %r13, %r12 jne .L30 leaq 32(%rsp), %rdi movl $0, %esi call gettimeofday@PLT movl $0, %r9d .L33: leaq 262144(%r15), %rcx movq %rbx, %r12 movq %r9, %r8 salq $10, %r8 leaq (%rbx,%r8), %rdi addq %r14, %r8 movl $0, %esi .L37: movss (%rdi,%rsi,4), %xmm1 leaq -262144(%rcx), %rax movq %r8, %rdx .L34: movss (%rdx), %xmm0 mulss (%rax), %xmm0 addss %xmm0, %xmm1 addq $4, %rdx addq $1024, %rax cmpq %rcx, %rax jne .L34 movss %xmm1, (%rdi,%rsi,4) addq $1, %rsi addq $4, %rcx cmpq $256, %rsi jne .L37 addq $1, %r9 cmpq $256, %r9 jne .L33 leaq 48(%rsp), %rdi movl $0, %esi call gettimeofday@PLT movq 48(%rsp), %rax subq 32(%rsp), %rax pxor %xmm1, %xmm1 cvtsi2sdq %rax, %xmm1 movsd .LC7(%rip), %xmm2 mulsd %xmm2, %xmm1 movq 56(%rsp), %rax subq 40(%rsp), %rax pxor %xmm0, %xmm0 cvtsi2sdq %rax, %xmm0 divsd %xmm2, %xmm0 addsd %xmm1, %xmm0 leaq .LC8(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movl $32, %ecx movq 16(%rsp), %rbp movq %rbp, %rdx movq %r15, %rsi movq %r14, %rdi call _Z6MatMulPfS_S_i movl $1, 12(%rsp) movl $0, %r13d movq %rbx, %rax jmp .L38 .L46: movl %ebx, %ecx cvtss2sd %xmm0, %xmm0 cvtss2sd %xmm1, %xmm1 movl %ebx, %r9d movl %r13d, %r8d movl %r13d, %edx leaq .LC9(%rip), %rsi movl $2, %edi movl $2, %eax call __printf_chk@PLT movl $0, 12(%rsp) .L39: addq $1, %rbx cmpq $256, %rbx je .L53 .L41: movss (%r12,%rbx,4), %xmm0 movss 0(%rbp,%rbx,4), %xmm1 ucomiss %xmm1, %xmm0 jp .L46 je .L39 jmp .L46 .L53: movq %r14, %rax movq 24(%rsp), %r14 addl $1, %r13d addq $1024, %r12 addq $1024, %rbp cmpl $256, %r13d je .L42 .L38: movl $0, %ebx movq %r14, 24(%rsp) movq %rax, %r14 jmp .L41 .L42: movq %rax, %rbx movq %r14, %rdi call free@PLT movq %r15, %rdi call free@PLT movq 16(%rsp), %rdi call free@PLT movq %rbx, %rdi call free@PLT cmpl $0, 12(%rsp) leaq .LC6(%rip), %rdx leaq .LC5(%rip), %rax cmovne %rax, %rdx leaq .LC10(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 72(%rsp), %rax subq %fs:40, %rax jne .L54 movl $0, %eax addq $88, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L54: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size main, .-main .section .rodata.str1.1 .LC11: .string "_Z12MatMulKernelPfS_S_i" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2087: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC11(%rip), %rdx movq %rdx, %rcx leaq _Z12MatMulKernelPfS_S_i(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2087: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .section .rodata.cst8,"aM",@progbits,8 .align 8 .LC7: .long 0 .long 1083129856 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include <stdio.h> #include <stdlib.h> #include <sys/time.h> #include <cuda_runtime.h> #define WIDTH 32 #define ROW_M 256 #define COL_M 256 #define ROW_N 256 #define COL_N 256 #define ROW_MxN ROW_M #define COL_MxN COL_N __global__ void MatMulKernel(float *Md, float *Nd, float *Pd, int width); void MatMul(float *M, float *N, float *P, int width); void printTwoDimDynamicArray(float *Array, const int col, const int row); int main(int argc, char *argv[]) { int i, j, k; int width = WIDTH; // Dynamic float *M = (float *)malloc(ROW_M * COL_M * sizeof(float)); float *N = (float *)malloc(ROW_N * COL_N * sizeof(float)); float *P = (float *)malloc(ROW_MxN * COL_MxN * sizeof(float)); float *MxN = (float *)malloc(ROW_MxN * COL_MxN * sizeof(float)); int pass = 1; // Initial for (i = 0; i < ROW_M; ++i) { for (j = 0; j < COL_M; ++j) { M[i*COL_M + j] = rand() % 5; } } for (i = 0; i < ROW_N; ++i) { for (j = 0; j < COL_N; ++j) { N[i*COL_N + j] = rand() % 5; } } struct timeval starttime, endtime; gettimeofday(&starttime, NULL); // CPU for (i = 0; i < ROW_M; ++i) { for (j = 0; j < COL_N; ++j) { for (k = 0; k < COL_M; ++k) { MxN[i*COL_N + j] += M[i*COL_M + k] * N[k*COL_N + j]; } } } gettimeofday(&endtime, NULL); double executime; executime = (endtime.tv_sec - starttime.tv_sec) * 1000.0; executime += (endtime.tv_usec - starttime.tv_usec) / 1000.0; printf("CPU time: %13lf msec\n", executime); /* printTwoDimDynamicArray(M, COL_M, ROW_M); printf("============================\n"); printTwoDimDynamicArray(N, COL_N, ROW_N); printf("============================\n"); printTwoDimDynamicArray(MxN, COL_MxN, ROW_MxN); */ // GPU MatMul((float *)M, (float *)N, (float *)P, width); // Compare for(i = 0; i < ROW_MxN; i++) { for(j = 0; j < COL_MxN; j++) { if(MxN[i*COL_MxN + j] != P[i*COL_MxN + j]) { printf("MxN[%d][%d] = %2.0f P[%d][%d] = %2.0f\n", i, j, MxN[i*COL_MxN + j], i, j, P[i*COL_MxN + j]); pass = 0; } } } free(M); free(N); free(P); free(MxN); printf("Test %s\n", (pass)?"PASSED":"FAILED"); return 0; } void printTwoDimDynamicArray(float *Array, const int col, const int row) { int x, y; for(y = 0; y != row; ++y) { for(x = 0; x != col; ++x) printf("%f ", Array[y*col + x]); printf("\n"); } } // Matrix multiplication kernel called by MatMul() __global__ void MatMulKernel(float *Md, float *Nd, float *Pd, int width){ int row = (blockIdx.y * blockDim.y) + threadIdx.y; int col = (blockIdx.x * blockDim.x) + threadIdx.x; float Pvalue = 0; // Multiply M and N if(row < ROW_MxN && col < COL_MxN){ for (int k = 0; k < COL_M; ++k) { float Melement = *(Md + row*COL_M + k); float Nelement = *(Nd + k*COL_N + col); Pvalue += Melement * Nelement; } *(Pd + row*COL_N + col) = Pvalue; } } // Matrix multiplication - Host code void MatMul(float *M, float *N, float *P, int width) { size_t size_M = ROW_M * COL_M * sizeof(float); size_t size_N = ROW_N * COL_N * sizeof(float); size_t size_P = ROW_MxN * COL_MxN * sizeof(float); float *Md, *Nd, *Pd; // Allocate and Load M, N to device memory cudaMalloc((void **)&Md, size_M); cudaMemcpy(Md, M, size_M, cudaMemcpyHostToDevice); cudaMalloc((void **)&Nd, size_N); cudaMemcpy(Nd, N, size_N, cudaMemcpyHostToDevice); // Allocate P on the device cudaMalloc((void **)&Pd, size_P); // Setup the execution configuration int gridDim_X = (ROW_MxN + 31)/32; int gridDim_Y = (COL_MxN + 31)/32; dim3 dimGrid(gridDim_X, gridDim_Y); dim3 dimBlock(width, width); printf("============================\n"); // Get start time event cudaEvent_t start, stop; cudaEventCreate(&start); cudaEventCreate(&stop); cudaEventRecord(start, 0); // Invoke kernel MatMulKernel<<<dimGrid, dimBlock>>>(Md, Nd, Pd, width); cudaError_t cuda_err = cudaGetLastError(); if ( cudaSuccess != cuda_err ){ printf("before kernel call: error = %s\n", cudaGetErrorString (cuda_err)); exit(1) ; } // Get stop time event cudaEventRecord(stop, 0); cudaEventSynchronize(stop); // Compute execution time float elapsedTime; cudaEventElapsedTime(&elapsedTime, start, stop); printf("GPU time: %13f msec\n", elapsedTime); cudaEventDestroy(start); cudaEventDestroy(stop); // Read P from device memory cudaMemcpy(P, Pd, size_P, cudaMemcpyDeviceToHost); // Free device memory cudaFree(Md); cudaFree(Nd); cudaFree(Pd); }
#include <stdio.h> #include <stdlib.h> #include <sys/time.h> #include <hip/hip_runtime.h> #define WIDTH 32 #define ROW_M 256 #define COL_M 256 #define ROW_N 256 #define COL_N 256 #define ROW_MxN ROW_M #define COL_MxN COL_N __global__ void MatMulKernel(float *Md, float *Nd, float *Pd, int width); void MatMul(float *M, float *N, float *P, int width); void printTwoDimDynamicArray(float *Array, const int col, const int row); int main(int argc, char *argv[]) { int i, j, k; int width = WIDTH; // Dynamic float *M = (float *)malloc(ROW_M * COL_M * sizeof(float)); float *N = (float *)malloc(ROW_N * COL_N * sizeof(float)); float *P = (float *)malloc(ROW_MxN * COL_MxN * sizeof(float)); float *MxN = (float *)malloc(ROW_MxN * COL_MxN * sizeof(float)); int pass = 1; // Initial for (i = 0; i < ROW_M; ++i) { for (j = 0; j < COL_M; ++j) { M[i*COL_M + j] = rand() % 5; } } for (i = 0; i < ROW_N; ++i) { for (j = 0; j < COL_N; ++j) { N[i*COL_N + j] = rand() % 5; } } struct timeval starttime, endtime; gettimeofday(&starttime, NULL); // CPU for (i = 0; i < ROW_M; ++i) { for (j = 0; j < COL_N; ++j) { for (k = 0; k < COL_M; ++k) { MxN[i*COL_N + j] += M[i*COL_M + k] * N[k*COL_N + j]; } } } gettimeofday(&endtime, NULL); double executime; executime = (endtime.tv_sec - starttime.tv_sec) * 1000.0; executime += (endtime.tv_usec - starttime.tv_usec) / 1000.0; printf("CPU time: %13lf msec\n", executime); /* printTwoDimDynamicArray(M, COL_M, ROW_M); printf("============================\n"); printTwoDimDynamicArray(N, COL_N, ROW_N); printf("============================\n"); printTwoDimDynamicArray(MxN, COL_MxN, ROW_MxN); */ // GPU MatMul((float *)M, (float *)N, (float *)P, width); // Compare for(i = 0; i < ROW_MxN; i++) { for(j = 0; j < COL_MxN; j++) { if(MxN[i*COL_MxN + j] != P[i*COL_MxN + j]) { printf("MxN[%d][%d] = %2.0f P[%d][%d] = %2.0f\n", i, j, MxN[i*COL_MxN + j], i, j, P[i*COL_MxN + j]); pass = 0; } } } free(M); free(N); free(P); free(MxN); printf("Test %s\n", (pass)?"PASSED":"FAILED"); return 0; } void printTwoDimDynamicArray(float *Array, const int col, const int row) { int x, y; for(y = 0; y != row; ++y) { for(x = 0; x != col; ++x) printf("%f ", Array[y*col + x]); printf("\n"); } } // Matrix multiplication kernel called by MatMul() __global__ void MatMulKernel(float *Md, float *Nd, float *Pd, int width){ int row = (blockIdx.y * blockDim.y) + threadIdx.y; int col = (blockIdx.x * blockDim.x) + threadIdx.x; float Pvalue = 0; // Multiply M and N if(row < ROW_MxN && col < COL_MxN){ for (int k = 0; k < COL_M; ++k) { float Melement = *(Md + row*COL_M + k); float Nelement = *(Nd + k*COL_N + col); Pvalue += Melement * Nelement; } *(Pd + row*COL_N + col) = Pvalue; } } // Matrix multiplication - Host code void MatMul(float *M, float *N, float *P, int width) { size_t size_M = ROW_M * COL_M * sizeof(float); size_t size_N = ROW_N * COL_N * sizeof(float); size_t size_P = ROW_MxN * COL_MxN * sizeof(float); float *Md, *Nd, *Pd; // Allocate and Load M, N to device memory hipMalloc((void **)&Md, size_M); hipMemcpy(Md, M, size_M, hipMemcpyHostToDevice); hipMalloc((void **)&Nd, size_N); hipMemcpy(Nd, N, size_N, hipMemcpyHostToDevice); // Allocate P on the device hipMalloc((void **)&Pd, size_P); // Setup the execution configuration int gridDim_X = (ROW_MxN + 31)/32; int gridDim_Y = (COL_MxN + 31)/32; dim3 dimGrid(gridDim_X, gridDim_Y); dim3 dimBlock(width, width); printf("============================\n"); // Get start time event hipEvent_t start, stop; hipEventCreate(&start); hipEventCreate(&stop); hipEventRecord(start, 0); // Invoke kernel MatMulKernel<<<dimGrid, dimBlock>>>(Md, Nd, Pd, width); hipError_t cuda_err = hipGetLastError(); if ( hipSuccess != cuda_err ){ printf("before kernel call: error = %s\n", hipGetErrorString (cuda_err)); exit(1) ; } // Get stop time event hipEventRecord(stop, 0); hipEventSynchronize(stop); // Compute execution time float elapsedTime; hipEventElapsedTime(&elapsedTime, start, stop); printf("GPU time: %13f msec\n", elapsedTime); hipEventDestroy(start); hipEventDestroy(stop); // Read P from device memory hipMemcpy(P, Pd, size_P, hipMemcpyDeviceToHost); // Free device memory hipFree(Md); hipFree(Nd); hipFree(Pd); }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <stdio.h> #include <stdlib.h> #include <sys/time.h> #include <hip/hip_runtime.h> #define WIDTH 32 #define ROW_M 256 #define COL_M 256 #define ROW_N 256 #define COL_N 256 #define ROW_MxN ROW_M #define COL_MxN COL_N __global__ void MatMulKernel(float *Md, float *Nd, float *Pd, int width); void MatMul(float *M, float *N, float *P, int width); void printTwoDimDynamicArray(float *Array, const int col, const int row); int main(int argc, char *argv[]) { int i, j, k; int width = WIDTH; // Dynamic float *M = (float *)malloc(ROW_M * COL_M * sizeof(float)); float *N = (float *)malloc(ROW_N * COL_N * sizeof(float)); float *P = (float *)malloc(ROW_MxN * COL_MxN * sizeof(float)); float *MxN = (float *)malloc(ROW_MxN * COL_MxN * sizeof(float)); int pass = 1; // Initial for (i = 0; i < ROW_M; ++i) { for (j = 0; j < COL_M; ++j) { M[i*COL_M + j] = rand() % 5; } } for (i = 0; i < ROW_N; ++i) { for (j = 0; j < COL_N; ++j) { N[i*COL_N + j] = rand() % 5; } } struct timeval starttime, endtime; gettimeofday(&starttime, NULL); // CPU for (i = 0; i < ROW_M; ++i) { for (j = 0; j < COL_N; ++j) { for (k = 0; k < COL_M; ++k) { MxN[i*COL_N + j] += M[i*COL_M + k] * N[k*COL_N + j]; } } } gettimeofday(&endtime, NULL); double executime; executime = (endtime.tv_sec - starttime.tv_sec) * 1000.0; executime += (endtime.tv_usec - starttime.tv_usec) / 1000.0; printf("CPU time: %13lf msec\n", executime); /* printTwoDimDynamicArray(M, COL_M, ROW_M); printf("============================\n"); printTwoDimDynamicArray(N, COL_N, ROW_N); printf("============================\n"); printTwoDimDynamicArray(MxN, COL_MxN, ROW_MxN); */ // GPU MatMul((float *)M, (float *)N, (float *)P, width); // Compare for(i = 0; i < ROW_MxN; i++) { for(j = 0; j < COL_MxN; j++) { if(MxN[i*COL_MxN + j] != P[i*COL_MxN + j]) { printf("MxN[%d][%d] = %2.0f P[%d][%d] = %2.0f\n", i, j, MxN[i*COL_MxN + j], i, j, P[i*COL_MxN + j]); pass = 0; } } } free(M); free(N); free(P); free(MxN); printf("Test %s\n", (pass)?"PASSED":"FAILED"); return 0; } void printTwoDimDynamicArray(float *Array, const int col, const int row) { int x, y; for(y = 0; y != row; ++y) { for(x = 0; x != col; ++x) printf("%f ", Array[y*col + x]); printf("\n"); } } // Matrix multiplication kernel called by MatMul() __global__ void MatMulKernel(float *Md, float *Nd, float *Pd, int width){ int row = (blockIdx.y * blockDim.y) + threadIdx.y; int col = (blockIdx.x * blockDim.x) + threadIdx.x; float Pvalue = 0; // Multiply M and N if(row < ROW_MxN && col < COL_MxN){ for (int k = 0; k < COL_M; ++k) { float Melement = *(Md + row*COL_M + k); float Nelement = *(Nd + k*COL_N + col); Pvalue += Melement * Nelement; } *(Pd + row*COL_N + col) = Pvalue; } } // Matrix multiplication - Host code void MatMul(float *M, float *N, float *P, int width) { size_t size_M = ROW_M * COL_M * sizeof(float); size_t size_N = ROW_N * COL_N * sizeof(float); size_t size_P = ROW_MxN * COL_MxN * sizeof(float); float *Md, *Nd, *Pd; // Allocate and Load M, N to device memory hipMalloc((void **)&Md, size_M); hipMemcpy(Md, M, size_M, hipMemcpyHostToDevice); hipMalloc((void **)&Nd, size_N); hipMemcpy(Nd, N, size_N, hipMemcpyHostToDevice); // Allocate P on the device hipMalloc((void **)&Pd, size_P); // Setup the execution configuration int gridDim_X = (ROW_MxN + 31)/32; int gridDim_Y = (COL_MxN + 31)/32; dim3 dimGrid(gridDim_X, gridDim_Y); dim3 dimBlock(width, width); printf("============================\n"); // Get start time event hipEvent_t start, stop; hipEventCreate(&start); hipEventCreate(&stop); hipEventRecord(start, 0); // Invoke kernel MatMulKernel<<<dimGrid, dimBlock>>>(Md, Nd, Pd, width); hipError_t cuda_err = hipGetLastError(); if ( hipSuccess != cuda_err ){ printf("before kernel call: error = %s\n", hipGetErrorString (cuda_err)); exit(1) ; } // Get stop time event hipEventRecord(stop, 0); hipEventSynchronize(stop); // Compute execution time float elapsedTime; hipEventElapsedTime(&elapsedTime, start, stop); printf("GPU time: %13f msec\n", elapsedTime); hipEventDestroy(start); hipEventDestroy(stop); // Read P from device memory hipMemcpy(P, Pd, size_P, hipMemcpyDeviceToHost); // Free device memory hipFree(Md); hipFree(Nd); hipFree(Pd); }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z12MatMulKernelPfS_S_i .globl _Z12MatMulKernelPfS_S_i .p2align 8 .type _Z12MatMulKernelPfS_S_i,@function _Z12MatMulKernelPfS_S_i: s_load_b32 s2, s[0:1], 0x2c v_bfe_u32 v1, v0, 10, 10 v_and_b32_e32 v4, 0x3ff, v0 s_waitcnt lgkmcnt(0) s_lshr_b32 s3, s2, 16 s_and_b32 s2, s2, 0xffff v_mad_u64_u32 v[2:3], null, s15, s3, v[1:2] v_mad_u64_u32 v[0:1], null, s14, s2, v[4:5] s_mov_b32 s2, exec_lo s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_max_i32_e32 v1, v2, v0 v_cmpx_gt_i32_e32 0x100, v1 s_cbranch_execz .LBB0_4 s_load_b128 s[4:7], s[0:1], 0x0 v_lshlrev_b32_e32 v3, 8, v2 v_ashrrev_i32_e32 v1, 31, v0 v_mov_b32_e32 v5, 0 s_mov_b64 s[2:3], 0 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_ashrrev_i32_e32 v4, 31, v3 v_lshlrev_b64 v[8:9], 2, v[0:1] s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[3:4], 2, v[3:4] s_waitcnt lgkmcnt(0) v_add_co_u32 v6, vcc_lo, s4, v3 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_ci_u32_e32 v7, vcc_lo, s5, v4, vcc_lo v_add_co_u32 v3, vcc_lo, s6, v8 v_add_co_ci_u32_e32 v4, vcc_lo, s7, v9, vcc_lo .LBB0_2: s_delay_alu instid0(VALU_DEP_4) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_u32 v8, vcc_lo, v6, s2 v_add_co_ci_u32_e32 v9, vcc_lo, s3, v7, vcc_lo s_add_u32 s2, s2, 4 s_addc_u32 s3, s3, 0 global_load_b32 v10, v[3:4], off global_load_b32 v8, v[8:9], off v_add_co_u32 v3, vcc_lo, v3, 0x400 v_add_co_ci_u32_e32 v4, vcc_lo, 0, v4, vcc_lo s_cmpk_eq_i32 s2, 0x400 s_waitcnt vmcnt(0) v_fmac_f32_e32 v5, v8, v10 s_cbranch_scc0 .LBB0_2 s_load_b64 s[0:1], s[0:1], 0x10 v_lshlrev_b32_e32 v2, 8, v2 v_lshlrev_b64 v[0:1], 2, v[0:1] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v3, 31, v2 v_lshlrev_b64 v[2:3], 2, v[2:3] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v2, vcc_lo, s0, v2 v_add_co_ci_u32_e32 v3, vcc_lo, s1, v3, vcc_lo s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v0, vcc_lo, v2, v0 v_add_co_ci_u32_e32 v1, vcc_lo, v3, v1, vcc_lo global_store_b32 v[0:1], v5, off .LBB0_4: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z12MatMulKernelPfS_S_i .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 288 .amdhsa_user_sgpr_count 14 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 1 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 1 .amdhsa_next_free_vgpr 11 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z12MatMulKernelPfS_S_i, .Lfunc_end0-_Z12MatMulKernelPfS_S_i .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 8 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .offset: 24 .size: 4 .value_kind: by_value - .offset: 32 .size: 4 .value_kind: hidden_block_count_x - .offset: 36 .size: 4 .value_kind: hidden_block_count_y - .offset: 40 .size: 4 .value_kind: hidden_block_count_z - .offset: 44 .size: 2 .value_kind: hidden_group_size_x - .offset: 46 .size: 2 .value_kind: hidden_group_size_y - .offset: 48 .size: 2 .value_kind: hidden_group_size_z - .offset: 50 .size: 2 .value_kind: hidden_remainder_x - .offset: 52 .size: 2 .value_kind: hidden_remainder_y - .offset: 54 .size: 2 .value_kind: hidden_remainder_z - .offset: 72 .size: 8 .value_kind: hidden_global_offset_x - .offset: 80 .size: 8 .value_kind: hidden_global_offset_y - .offset: 88 .size: 8 .value_kind: hidden_global_offset_z - .offset: 96 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 288 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z12MatMulKernelPfS_S_i .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z12MatMulKernelPfS_S_i.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 11 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <stdio.h> #include <stdlib.h> #include <sys/time.h> #include <hip/hip_runtime.h> #define WIDTH 32 #define ROW_M 256 #define COL_M 256 #define ROW_N 256 #define COL_N 256 #define ROW_MxN ROW_M #define COL_MxN COL_N __global__ void MatMulKernel(float *Md, float *Nd, float *Pd, int width); void MatMul(float *M, float *N, float *P, int width); void printTwoDimDynamicArray(float *Array, const int col, const int row); int main(int argc, char *argv[]) { int i, j, k; int width = WIDTH; // Dynamic float *M = (float *)malloc(ROW_M * COL_M * sizeof(float)); float *N = (float *)malloc(ROW_N * COL_N * sizeof(float)); float *P = (float *)malloc(ROW_MxN * COL_MxN * sizeof(float)); float *MxN = (float *)malloc(ROW_MxN * COL_MxN * sizeof(float)); int pass = 1; // Initial for (i = 0; i < ROW_M; ++i) { for (j = 0; j < COL_M; ++j) { M[i*COL_M + j] = rand() % 5; } } for (i = 0; i < ROW_N; ++i) { for (j = 0; j < COL_N; ++j) { N[i*COL_N + j] = rand() % 5; } } struct timeval starttime, endtime; gettimeofday(&starttime, NULL); // CPU for (i = 0; i < ROW_M; ++i) { for (j = 0; j < COL_N; ++j) { for (k = 0; k < COL_M; ++k) { MxN[i*COL_N + j] += M[i*COL_M + k] * N[k*COL_N + j]; } } } gettimeofday(&endtime, NULL); double executime; executime = (endtime.tv_sec - starttime.tv_sec) * 1000.0; executime += (endtime.tv_usec - starttime.tv_usec) / 1000.0; printf("CPU time: %13lf msec\n", executime); /* printTwoDimDynamicArray(M, COL_M, ROW_M); printf("============================\n"); printTwoDimDynamicArray(N, COL_N, ROW_N); printf("============================\n"); printTwoDimDynamicArray(MxN, COL_MxN, ROW_MxN); */ // GPU MatMul((float *)M, (float *)N, (float *)P, width); // Compare for(i = 0; i < ROW_MxN; i++) { for(j = 0; j < COL_MxN; j++) { if(MxN[i*COL_MxN + j] != P[i*COL_MxN + j]) { printf("MxN[%d][%d] = %2.0f P[%d][%d] = %2.0f\n", i, j, MxN[i*COL_MxN + j], i, j, P[i*COL_MxN + j]); pass = 0; } } } free(M); free(N); free(P); free(MxN); printf("Test %s\n", (pass)?"PASSED":"FAILED"); return 0; } void printTwoDimDynamicArray(float *Array, const int col, const int row) { int x, y; for(y = 0; y != row; ++y) { for(x = 0; x != col; ++x) printf("%f ", Array[y*col + x]); printf("\n"); } } // Matrix multiplication kernel called by MatMul() __global__ void MatMulKernel(float *Md, float *Nd, float *Pd, int width){ int row = (blockIdx.y * blockDim.y) + threadIdx.y; int col = (blockIdx.x * blockDim.x) + threadIdx.x; float Pvalue = 0; // Multiply M and N if(row < ROW_MxN && col < COL_MxN){ for (int k = 0; k < COL_M; ++k) { float Melement = *(Md + row*COL_M + k); float Nelement = *(Nd + k*COL_N + col); Pvalue += Melement * Nelement; } *(Pd + row*COL_N + col) = Pvalue; } } // Matrix multiplication - Host code void MatMul(float *M, float *N, float *P, int width) { size_t size_M = ROW_M * COL_M * sizeof(float); size_t size_N = ROW_N * COL_N * sizeof(float); size_t size_P = ROW_MxN * COL_MxN * sizeof(float); float *Md, *Nd, *Pd; // Allocate and Load M, N to device memory hipMalloc((void **)&Md, size_M); hipMemcpy(Md, M, size_M, hipMemcpyHostToDevice); hipMalloc((void **)&Nd, size_N); hipMemcpy(Nd, N, size_N, hipMemcpyHostToDevice); // Allocate P on the device hipMalloc((void **)&Pd, size_P); // Setup the execution configuration int gridDim_X = (ROW_MxN + 31)/32; int gridDim_Y = (COL_MxN + 31)/32; dim3 dimGrid(gridDim_X, gridDim_Y); dim3 dimBlock(width, width); printf("============================\n"); // Get start time event hipEvent_t start, stop; hipEventCreate(&start); hipEventCreate(&stop); hipEventRecord(start, 0); // Invoke kernel MatMulKernel<<<dimGrid, dimBlock>>>(Md, Nd, Pd, width); hipError_t cuda_err = hipGetLastError(); if ( hipSuccess != cuda_err ){ printf("before kernel call: error = %s\n", hipGetErrorString (cuda_err)); exit(1) ; } // Get stop time event hipEventRecord(stop, 0); hipEventSynchronize(stop); // Compute execution time float elapsedTime; hipEventElapsedTime(&elapsedTime, start, stop); printf("GPU time: %13f msec\n", elapsedTime); hipEventDestroy(start); hipEventDestroy(stop); // Read P from device memory hipMemcpy(P, Pd, size_P, hipMemcpyDeviceToHost); // Free device memory hipFree(Md); hipFree(Nd); hipFree(Pd); }
.text .file "matrix.hip" .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 # -- Begin function main .LCPI0_0: .quad 0x408f400000000000 # double 1000 .text .globl main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 subq $56, %rsp .cfi_def_cfa_offset 112 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl $262144, %edi # imm = 0x40000 callq malloc movq %rax, %rbp movl $262144, %edi # imm = 0x40000 callq malloc movq %rax, %r13 movl $262144, %edi # imm = 0x40000 callq malloc movq %rax, (%rsp) # 8-byte Spill movl $262144, %edi # imm = 0x40000 callq malloc movq %rax, %rbx xorl %r14d, %r14d movq %rbp, %r15 .p2align 4, 0x90 .LBB0_1: # %.preheader73 # =>This Loop Header: Depth=1 # Child Loop BB0_2 Depth 2 xorl %r12d, %r12d .p2align 4, 0x90 .LBB0_2: # Parent Loop BB0_1 Depth=1 # => This Inner Loop Header: Depth=2 callq rand cltq imulq $1717986919, %rax, %rcx # imm = 0x66666667 movq %rcx, %rdx shrq $63, %rdx sarq $33, %rcx addl %edx, %ecx leal (%rcx,%rcx,4), %ecx subl %ecx, %eax xorps %xmm0, %xmm0 cvtsi2ss %eax, %xmm0 movss %xmm0, (%r15,%r12,4) incq %r12 cmpq $256, %r12 # imm = 0x100 jne .LBB0_2 # %bb.3: # in Loop: Header=BB0_1 Depth=1 incq %r14 addq $1024, %r15 # imm = 0x400 cmpq $256, %r14 # imm = 0x100 jne .LBB0_1 # %bb.4: # %.preheader71.preheader xorl %r14d, %r14d movq %r13, %r15 .p2align 4, 0x90 .LBB0_5: # %.preheader71 # =>This Loop Header: Depth=1 # Child Loop BB0_6 Depth 2 xorl %r12d, %r12d .p2align 4, 0x90 .LBB0_6: # Parent Loop BB0_5 Depth=1 # => This Inner Loop Header: Depth=2 callq rand cltq imulq $1717986919, %rax, %rcx # imm = 0x66666667 movq %rcx, %rdx shrq $63, %rdx sarq $33, %rcx addl %edx, %ecx leal (%rcx,%rcx,4), %ecx subl %ecx, %eax xorps %xmm0, %xmm0 cvtsi2ss %eax, %xmm0 movss %xmm0, (%r15,%r12,4) incq %r12 cmpq $256, %r12 # imm = 0x100 jne .LBB0_6 # %bb.7: # in Loop: Header=BB0_5 Depth=1 incq %r14 addq $1024, %r15 # imm = 0x400 cmpq $256, %r14 # imm = 0x100 jne .LBB0_5 # %bb.8: xorl %r15d, %r15d leaq 40(%rsp), %rdi xorl %esi, %esi callq gettimeofday movq %rbp, %rax .p2align 4, 0x90 .LBB0_9: # %.preheader70 # =>This Loop Header: Depth=1 # Child Loop BB0_10 Depth 2 # Child Loop BB0_11 Depth 3 movq %r15, %rcx shlq $8, %rcx movq %r13, %rdx xorl %esi, %esi .p2align 4, 0x90 .LBB0_10: # %.preheader69 # Parent Loop BB0_9 Depth=1 # => This Loop Header: Depth=2 # Child Loop BB0_11 Depth 3 movq %rsi, %rdi orq %rcx, %rdi movss (%rbx,%rdi,4), %xmm0 # xmm0 = mem[0],zero,zero,zero movq %rdx, %r8 xorl %r9d, %r9d .p2align 4, 0x90 .LBB0_11: # Parent Loop BB0_9 Depth=1 # Parent Loop BB0_10 Depth=2 # => This Inner Loop Header: Depth=3 movss (%rax,%r9,4), %xmm1 # xmm1 = mem[0],zero,zero,zero mulss (%r8), %xmm1 addss %xmm1, %xmm0 incq %r9 addq $1024, %r8 # imm = 0x400 cmpq $256, %r9 # imm = 0x100 jne .LBB0_11 # %bb.12: # in Loop: Header=BB0_10 Depth=2 movss %xmm0, (%rbx,%rdi,4) incq %rsi addq $4, %rdx cmpq $256, %rsi # imm = 0x100 jne .LBB0_10 # %bb.13: # in Loop: Header=BB0_9 Depth=1 incq %r15 addq $1024, %rax # imm = 0x400 cmpq $256, %r15 # imm = 0x100 jne .LBB0_9 # %bb.14: movq %r13, %r14 xorl %r13d, %r13d leaq 24(%rsp), %rdi xorl %esi, %esi callq gettimeofday movq 24(%rsp), %rax movq 32(%rsp), %rcx subq 40(%rsp), %rax xorps %xmm1, %xmm1 cvtsi2sd %rax, %xmm1 movsd .LCPI0_0(%rip), %xmm2 # xmm2 = mem[0],zero mulsd %xmm2, %xmm1 subq 48(%rsp), %rcx xorps %xmm0, %xmm0 cvtsi2sd %rcx, %xmm0 divsd %xmm2, %xmm0 addsd %xmm1, %xmm0 movl $.L.str, %edi movb $1, %al callq printf movq %rbp, 16(%rsp) # 8-byte Spill movq %rbp, %rdi movq %r14, 8(%rsp) # 8-byte Spill movq %r14, %rsi movq (%rsp), %r14 # 8-byte Reload movq %r14, %rdx movl $32, %ecx callq _Z6MatMulPfS_S_i movl $1, %r15d movq %rbx, %r12 jmp .LBB0_15 .p2align 4, 0x90 .LBB0_19: # in Loop: Header=BB0_15 Depth=1 incq %r13 addq $1024, %r14 # imm = 0x400 addq $1024, %r12 # imm = 0x400 cmpq $256, %r13 # imm = 0x100 je .LBB0_20 .LBB0_15: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB0_16 Depth 2 xorl %ebp, %ebp jmp .LBB0_16 .p2align 4, 0x90 .LBB0_18: # in Loop: Header=BB0_16 Depth=2 incq %rbp cmpq $256, %rbp # imm = 0x100 je .LBB0_19 .LBB0_16: # Parent Loop BB0_15 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r12,%rbp,4), %xmm0 # xmm0 = mem[0],zero,zero,zero movss (%r14,%rbp,4), %xmm1 # xmm1 = mem[0],zero,zero,zero ucomiss %xmm1, %xmm0 jne .LBB0_17 jnp .LBB0_18 .LBB0_17: # in Loop: Header=BB0_16 Depth=2 cvtss2sd %xmm0, %xmm0 cvtss2sd %xmm1, %xmm1 movl $.L.str.1, %edi movl %r13d, %esi movl %ebp, %edx movl %r13d, %ecx movl %ebp, %r8d movb $2, %al callq printf xorl %r15d, %r15d jmp .LBB0_18 .LBB0_20: movq 16(%rsp), %rdi # 8-byte Reload callq free movq 8(%rsp), %rdi # 8-byte Reload callq free movq (%rsp), %rdi # 8-byte Reload callq free movq %rbx, %rdi callq free testl %r15d, %r15d movl $.L.str.4, %eax movl $.L.str.3, %esi cmoveq %rax, %rsi movl $.L.str.2, %edi xorl %eax, %eax callq printf xorl %eax, %eax addq $56, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end0: .size main, .Lfunc_end0-main .cfi_endproc # -- End function .globl _Z6MatMulPfS_S_i # -- Begin function _Z6MatMulPfS_S_i .p2align 4, 0x90 .type _Z6MatMulPfS_S_i,@function _Z6MatMulPfS_S_i: # @_Z6MatMulPfS_S_i .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %rbx .cfi_def_cfa_offset 40 subq $168, %rsp .cfi_def_cfa_offset 208 .cfi_offset %rbx, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl %ecx, %ebp movq %rdx, %rbx movq %rsi, %r14 movq %rdi, %r15 leaq 40(%rsp), %rdi movl $262144, %esi # imm = 0x40000 callq hipMalloc movq 40(%rsp), %rdi movl $262144, %edx # imm = 0x40000 movq %r15, %rsi movl $1, %ecx callq hipMemcpy leaq 32(%rsp), %rdi movl $262144, %esi # imm = 0x40000 callq hipMalloc movq 32(%rsp), %rdi movl $262144, %edx # imm = 0x40000 movq %r14, %rsi movl $1, %ecx callq hipMemcpy leaq 24(%rsp), %rdi movl $262144, %esi # imm = 0x40000 callq hipMalloc movl %ebp, %eax movq %rax, %r14 shlq $32, %r14 orq %rax, %r14 movl $.Lstr, %edi callq puts@PLT leaq 16(%rsp), %rdi callq hipEventCreate leaq 8(%rsp), %rdi callq hipEventCreate movq 16(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movabsq $34359738376, %rdi # imm = 0x800000008 movl $1, %esi movq %r14, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_2 # %bb.1: movq 40(%rsp), %rax movq 32(%rsp), %rcx movq 24(%rsp), %rdx movq %rax, 120(%rsp) movq %rcx, 112(%rsp) movq %rdx, 104(%rsp) movl %ebp, 52(%rsp) leaq 120(%rsp), %rax movq %rax, 128(%rsp) leaq 112(%rsp), %rax movq %rax, 136(%rsp) leaq 104(%rsp), %rax movq %rax, 144(%rsp) leaq 52(%rsp), %rax movq %rax, 152(%rsp) leaq 88(%rsp), %rdi leaq 72(%rsp), %rsi leaq 64(%rsp), %rdx leaq 56(%rsp), %rcx callq __hipPopCallConfiguration movq 88(%rsp), %rsi movl 96(%rsp), %edx movq 72(%rsp), %rcx movl 80(%rsp), %r8d leaq 128(%rsp), %r9 movl $_Z12MatMulKernelPfS_S_i, %edi pushq 56(%rsp) .cfi_adjust_cfa_offset 8 pushq 72(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_2: callq hipGetLastError testl %eax, %eax jne .LBB1_4 # %bb.3: movq 8(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movq 8(%rsp), %rdi callq hipEventSynchronize movq 16(%rsp), %rsi movq 8(%rsp), %rdx leaq 128(%rsp), %rdi callq hipEventElapsedTime movss 128(%rsp), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str.9, %edi movb $1, %al callq printf movq 16(%rsp), %rdi callq hipEventDestroy movq 8(%rsp), %rdi callq hipEventDestroy movq 24(%rsp), %rsi movl $262144, %edx # imm = 0x40000 movq %rbx, %rdi movl $2, %ecx callq hipMemcpy movq 40(%rsp), %rdi callq hipFree movq 32(%rsp), %rdi callq hipFree movq 24(%rsp), %rdi callq hipFree addq $168, %rsp .cfi_def_cfa_offset 40 popq %rbx .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .LBB1_4: .cfi_def_cfa_offset 208 movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %edi movq %rax, %rsi xorl %eax, %eax callq printf movl $1, %edi callq exit .Lfunc_end1: .size _Z6MatMulPfS_S_i, .Lfunc_end1-_Z6MatMulPfS_S_i .cfi_endproc # -- End function .globl _Z23printTwoDimDynamicArrayPfii # -- Begin function _Z23printTwoDimDynamicArrayPfii .p2align 4, 0x90 .type _Z23printTwoDimDynamicArrayPfii,@function _Z23printTwoDimDynamicArrayPfii: # @_Z23printTwoDimDynamicArrayPfii .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 pushq %rax .cfi_def_cfa_offset 64 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl %esi, 4(%rsp) # 4-byte Spill testl %edx, %edx je .LBB2_6 # %bb.1: # %.preheader.lr.ph movq %rdi, %r14 movslq 4(%rsp), %r15 # 4-byte Folded Reload movl %edx, %r12d movl %r15d, %r13d shlq $2, %r15 xorl %ebp, %ebp jmp .LBB2_2 .p2align 4, 0x90 .LBB2_5: # %._crit_edge # in Loop: Header=BB2_2 Depth=1 movl $10, %edi callq putchar@PLT incq %rbp addq %r15, %r14 cmpq %r12, %rbp je .LBB2_6 .LBB2_2: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB2_4 Depth 2 cmpl $0, 4(%rsp) # 4-byte Folded Reload je .LBB2_5 # %bb.3: # %.lr.ph # in Loop: Header=BB2_2 Depth=1 xorl %ebx, %ebx .p2align 4, 0x90 .LBB2_4: # Parent Loop BB2_2 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r14,%rbx,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str.5, %edi movb $1, %al callq printf incq %rbx cmpq %rbx, %r13 jne .LBB2_4 jmp .LBB2_5 .LBB2_6: # %._crit_edge16 addq $8, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end2: .size _Z23printTwoDimDynamicArrayPfii, .Lfunc_end2-_Z23printTwoDimDynamicArrayPfii .cfi_endproc # -- End function .globl _Z27__device_stub__MatMulKernelPfS_S_i # -- Begin function _Z27__device_stub__MatMulKernelPfS_S_i .p2align 4, 0x90 .type _Z27__device_stub__MatMulKernelPfS_S_i,@function _Z27__device_stub__MatMulKernelPfS_S_i: # @_Z27__device_stub__MatMulKernelPfS_S_i .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movq %rdx, 56(%rsp) movl %ecx, 4(%rsp) leaq 72(%rsp), %rax movq %rax, 80(%rsp) leaq 64(%rsp), %rax movq %rax, 88(%rsp) leaq 56(%rsp), %rax movq %rax, 96(%rsp) leaq 4(%rsp), %rax movq %rax, 104(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 80(%rsp), %r9 movl $_Z12MatMulKernelPfS_S_i, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $136, %rsp .cfi_adjust_cfa_offset -136 retq .Lfunc_end3: .size _Z27__device_stub__MatMulKernelPfS_S_i, .Lfunc_end3-_Z27__device_stub__MatMulKernelPfS_S_i .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB4_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB4_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z12MatMulKernelPfS_S_i, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end4: .size __hip_module_ctor, .Lfunc_end4-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB5_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB5_2: retq .Lfunc_end5: .size __hip_module_dtor, .Lfunc_end5-__hip_module_dtor .cfi_endproc # -- End function .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "CPU time: %13lf msec\n" .size .L.str, 22 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "MxN[%d][%d] = %2.0f P[%d][%d] = %2.0f\n" .size .L.str.1, 41 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "Test %s\n" .size .L.str.2, 9 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "PASSED" .size .L.str.3, 7 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "FAILED" .size .L.str.4, 7 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "%f " .size .L.str.5, 4 .type _Z12MatMulKernelPfS_S_i,@object # @_Z12MatMulKernelPfS_S_i .section .rodata,"a",@progbits .globl _Z12MatMulKernelPfS_S_i .p2align 3, 0x0 _Z12MatMulKernelPfS_S_i: .quad _Z27__device_stub__MatMulKernelPfS_S_i .size _Z12MatMulKernelPfS_S_i, 8 .type .L.str.8,@object # @.str.8 .section .rodata.str1.1,"aMS",@progbits,1 .L.str.8: .asciz "before kernel call: error = %s\n" .size .L.str.8, 32 .type .L.str.9,@object # @.str.9 .L.str.9: .asciz "GPU time: %13f msec\n" .size .L.str.9, 21 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z12MatMulKernelPfS_S_i" .size .L__unnamed_1, 24 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .type .Lstr,@object # @str .section .rodata.str1.1,"aMS",@progbits,1 .Lstr: .asciz "============================" .size .Lstr, 29 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z27__device_stub__MatMulKernelPfS_S_i .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z12MatMulKernelPfS_S_i .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA device assembly to AMD device assembly.
code for sm_80 Function : _Z12MatMulKernelPfS_S_i .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R0, SR_CTAID.X ; /* 0x0000000000007919 */ /* 0x000e280000002500 */ /*0020*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e280000002100 */ /*0030*/ S2R R7, SR_CTAID.Y ; /* 0x0000000000077919 */ /* 0x000e680000002600 */ /*0040*/ S2R R2, SR_TID.Y ; /* 0x0000000000027919 */ /* 0x000e620000002200 */ /*0050*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x001fca00078e0203 */ /*0060*/ ISETP.GT.AND P0, PT, R0, 0xff, PT ; /* 0x000000ff0000780c */ /* 0x000fe20003f04270 */ /*0070*/ IMAD R7, R7, c[0x0][0x4], R2 ; /* 0x0000010007077a24 */ /* 0x002fca00078e0202 */ /*0080*/ ISETP.GT.OR P0, PT, R7, 0xff, P0 ; /* 0x000000ff0700780c */ /* 0x000fda0000704670 */ /*0090*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*00a0*/ HFMA2.MMA R4, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff047435 */ /* 0x000fe200000001ff */ /*00b0*/ SHF.L.U32 R7, R7, 0x8, RZ ; /* 0x0000000807077819 */ /* 0x000fe200000006ff */ /*00c0*/ HFMA2.MMA R8, -RZ, RZ, 0, 0 ; /* 0x00000000ff087435 */ /* 0x000fe200000001ff */ /*00d0*/ SHF.R.S32.HI R6, RZ, 0x1f, R0 ; /* 0x0000001fff067819 */ /* 0x000fe20000011400 */ /*00e0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*00f0*/ MOV R15, RZ ; /* 0x000000ff000f7202 */ /* 0x000fca0000000f00 */ /*0100*/ IMAD.WIDE R2, R0, R4, c[0x0][0x168] ; /* 0x00005a0000027625 */ /* 0x000fc800078e0204 */ /*0110*/ IMAD.WIDE R4, R7, R4, c[0x0][0x160] ; /* 0x0000580007047625 */ /* 0x000fe200078e0204 */ /*0120*/ IADD3 R10, P0, R2, 0x2000, RZ ; /* 0x00002000020a7810 */ /* 0x000fc80007f1e0ff */ /*0130*/ IADD3 R2, P1, R4, 0x20, RZ ; /* 0x0000002004027810 */ /* 0x000fe40007f3e0ff */ /*0140*/ IADD3.X R11, RZ, R3, RZ, P0, !PT ; /* 0x00000003ff0b7210 */ /* 0x000fc600007fe4ff */ /*0150*/ IMAD.X R3, RZ, RZ, R5, P1 ; /* 0x000000ffff037224 */ /* 0x000fc600008e0605 */ /*0160*/ IMAD.MOV.U32 R4, RZ, RZ, R10 ; /* 0x000000ffff047224 */ /* 0x000fe200078e000a */ /*0170*/ MOV R5, R11 ; /* 0x0000000b00057202 */ /* 0x000fe20000000f00 */ /*0180*/ LDG.E R17, [R2.64+-0x20] ; /* 0xffffe00402117981 */ /* 0x000ea8000c1e1900 */ /*0190*/ LDG.E R16, [R4.64+-0x2000] ; /* 0xffe0000404107981 */ /* 0x000ea8000c1e1900 */ /*01a0*/ LDG.E R18, [R4.64+-0x1c00] ; /* 0xffe4000404127981 */ /* 0x000ee8000c1e1900 */ /*01b0*/ LDG.E R25, [R2.64+-0x1c] ; /* 0xffffe40402197981 */ /* 0x000ee8000c1e1900 */ /*01c0*/ LDG.E R19, [R4.64+-0x1800] ; /* 0xffe8000404137981 */ /* 0x000f28000c1e1900 */ /*01d0*/ LDG.E R20, [R2.64+-0x18] ; /* 0xffffe80402147981 */ /* 0x000f28000c1e1900 */ /*01e0*/ LDG.E R22, [R4.64+-0x1400] ; /* 0xffec000404167981 */ /* 0x000f68000c1e1900 */ /*01f0*/ LDG.E R21, [R2.64+-0x14] ; /* 0xffffec0402157981 */ /* 0x000f68000c1e1900 */ /*0200*/ LDG.E R23, [R4.64+-0x1000] ; /* 0xfff0000404177981 */ /* 0x000f68000c1e1900 */ /*0210*/ LDG.E R24, [R2.64+-0x10] ; /* 0xfffff00402187981 */ /* 0x000f68000c1e1900 */ /*0220*/ LDG.E R9, [R4.64+-0xc00] ; /* 0xfff4000404097981 */ /* 0x000f68000c1e1900 */ /*0230*/ LDG.E R10, [R2.64+-0xc] ; /* 0xfffff404020a7981 */ /* 0x000f68000c1e1900 */ /*0240*/ LDG.E R11, [R4.64+-0x800] ; /* 0xfff80004040b7981 */ /* 0x000f68000c1e1900 */ /*0250*/ LDG.E R12, [R2.64+-0x8] ; /* 0xfffff804020c7981 */ /* 0x000f68000c1e1900 */ /*0260*/ LDG.E R13, [R4.64+-0x400] ; /* 0xfffc0004040d7981 */ /* 0x000f68000c1e1900 */ /*0270*/ LDG.E R14, [R2.64+-0x4] ; /* 0xfffffc04020e7981 */ /* 0x000f62000c1e1900 */ /*0280*/ FFMA R17, R16, R17, R15 ; /* 0x0000001110117223 */ /* 0x004fc6000000000f */ /*0290*/ LDG.E R15, [R4.64] ; /* 0x00000004040f7981 */ /* 0x000ea8000c1e1900 */ /*02a0*/ LDG.E R16, [R2.64] ; /* 0x0000000402107981 */ /* 0x0000a2000c1e1900 */ /*02b0*/ FFMA R25, R18, R25, R17 ; /* 0x0000001912197223 */ /* 0x008fc60000000011 */ /*02c0*/ LDG.E R17, [R4.64+0x400] ; /* 0x0004000404117981 */ /* 0x000ee8000c1e1900 */ /*02d0*/ LDG.E R18, [R2.64+0x4] ; /* 0x0000040402127981 */ /* 0x0000e2000c1e1900 */ /*02e0*/ FFMA R25, R19, R20, R25 ; /* 0x0000001413197223 */ /* 0x010fc60000000019 */ /*02f0*/ LDG.E R19, [R4.64+0x800] ; /* 0x0008000404137981 */ /* 0x000f28000c1e1900 */ /*0300*/ LDG.E R20, [R2.64+0x8] ; /* 0x0000080402147981 */ /* 0x000122000c1e1900 */ /*0310*/ FFMA R25, R22, R21, R25 ; /* 0x0000001516197223 */ /* 0x020fc60000000019 */ /*0320*/ LDG.E R21, [R4.64+0xc00] ; /* 0x000c000404157981 */ /* 0x000f68000c1e1900 */ /*0330*/ LDG.E R22, [R2.64+0xc] ; /* 0x00000c0402167981 */ /* 0x000162000c1e1900 */ /*0340*/ FFMA R25, R23, R24, R25 ; /* 0x0000001817197223 */ /* 0x000fc60000000019 */ /*0350*/ LDG.E R23, [R4.64+0x1000] ; /* 0x0010000404177981 */ /* 0x000f68000c1e1900 */ /*0360*/ LDG.E R24, [R2.64+0x10] ; /* 0x0000100402187981 */ /* 0x000162000c1e1900 */ /*0370*/ FFMA R25, R9, R10, R25 ; /* 0x0000000a09197223 */ /* 0x000fc60000000019 */ /*0380*/ LDG.E R9, [R4.64+0x1400] ; /* 0x0014000404097981 */ /* 0x000f68000c1e1900 */ /*0390*/ LDG.E R10, [R2.64+0x14] ; /* 0x00001404020a7981 */ /* 0x000162000c1e1900 */ /*03a0*/ FFMA R25, R11, R12, R25 ; /* 0x0000000c0b197223 */ /* 0x000fc60000000019 */ /*03b0*/ LDG.E R12, [R4.64+0x1800] ; /* 0x00180004040c7981 */ /* 0x000f68000c1e1900 */ /*03c0*/ LDG.E R11, [R2.64+0x18] ; /* 0x00001804020b7981 */ /* 0x000162000c1e1900 */ /*03d0*/ FFMA R25, R13, R14, R25 ; /* 0x0000000e0d197223 */ /* 0x000fc60000000019 */ /*03e0*/ LDG.E R13, [R4.64+0x1c00] ; /* 0x001c0004040d7981 */ /* 0x000f68000c1e1900 */ /*03f0*/ LDG.E R14, [R2.64+0x1c] ; /* 0x00001c04020e7981 */ /* 0x000162000c1e1900 */ /*0400*/ IADD3 R8, R8, 0x10, RZ ; /* 0x0000001008087810 */ /* 0x000fc80007ffe0ff */ /*0410*/ ISETP.NE.AND P0, PT, R8, 0x100, PT ; /* 0x000001000800780c */ /* 0x000fe40003f05270 */ /*0420*/ IADD3 R2, P2, R2, 0x40, RZ ; /* 0x0000004002027810 */ /* 0x001fc80007f5e0ff */ /*0430*/ IADD3.X R3, RZ, R3, RZ, P2, !PT ; /* 0x00000003ff037210 */ /* 0x000fe200017fe4ff */ /*0440*/ FFMA R15, R15, R16, R25 ; /* 0x000000100f0f7223 */ /* 0x004fc80000000019 */ /*0450*/ FFMA R15, R17, R18, R15 ; /* 0x00000012110f7223 */ /* 0x008fc8000000000f */ /*0460*/ FFMA R15, R19, R20, R15 ; /* 0x00000014130f7223 */ /* 0x010fc8000000000f */ /*0470*/ FFMA R15, R21, R22, R15 ; /* 0x00000016150f7223 */ /* 0x020fc8000000000f */ /*0480*/ FFMA R15, R23, R24, R15 ; /* 0x00000018170f7223 */ /* 0x000fc8000000000f */ /*0490*/ FFMA R9, R9, R10, R15 ; /* 0x0000000a09097223 */ /* 0x000fe2000000000f */ /*04a0*/ IADD3 R10, P1, R4, 0x4000, RZ ; /* 0x00004000040a7810 */ /* 0x000fc60007f3e0ff */ /*04b0*/ FFMA R9, R12, R11, R9 ; /* 0x0000000b0c097223 */ /* 0x000fe20000000009 */ /*04c0*/ IADD3.X R11, RZ, R5, RZ, P1, !PT ; /* 0x00000005ff0b7210 */ /* 0x000fc60000ffe4ff */ /*04d0*/ FFMA R15, R13, R14, R9 ; /* 0x0000000e0d0f7223 */ /* 0x000fe20000000009 */ /*04e0*/ @P0 BRA 0x160 ; /* 0xfffffc7000000947 */ /* 0x000fea000383ffff */ /*04f0*/ IADD3 R0, P0, R0, R7, RZ ; /* 0x0000000700007210 */ /* 0x000fc80007f1e0ff */ /*0500*/ LEA.HI.X.SX32 R7, R7, R6, 0x1, P0 ; /* 0x0000000607077211 */ /* 0x000fe400000f0eff */ /*0510*/ LEA R2, P0, R0, c[0x0][0x170], 0x2 ; /* 0x00005c0000027a11 */ /* 0x000fc800078010ff */ /*0520*/ LEA.HI.X R3, R0, c[0x0][0x174], R7, 0x2, P0 ; /* 0x00005d0000037a11 */ /* 0x000fca00000f1407 */ /*0530*/ STG.E [R2.64], R15 ; /* 0x0000000f02007986 */ /* 0x000fe2000c101904 */ /*0540*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0550*/ BRA 0x550; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0560*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0570*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0580*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0590*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z12MatMulKernelPfS_S_i .globl _Z12MatMulKernelPfS_S_i .p2align 8 .type _Z12MatMulKernelPfS_S_i,@function _Z12MatMulKernelPfS_S_i: s_load_b32 s2, s[0:1], 0x2c v_bfe_u32 v1, v0, 10, 10 v_and_b32_e32 v4, 0x3ff, v0 s_waitcnt lgkmcnt(0) s_lshr_b32 s3, s2, 16 s_and_b32 s2, s2, 0xffff v_mad_u64_u32 v[2:3], null, s15, s3, v[1:2] v_mad_u64_u32 v[0:1], null, s14, s2, v[4:5] s_mov_b32 s2, exec_lo s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_max_i32_e32 v1, v2, v0 v_cmpx_gt_i32_e32 0x100, v1 s_cbranch_execz .LBB0_4 s_load_b128 s[4:7], s[0:1], 0x0 v_lshlrev_b32_e32 v3, 8, v2 v_ashrrev_i32_e32 v1, 31, v0 v_mov_b32_e32 v5, 0 s_mov_b64 s[2:3], 0 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_ashrrev_i32_e32 v4, 31, v3 v_lshlrev_b64 v[8:9], 2, v[0:1] s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[3:4], 2, v[3:4] s_waitcnt lgkmcnt(0) v_add_co_u32 v6, vcc_lo, s4, v3 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_ci_u32_e32 v7, vcc_lo, s5, v4, vcc_lo v_add_co_u32 v3, vcc_lo, s6, v8 v_add_co_ci_u32_e32 v4, vcc_lo, s7, v9, vcc_lo .LBB0_2: s_delay_alu instid0(VALU_DEP_4) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_u32 v8, vcc_lo, v6, s2 v_add_co_ci_u32_e32 v9, vcc_lo, s3, v7, vcc_lo s_add_u32 s2, s2, 4 s_addc_u32 s3, s3, 0 global_load_b32 v10, v[3:4], off global_load_b32 v8, v[8:9], off v_add_co_u32 v3, vcc_lo, v3, 0x400 v_add_co_ci_u32_e32 v4, vcc_lo, 0, v4, vcc_lo s_cmpk_eq_i32 s2, 0x400 s_waitcnt vmcnt(0) v_fmac_f32_e32 v5, v8, v10 s_cbranch_scc0 .LBB0_2 s_load_b64 s[0:1], s[0:1], 0x10 v_lshlrev_b32_e32 v2, 8, v2 v_lshlrev_b64 v[0:1], 2, v[0:1] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v3, 31, v2 v_lshlrev_b64 v[2:3], 2, v[2:3] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v2, vcc_lo, s0, v2 v_add_co_ci_u32_e32 v3, vcc_lo, s1, v3, vcc_lo s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v0, vcc_lo, v2, v0 v_add_co_ci_u32_e32 v1, vcc_lo, v3, v1, vcc_lo global_store_b32 v[0:1], v5, off .LBB0_4: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z12MatMulKernelPfS_S_i .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 288 .amdhsa_user_sgpr_count 14 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 1 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 1 .amdhsa_next_free_vgpr 11 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z12MatMulKernelPfS_S_i, .Lfunc_end0-_Z12MatMulKernelPfS_S_i .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 8 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .offset: 24 .size: 4 .value_kind: by_value - .offset: 32 .size: 4 .value_kind: hidden_block_count_x - .offset: 36 .size: 4 .value_kind: hidden_block_count_y - .offset: 40 .size: 4 .value_kind: hidden_block_count_z - .offset: 44 .size: 2 .value_kind: hidden_group_size_x - .offset: 46 .size: 2 .value_kind: hidden_group_size_y - .offset: 48 .size: 2 .value_kind: hidden_group_size_z - .offset: 50 .size: 2 .value_kind: hidden_remainder_x - .offset: 52 .size: 2 .value_kind: hidden_remainder_y - .offset: 54 .size: 2 .value_kind: hidden_remainder_z - .offset: 72 .size: 8 .value_kind: hidden_global_offset_x - .offset: 80 .size: 8 .value_kind: hidden_global_offset_y - .offset: 88 .size: 8 .value_kind: hidden_global_offset_z - .offset: 96 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 288 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z12MatMulKernelPfS_S_i .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z12MatMulKernelPfS_S_i.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 11 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_001bb9c0_00000000-6_matrix.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2062: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2062: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "%f " .LC1: .string "\n" .text .globl _Z23printTwoDimDynamicArrayPfii .type _Z23printTwoDimDynamicArrayPfii, @function _Z23printTwoDimDynamicArrayPfii: .LFB2058: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $40, %rsp .cfi_def_cfa_offset 96 movq %rdi, 16(%rsp) testl %edx, %edx je .L3 movl %esi, %r15d movslq %edx, %rax movq %rax, 8(%rsp) movl $0, %r14d movl $0, %r13d movslq %esi, %rax movq %rax, 24(%rsp) leaq .LC0(%rip), %r12 jmp .L5 .L7: movslq %r14d, %rax movq 16(%rsp), %rcx leaq (%rcx,%rax,4), %rbx movq 24(%rsp), %rdx addq %rdx, %rax leaq (%rcx,%rax,4), %rbp .L6: pxor %xmm0, %xmm0 cvtss2sd (%rbx), %xmm0 movq %r12, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT addq $4, %rbx cmpq %rbp, %rbx jne .L6 .L8: leaq .LC1(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT addq $1, %r13 addl %r15d, %r14d movq 8(%rsp), %rax cmpq %rax, %r13 je .L3 .L5: testl %r15d, %r15d jne .L7 jmp .L8 .L3: addq $40, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2058: .size _Z23printTwoDimDynamicArrayPfii, .-_Z23printTwoDimDynamicArrayPfii .globl _Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i .type _Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i, @function _Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i: .LFB2084: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 24(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movl %ecx, 4(%rsp) movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax leaq 24(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) leaq 4(%rsp), %rax movq %rax, 120(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 56(%rsp) movl $1, 60(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) leaq 40(%rsp), %rcx leaq 32(%rsp), %rdx leaq 60(%rsp), %rsi leaq 48(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L15 .L11: movq 136(%rsp), %rax subq %fs:40, %rax jne .L16 addq $152, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L15: .cfi_restore_state pushq 40(%rsp) .cfi_def_cfa_offset 168 pushq 40(%rsp) .cfi_def_cfa_offset 176 leaq 112(%rsp), %r9 movq 76(%rsp), %rcx movl 84(%rsp), %r8d movq 64(%rsp), %rsi movl 72(%rsp), %edx leaq _Z12MatMulKernelPfS_S_i(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L11 .L16: call __stack_chk_fail@PLT .cfi_endproc .LFE2084: .size _Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i, .-_Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i .globl _Z12MatMulKernelPfS_S_i .type _Z12MatMulKernelPfS_S_i, @function _Z12MatMulKernelPfS_S_i: .LFB2085: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2085: .size _Z12MatMulKernelPfS_S_i, .-_Z12MatMulKernelPfS_S_i .section .rodata.str1.1 .LC2: .string "============================\n" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC3: .string "before kernel call: error = %s\n" .section .rodata.str1.1 .LC4: .string "GPU time: %13f msec\n" .text .globl _Z6MatMulPfS_S_i .type _Z6MatMulPfS_S_i, @function _Z6MatMulPfS_S_i: .LFB2059: .cfi_startproc endbr64 pushq %r13 .cfi_def_cfa_offset 16 .cfi_offset 13, -16 pushq %r12 .cfi_def_cfa_offset 24 .cfi_offset 12, -24 pushq %rbp .cfi_def_cfa_offset 32 .cfi_offset 6, -32 pushq %rbx .cfi_def_cfa_offset 40 .cfi_offset 3, -40 subq $88, %rsp .cfi_def_cfa_offset 128 movq %rdi, %r12 movq %rsi, %rbp movq %rdx, %r13 movl %ecx, %ebx movq %fs:40, %rax movq %rax, 72(%rsp) xorl %eax, %eax leaq 8(%rsp), %rdi movl $262144, %esi call cudaMalloc@PLT movl $1, %ecx movl $262144, %edx movq %r12, %rsi movq 8(%rsp), %rdi call cudaMemcpy@PLT leaq 16(%rsp), %rdi movl $262144, %esi call cudaMalloc@PLT movl $1, %ecx movl $262144, %edx movq %rbp, %rsi movq 16(%rsp), %rdi call cudaMemcpy@PLT leaq 24(%rsp), %rdi movl $262144, %esi call cudaMalloc@PLT movl $8, 48(%rsp) movl $8, 52(%rsp) movl $1, 56(%rsp) movl %ebx, 60(%rsp) movl %ebx, 64(%rsp) movl $1, 68(%rsp) leaq .LC2(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT leaq 32(%rsp), %rdi call cudaEventCreate@PLT leaq 40(%rsp), %rdi call cudaEventCreate@PLT movl $0, %esi movq 32(%rsp), %rdi call cudaEventRecord@PLT movl 68(%rsp), %ecx movl $0, %r9d movl $0, %r8d movq 60(%rsp), %rdx movq 48(%rsp), %rdi movl 56(%rsp), %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L24 .L20: call cudaGetLastError@PLT testl %eax, %eax jne .L25 movl $0, %esi movq 40(%rsp), %rdi call cudaEventRecord@PLT movq 40(%rsp), %rdi call cudaEventSynchronize@PLT leaq 4(%rsp), %rdi movq 40(%rsp), %rdx movq 32(%rsp), %rsi call cudaEventElapsedTime@PLT pxor %xmm0, %xmm0 cvtss2sd 4(%rsp), %xmm0 leaq .LC4(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movq 32(%rsp), %rdi call cudaEventDestroy@PLT movq 40(%rsp), %rdi call cudaEventDestroy@PLT movl $2, %ecx movl $262144, %edx movq 24(%rsp), %rsi movq %r13, %rdi call cudaMemcpy@PLT movq 8(%rsp), %rdi call cudaFree@PLT movq 16(%rsp), %rdi call cudaFree@PLT movq 24(%rsp), %rdi call cudaFree@PLT movq 72(%rsp), %rax subq %fs:40, %rax jne .L26 addq $88, %rsp .cfi_remember_state .cfi_def_cfa_offset 40 popq %rbx .cfi_def_cfa_offset 32 popq %rbp .cfi_def_cfa_offset 24 popq %r12 .cfi_def_cfa_offset 16 popq %r13 .cfi_def_cfa_offset 8 ret .L24: .cfi_restore_state movl %ebx, %ecx movq 24(%rsp), %rdx movq 16(%rsp), %rsi movq 8(%rsp), %rdi call _Z37__device_stub__Z12MatMulKernelPfS_S_iPfS_S_i jmp .L20 .L25: movl %eax, %edi call cudaGetErrorString@PLT movq %rax, %rdx leaq .LC3(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $1, %edi call exit@PLT .L26: call __stack_chk_fail@PLT .cfi_endproc .LFE2059: .size _Z6MatMulPfS_S_i, .-_Z6MatMulPfS_S_i .section .rodata.str1.1 .LC5: .string "PASSED" .LC6: .string "FAILED" .LC8: .string "CPU time: %13lf msec\n" .section .rodata.str1.8 .align 8 .LC9: .string "MxN[%d][%d] = %2.0f P[%d][%d] = %2.0f\n" .section .rodata.str1.1 .LC10: .string "Test %s\n" .text .globl main .type main, @function main: .LFB2057: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $88, %rsp .cfi_def_cfa_offset 144 movq %fs:40, %rax movq %rax, 72(%rsp) xorl %eax, %eax movl $262144, %edi call malloc@PLT movq %rax, %r14 movl $262144, %edi call malloc@PLT movq %rax, %r15 movl $262144, %edi call malloc@PLT movq %rax, 16(%rsp) movl $262144, %edi call malloc@PLT movq %rax, %rbx leaq 1024(%r14), %r12 leaq 263168(%r14), %r13 .L28: leaq -1024(%r12), %rbp .L29: call rand@PLT movslq %eax, %rdx imulq $1717986919, %rdx, %rdx sarq $33, %rdx movl %eax, %ecx sarl $31, %ecx subl %ecx, %edx leal (%rdx,%rdx,4), %edx subl %edx, %eax pxor %xmm0, %xmm0 cvtsi2ssl %eax, %xmm0 movss %xmm0, 0(%rbp) addq $4, %rbp cmpq %r12, %rbp jne .L29 addq $1024, %r12 cmpq %r13, %r12 jne .L28 leaq 1024(%r15), %r12 leaq 263168(%r15), %r13 .L30: leaq -1024(%r12), %rbp .L31: call rand@PLT movslq %eax, %rdx imulq $1717986919, %rdx, %rdx sarq $33, %rdx movl %eax, %ecx sarl $31, %ecx subl %ecx, %edx leal (%rdx,%rdx,4), %edx subl %edx, %eax pxor %xmm0, %xmm0 cvtsi2ssl %eax, %xmm0 movss %xmm0, 0(%rbp) addq $4, %rbp cmpq %r12, %rbp jne .L31 addq $1024, %r12 cmpq %r13, %r12 jne .L30 leaq 32(%rsp), %rdi movl $0, %esi call gettimeofday@PLT movl $0, %r9d .L33: leaq 262144(%r15), %rcx movq %rbx, %r12 movq %r9, %r8 salq $10, %r8 leaq (%rbx,%r8), %rdi addq %r14, %r8 movl $0, %esi .L37: movss (%rdi,%rsi,4), %xmm1 leaq -262144(%rcx), %rax movq %r8, %rdx .L34: movss (%rdx), %xmm0 mulss (%rax), %xmm0 addss %xmm0, %xmm1 addq $4, %rdx addq $1024, %rax cmpq %rcx, %rax jne .L34 movss %xmm1, (%rdi,%rsi,4) addq $1, %rsi addq $4, %rcx cmpq $256, %rsi jne .L37 addq $1, %r9 cmpq $256, %r9 jne .L33 leaq 48(%rsp), %rdi movl $0, %esi call gettimeofday@PLT movq 48(%rsp), %rax subq 32(%rsp), %rax pxor %xmm1, %xmm1 cvtsi2sdq %rax, %xmm1 movsd .LC7(%rip), %xmm2 mulsd %xmm2, %xmm1 movq 56(%rsp), %rax subq 40(%rsp), %rax pxor %xmm0, %xmm0 cvtsi2sdq %rax, %xmm0 divsd %xmm2, %xmm0 addsd %xmm1, %xmm0 leaq .LC8(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movl $32, %ecx movq 16(%rsp), %rbp movq %rbp, %rdx movq %r15, %rsi movq %r14, %rdi call _Z6MatMulPfS_S_i movl $1, 12(%rsp) movl $0, %r13d movq %rbx, %rax jmp .L38 .L46: movl %ebx, %ecx cvtss2sd %xmm0, %xmm0 cvtss2sd %xmm1, %xmm1 movl %ebx, %r9d movl %r13d, %r8d movl %r13d, %edx leaq .LC9(%rip), %rsi movl $2, %edi movl $2, %eax call __printf_chk@PLT movl $0, 12(%rsp) .L39: addq $1, %rbx cmpq $256, %rbx je .L53 .L41: movss (%r12,%rbx,4), %xmm0 movss 0(%rbp,%rbx,4), %xmm1 ucomiss %xmm1, %xmm0 jp .L46 je .L39 jmp .L46 .L53: movq %r14, %rax movq 24(%rsp), %r14 addl $1, %r13d addq $1024, %r12 addq $1024, %rbp cmpl $256, %r13d je .L42 .L38: movl $0, %ebx movq %r14, 24(%rsp) movq %rax, %r14 jmp .L41 .L42: movq %rax, %rbx movq %r14, %rdi call free@PLT movq %r15, %rdi call free@PLT movq 16(%rsp), %rdi call free@PLT movq %rbx, %rdi call free@PLT cmpl $0, 12(%rsp) leaq .LC6(%rip), %rdx leaq .LC5(%rip), %rax cmovne %rax, %rdx leaq .LC10(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 72(%rsp), %rax subq %fs:40, %rax jne .L54 movl $0, %eax addq $88, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L54: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size main, .-main .section .rodata.str1.1 .LC11: .string "_Z12MatMulKernelPfS_S_i" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2087: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC11(%rip), %rdx movq %rdx, %rcx leaq _Z12MatMulKernelPfS_S_i(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2087: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .section .rodata.cst8,"aM",@progbits,8 .align 8 .LC7: .long 0 .long 1083129856 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "matrix.hip" .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 # -- Begin function main .LCPI0_0: .quad 0x408f400000000000 # double 1000 .text .globl main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 subq $56, %rsp .cfi_def_cfa_offset 112 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl $262144, %edi # imm = 0x40000 callq malloc movq %rax, %rbp movl $262144, %edi # imm = 0x40000 callq malloc movq %rax, %r13 movl $262144, %edi # imm = 0x40000 callq malloc movq %rax, (%rsp) # 8-byte Spill movl $262144, %edi # imm = 0x40000 callq malloc movq %rax, %rbx xorl %r14d, %r14d movq %rbp, %r15 .p2align 4, 0x90 .LBB0_1: # %.preheader73 # =>This Loop Header: Depth=1 # Child Loop BB0_2 Depth 2 xorl %r12d, %r12d .p2align 4, 0x90 .LBB0_2: # Parent Loop BB0_1 Depth=1 # => This Inner Loop Header: Depth=2 callq rand cltq imulq $1717986919, %rax, %rcx # imm = 0x66666667 movq %rcx, %rdx shrq $63, %rdx sarq $33, %rcx addl %edx, %ecx leal (%rcx,%rcx,4), %ecx subl %ecx, %eax xorps %xmm0, %xmm0 cvtsi2ss %eax, %xmm0 movss %xmm0, (%r15,%r12,4) incq %r12 cmpq $256, %r12 # imm = 0x100 jne .LBB0_2 # %bb.3: # in Loop: Header=BB0_1 Depth=1 incq %r14 addq $1024, %r15 # imm = 0x400 cmpq $256, %r14 # imm = 0x100 jne .LBB0_1 # %bb.4: # %.preheader71.preheader xorl %r14d, %r14d movq %r13, %r15 .p2align 4, 0x90 .LBB0_5: # %.preheader71 # =>This Loop Header: Depth=1 # Child Loop BB0_6 Depth 2 xorl %r12d, %r12d .p2align 4, 0x90 .LBB0_6: # Parent Loop BB0_5 Depth=1 # => This Inner Loop Header: Depth=2 callq rand cltq imulq $1717986919, %rax, %rcx # imm = 0x66666667 movq %rcx, %rdx shrq $63, %rdx sarq $33, %rcx addl %edx, %ecx leal (%rcx,%rcx,4), %ecx subl %ecx, %eax xorps %xmm0, %xmm0 cvtsi2ss %eax, %xmm0 movss %xmm0, (%r15,%r12,4) incq %r12 cmpq $256, %r12 # imm = 0x100 jne .LBB0_6 # %bb.7: # in Loop: Header=BB0_5 Depth=1 incq %r14 addq $1024, %r15 # imm = 0x400 cmpq $256, %r14 # imm = 0x100 jne .LBB0_5 # %bb.8: xorl %r15d, %r15d leaq 40(%rsp), %rdi xorl %esi, %esi callq gettimeofday movq %rbp, %rax .p2align 4, 0x90 .LBB0_9: # %.preheader70 # =>This Loop Header: Depth=1 # Child Loop BB0_10 Depth 2 # Child Loop BB0_11 Depth 3 movq %r15, %rcx shlq $8, %rcx movq %r13, %rdx xorl %esi, %esi .p2align 4, 0x90 .LBB0_10: # %.preheader69 # Parent Loop BB0_9 Depth=1 # => This Loop Header: Depth=2 # Child Loop BB0_11 Depth 3 movq %rsi, %rdi orq %rcx, %rdi movss (%rbx,%rdi,4), %xmm0 # xmm0 = mem[0],zero,zero,zero movq %rdx, %r8 xorl %r9d, %r9d .p2align 4, 0x90 .LBB0_11: # Parent Loop BB0_9 Depth=1 # Parent Loop BB0_10 Depth=2 # => This Inner Loop Header: Depth=3 movss (%rax,%r9,4), %xmm1 # xmm1 = mem[0],zero,zero,zero mulss (%r8), %xmm1 addss %xmm1, %xmm0 incq %r9 addq $1024, %r8 # imm = 0x400 cmpq $256, %r9 # imm = 0x100 jne .LBB0_11 # %bb.12: # in Loop: Header=BB0_10 Depth=2 movss %xmm0, (%rbx,%rdi,4) incq %rsi addq $4, %rdx cmpq $256, %rsi # imm = 0x100 jne .LBB0_10 # %bb.13: # in Loop: Header=BB0_9 Depth=1 incq %r15 addq $1024, %rax # imm = 0x400 cmpq $256, %r15 # imm = 0x100 jne .LBB0_9 # %bb.14: movq %r13, %r14 xorl %r13d, %r13d leaq 24(%rsp), %rdi xorl %esi, %esi callq gettimeofday movq 24(%rsp), %rax movq 32(%rsp), %rcx subq 40(%rsp), %rax xorps %xmm1, %xmm1 cvtsi2sd %rax, %xmm1 movsd .LCPI0_0(%rip), %xmm2 # xmm2 = mem[0],zero mulsd %xmm2, %xmm1 subq 48(%rsp), %rcx xorps %xmm0, %xmm0 cvtsi2sd %rcx, %xmm0 divsd %xmm2, %xmm0 addsd %xmm1, %xmm0 movl $.L.str, %edi movb $1, %al callq printf movq %rbp, 16(%rsp) # 8-byte Spill movq %rbp, %rdi movq %r14, 8(%rsp) # 8-byte Spill movq %r14, %rsi movq (%rsp), %r14 # 8-byte Reload movq %r14, %rdx movl $32, %ecx callq _Z6MatMulPfS_S_i movl $1, %r15d movq %rbx, %r12 jmp .LBB0_15 .p2align 4, 0x90 .LBB0_19: # in Loop: Header=BB0_15 Depth=1 incq %r13 addq $1024, %r14 # imm = 0x400 addq $1024, %r12 # imm = 0x400 cmpq $256, %r13 # imm = 0x100 je .LBB0_20 .LBB0_15: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB0_16 Depth 2 xorl %ebp, %ebp jmp .LBB0_16 .p2align 4, 0x90 .LBB0_18: # in Loop: Header=BB0_16 Depth=2 incq %rbp cmpq $256, %rbp # imm = 0x100 je .LBB0_19 .LBB0_16: # Parent Loop BB0_15 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r12,%rbp,4), %xmm0 # xmm0 = mem[0],zero,zero,zero movss (%r14,%rbp,4), %xmm1 # xmm1 = mem[0],zero,zero,zero ucomiss %xmm1, %xmm0 jne .LBB0_17 jnp .LBB0_18 .LBB0_17: # in Loop: Header=BB0_16 Depth=2 cvtss2sd %xmm0, %xmm0 cvtss2sd %xmm1, %xmm1 movl $.L.str.1, %edi movl %r13d, %esi movl %ebp, %edx movl %r13d, %ecx movl %ebp, %r8d movb $2, %al callq printf xorl %r15d, %r15d jmp .LBB0_18 .LBB0_20: movq 16(%rsp), %rdi # 8-byte Reload callq free movq 8(%rsp), %rdi # 8-byte Reload callq free movq (%rsp), %rdi # 8-byte Reload callq free movq %rbx, %rdi callq free testl %r15d, %r15d movl $.L.str.4, %eax movl $.L.str.3, %esi cmoveq %rax, %rsi movl $.L.str.2, %edi xorl %eax, %eax callq printf xorl %eax, %eax addq $56, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end0: .size main, .Lfunc_end0-main .cfi_endproc # -- End function .globl _Z6MatMulPfS_S_i # -- Begin function _Z6MatMulPfS_S_i .p2align 4, 0x90 .type _Z6MatMulPfS_S_i,@function _Z6MatMulPfS_S_i: # @_Z6MatMulPfS_S_i .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %rbx .cfi_def_cfa_offset 40 subq $168, %rsp .cfi_def_cfa_offset 208 .cfi_offset %rbx, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl %ecx, %ebp movq %rdx, %rbx movq %rsi, %r14 movq %rdi, %r15 leaq 40(%rsp), %rdi movl $262144, %esi # imm = 0x40000 callq hipMalloc movq 40(%rsp), %rdi movl $262144, %edx # imm = 0x40000 movq %r15, %rsi movl $1, %ecx callq hipMemcpy leaq 32(%rsp), %rdi movl $262144, %esi # imm = 0x40000 callq hipMalloc movq 32(%rsp), %rdi movl $262144, %edx # imm = 0x40000 movq %r14, %rsi movl $1, %ecx callq hipMemcpy leaq 24(%rsp), %rdi movl $262144, %esi # imm = 0x40000 callq hipMalloc movl %ebp, %eax movq %rax, %r14 shlq $32, %r14 orq %rax, %r14 movl $.Lstr, %edi callq puts@PLT leaq 16(%rsp), %rdi callq hipEventCreate leaq 8(%rsp), %rdi callq hipEventCreate movq 16(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movabsq $34359738376, %rdi # imm = 0x800000008 movl $1, %esi movq %r14, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_2 # %bb.1: movq 40(%rsp), %rax movq 32(%rsp), %rcx movq 24(%rsp), %rdx movq %rax, 120(%rsp) movq %rcx, 112(%rsp) movq %rdx, 104(%rsp) movl %ebp, 52(%rsp) leaq 120(%rsp), %rax movq %rax, 128(%rsp) leaq 112(%rsp), %rax movq %rax, 136(%rsp) leaq 104(%rsp), %rax movq %rax, 144(%rsp) leaq 52(%rsp), %rax movq %rax, 152(%rsp) leaq 88(%rsp), %rdi leaq 72(%rsp), %rsi leaq 64(%rsp), %rdx leaq 56(%rsp), %rcx callq __hipPopCallConfiguration movq 88(%rsp), %rsi movl 96(%rsp), %edx movq 72(%rsp), %rcx movl 80(%rsp), %r8d leaq 128(%rsp), %r9 movl $_Z12MatMulKernelPfS_S_i, %edi pushq 56(%rsp) .cfi_adjust_cfa_offset 8 pushq 72(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_2: callq hipGetLastError testl %eax, %eax jne .LBB1_4 # %bb.3: movq 8(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movq 8(%rsp), %rdi callq hipEventSynchronize movq 16(%rsp), %rsi movq 8(%rsp), %rdx leaq 128(%rsp), %rdi callq hipEventElapsedTime movss 128(%rsp), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str.9, %edi movb $1, %al callq printf movq 16(%rsp), %rdi callq hipEventDestroy movq 8(%rsp), %rdi callq hipEventDestroy movq 24(%rsp), %rsi movl $262144, %edx # imm = 0x40000 movq %rbx, %rdi movl $2, %ecx callq hipMemcpy movq 40(%rsp), %rdi callq hipFree movq 32(%rsp), %rdi callq hipFree movq 24(%rsp), %rdi callq hipFree addq $168, %rsp .cfi_def_cfa_offset 40 popq %rbx .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .LBB1_4: .cfi_def_cfa_offset 208 movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %edi movq %rax, %rsi xorl %eax, %eax callq printf movl $1, %edi callq exit .Lfunc_end1: .size _Z6MatMulPfS_S_i, .Lfunc_end1-_Z6MatMulPfS_S_i .cfi_endproc # -- End function .globl _Z23printTwoDimDynamicArrayPfii # -- Begin function _Z23printTwoDimDynamicArrayPfii .p2align 4, 0x90 .type _Z23printTwoDimDynamicArrayPfii,@function _Z23printTwoDimDynamicArrayPfii: # @_Z23printTwoDimDynamicArrayPfii .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 pushq %rax .cfi_def_cfa_offset 64 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl %esi, 4(%rsp) # 4-byte Spill testl %edx, %edx je .LBB2_6 # %bb.1: # %.preheader.lr.ph movq %rdi, %r14 movslq 4(%rsp), %r15 # 4-byte Folded Reload movl %edx, %r12d movl %r15d, %r13d shlq $2, %r15 xorl %ebp, %ebp jmp .LBB2_2 .p2align 4, 0x90 .LBB2_5: # %._crit_edge # in Loop: Header=BB2_2 Depth=1 movl $10, %edi callq putchar@PLT incq %rbp addq %r15, %r14 cmpq %r12, %rbp je .LBB2_6 .LBB2_2: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB2_4 Depth 2 cmpl $0, 4(%rsp) # 4-byte Folded Reload je .LBB2_5 # %bb.3: # %.lr.ph # in Loop: Header=BB2_2 Depth=1 xorl %ebx, %ebx .p2align 4, 0x90 .LBB2_4: # Parent Loop BB2_2 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r14,%rbx,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str.5, %edi movb $1, %al callq printf incq %rbx cmpq %rbx, %r13 jne .LBB2_4 jmp .LBB2_5 .LBB2_6: # %._crit_edge16 addq $8, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end2: .size _Z23printTwoDimDynamicArrayPfii, .Lfunc_end2-_Z23printTwoDimDynamicArrayPfii .cfi_endproc # -- End function .globl _Z27__device_stub__MatMulKernelPfS_S_i # -- Begin function _Z27__device_stub__MatMulKernelPfS_S_i .p2align 4, 0x90 .type _Z27__device_stub__MatMulKernelPfS_S_i,@function _Z27__device_stub__MatMulKernelPfS_S_i: # @_Z27__device_stub__MatMulKernelPfS_S_i .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movq %rdx, 56(%rsp) movl %ecx, 4(%rsp) leaq 72(%rsp), %rax movq %rax, 80(%rsp) leaq 64(%rsp), %rax movq %rax, 88(%rsp) leaq 56(%rsp), %rax movq %rax, 96(%rsp) leaq 4(%rsp), %rax movq %rax, 104(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 80(%rsp), %r9 movl $_Z12MatMulKernelPfS_S_i, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $136, %rsp .cfi_adjust_cfa_offset -136 retq .Lfunc_end3: .size _Z27__device_stub__MatMulKernelPfS_S_i, .Lfunc_end3-_Z27__device_stub__MatMulKernelPfS_S_i .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB4_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB4_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z12MatMulKernelPfS_S_i, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end4: .size __hip_module_ctor, .Lfunc_end4-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB5_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB5_2: retq .Lfunc_end5: .size __hip_module_dtor, .Lfunc_end5-__hip_module_dtor .cfi_endproc # -- End function .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "CPU time: %13lf msec\n" .size .L.str, 22 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "MxN[%d][%d] = %2.0f P[%d][%d] = %2.0f\n" .size .L.str.1, 41 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "Test %s\n" .size .L.str.2, 9 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "PASSED" .size .L.str.3, 7 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "FAILED" .size .L.str.4, 7 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "%f " .size .L.str.5, 4 .type _Z12MatMulKernelPfS_S_i,@object # @_Z12MatMulKernelPfS_S_i .section .rodata,"a",@progbits .globl _Z12MatMulKernelPfS_S_i .p2align 3, 0x0 _Z12MatMulKernelPfS_S_i: .quad _Z27__device_stub__MatMulKernelPfS_S_i .size _Z12MatMulKernelPfS_S_i, 8 .type .L.str.8,@object # @.str.8 .section .rodata.str1.1,"aMS",@progbits,1 .L.str.8: .asciz "before kernel call: error = %s\n" .size .L.str.8, 32 .type .L.str.9,@object # @.str.9 .L.str.9: .asciz "GPU time: %13f msec\n" .size .L.str.9, 21 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z12MatMulKernelPfS_S_i" .size .L__unnamed_1, 24 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .type .Lstr,@object # @str .section .rodata.str1.1,"aMS",@progbits,1 .Lstr: .asciz "============================" .size .Lstr, 29 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z27__device_stub__MatMulKernelPfS_S_i .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z12MatMulKernelPfS_S_i .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
#include <stdio.h> #include <stdlib.h> #include <cuda_runtime.h> #define BLOCK_DIM 128 const int size_x = 128; const int size_y = 1; __global__ static void threadDemo(unsigned int * ret) { unsigned int xIndex = blockDim.x *blockIdx.x + threadIdx.x; unsigned int yIndex = blockDim.y * blockIdx.y + threadIdx.y; printf("block Dim (%d, %d) \n", blockDim.x, blockDim.y); printf("block id (%d, %d)\n", blockIdx.x, blockIdx.y); printf("thread id (%d, %d) \n", threadIdx.x, threadIdx.y); if (xIndex < size_x && yIndex < size_y) { unsigned int index = xIndex + size_x * yIndex; ret[index] = xIndex; ret[index + size_x * size_y] = yIndex; } } void ThreadDemo(void) { unsigned int * ret = 0; unsigned int host_ret[size_x*size_y*2] = {0}; int i = 0; cudaMalloc((void**) &ret, sizeof(unsigned int)*(size_x*size_y*2)); dim3 grid(size_x / BLOCK_DIM, 1); dim3 block(BLOCK_DIM, 1, 1); threadDemo<<<grid,block>>>(ret); cudaMemcpy(&host_ret, ret, sizeof(unsigned int)*(size_x*size_y*2), cudaMemcpyDeviceToHost); for (i = 0; i < size_x*size_y; i++) { printf("(%u,%u)", host_ret[i], host_ret[size_x*size_y+i]); } cudaFree(ret); }
code for sm_80 Function : _Z10threadDemoPj .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ IMAD.MOV.U32 R1, RZ, RZ, c[0x0][0x28] ; /* 0x00000a00ff017624 */ /* 0x000fca00078e00ff */ /*0010*/ IADD3 R1, R1, -0x8, RZ ; /* 0xfffffff801017810 */ /* 0x000fe20007ffe0ff */ /*0020*/ IMAD.MOV.U32 R8, RZ, RZ, c[0x0][0x0] ; /* 0x00000000ff087624 */ /* 0x000fe200078e00ff */ /*0030*/ S2R R16, SR_CTAID.X ; /* 0x0000000000107919 */ /* 0x000e220000002500 */ /*0040*/ IMAD.MOV.U32 R9, RZ, RZ, c[0x0][0x4] ; /* 0x00000100ff097624 */ /* 0x000fe200078e00ff */ /*0050*/ MOV R26, 0x0 ; /* 0x00000000001a7802 */ /* 0x000fe20000000f00 */ /*0060*/ IMAD.MOV.U32 R4, RZ, RZ, c[0x4][0x8] ; /* 0x01000200ff047624 */ /* 0x000fe200078e00ff */ /*0070*/ S2R R18, SR_TID.X ; /* 0x0000000000127919 */ /* 0x000e220000002100 */ /*0080*/ IADD3 R25, P0, R1, c[0x0][0x20], RZ ; /* 0x0000080001197a10 */ /* 0x000fe20007f1e0ff */ /*0090*/ LDC.64 R2, c[0x4][R26] ; /* 0x010000001a027b82 */ /* 0x0002a20000000a00 */ /*00a0*/ IMAD.MOV.U32 R5, RZ, RZ, c[0x4][0xc] ; /* 0x01000300ff057624 */ /* 0x000fe200078e00ff */ /*00b0*/ S2R R17, SR_CTAID.Y ; /* 0x0000000000117919 */ /* 0x000ee40000002600 */ /*00c0*/ IMAD.X R24, RZ, RZ, c[0x0][0x24], P0 ; /* 0x00000900ff187624 */ /* 0x000fc400000e06ff */ /*00d0*/ STL.64 [R1], R8 ; /* 0x0000000801007387 */ /* 0x0003e20000100a00 */ /*00e0*/ IMAD.MOV.U32 R6, RZ, RZ, R25 ; /* 0x000000ffff067224 */ /* 0x000fe400078e0019 */ /*00f0*/ IMAD.MOV.U32 R7, RZ, RZ, R24 ; /* 0x000000ffff077224 */ /* 0x000fe400078e0018 */ /*0100*/ IMAD R23, R16, c[0x0][0x0], R18 ; /* 0x0000000010177a24 */ /* 0x001fe400078e0212 */ /*0110*/ IMAD R22, R17, c[0x0][0x4], RZ ; /* 0x0000010011167a24 */ /* 0x008fe400078e02ff */ /*0120*/ LEPC R8 ; /* 0x000000000008734e */ /* 0x006fca0000000000 */ /*0130*/ MOV R11, 0x1a0 ; /* 0x000001a0000b7802 */ /* 0x000fe40000000f00 */ /*0140*/ MOV R20, 0x120 ; /* 0x0000012000147802 */ /* 0x000fe40000000f00 */ /*0150*/ MOV R21, 0x0 ; /* 0x0000000000157802 */ /* 0x000fc40000000f00 */ /*0160*/ MOV R0, 0x0 ; /* 0x0000000000007802 */ /* 0x000fe40000000f00 */ /*0170*/ IADD3 R20, P0, P1, -R20, R11, R8 ; /* 0x0000000b14147210 */ /* 0x000fc8000791e108 */ /*0180*/ IADD3.X R21, ~R0, R21, R9, P0, P1 ; /* 0x0000001500157210 */ /* 0x000fc800007e2509 */ /*0190*/ CALL.ABS.NOINC R2 ; /* 0x0000000002007343 */ /* 0x000fea0003c00000 */ /*01a0*/ STL.64 [R1], R16 ; /* 0x0000001001007387 */ /* 0x0001e20000100a00 */ /*01b0*/ LDC.64 R2, c[0x4][R26] ; /* 0x010000001a027b82 */ /* 0x0000620000000a00 */ /*01c0*/ IMAD.MOV.U32 R4, RZ, RZ, c[0x4][0x10] ; /* 0x01000400ff047624 */ /* 0x000fe400078e00ff */ /*01d0*/ IMAD.MOV.U32 R5, RZ, RZ, c[0x4][0x14] ; /* 0x01000500ff057624 */ /* 0x000fe400078e00ff */ /*01e0*/ IMAD.MOV.U32 R6, RZ, RZ, R25 ; /* 0x000000ffff067224 */ /* 0x000fe400078e0019 */ /*01f0*/ IMAD.MOV.U32 R7, RZ, RZ, R24 ; /* 0x000000ffff077224 */ /* 0x000fe400078e0018 */ /*0200*/ LEPC R8 ; /* 0x000000000008734e */ /* 0x000fca0000000000 */ /*0210*/ MOV R11, 0x280 ; /* 0x00000280000b7802 */ /* 0x000fe40000000f00 */ /*0220*/ MOV R20, 0x200 ; /* 0x0000020000147802 */ /* 0x000fc40000000f00 */ /*0230*/ MOV R21, 0x0 ; /* 0x0000000000157802 */ /* 0x000fe40000000f00 */ /*0240*/ MOV R0, 0x0 ; /* 0x0000000000007802 */ /* 0x000fe40000000f00 */ /*0250*/ IADD3 R20, P0, P1, -R20, R11, R8 ; /* 0x0000000b14147210 */ /* 0x000fc8000791e108 */ /*0260*/ IADD3.X R21, ~R0, R21, R9, P0, P1 ; /* 0x0000001500157210 */ /* 0x000fc800007e2509 */ /*0270*/ CALL.ABS.NOINC R2 ; /* 0x0000000002007343 */ /* 0x003fea0003c00000 */ /*0280*/ S2R R19, SR_TID.Y ; /* 0x0000000000137919 */ /* 0x000e220000002200 */ /*0290*/ LDC.64 R26, c[0x4][R26] ; /* 0x010000001a1a7b82 */ /* 0x000e620000000a00 */ /*02a0*/ IMAD.MOV.U32 R6, RZ, RZ, R25 ; /* 0x000000ffff067224 */ /* 0x000fe400078e0019 */ /*02b0*/ IMAD.MOV.U32 R7, RZ, RZ, R24 ; /* 0x000000ffff077224 */ /* 0x000fe400078e0018 */ /*02c0*/ IMAD.MOV.U32 R4, RZ, RZ, c[0x4][0x18] ; /* 0x01000600ff047624 */ /* 0x000fe400078e00ff */ /*02d0*/ IMAD.MOV.U32 R5, RZ, RZ, c[0x4][0x1c] ; /* 0x01000700ff057624 */ /* 0x000fe200078e00ff */ /*02e0*/ STL.64 [R1], R18 ; /* 0x0000001201007387 */ /* 0x0011e80000100a00 */ /*02f0*/ LEPC R2 ; /* 0x000000000002734e */ /* 0x000fe40000000000 */ /*0300*/ MOV R9, 0x370 ; /* 0x0000037000097802 */ /* 0x000fe40000000f00 */ /*0310*/ MOV R20, 0x2f0 ; /* 0x000002f000147802 */ /* 0x000fc40000000f00 */ /*0320*/ MOV R21, 0x0 ; /* 0x0000000000157802 */ /* 0x000fe40000000f00 */ /*0330*/ MOV R0, 0x0 ; /* 0x0000000000007802 */ /* 0x000fe40000000f00 */ /*0340*/ IADD3 R20, P0, P1, -R20, R9, R2 ; /* 0x0000000914147210 */ /* 0x000fc8000791e102 */ /*0350*/ IADD3.X R21, ~R0, R21, R3, P0, P1 ; /* 0x0000001500157210 */ /* 0x000fc800007e2503 */ /*0360*/ CALL.ABS.NOINC R26 ; /* 0x000000001a007343 */ /* 0x003fea0003c00000 */ /*0370*/ IMAD.MOV R19, RZ, RZ, -R19 ; /* 0x000000ffff137224 */ /* 0x000fe200078e0a13 */ /*0380*/ ISETP.GT.U32.AND P0, PT, R23, 0x7f, PT ; /* 0x0000007f1700780c */ /* 0x000fc80003f04070 */ /*0390*/ ISETP.NE.OR P0, PT, R22, R19, P0 ; /* 0x000000131600720c */ /* 0x000fda0000705670 */ /*03a0*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*03b0*/ IMAD.MOV.U32 R5, RZ, RZ, 0x4 ; /* 0x00000004ff057424 */ /* 0x000fe200078e00ff */ /*03c0*/ IADD3 R4, R23.reuse, 0x80, RZ ; /* 0x0000008017047810 */ /* 0x040fe20007ffe0ff */ /*03d0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe40000000a00 */ /*03e0*/ IMAD.WIDE.U32 R2, R23, R5, c[0x0][0x160] ; /* 0x0000580017027625 */ /* 0x000fc800078e0005 */ /*03f0*/ IMAD.WIDE.U32 R4, R4, R5, c[0x0][0x160] ; /* 0x0000580004047625 */ /* 0x000fe200078e0005 */ /*0400*/ STG.E [R2.64], R23 ; /* 0x0000001702007986 */ /* 0x000fe8000c101904 */ /*0410*/ STG.E [R4.64], RZ ; /* 0x000000ff04007986 */ /* 0x000fe2000c101904 */ /*0420*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0430*/ BRA 0x430; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0440*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0450*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0460*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0470*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0480*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0490*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*04f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include <stdio.h> #include <stdlib.h> #include <cuda_runtime.h> #define BLOCK_DIM 128 const int size_x = 128; const int size_y = 1; __global__ static void threadDemo(unsigned int * ret) { unsigned int xIndex = blockDim.x *blockIdx.x + threadIdx.x; unsigned int yIndex = blockDim.y * blockIdx.y + threadIdx.y; printf("block Dim (%d, %d) \n", blockDim.x, blockDim.y); printf("block id (%d, %d)\n", blockIdx.x, blockIdx.y); printf("thread id (%d, %d) \n", threadIdx.x, threadIdx.y); if (xIndex < size_x && yIndex < size_y) { unsigned int index = xIndex + size_x * yIndex; ret[index] = xIndex; ret[index + size_x * size_y] = yIndex; } } void ThreadDemo(void) { unsigned int * ret = 0; unsigned int host_ret[size_x*size_y*2] = {0}; int i = 0; cudaMalloc((void**) &ret, sizeof(unsigned int)*(size_x*size_y*2)); dim3 grid(size_x / BLOCK_DIM, 1); dim3 block(BLOCK_DIM, 1, 1); threadDemo<<<grid,block>>>(ret); cudaMemcpy(&host_ret, ret, sizeof(unsigned int)*(size_x*size_y*2), cudaMemcpyDeviceToHost); for (i = 0; i < size_x*size_y; i++) { printf("(%u,%u)", host_ret[i], host_ret[size_x*size_y+i]); } cudaFree(ret); }
.file "tmpxft_001a1b3a_00000000-6_threadDemo.cudafe1.cpp" .text #APP #NO_APP .type _ZL30__device_stub__Z10threadDemoPjPj, @function _ZL30__device_stub__Z10threadDemoPjPj: .LFB2082: .cfi_startproc subq $104, %rsp .cfi_def_cfa_offset 112 movq %rdi, 8(%rsp) movq %fs:40, %rax movq %rax, 88(%rsp) xorl %eax, %eax leaq 8(%rsp), %rax movq %rax, 80(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) movl $1, 40(%rsp) movl $1, 44(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) leaq 24(%rsp), %rcx leaq 16(%rsp), %rdx leaq 44(%rsp), %rsi leaq 32(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L5 .L1: movq 88(%rsp), %rax subq %fs:40, %rax jne .L6 addq $104, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L5: .cfi_restore_state pushq 24(%rsp) .cfi_def_cfa_offset 120 pushq 24(%rsp) .cfi_def_cfa_offset 128 leaq 96(%rsp), %r9 movq 60(%rsp), %rcx movl 68(%rsp), %r8d movq 48(%rsp), %rsi movl 56(%rsp), %edx leaq _ZL10threadDemoPj(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 112 jmp .L1 .L6: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _ZL30__device_stub__Z10threadDemoPjPj, .-_ZL30__device_stub__Z10threadDemoPjPj .type _ZL10threadDemoPj, @function _ZL10threadDemoPj: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _ZL30__device_stub__Z10threadDemoPjPj addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _ZL10threadDemoPj, .-_ZL10threadDemoPj .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2060: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2060: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "(%u,%u)" .text .globl _Z10ThreadDemov .type _Z10ThreadDemov, @function _Z10ThreadDemov: .LFB2057: .cfi_startproc endbr64 pushq %r12 .cfi_def_cfa_offset 16 .cfi_offset 12, -16 pushq %rbp .cfi_def_cfa_offset 24 .cfi_offset 6, -24 pushq %rbx .cfi_def_cfa_offset 32 .cfi_offset 3, -32 subq $1072, %rsp .cfi_def_cfa_offset 1104 movq %fs:40, %rax movq %rax, 1064(%rsp) xorl %eax, %eax movq $0, (%rsp) leaq 32(%rsp), %rdi movl $128, %ecx rep stosq movq %rsp, %rdi movl $1024, %esi call cudaMalloc@PLT movl $1, 8(%rsp) movl $1, 12(%rsp) movl $1, 16(%rsp) movl $128, 20(%rsp) movl $1, 24(%rsp) movl $1, 28(%rsp) movl $0, %r9d movl $0, %r8d movq 20(%rsp), %rdx movl $1, %ecx movq 8(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L17 .L12: leaq 32(%rsp), %rbx movl $2, %ecx movl $1024, %edx movq (%rsp), %rsi movq %rbx, %rdi call cudaMemcpy@PLT leaq 544(%rsp), %r12 leaq .LC0(%rip), %rbp .L13: movl 512(%rbx), %ecx movl (%rbx), %edx movq %rbp, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT addq $4, %rbx cmpq %r12, %rbx jne .L13 movq (%rsp), %rdi call cudaFree@PLT movq 1064(%rsp), %rax subq %fs:40, %rax jne .L18 addq $1072, %rsp .cfi_remember_state .cfi_def_cfa_offset 32 popq %rbx .cfi_def_cfa_offset 24 popq %rbp .cfi_def_cfa_offset 16 popq %r12 .cfi_def_cfa_offset 8 ret .L17: .cfi_restore_state movq (%rsp), %rdi call _ZL30__device_stub__Z10threadDemoPjPj jmp .L12 .L18: call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size _Z10ThreadDemov, .-_Z10ThreadDemov .section .rodata.str1.1 .LC1: .string "_Z10threadDemoPj" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2085: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC1(%rip), %rdx movq %rdx, %rcx leaq _ZL10threadDemoPj(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2085: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include <stdio.h> #include <stdlib.h> #include <cuda_runtime.h> #define BLOCK_DIM 128 const int size_x = 128; const int size_y = 1; __global__ static void threadDemo(unsigned int * ret) { unsigned int xIndex = blockDim.x *blockIdx.x + threadIdx.x; unsigned int yIndex = blockDim.y * blockIdx.y + threadIdx.y; printf("block Dim (%d, %d) \n", blockDim.x, blockDim.y); printf("block id (%d, %d)\n", blockIdx.x, blockIdx.y); printf("thread id (%d, %d) \n", threadIdx.x, threadIdx.y); if (xIndex < size_x && yIndex < size_y) { unsigned int index = xIndex + size_x * yIndex; ret[index] = xIndex; ret[index + size_x * size_y] = yIndex; } } void ThreadDemo(void) { unsigned int * ret = 0; unsigned int host_ret[size_x*size_y*2] = {0}; int i = 0; cudaMalloc((void**) &ret, sizeof(unsigned int)*(size_x*size_y*2)); dim3 grid(size_x / BLOCK_DIM, 1); dim3 block(BLOCK_DIM, 1, 1); threadDemo<<<grid,block>>>(ret); cudaMemcpy(&host_ret, ret, sizeof(unsigned int)*(size_x*size_y*2), cudaMemcpyDeviceToHost); for (i = 0; i < size_x*size_y; i++) { printf("(%u,%u)", host_ret[i], host_ret[size_x*size_y+i]); } cudaFree(ret); }
#include <stdio.h> #include <stdlib.h> #include <hip/hip_runtime.h> #define BLOCK_DIM 128 const int size_x = 128; const int size_y = 1; __global__ static void threadDemo(unsigned int * ret) { unsigned int xIndex = blockDim.x *blockIdx.x + threadIdx.x; unsigned int yIndex = blockDim.y * blockIdx.y + threadIdx.y; printf("block Dim (%d, %d) \n", blockDim.x, blockDim.y); printf("block id (%d, %d)\n", blockIdx.x, blockIdx.y); printf("thread id (%d, %d) \n", threadIdx.x, threadIdx.y); if (xIndex < size_x && yIndex < size_y) { unsigned int index = xIndex + size_x * yIndex; ret[index] = xIndex; ret[index + size_x * size_y] = yIndex; } } void ThreadDemo(void) { unsigned int * ret = 0; unsigned int host_ret[size_x*size_y*2] = {0}; int i = 0; hipMalloc((void**) &ret, sizeof(unsigned int)*(size_x*size_y*2)); dim3 grid(size_x / BLOCK_DIM, 1); dim3 block(BLOCK_DIM, 1, 1); threadDemo<<<grid,block>>>(ret); hipMemcpy(&host_ret, ret, sizeof(unsigned int)*(size_x*size_y*2), hipMemcpyDeviceToHost); for (i = 0; i < size_x*size_y; i++) { printf("(%u,%u)", host_ret[i], host_ret[size_x*size_y+i]); } hipFree(ret); }
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <stdio.h> #include <stdlib.h> #include <hip/hip_runtime.h> #define BLOCK_DIM 128 const int size_x = 128; const int size_y = 1; __global__ static void threadDemo(unsigned int * ret) { unsigned int xIndex = blockDim.x *blockIdx.x + threadIdx.x; unsigned int yIndex = blockDim.y * blockIdx.y + threadIdx.y; printf("block Dim (%d, %d) \n", blockDim.x, blockDim.y); printf("block id (%d, %d)\n", blockIdx.x, blockIdx.y); printf("thread id (%d, %d) \n", threadIdx.x, threadIdx.y); if (xIndex < size_x && yIndex < size_y) { unsigned int index = xIndex + size_x * yIndex; ret[index] = xIndex; ret[index + size_x * size_y] = yIndex; } } void ThreadDemo(void) { unsigned int * ret = 0; unsigned int host_ret[size_x*size_y*2] = {0}; int i = 0; hipMalloc((void**) &ret, sizeof(unsigned int)*(size_x*size_y*2)); dim3 grid(size_x / BLOCK_DIM, 1); dim3 block(BLOCK_DIM, 1, 1); threadDemo<<<grid,block>>>(ret); hipMemcpy(&host_ret, ret, sizeof(unsigned int)*(size_x*size_y*2), hipMemcpyDeviceToHost); for (i = 0; i < size_x*size_y; i++) { printf("(%u,%u)", host_ret[i], host_ret[size_x*size_y+i]); } hipFree(ret); }
.text .file "threadDemo.hip" .globl _Z10ThreadDemov # -- Begin function _Z10ThreadDemov .p2align 4, 0x90 .type _Z10ThreadDemov,@function _Z10ThreadDemov: # @_Z10ThreadDemov .cfi_startproc # %bb.0: pushq %rbx .cfi_def_cfa_offset 16 subq $1104, %rsp # imm = 0x450 .cfi_def_cfa_offset 1120 .cfi_offset %rbx, -16 movq $0, 8(%rsp) leaq 80(%rsp), %rdi movl $1024, %edx # imm = 0x400 xorl %esi, %esi callq memset@PLT leaq 8(%rsp), %rdi movl $1024, %esi # imm = 0x400 callq hipMalloc movabsq $4294967297, %rdi # imm = 0x100000001 leaq 127(%rdi), %rdx movl $1, %esi movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB0_2 # %bb.1: movq 8(%rsp), %rax movq %rax, 72(%rsp) leaq 72(%rsp), %rax movq %rax, 16(%rsp) leaq 56(%rsp), %rdi leaq 40(%rsp), %rsi leaq 32(%rsp), %rdx leaq 24(%rsp), %rcx callq __hipPopCallConfiguration movq 56(%rsp), %rsi movl 64(%rsp), %edx movq 40(%rsp), %rcx movl 48(%rsp), %r8d leaq 16(%rsp), %r9 movl $_ZL10threadDemoPj, %edi pushq 24(%rsp) .cfi_adjust_cfa_offset 8 pushq 40(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB0_2: movq 8(%rsp), %rsi leaq 80(%rsp), %rdi movl $1024, %edx # imm = 0x400 movl $2, %ecx callq hipMemcpy xorl %ebx, %ebx .p2align 4, 0x90 .LBB0_3: # =>This Inner Loop Header: Depth=1 movl 80(%rsp,%rbx,4), %esi movl 592(%rsp,%rbx,4), %edx movl $.L.str, %edi xorl %eax, %eax callq printf incq %rbx cmpq $128, %rbx jne .LBB0_3 # %bb.4: movq 8(%rsp), %rdi callq hipFree addq $1104, %rsp # imm = 0x450 .cfi_def_cfa_offset 16 popq %rbx .cfi_def_cfa_offset 8 retq .Lfunc_end0: .size _Z10ThreadDemov, .Lfunc_end0-_Z10ThreadDemov .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function _ZL25__device_stub__threadDemoPj .type _ZL25__device_stub__threadDemoPj,@function _ZL25__device_stub__threadDemoPj: # @_ZL25__device_stub__threadDemoPj .cfi_startproc # %bb.0: subq $72, %rsp .cfi_def_cfa_offset 80 movq %rdi, 64(%rsp) leaq 64(%rsp), %rax movq %rax, (%rsp) leaq 48(%rsp), %rdi leaq 32(%rsp), %rsi leaq 24(%rsp), %rdx leaq 16(%rsp), %rcx callq __hipPopCallConfiguration movq 48(%rsp), %rsi movl 56(%rsp), %edx movq 32(%rsp), %rcx movl 40(%rsp), %r8d movq %rsp, %r9 movl $_ZL10threadDemoPj, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $88, %rsp .cfi_adjust_cfa_offset -88 retq .Lfunc_end1: .size _ZL25__device_stub__threadDemoPj, .Lfunc_end1-_ZL25__device_stub__threadDemoPj .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB2_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB2_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_ZL10threadDemoPj, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end2: .size __hip_module_ctor, .Lfunc_end2-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB3_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB3_2: retq .Lfunc_end3: .size __hip_module_dtor, .Lfunc_end3-__hip_module_dtor .cfi_endproc # -- End function .type _ZL10threadDemoPj,@object # @_ZL10threadDemoPj .section .rodata,"a",@progbits .p2align 3, 0x0 _ZL10threadDemoPj: .quad _ZL25__device_stub__threadDemoPj .size _ZL10threadDemoPj, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "(%u,%u)" .size .L.str, 8 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_ZL10threadDemoPj" .size .L__unnamed_1, 18 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _ZL25__device_stub__threadDemoPj .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _ZL10threadDemoPj .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_001a1b3a_00000000-6_threadDemo.cudafe1.cpp" .text #APP #NO_APP .type _ZL30__device_stub__Z10threadDemoPjPj, @function _ZL30__device_stub__Z10threadDemoPjPj: .LFB2082: .cfi_startproc subq $104, %rsp .cfi_def_cfa_offset 112 movq %rdi, 8(%rsp) movq %fs:40, %rax movq %rax, 88(%rsp) xorl %eax, %eax leaq 8(%rsp), %rax movq %rax, 80(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) movl $1, 40(%rsp) movl $1, 44(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) leaq 24(%rsp), %rcx leaq 16(%rsp), %rdx leaq 44(%rsp), %rsi leaq 32(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L5 .L1: movq 88(%rsp), %rax subq %fs:40, %rax jne .L6 addq $104, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L5: .cfi_restore_state pushq 24(%rsp) .cfi_def_cfa_offset 120 pushq 24(%rsp) .cfi_def_cfa_offset 128 leaq 96(%rsp), %r9 movq 60(%rsp), %rcx movl 68(%rsp), %r8d movq 48(%rsp), %rsi movl 56(%rsp), %edx leaq _ZL10threadDemoPj(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 112 jmp .L1 .L6: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _ZL30__device_stub__Z10threadDemoPjPj, .-_ZL30__device_stub__Z10threadDemoPjPj .type _ZL10threadDemoPj, @function _ZL10threadDemoPj: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _ZL30__device_stub__Z10threadDemoPjPj addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _ZL10threadDemoPj, .-_ZL10threadDemoPj .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2060: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2060: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "(%u,%u)" .text .globl _Z10ThreadDemov .type _Z10ThreadDemov, @function _Z10ThreadDemov: .LFB2057: .cfi_startproc endbr64 pushq %r12 .cfi_def_cfa_offset 16 .cfi_offset 12, -16 pushq %rbp .cfi_def_cfa_offset 24 .cfi_offset 6, -24 pushq %rbx .cfi_def_cfa_offset 32 .cfi_offset 3, -32 subq $1072, %rsp .cfi_def_cfa_offset 1104 movq %fs:40, %rax movq %rax, 1064(%rsp) xorl %eax, %eax movq $0, (%rsp) leaq 32(%rsp), %rdi movl $128, %ecx rep stosq movq %rsp, %rdi movl $1024, %esi call cudaMalloc@PLT movl $1, 8(%rsp) movl $1, 12(%rsp) movl $1, 16(%rsp) movl $128, 20(%rsp) movl $1, 24(%rsp) movl $1, 28(%rsp) movl $0, %r9d movl $0, %r8d movq 20(%rsp), %rdx movl $1, %ecx movq 8(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L17 .L12: leaq 32(%rsp), %rbx movl $2, %ecx movl $1024, %edx movq (%rsp), %rsi movq %rbx, %rdi call cudaMemcpy@PLT leaq 544(%rsp), %r12 leaq .LC0(%rip), %rbp .L13: movl 512(%rbx), %ecx movl (%rbx), %edx movq %rbp, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT addq $4, %rbx cmpq %r12, %rbx jne .L13 movq (%rsp), %rdi call cudaFree@PLT movq 1064(%rsp), %rax subq %fs:40, %rax jne .L18 addq $1072, %rsp .cfi_remember_state .cfi_def_cfa_offset 32 popq %rbx .cfi_def_cfa_offset 24 popq %rbp .cfi_def_cfa_offset 16 popq %r12 .cfi_def_cfa_offset 8 ret .L17: .cfi_restore_state movq (%rsp), %rdi call _ZL30__device_stub__Z10threadDemoPjPj jmp .L12 .L18: call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size _Z10ThreadDemov, .-_Z10ThreadDemov .section .rodata.str1.1 .LC1: .string "_Z10threadDemoPj" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2085: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC1(%rip), %rdx movq %rdx, %rcx leaq _ZL10threadDemoPj(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2085: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "threadDemo.hip" .globl _Z10ThreadDemov # -- Begin function _Z10ThreadDemov .p2align 4, 0x90 .type _Z10ThreadDemov,@function _Z10ThreadDemov: # @_Z10ThreadDemov .cfi_startproc # %bb.0: pushq %rbx .cfi_def_cfa_offset 16 subq $1104, %rsp # imm = 0x450 .cfi_def_cfa_offset 1120 .cfi_offset %rbx, -16 movq $0, 8(%rsp) leaq 80(%rsp), %rdi movl $1024, %edx # imm = 0x400 xorl %esi, %esi callq memset@PLT leaq 8(%rsp), %rdi movl $1024, %esi # imm = 0x400 callq hipMalloc movabsq $4294967297, %rdi # imm = 0x100000001 leaq 127(%rdi), %rdx movl $1, %esi movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB0_2 # %bb.1: movq 8(%rsp), %rax movq %rax, 72(%rsp) leaq 72(%rsp), %rax movq %rax, 16(%rsp) leaq 56(%rsp), %rdi leaq 40(%rsp), %rsi leaq 32(%rsp), %rdx leaq 24(%rsp), %rcx callq __hipPopCallConfiguration movq 56(%rsp), %rsi movl 64(%rsp), %edx movq 40(%rsp), %rcx movl 48(%rsp), %r8d leaq 16(%rsp), %r9 movl $_ZL10threadDemoPj, %edi pushq 24(%rsp) .cfi_adjust_cfa_offset 8 pushq 40(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB0_2: movq 8(%rsp), %rsi leaq 80(%rsp), %rdi movl $1024, %edx # imm = 0x400 movl $2, %ecx callq hipMemcpy xorl %ebx, %ebx .p2align 4, 0x90 .LBB0_3: # =>This Inner Loop Header: Depth=1 movl 80(%rsp,%rbx,4), %esi movl 592(%rsp,%rbx,4), %edx movl $.L.str, %edi xorl %eax, %eax callq printf incq %rbx cmpq $128, %rbx jne .LBB0_3 # %bb.4: movq 8(%rsp), %rdi callq hipFree addq $1104, %rsp # imm = 0x450 .cfi_def_cfa_offset 16 popq %rbx .cfi_def_cfa_offset 8 retq .Lfunc_end0: .size _Z10ThreadDemov, .Lfunc_end0-_Z10ThreadDemov .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function _ZL25__device_stub__threadDemoPj .type _ZL25__device_stub__threadDemoPj,@function _ZL25__device_stub__threadDemoPj: # @_ZL25__device_stub__threadDemoPj .cfi_startproc # %bb.0: subq $72, %rsp .cfi_def_cfa_offset 80 movq %rdi, 64(%rsp) leaq 64(%rsp), %rax movq %rax, (%rsp) leaq 48(%rsp), %rdi leaq 32(%rsp), %rsi leaq 24(%rsp), %rdx leaq 16(%rsp), %rcx callq __hipPopCallConfiguration movq 48(%rsp), %rsi movl 56(%rsp), %edx movq 32(%rsp), %rcx movl 40(%rsp), %r8d movq %rsp, %r9 movl $_ZL10threadDemoPj, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $88, %rsp .cfi_adjust_cfa_offset -88 retq .Lfunc_end1: .size _ZL25__device_stub__threadDemoPj, .Lfunc_end1-_ZL25__device_stub__threadDemoPj .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB2_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB2_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_ZL10threadDemoPj, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end2: .size __hip_module_ctor, .Lfunc_end2-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB3_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB3_2: retq .Lfunc_end3: .size __hip_module_dtor, .Lfunc_end3-__hip_module_dtor .cfi_endproc # -- End function .type _ZL10threadDemoPj,@object # @_ZL10threadDemoPj .section .rodata,"a",@progbits .p2align 3, 0x0 _ZL10threadDemoPj: .quad _ZL25__device_stub__threadDemoPj .size _ZL10threadDemoPj, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "(%u,%u)" .size .L.str, 8 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_ZL10threadDemoPj" .size .L__unnamed_1, 18 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _ZL25__device_stub__threadDemoPj .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _ZL10threadDemoPj .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
#include <cstdio> int main(void) { int count; cudaGetDeviceCount(&count); printf("%d devices found supporting CUDA\n", count); char split[] = "----------------------------------\n"; cudaDeviceProp p; for(int d = 0; d < count; d++){ cudaGetDeviceProperties(&p, d); printf("%s", split); printf("Device %s\n", p.name); printf("%s", split); printf(" Device memory: \t%zu\n", p.totalGlobalMem); printf(" Memory per-block: \t%lu\n", p.sharedMemPerBlock); printf(" Register per-block: \t%d\n", p.regsPerBlock); printf(" Warp size: \t\t%d\n", p.warpSize); printf(" Memory pitch: \t\t%lu\n", p.memPitch); printf(" Constant Memory: \t%lu\n", p.totalConstMem); printf(" Max thread per-block: \t%d\n", p.maxThreadsPerBlock); printf(" Max thread dim: \t%d / %d / %d\n", p.maxThreadsDim[0], p.maxThreadsDim[1], p.maxThreadsDim[2]); printf(" Max grid size: \t%d / %d / %d\n", p.maxGridSize[0], p.maxGridSize[1], p.maxGridSize[2]); printf(" Ver: \t\t\t%d.%d\n", p.major, p.minor); printf(" Clock: \t\t%d\n", p.clockRate); printf(" Texture Alignment: \t%lu\n", p.textureAlignment); } return 0; }
code for sm_80
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include <cstdio> int main(void) { int count; cudaGetDeviceCount(&count); printf("%d devices found supporting CUDA\n", count); char split[] = "----------------------------------\n"; cudaDeviceProp p; for(int d = 0; d < count; d++){ cudaGetDeviceProperties(&p, d); printf("%s", split); printf("Device %s\n", p.name); printf("%s", split); printf(" Device memory: \t%zu\n", p.totalGlobalMem); printf(" Memory per-block: \t%lu\n", p.sharedMemPerBlock); printf(" Register per-block: \t%d\n", p.regsPerBlock); printf(" Warp size: \t\t%d\n", p.warpSize); printf(" Memory pitch: \t\t%lu\n", p.memPitch); printf(" Constant Memory: \t%lu\n", p.totalConstMem); printf(" Max thread per-block: \t%d\n", p.maxThreadsPerBlock); printf(" Max thread dim: \t%d / %d / %d\n", p.maxThreadsDim[0], p.maxThreadsDim[1], p.maxThreadsDim[2]); printf(" Max grid size: \t%d / %d / %d\n", p.maxGridSize[0], p.maxGridSize[1], p.maxGridSize[2]); printf(" Ver: \t\t\t%d.%d\n", p.major, p.minor); printf(" Clock: \t\t%d\n", p.clockRate); printf(" Texture Alignment: \t%lu\n", p.textureAlignment); } return 0; }
.file "tmpxft_001b3b92_00000000-6_hello.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2060: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2060: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "%d devices found supporting CUDA\n" .section .rodata.str1.1,"aMS",@progbits,1 .LC1: .string "%s" .LC2: .string "Device %s\n" .LC3: .string " Device memory: \t%zu\n" .LC4: .string " Memory per-block: \t%lu\n" .LC5: .string " Register per-block: \t%d\n" .LC6: .string " Warp size: \t\t%d\n" .LC7: .string " Memory pitch: \t\t%lu\n" .LC8: .string " Constant Memory: \t%lu\n" .LC9: .string " Max thread per-block: \t%d\n" .section .rodata.str1.8 .align 8 .LC10: .string " Max thread dim: \t%d / %d / %d\n" .align 8 .LC11: .string " Max grid size: \t%d / %d / %d\n" .section .rodata.str1.1 .LC12: .string " Ver: \t\t\t%d.%d\n" .LC13: .string " Clock: \t\t%d\n" .LC14: .string " Texture Alignment: \t%lu\n" .text .globl main .type main, @function main: .LFB2057: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $1112, %rsp .cfi_def_cfa_offset 1168 movq %fs:40, %rax movq %rax, 1096(%rsp) xorl %eax, %eax leaq 12(%rsp), %rdi call cudaGetDeviceCount@PLT movl 12(%rsp), %edx leaq .LC0(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movabsq $3255307777713450285, %rax movabsq $3255307777713450285, %rdx movq %rax, 1056(%rsp) movq %rdx, 1064(%rsp) movq %rax, 1072(%rsp) movq %rdx, 1080(%rsp) movl $666925, 1088(%rsp) cmpl $0, 12(%rsp) jle .L4 movl $0, %ebx leaq .LC1(%rip), %r13 leaq .LC2(%rip), %r15 leaq .LC3(%rip), %r14 .L5: leaq 16(%rsp), %r12 movl %ebx, %esi movq %r12, %rdi call cudaGetDeviceProperties_v2@PLT leaq 1056(%rsp), %rbp movq %rbp, %rdx movq %r13, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq %r12, %rdx movq %r15, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq %rbp, %rdx movq %r13, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 304(%rsp), %rdx movq %r14, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 312(%rsp), %rdx leaq .LC4(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 320(%rsp), %edx leaq .LC5(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 324(%rsp), %edx leaq .LC6(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 328(%rsp), %rdx leaq .LC7(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 368(%rsp), %rdx leaq .LC8(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 336(%rsp), %edx leaq .LC9(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 348(%rsp), %r8d movl 344(%rsp), %ecx movl 340(%rsp), %edx leaq .LC10(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 360(%rsp), %r8d movl 356(%rsp), %ecx movl 352(%rsp), %edx leaq .LC11(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 380(%rsp), %ecx movl 376(%rsp), %edx leaq .LC12(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 364(%rsp), %edx leaq .LC13(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 384(%rsp), %rdx leaq .LC14(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT addl $1, %ebx cmpl %ebx, 12(%rsp) jg .L5 .L4: movq 1096(%rsp), %rax subq %fs:40, %rax jne .L9 movl $0, %eax addq $1112, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L9: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size main, .-main .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include <cstdio> int main(void) { int count; cudaGetDeviceCount(&count); printf("%d devices found supporting CUDA\n", count); char split[] = "----------------------------------\n"; cudaDeviceProp p; for(int d = 0; d < count; d++){ cudaGetDeviceProperties(&p, d); printf("%s", split); printf("Device %s\n", p.name); printf("%s", split); printf(" Device memory: \t%zu\n", p.totalGlobalMem); printf(" Memory per-block: \t%lu\n", p.sharedMemPerBlock); printf(" Register per-block: \t%d\n", p.regsPerBlock); printf(" Warp size: \t\t%d\n", p.warpSize); printf(" Memory pitch: \t\t%lu\n", p.memPitch); printf(" Constant Memory: \t%lu\n", p.totalConstMem); printf(" Max thread per-block: \t%d\n", p.maxThreadsPerBlock); printf(" Max thread dim: \t%d / %d / %d\n", p.maxThreadsDim[0], p.maxThreadsDim[1], p.maxThreadsDim[2]); printf(" Max grid size: \t%d / %d / %d\n", p.maxGridSize[0], p.maxGridSize[1], p.maxGridSize[2]); printf(" Ver: \t\t\t%d.%d\n", p.major, p.minor); printf(" Clock: \t\t%d\n", p.clockRate); printf(" Texture Alignment: \t%lu\n", p.textureAlignment); } return 0; }
#include <hip/hip_runtime.h> #include <cstdio> int main(void) { int count; hipGetDeviceCount(&count); printf("%d devices found supporting CUDA\n", count); char split[] = "----------------------------------\n"; hipDeviceProp_t p; for(int d = 0; d < count; d++){ hipGetDeviceProperties(&p, d); printf("%s", split); printf("Device %s\n", p.name); printf("%s", split); printf(" Device memory: \t%zu\n", p.totalGlobalMem); printf(" Memory per-block: \t%lu\n", p.sharedMemPerBlock); printf(" Register per-block: \t%d\n", p.regsPerBlock); printf(" Warp size: \t\t%d\n", p.warpSize); printf(" Memory pitch: \t\t%lu\n", p.memPitch); printf(" Constant Memory: \t%lu\n", p.totalConstMem); printf(" Max thread per-block: \t%d\n", p.maxThreadsPerBlock); printf(" Max thread dim: \t%d / %d / %d\n", p.maxThreadsDim[0], p.maxThreadsDim[1], p.maxThreadsDim[2]); printf(" Max grid size: \t%d / %d / %d\n", p.maxGridSize[0], p.maxGridSize[1], p.maxGridSize[2]); printf(" Ver: \t\t\t%d.%d\n", p.major, p.minor); printf(" Clock: \t\t%d\n", p.clockRate); printf(" Texture Alignment: \t%lu\n", p.textureAlignment); } return 0; }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include <cstdio> int main(void) { int count; hipGetDeviceCount(&count); printf("%d devices found supporting CUDA\n", count); char split[] = "----------------------------------\n"; hipDeviceProp_t p; for(int d = 0; d < count; d++){ hipGetDeviceProperties(&p, d); printf("%s", split); printf("Device %s\n", p.name); printf("%s", split); printf(" Device memory: \t%zu\n", p.totalGlobalMem); printf(" Memory per-block: \t%lu\n", p.sharedMemPerBlock); printf(" Register per-block: \t%d\n", p.regsPerBlock); printf(" Warp size: \t\t%d\n", p.warpSize); printf(" Memory pitch: \t\t%lu\n", p.memPitch); printf(" Constant Memory: \t%lu\n", p.totalConstMem); printf(" Max thread per-block: \t%d\n", p.maxThreadsPerBlock); printf(" Max thread dim: \t%d / %d / %d\n", p.maxThreadsDim[0], p.maxThreadsDim[1], p.maxThreadsDim[2]); printf(" Max grid size: \t%d / %d / %d\n", p.maxGridSize[0], p.maxGridSize[1], p.maxGridSize[2]); printf(" Ver: \t\t\t%d.%d\n", p.major, p.minor); printf(" Clock: \t\t%d\n", p.clockRate); printf(" Texture Alignment: \t%lu\n", p.textureAlignment); } return 0; }
.text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .amdgpu_metadata --- amdhsa.kernels: [] amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <hip/hip_runtime.h> #include <cstdio> int main(void) { int count; hipGetDeviceCount(&count); printf("%d devices found supporting CUDA\n", count); char split[] = "----------------------------------\n"; hipDeviceProp_t p; for(int d = 0; d < count; d++){ hipGetDeviceProperties(&p, d); printf("%s", split); printf("Device %s\n", p.name); printf("%s", split); printf(" Device memory: \t%zu\n", p.totalGlobalMem); printf(" Memory per-block: \t%lu\n", p.sharedMemPerBlock); printf(" Register per-block: \t%d\n", p.regsPerBlock); printf(" Warp size: \t\t%d\n", p.warpSize); printf(" Memory pitch: \t\t%lu\n", p.memPitch); printf(" Constant Memory: \t%lu\n", p.totalConstMem); printf(" Max thread per-block: \t%d\n", p.maxThreadsPerBlock); printf(" Max thread dim: \t%d / %d / %d\n", p.maxThreadsDim[0], p.maxThreadsDim[1], p.maxThreadsDim[2]); printf(" Max grid size: \t%d / %d / %d\n", p.maxGridSize[0], p.maxGridSize[1], p.maxGridSize[2]); printf(" Ver: \t\t\t%d.%d\n", p.major, p.minor); printf(" Clock: \t\t%d\n", p.clockRate); printf(" Texture Alignment: \t%lu\n", p.textureAlignment); } return 0; }
.text .file "hello.hip" .section .rodata.cst16,"aM",@progbits,16 .p2align 4, 0x0 # -- Begin function main .LCPI0_0: .zero 16,45 .text .globl main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r14 .cfi_def_cfa_offset 24 pushq %rbx .cfi_def_cfa_offset 32 subq $1536, %rsp # imm = 0x600 .cfi_def_cfa_offset 1568 .cfi_offset %rbx, -32 .cfi_offset %r14, -24 .cfi_offset %rbp, -16 leaq 12(%rsp), %rdi callq hipGetDeviceCount movl 12(%rsp), %esi movl $.L.str, %edi xorl %eax, %eax callq printf movaps .LCPI0_0(%rip), %xmm0 # xmm0 = [45,45,45,45,45,45,45,45,45,45,45,45,45,45,45,45] movaps %xmm0, 32(%rsp) movaps %xmm0, 16(%rsp) movl $666925, 48(%rsp) # imm = 0xA2D2D cmpl $0, 12(%rsp) jle .LBB0_3 # %bb.1: # %.lr.ph leaq 64(%rsp), %rbx leaq 16(%rsp), %r14 xorl %ebp, %ebp .p2align 4, 0x90 .LBB0_2: # =>This Inner Loop Header: Depth=1 movq %rbx, %rdi movl %ebp, %esi callq hipGetDevicePropertiesR0600 movl $.L.str.1, %edi movq %r14, %rsi xorl %eax, %eax callq printf movl $.L.str.2, %edi movq %rbx, %rsi xorl %eax, %eax callq printf movl $.L.str.1, %edi movq %r14, %rsi xorl %eax, %eax callq printf movq 352(%rsp), %rsi movl $.L.str.3, %edi xorl %eax, %eax callq printf movq 360(%rsp), %rsi movl $.L.str.4, %edi xorl %eax, %eax callq printf movl 368(%rsp), %esi movl $.L.str.5, %edi xorl %eax, %eax callq printf movl 372(%rsp), %esi movl $.L.str.6, %edi xorl %eax, %eax callq printf movq 376(%rsp), %rsi movl $.L.str.7, %edi xorl %eax, %eax callq printf movq 416(%rsp), %rsi movl $.L.str.8, %edi xorl %eax, %eax callq printf movl 384(%rsp), %esi movl $.L.str.9, %edi xorl %eax, %eax callq printf movl 388(%rsp), %esi movl 392(%rsp), %edx movl 396(%rsp), %ecx movl $.L.str.10, %edi xorl %eax, %eax callq printf movl 400(%rsp), %esi movl 404(%rsp), %edx movl 408(%rsp), %ecx movl $.L.str.11, %edi xorl %eax, %eax callq printf movl 424(%rsp), %esi movl 428(%rsp), %edx movl $.L.str.12, %edi xorl %eax, %eax callq printf movl 412(%rsp), %esi movl $.L.str.13, %edi xorl %eax, %eax callq printf movq 432(%rsp), %rsi movl $.L.str.14, %edi xorl %eax, %eax callq printf incl %ebp cmpl 12(%rsp), %ebp jl .LBB0_2 .LBB0_3: # %._crit_edge xorl %eax, %eax addq $1536, %rsp # imm = 0x600 .cfi_def_cfa_offset 32 popq %rbx .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end0: .size main, .Lfunc_end0-main .cfi_endproc # -- End function .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "%d devices found supporting CUDA\n" .size .L.str, 34 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "%s" .size .L.str.1, 3 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "Device %s\n" .size .L.str.2, 11 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz " Device memory: \t%zu\n" .size .L.str.3, 22 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz " Memory per-block: \t%lu\n" .size .L.str.4, 25 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz " Register per-block: \t%d\n" .size .L.str.5, 26 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz " Warp size: \t\t%d\n" .size .L.str.6, 18 .type .L.str.7,@object # @.str.7 .L.str.7: .asciz " Memory pitch: \t\t%lu\n" .size .L.str.7, 22 .type .L.str.8,@object # @.str.8 .L.str.8: .asciz " Constant Memory: \t%lu\n" .size .L.str.8, 24 .type .L.str.9,@object # @.str.9 .L.str.9: .asciz " Max thread per-block: \t%d\n" .size .L.str.9, 28 .type .L.str.10,@object # @.str.10 .L.str.10: .asciz " Max thread dim: \t%d / %d / %d\n" .size .L.str.10, 32 .type .L.str.11,@object # @.str.11 .L.str.11: .asciz " Max grid size: \t%d / %d / %d\n" .size .L.str.11, 31 .type .L.str.12,@object # @.str.12 .L.str.12: .asciz " Ver: \t\t\t%d.%d\n" .size .L.str.12, 16 .type .L.str.13,@object # @.str.13 .L.str.13: .asciz " Clock: \t\t%d\n" .size .L.str.13, 14 .type .L.str.14,@object # @.str.14 .L.str.14: .asciz " Texture Alignment: \t%lu\n" .size .L.str.14, 26 .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA device assembly to AMD device assembly.
code for sm_80
.text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .amdgpu_metadata --- amdhsa.kernels: [] amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_001b3b92_00000000-6_hello.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2060: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2060: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "%d devices found supporting CUDA\n" .section .rodata.str1.1,"aMS",@progbits,1 .LC1: .string "%s" .LC2: .string "Device %s\n" .LC3: .string " Device memory: \t%zu\n" .LC4: .string " Memory per-block: \t%lu\n" .LC5: .string " Register per-block: \t%d\n" .LC6: .string " Warp size: \t\t%d\n" .LC7: .string " Memory pitch: \t\t%lu\n" .LC8: .string " Constant Memory: \t%lu\n" .LC9: .string " Max thread per-block: \t%d\n" .section .rodata.str1.8 .align 8 .LC10: .string " Max thread dim: \t%d / %d / %d\n" .align 8 .LC11: .string " Max grid size: \t%d / %d / %d\n" .section .rodata.str1.1 .LC12: .string " Ver: \t\t\t%d.%d\n" .LC13: .string " Clock: \t\t%d\n" .LC14: .string " Texture Alignment: \t%lu\n" .text .globl main .type main, @function main: .LFB2057: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $1112, %rsp .cfi_def_cfa_offset 1168 movq %fs:40, %rax movq %rax, 1096(%rsp) xorl %eax, %eax leaq 12(%rsp), %rdi call cudaGetDeviceCount@PLT movl 12(%rsp), %edx leaq .LC0(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movabsq $3255307777713450285, %rax movabsq $3255307777713450285, %rdx movq %rax, 1056(%rsp) movq %rdx, 1064(%rsp) movq %rax, 1072(%rsp) movq %rdx, 1080(%rsp) movl $666925, 1088(%rsp) cmpl $0, 12(%rsp) jle .L4 movl $0, %ebx leaq .LC1(%rip), %r13 leaq .LC2(%rip), %r15 leaq .LC3(%rip), %r14 .L5: leaq 16(%rsp), %r12 movl %ebx, %esi movq %r12, %rdi call cudaGetDeviceProperties_v2@PLT leaq 1056(%rsp), %rbp movq %rbp, %rdx movq %r13, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq %r12, %rdx movq %r15, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq %rbp, %rdx movq %r13, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 304(%rsp), %rdx movq %r14, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 312(%rsp), %rdx leaq .LC4(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 320(%rsp), %edx leaq .LC5(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 324(%rsp), %edx leaq .LC6(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 328(%rsp), %rdx leaq .LC7(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 368(%rsp), %rdx leaq .LC8(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 336(%rsp), %edx leaq .LC9(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 348(%rsp), %r8d movl 344(%rsp), %ecx movl 340(%rsp), %edx leaq .LC10(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 360(%rsp), %r8d movl 356(%rsp), %ecx movl 352(%rsp), %edx leaq .LC11(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 380(%rsp), %ecx movl 376(%rsp), %edx leaq .LC12(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 364(%rsp), %edx leaq .LC13(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 384(%rsp), %rdx leaq .LC14(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT addl $1, %ebx cmpl %ebx, 12(%rsp) jg .L5 .L4: movq 1096(%rsp), %rax subq %fs:40, %rax jne .L9 movl $0, %eax addq $1112, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L9: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size main, .-main .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "hello.hip" .section .rodata.cst16,"aM",@progbits,16 .p2align 4, 0x0 # -- Begin function main .LCPI0_0: .zero 16,45 .text .globl main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r14 .cfi_def_cfa_offset 24 pushq %rbx .cfi_def_cfa_offset 32 subq $1536, %rsp # imm = 0x600 .cfi_def_cfa_offset 1568 .cfi_offset %rbx, -32 .cfi_offset %r14, -24 .cfi_offset %rbp, -16 leaq 12(%rsp), %rdi callq hipGetDeviceCount movl 12(%rsp), %esi movl $.L.str, %edi xorl %eax, %eax callq printf movaps .LCPI0_0(%rip), %xmm0 # xmm0 = [45,45,45,45,45,45,45,45,45,45,45,45,45,45,45,45] movaps %xmm0, 32(%rsp) movaps %xmm0, 16(%rsp) movl $666925, 48(%rsp) # imm = 0xA2D2D cmpl $0, 12(%rsp) jle .LBB0_3 # %bb.1: # %.lr.ph leaq 64(%rsp), %rbx leaq 16(%rsp), %r14 xorl %ebp, %ebp .p2align 4, 0x90 .LBB0_2: # =>This Inner Loop Header: Depth=1 movq %rbx, %rdi movl %ebp, %esi callq hipGetDevicePropertiesR0600 movl $.L.str.1, %edi movq %r14, %rsi xorl %eax, %eax callq printf movl $.L.str.2, %edi movq %rbx, %rsi xorl %eax, %eax callq printf movl $.L.str.1, %edi movq %r14, %rsi xorl %eax, %eax callq printf movq 352(%rsp), %rsi movl $.L.str.3, %edi xorl %eax, %eax callq printf movq 360(%rsp), %rsi movl $.L.str.4, %edi xorl %eax, %eax callq printf movl 368(%rsp), %esi movl $.L.str.5, %edi xorl %eax, %eax callq printf movl 372(%rsp), %esi movl $.L.str.6, %edi xorl %eax, %eax callq printf movq 376(%rsp), %rsi movl $.L.str.7, %edi xorl %eax, %eax callq printf movq 416(%rsp), %rsi movl $.L.str.8, %edi xorl %eax, %eax callq printf movl 384(%rsp), %esi movl $.L.str.9, %edi xorl %eax, %eax callq printf movl 388(%rsp), %esi movl 392(%rsp), %edx movl 396(%rsp), %ecx movl $.L.str.10, %edi xorl %eax, %eax callq printf movl 400(%rsp), %esi movl 404(%rsp), %edx movl 408(%rsp), %ecx movl $.L.str.11, %edi xorl %eax, %eax callq printf movl 424(%rsp), %esi movl 428(%rsp), %edx movl $.L.str.12, %edi xorl %eax, %eax callq printf movl 412(%rsp), %esi movl $.L.str.13, %edi xorl %eax, %eax callq printf movq 432(%rsp), %rsi movl $.L.str.14, %edi xorl %eax, %eax callq printf incl %ebp cmpl 12(%rsp), %ebp jl .LBB0_2 .LBB0_3: # %._crit_edge xorl %eax, %eax addq $1536, %rsp # imm = 0x600 .cfi_def_cfa_offset 32 popq %rbx .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end0: .size main, .Lfunc_end0-main .cfi_endproc # -- End function .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "%d devices found supporting CUDA\n" .size .L.str, 34 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "%s" .size .L.str.1, 3 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "Device %s\n" .size .L.str.2, 11 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz " Device memory: \t%zu\n" .size .L.str.3, 22 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz " Memory per-block: \t%lu\n" .size .L.str.4, 25 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz " Register per-block: \t%d\n" .size .L.str.5, 26 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz " Warp size: \t\t%d\n" .size .L.str.6, 18 .type .L.str.7,@object # @.str.7 .L.str.7: .asciz " Memory pitch: \t\t%lu\n" .size .L.str.7, 22 .type .L.str.8,@object # @.str.8 .L.str.8: .asciz " Constant Memory: \t%lu\n" .size .L.str.8, 24 .type .L.str.9,@object # @.str.9 .L.str.9: .asciz " Max thread per-block: \t%d\n" .size .L.str.9, 28 .type .L.str.10,@object # @.str.10 .L.str.10: .asciz " Max thread dim: \t%d / %d / %d\n" .size .L.str.10, 32 .type .L.str.11,@object # @.str.11 .L.str.11: .asciz " Max grid size: \t%d / %d / %d\n" .size .L.str.11, 31 .type .L.str.12,@object # @.str.12 .L.str.12: .asciz " Ver: \t\t\t%d.%d\n" .size .L.str.12, 16 .type .L.str.13,@object # @.str.13 .L.str.13: .asciz " Clock: \t\t%d\n" .size .L.str.13, 14 .type .L.str.14,@object # @.str.14 .L.str.14: .asciz " Texture Alignment: \t%lu\n" .size .L.str.14, 26 .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include <cuda.h> #include <cuda_runtime.h> #include <math.h> #include <math_constants.h> #include <stdlib.h> #include <vector> #include <algorithm> #include <thrust/host_vector.h> #include <thrust/device_vector.h> #include <thrust/generate.h> #include <thrust/sort.h> #include <thrust/copy.h> #include <cstdlib> #define BINNUM 9 #define CELLSIZE 6 #define CELLBLOCKSIZE 3 #define STEPSIZE 3 #define K 5 #define KERNEL_WIDTH 31 __global__ void computeGradient(float* outputGr,int* outputTag,float* input,unsigned int width,unsigned int height) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; float gradientX, gradientY; float angle; int nAngle; int indTag; if(tx < width && ty < height) { if(tx == 0) //image boundary gradientX = input[tx+1 + ty*width] - input[tx + ty*width]; else if(tx == width-1) gradientX = input[tx + ty*width] - input[tx-1 + ty*width]; else if(ty == 0 || ty == height-1) gradientX = 0; else gradientX = input[tx+1 + ty*width] - input[tx-1 + ty*width]; if(ty == 0) //image boundary gradientY = input[tx + (ty+1)*width] - input[tx + (ty)*width]; else if (ty == height-1) gradientY = input[tx + (ty)*width] - input[tx + (ty-1)*width]; else if (tx ==0 || tx == width-1) gradientY = 0; else gradientY = input[tx + (ty+1)*width] - input[tx + (ty-1)*width]; outputGr[tx + ty*width] = sqrt(gradientX*gradientX + gradientY*gradientY); //outputGr[tx + ty*width] = gradientX; if(gradientX == 0) gradientX = 1e-5; //angle = ((atan(gradientY/gradientX))); angle = ((atan(gradientY/gradientX)+(CUDART_PI_F/2))*180)/CUDART_PI_F; nAngle = 180/BINNUM; indTag = ceil(angle/nAngle); if(indTag == 0) indTag=1; else if(indTag==10) indTag=9; outputTag[tx + ty*width] = indTag; } //outputX[tx + ty*width] = gradientX; //outputY[tx + ty*width] = gradientY; } __global__ void computeBinHOG(float* outputHOGFeature,float* outputWeight,int *outputPosition,int* outputOffset,float* Gr,int* Tag,unsigned int xStepNum,unsigned int yStepNum, unsigned int width,unsigned int height) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; int centerX = floor(float(width)/2.0+0.5) - 1; int centerY = floor(float(height)/2.0+0.5) - 1; if(tx < xStepNum && ty < yStepNum) { //pixel position of center int x = tx*STEPSIZE+1.5*CELLSIZE-1; int y = ty*STEPSIZE+1.5*CELLSIZE-1; int leftupperX = x - 1.5*CELLSIZE +1 ; int leftupperY = y - 1.5*CELLSIZE +1 ; //__shared__ float tmp[CELLSIZE][CELLSIZE]; float tmp[CELLBLOCKSIZE*CELLBLOCKSIZE][BINNUM] = {0.0f}; int inc = 0; for(int i = 0; i < CELLBLOCKSIZE;i++) { for(int j=0; j<CELLBLOCKSIZE;j++) { int indX = leftupperX + j*CELLSIZE; int indY = leftupperY + i*CELLSIZE; for(int p=0;p<CELLSIZE;p++) { for(int q=0;q<CELLSIZE;q++) { //int currendindX = indX+p; //int currendindY = indY+q; int binind = Tag[indX+q + width*(indY+p)]-1; float tmpdebug = Gr[indX+q + width*(indY+p)]; tmp[inc][binind] +=tmpdebug; //tmp[inc][binind] += Gr[indX+q + width*(indY+p)]; } } inc++; } } float norm = 0.0f; int nonempty = 0; for(int p=0;p<CELLBLOCKSIZE*CELLBLOCKSIZE;p++) { for(int q=0;q<BINNUM;q++) { norm+= (tmp[p][q]*tmp[p][q]); if(tmp[p][q]>0) nonempty ++; } } norm = sqrt(norm); norm +=1e-5; //if(norm <= 0.0f) // norm = 1e-5; //float acc = 0.0; for(int p=0;p<CELLBLOCKSIZE*CELLBLOCKSIZE;p++) { for(int q=0;q<BINNUM;q++) { //outputHOGFeature[ (tx+ty*xStepNum)*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM +(q+p*BINNUM)] = tmp[p][q]/norm; outputHOGFeature[ (ty+tx*yStepNum)*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM +(q+p*BINNUM)] = tmp[p][q]/norm; //acc +=tmp[p][q]/norm; } } /*outputPosition[(tx + ty*xStepNum)*2] = x; outputPosition[(tx + ty*xStepNum)*2 + 1] = y; outputOffset[(tx + ty*xStepNum)*2] = centerX - x; outputOffset[(tx + ty*xStepNum)*2 + 1] = centerY - y; outputWeight[tx + ty*xStepNum] = float(nonempty)/81.0f;*/ outputPosition[(ty+tx*yStepNum)*2] = x; outputPosition[(ty+tx*yStepNum)*2 + 1] = y; outputOffset[(ty+tx*yStepNum)*2] = centerX - x; outputOffset[(ty+tx*yStepNum)*2 + 1] = centerY - y; outputWeight[(ty+tx*yStepNum)] = float(nonempty)/81.0f; //outputWeight[(ty+tx*yStepNum)] = float(acc)/1.0f; } } __global__ void computeBinHOGTar(float* outputHOGFeature,int *outputPosition,float* Gr,int* Tag,unsigned int xStepNum,unsigned int yStepNum, unsigned int width,unsigned int height) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; if(tx < xStepNum && ty < yStepNum) { //pixel position of center int x = tx*STEPSIZE+1.5*CELLSIZE-1; int y = ty*STEPSIZE+1.5*CELLSIZE-1; int leftupperX = x - 1.5*CELLSIZE +1 ; int leftupperY = y - 1.5*CELLSIZE +1 ; //__shared__ float tmp[CELLSIZE][CELLSIZE]; float tmp[CELLBLOCKSIZE*CELLBLOCKSIZE][BINNUM] = {0.0f}; int inc = 0; for(int i = 0; i < CELLBLOCKSIZE;i++) { for(int j=0; j<CELLBLOCKSIZE;j++) { int indX = leftupperX + j*CELLSIZE; int indY = leftupperY + i*CELLSIZE; for(int p=0;p<CELLSIZE;p++) { for(int q=0;q<CELLSIZE;q++) { //int currendindX = indX+p; //int currendindY = indY+q; int binind = Tag[indX+q + width*(indY+p)]-1; tmp[inc][binind] += Gr[indX+q + width*(indY+p)]; } } inc++; } } float norm = 0.0f; int nonempty = 0; for(int p=0;p<CELLBLOCKSIZE*CELLBLOCKSIZE;p++) { for(int q=0;q<BINNUM;q++) { norm+= (tmp[p][q]*tmp[p][q]); if(tmp[p][q]>0) nonempty ++; } } norm = sqrt(norm); norm +=1e-5; //if(norm <= 0.0f) // norm = 1e-5; for(int p=0;p<CELLBLOCKSIZE*CELLBLOCKSIZE;p++) { for(int q=0;q<BINNUM;q++) { //outputHOGFeature[ (tx+ty*xStepNum)*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM +(q+p*BINNUM)] = tmp[p][q]/norm; outputHOGFeature[ (ty+tx*yStepNum)*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM +(q+p*BINNUM)] = tmp[p][q]/norm; } } //outputPosition[(tx + ty*xStepNum)*2] = x; //outputPosition[(tx + ty*xStepNum)*2 + 1] = y; outputPosition[(ty+tx*yStepNum)*2] = x; outputPosition[(ty+tx*yStepNum)*2 + 1] = y; } } __global__ void computeDistance(float* output,float* hog,float* hogTar,unsigned int width,unsigned int height) { /*__shared__ float tmp[2][81]; unsigned int bx = blockIdx.x * 1; unsigned int by = blockIdx.y * 1; unsigned int tx = threadIdx.x * 1; if(tx < 81) { tmp[0][tx] = hog[tx + bx*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM]; tmp[1][tx] = hogTar[tx + by*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM]; } __syncthreads(); float perDistance=0.0; float tmp1 = 0.0; float tmp2 = 0.0; for(int i=0;i<81;i++) { tmp1 = tmp[0][i]; tmp2 = tmp[1][i]; perDistance += ((tmp1-tmp2)*(tmp1-tmp2))/(tmp1 + tmp2 + 1e-5); } output[bx + by*width] = 0.5 * perDistance;*/ unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; if(tx < width && ty < height) { float perDistance = 0.0; //float tmp[CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM]= {0.0f}; //float tmpTar[CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM]= {0.0f}; float tmp = 0.0; float tmpTar = 0.0; //int i=0; for(int i =0;i<CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM;i++) { tmp = hog[i + tx*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM]; tmpTar = hogTar[i + ty*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM]; //float tmp = hog[i*(width) + tx]; //float tmpTar = hogTar[i*height + ty]; perDistance = perDistance + ((tmp-tmpTar)*(tmp-tmpTar))/(tmp + tmpTar + 1e-5); //perDistance=1; } output[tx + ty*width] = 0.5 * perDistance; //output[ty + tx*height] = 0.5 * perDistance; } } __global__ void vote(float *OutputVoteMatrix,float *HOGDistance, float *weight,int *offset,int *samplepixel, unsigned int width,unsigned int height,unsigned int voteWidth,unsigned int voteHeight) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; if(tx < width && ty < height) { //if (HOGDistance[tx + ty*width] <= threshold[ty]) if (HOGDistance[tx + ty*width] <= 2)// { int x = samplepixel[ty*2]; int y = samplepixel[ty*2+1]; int offsetx = offset[tx*2]; int offsety = offset[tx*2+1]; int centerx = x + offsetx; int centery = y + offsety; if(centerx >= 0 && centery >= 0 && centerx < voteWidth && centery< voteHeight) { OutputVoteMatrix[centerx + centery*voteWidth] += 1/(HOGDistance[tx + ty*width]+1e-5 + 1) * weight[tx]; } } } } __global__ void genGaussianFilter(float* output) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; __shared__ float gaussianKernel[KERNEL_WIDTH][KERNEL_WIDTH]; float sigma = 5.0; float gaussianSum = 0.0; gaussianKernel[ty][tx] = powf(2.71828,-((tx-15)*(tx-15)/(2*sigma*sigma) + (ty-15)*(ty-15)/(2*sigma*sigma))); __syncthreads(); for(int i=0;i<KERNEL_WIDTH;i++) for(int j=0;j<KERNEL_WIDTH;j++) { gaussianSum += gaussianKernel[i][j]; } output[ty*KERNEL_WIDTH + tx] = gaussianKernel[ty][tx]/gaussianSum; } __global__ void gaussianFilter(float* output,float* input, float* gaussianKernel,unsigned int width,unsigned int height) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; if(tx < width && ty < height) { float accum = 0.0; float offset = 15; for (int i = 0; i < KERNEL_WIDTH; ++i) { for (int j = 0; j < KERNEL_WIDTH; ++j) { int ind = tx+j-offset + (ty+i-offset) * width; if(ind > 0 && ind < width*height) accum += gaussianKernel[j + KERNEL_WIDTH*i] * input[ind]; } output[tx + ty*width] = accum; } } } __global__ void detectMaximal(float* output,float* input,unsigned int width,unsigned int height,float maximal) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; if(tx < width-1 && ty < height-1 && tx >0 && ty>0) { if (input[tx + ty*width] > 0.8*maximal && input[tx + ty*width] > input[tx-1 + (ty-1)*width] && input[tx + ty*width] > input[tx + (ty-1)*width] && input[tx + ty*width] > input[tx+1 + (ty-1)*width] && input[tx + ty*width] > input[tx-1 + (ty)*width] && input[tx + ty*width] > input[tx+1 + (ty)*width] && input[tx + ty*width] > input[tx-1 + (ty+1)*width] && input[tx + ty*width] > input[tx + (ty+1)*width] && input[tx + ty*width] > input[tx+1 + (ty+1)*width]) { output[tx + ty*width] = (tx+1) + (ty)*width; } } } __global__ void computeThreshold(float* output,float* input) { unsigned int tx = threadIdx.x; unsigned int bx = blockIdx.x; __shared__ float partialMin[500]; partialMin[tx] = input[tx + bx*blockDim.x]; unsigned int stride = 1; for (stride = 1; stride < blockDim.x; stride *= 2) { __syncthreads(); if (tx % (2 * stride) == 0 && tx + stride < blockDim.x) partialMin[tx] = partialMin[tx] < partialMin[tx + stride] ? partialMin[tx]:partialMin[tx + stride]; } /*unsigned int stridenew = stride/2; float kmax = partialMin[0]; for(int i=0;i<blockDim.x;i+=stridenew) { if(kmax < partialMin[i]) kmax = partialMin[i]; }*/ output[0] = partialMin[0]; } extern "C" float* computeHOG(float *img,float* imgTar,unsigned int width,unsigned int height,unsigned int widthTar,unsigned int heightTar) { //For performance analysis cudaEvent_t start, stop0,stop1, stop2, stop3, stop4, stop5, stop6,stop7,stop8,stop9,stop10,stop11,stop12,stop13,stop14; cudaEventCreate(&start); cudaEventCreate(&stop0); cudaEventCreate(&stop1); cudaEventCreate(&stop2); cudaEventCreate(&stop3); cudaEventCreate(&stop4); cudaEventCreate(&stop5); cudaEventCreate(&stop6); cudaEventCreate(&stop7); cudaEventCreate(&stop8); cudaEventCreate(&stop9); cudaEventCreate(&stop10); cudaEventCreate(&stop11); cudaEventCreate(&stop12); cudaEventCreate(&stop13); cudaEventCreate(&stop14); cudaEventRecord(start, 0); int xStepNum = floor((width-CELLSIZE*CELLBLOCKSIZE)/float(STEPSIZE)); int yStepNum = floor((height-CELLSIZE*CELLBLOCKSIZE)/float(STEPSIZE)); int xStepNumTar = floor((widthTar-CELLSIZE*CELLBLOCKSIZE)/float(STEPSIZE)); int yStepNumTar = floor((heightTar-CELLSIZE*CELLBLOCKSIZE)/float(STEPSIZE)); float *Gr; cudaMalloc((void**)&Gr , width*height*sizeof(float)); int *Tag; cudaMalloc((void**)&Tag , width*height*sizeof(int)); float *HOGFeature; cudaMalloc((void**)&HOGFeature , xStepNum*yStepNum*BINNUM*CELLBLOCKSIZE*CELLBLOCKSIZE*sizeof(float)); cudaMemset(HOGFeature,0,xStepNum*yStepNum*BINNUM*CELLBLOCKSIZE*CELLBLOCKSIZE*sizeof(float)); int *centerPosition; cudaMalloc((void**)&centerPosition , 2*xStepNum*yStepNum*sizeof(int)); int *offset; cudaMalloc((void**)&offset , 2*xStepNum*yStepNum*sizeof(int)); float *weight; cudaMalloc((void**)&weight , xStepNum*yStepNum*sizeof(float)); float *GrTar; cudaMalloc((void**)&GrTar , widthTar*heightTar*sizeof(float)); int *TagTar; cudaMalloc((void**)&TagTar , widthTar*heightTar*sizeof(int)); int *samplepixel; cudaMalloc((void**)&samplepixel , 2*xStepNum*yStepNum*sizeof(int)); float *HOGFeatureTar; cudaMalloc((void**)&HOGFeatureTar , xStepNumTar*yStepNumTar*BINNUM*CELLBLOCKSIZE*CELLBLOCKSIZE*sizeof(float)); cudaMemset(HOGFeatureTar,0,xStepNumTar*yStepNumTar*BINNUM*CELLBLOCKSIZE*CELLBLOCKSIZE*sizeof(float)); float *HOGDistance; cudaMalloc((void**)&HOGDistance , xStepNumTar*yStepNumTar*xStepNum*yStepNum*sizeof(float)); cudaMemset(HOGDistance,0,xStepNumTar*yStepNumTar*xStepNum*yStepNum*sizeof(float)); float *votematrix; cudaMalloc((void**)&votematrix , widthTar*heightTar* sizeof(float)); cudaMemset(votematrix,0,widthTar*heightTar*sizeof(float)); float *gaussianKernel_dev; cudaMalloc((void**)&gaussianKernel_dev , KERNEL_WIDTH*KERNEL_WIDTH* sizeof(float)); cudaMemset(gaussianKernel_dev,0,KERNEL_WIDTH*KERNEL_WIDTH*sizeof(float)); float *votematrix_smoothed; cudaMalloc((void**)&votematrix_smoothed , widthTar*heightTar* sizeof(float)); cudaMemset(votematrix_smoothed,0,widthTar*heightTar*sizeof(float)); float *vote_maxposition; cudaMalloc((void**)&vote_maxposition , widthTar*heightTar* sizeof(float)); cudaMemset(vote_maxposition,0,widthTar*heightTar*sizeof(float)); //******************** cudaEventRecord(stop0, 0); cudaEventSynchronize(stop0); float elapsedTime0; cudaEventElapsedTime(&elapsedTime0, start, stop0); printf("Initialization :%f ms\n", elapsedTime0); //******************** dim3 dimBlock(16,16,1); dim3 dimGrid(width/dimBlock.x + 1, height/dimBlock.y + 1,1); computeGradient<<<dimGrid,dimBlock>>>(Gr,Tag,img,width,height); //******************** cudaEventRecord(stop1, 0); cudaEventSynchronize(stop1); float elapsedTime1; cudaEventElapsedTime(&elapsedTime1, start, stop1); printf("Gradient Computation for Ref Image :%f ms\n", elapsedTime1); //******************** dimBlock = dim3(16,16,1); dimGrid = dim3(xStepNum/dimBlock.x + 1, yStepNum/dimBlock.y + 1,1); computeBinHOG<<<dimGrid,dimBlock>>>(HOGFeature,weight,centerPosition,offset,Gr,Tag,xStepNum,yStepNum,width,height); //******************** cudaEventRecord(stop2, 0); cudaEventSynchronize(stop2); float elapsedTime2; cudaEventElapsedTime(&elapsedTime2, start, stop2); printf("HOG Computation for Ref Image :%f ms\n", elapsedTime2); //******************** dimBlock = dim3(16,16,1); dimGrid = dim3(widthTar/dimBlock.x + 1, heightTar/dimBlock.y + 1,1); computeGradient<<<dimGrid,dimBlock>>>(GrTar,TagTar,imgTar,widthTar,heightTar); //******************** cudaEventRecord(stop3, 0); cudaEventSynchronize(stop3); float elapsedTime3; cudaEventElapsedTime(&elapsedTime3, start, stop3); printf("Gradient Computation for Target Image :%f ms\n", elapsedTime3); //******************** dimBlock = dim3(16,16,1); dimGrid = dim3(xStepNumTar/dimBlock.x + 1, yStepNumTar/dimBlock.y + 1,1); computeBinHOGTar<<<dimGrid,dimBlock>>>(HOGFeatureTar,samplepixel,GrTar,TagTar,xStepNumTar,yStepNumTar,widthTar,heightTar); //******************** cudaEventRecord(stop4, 0); cudaEventSynchronize(stop4); float elapsedTime4; cudaEventElapsedTime(&elapsedTime4, start, stop4); printf("HOG Computation for Target Image :%f ms\n", elapsedTime4); //******************** dimBlock = dim3(16,16,1); dimGrid = dim3((xStepNum*yStepNum)/dimBlock.x + 1, (xStepNumTar*yStepNumTar)/dimBlock.y + 1,1); computeDistance<<<dimGrid,dimBlock>>>(HOGDistance,HOGFeature,HOGFeatureTar,xStepNum*yStepNum,xStepNumTar*yStepNumTar); //******************** cudaEventRecord(stop5, 0); cudaEventSynchronize(stop5); float elapsedTime5; cudaEventElapsedTime(&elapsedTime5, start, stop5); printf("Distance Matrix Computation:%f ms\n", elapsedTime5); //******************** //float *HOGDistanceSorted = new float[xStepNumTar*yStepNumTar*xStepNum*yStepNum]; //cudaMemcpy(HOGDistanceSorted,HOGDistance, xStepNumTar*yStepNumTar*xStepNum*yStepNum*sizeof(float), cudaMemcpyDeviceToHost); /*float *threshold_host = new float[xStepNumTar*yStepNumTar]; for(int i=0;i<xStepNumTar*yStepNumTar;i++) { thrust::sort(HOGDistanceSorted + i*(xStepNum*yStepNum), HOGDistanceSorted +i*(xStepNum*yStepNum) +xStepNum*yStepNum); threshold_host[i] = HOGDistanceSorted[i*(xStepNum*yStepNum) + K - 1]; }*/ /*float *threshold_dev; cudaMalloc((void**)&threshold_dev , xStepNumTar*yStepNumTar* sizeof(float)); cudaMemcpy(threshold_dev,threshold_host, xStepNumTar*yStepNumTar*sizeof(float), cudaMemcpyHostToDevice);*/ //cudaMemcpy(threshold_dev,threshold_host, xStepNumTar*yStepNumTar*sizeof(float), cudaMemcpyHostToDevice); /*dimBlock = dim3(xStepNum*yStepNum,1,1); dimGrid = dim3(xStepNumTar*yStepNumTar,1,1); float *threshold_dev; cudaMalloc((void**)&threshold_dev , xStepNumTar*yStepNumTar* sizeof(float)); cudaMemset(threshold_dev,0,xStepNumTar*yStepNumTar*sizeof(float)); computeThreshold<<<dimGrid,dimBlock>>>(threshold_dev,HOGDistance);*/ dimBlock = dim3(16,16,1); dimGrid = dim3((xStepNum*yStepNum)/dimBlock.x + 1, (xStepNumTar*yStepNumTar)/dimBlock.y + 1,1); //vote<<<dimGrid,dimBlock>>>(votematrix,threshold_dev,HOGDistance,weight,offset,samplepixel,xStepNum*yStepNum,xStepNumTar*yStepNumTar,widthTar,heightTar); vote<<<dimGrid,dimBlock>>>(votematrix,HOGDistance,weight,offset,samplepixel,xStepNum*yStepNum,xStepNumTar*yStepNumTar,widthTar,heightTar); //******************** cudaEventRecord(stop6, 0); cudaEventSynchronize(stop6); float elapsedTime6; cudaEventElapsedTime(&elapsedTime6, start, stop6); printf("Threshold and Voting:%f ms\n", elapsedTime6); //******************** dimBlock = dim3(KERNEL_WIDTH,KERNEL_WIDTH,1); dimGrid = dim3(1,1,1); genGaussianFilter<<<dimGrid,dimBlock>>>(gaussianKernel_dev); //******************** cudaEventRecord(stop7, 0); cudaEventSynchronize(stop7); float elapsedTime7; cudaEventElapsedTime(&elapsedTime7, start, stop7); printf("Gaussian Kernel Generation:%f ms\n", elapsedTime7); //******************** dimBlock = dim3(16,16,1); dimGrid = dim3(widthTar/dimBlock.x + 1, heightTar/dimBlock.y + 1,1); gaussianFilter<<<dimGrid,dimBlock>>>(votematrix_smoothed,votematrix, gaussianKernel_dev,widthTar,heightTar); //******************** cudaEventRecord(stop8, 0); cudaEventSynchronize(stop8); float elapsedTime8; cudaEventElapsedTime(&elapsedTime8, start, stop8); printf("Gaussian Blur:%f ms\n", elapsedTime8); //******************** float *votematrix_smoothed_sorted = new float[widthTar*heightTar]; cudaMemcpy(votematrix_smoothed_sorted,votematrix_smoothed, widthTar*heightTar*sizeof(float), cudaMemcpyDeviceToHost); thrust::sort(votematrix_smoothed_sorted, votematrix_smoothed_sorted +widthTar*heightTar); float maxvalue = votematrix_smoothed_sorted[widthTar*heightTar-1]; //******************** cudaEventRecord(stop9, 0); cudaEventSynchronize(stop9); float elapsedTime9; cudaEventElapsedTime(&elapsedTime9, start, stop9); printf("Maximal Voting Value Computation:%f ms\n", elapsedTime9); //******************** dimBlock = dim3(16,16,1); dimGrid = dim3(widthTar/dimBlock.x + 1, heightTar/dimBlock.y + 1,1); detectMaximal<<<dimGrid,dimBlock>>>(vote_maxposition,votematrix_smoothed,widthTar,heightTar,maxvalue); //******************** cudaEventRecord(stop10, 0); cudaEventSynchronize(stop10); float elapsedTime10; cudaEventElapsedTime(&elapsedTime10, start, stop10); printf("Local Maximal Detection:%f ms\n", elapsedTime10); //******************** float *vote_maxposition_sorted = new float[widthTar*heightTar]; cudaMemcpy(vote_maxposition_sorted,vote_maxposition, widthTar*heightTar*sizeof(float), cudaMemcpyDeviceToHost); thrust::sort(vote_maxposition_sorted, vote_maxposition_sorted +widthTar*heightTar); int centernum = 0; for(int i=widthTar*heightTar-1;i>=0;i--) { if(vote_maxposition_sorted[i] > 0) centernum++; else break; } //******************** cudaEventRecord(stop11, 0); cudaEventSynchronize(stop11); float elapsedTime11; cudaEventElapsedTime(&elapsedTime11, start, stop11); printf("Transfering Detected Logo Position to HOST:%f ms\n", elapsedTime11); //******************** //float *ret = new float[centernum]; //cudaMemcpy(ret,vote_maxposition_sorted + widthTar*heightTar - centernum, centernum*sizeof(float), cudaMemcpyHostToHost); float *ret = new float[centernum+1]; ret[0] = float(centernum); cudaMemcpy(ret+1,vote_maxposition_sorted + widthTar*heightTar - centernum, centernum*sizeof(float), cudaMemcpyHostToHost); //float *ret = new float[widthTar*heightTar]; //cudaMemcpy(ret,votematrix_smoothed, widthTar*heightTar*sizeof(float), cudaMemcpyDeviceToHost); return ret; }
#include <hip/hip_runtime.h> #include <math.h> #include <hip/hip_math_constants.h> #include <stdlib.h> #include <vector> #include <algorithm> #include <thrust/host_vector.h> #include <thrust/device_vector.h> #include <thrust/generate.h> #include <thrust/sort.h> #include <thrust/copy.h> #include <cstdlib> #define BINNUM 9 #define CELLSIZE 6 #define CELLBLOCKSIZE 3 #define STEPSIZE 3 #define K 5 #define KERNEL_WIDTH 31 __global__ void computeGradient(float* outputGr,int* outputTag,float* input,unsigned int width,unsigned int height) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; float gradientX, gradientY; float angle; int nAngle; int indTag; if(tx < width && ty < height) { if(tx == 0) //image boundary gradientX = input[tx+1 + ty*width] - input[tx + ty*width]; else if(tx == width-1) gradientX = input[tx + ty*width] - input[tx-1 + ty*width]; else if(ty == 0 || ty == height-1) gradientX = 0; else gradientX = input[tx+1 + ty*width] - input[tx-1 + ty*width]; if(ty == 0) //image boundary gradientY = input[tx + (ty+1)*width] - input[tx + (ty)*width]; else if (ty == height-1) gradientY = input[tx + (ty)*width] - input[tx + (ty-1)*width]; else if (tx ==0 || tx == width-1) gradientY = 0; else gradientY = input[tx + (ty+1)*width] - input[tx + (ty-1)*width]; outputGr[tx + ty*width] = sqrt(gradientX*gradientX + gradientY*gradientY); //outputGr[tx + ty*width] = gradientX; if(gradientX == 0) gradientX = 1e-5; //angle = ((atan(gradientY/gradientX))); angle = ((atan(gradientY/gradientX)+(HIP_PI_F/2))*180)/HIP_PI_F; nAngle = 180/BINNUM; indTag = ceil(angle/nAngle); if(indTag == 0) indTag=1; else if(indTag==10) indTag=9; outputTag[tx + ty*width] = indTag; } //outputX[tx + ty*width] = gradientX; //outputY[tx + ty*width] = gradientY; } __global__ void computeBinHOG(float* outputHOGFeature,float* outputWeight,int *outputPosition,int* outputOffset,float* Gr,int* Tag,unsigned int xStepNum,unsigned int yStepNum, unsigned int width,unsigned int height) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; int centerX = floor(float(width)/2.0+0.5) - 1; int centerY = floor(float(height)/2.0+0.5) - 1; if(tx < xStepNum && ty < yStepNum) { //pixel position of center int x = tx*STEPSIZE+1.5*CELLSIZE-1; int y = ty*STEPSIZE+1.5*CELLSIZE-1; int leftupperX = x - 1.5*CELLSIZE +1 ; int leftupperY = y - 1.5*CELLSIZE +1 ; //__shared__ float tmp[CELLSIZE][CELLSIZE]; float tmp[CELLBLOCKSIZE*CELLBLOCKSIZE][BINNUM] = {0.0f}; int inc = 0; for(int i = 0; i < CELLBLOCKSIZE;i++) { for(int j=0; j<CELLBLOCKSIZE;j++) { int indX = leftupperX + j*CELLSIZE; int indY = leftupperY + i*CELLSIZE; for(int p=0;p<CELLSIZE;p++) { for(int q=0;q<CELLSIZE;q++) { //int currendindX = indX+p; //int currendindY = indY+q; int binind = Tag[indX+q + width*(indY+p)]-1; float tmpdebug = Gr[indX+q + width*(indY+p)]; tmp[inc][binind] +=tmpdebug; //tmp[inc][binind] += Gr[indX+q + width*(indY+p)]; } } inc++; } } float norm = 0.0f; int nonempty = 0; for(int p=0;p<CELLBLOCKSIZE*CELLBLOCKSIZE;p++) { for(int q=0;q<BINNUM;q++) { norm+= (tmp[p][q]*tmp[p][q]); if(tmp[p][q]>0) nonempty ++; } } norm = sqrt(norm); norm +=1e-5; //if(norm <= 0.0f) // norm = 1e-5; //float acc = 0.0; for(int p=0;p<CELLBLOCKSIZE*CELLBLOCKSIZE;p++) { for(int q=0;q<BINNUM;q++) { //outputHOGFeature[ (tx+ty*xStepNum)*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM +(q+p*BINNUM)] = tmp[p][q]/norm; outputHOGFeature[ (ty+tx*yStepNum)*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM +(q+p*BINNUM)] = tmp[p][q]/norm; //acc +=tmp[p][q]/norm; } } /*outputPosition[(tx + ty*xStepNum)*2] = x; outputPosition[(tx + ty*xStepNum)*2 + 1] = y; outputOffset[(tx + ty*xStepNum)*2] = centerX - x; outputOffset[(tx + ty*xStepNum)*2 + 1] = centerY - y; outputWeight[tx + ty*xStepNum] = float(nonempty)/81.0f;*/ outputPosition[(ty+tx*yStepNum)*2] = x; outputPosition[(ty+tx*yStepNum)*2 + 1] = y; outputOffset[(ty+tx*yStepNum)*2] = centerX - x; outputOffset[(ty+tx*yStepNum)*2 + 1] = centerY - y; outputWeight[(ty+tx*yStepNum)] = float(nonempty)/81.0f; //outputWeight[(ty+tx*yStepNum)] = float(acc)/1.0f; } } __global__ void computeBinHOGTar(float* outputHOGFeature,int *outputPosition,float* Gr,int* Tag,unsigned int xStepNum,unsigned int yStepNum, unsigned int width,unsigned int height) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; if(tx < xStepNum && ty < yStepNum) { //pixel position of center int x = tx*STEPSIZE+1.5*CELLSIZE-1; int y = ty*STEPSIZE+1.5*CELLSIZE-1; int leftupperX = x - 1.5*CELLSIZE +1 ; int leftupperY = y - 1.5*CELLSIZE +1 ; //__shared__ float tmp[CELLSIZE][CELLSIZE]; float tmp[CELLBLOCKSIZE*CELLBLOCKSIZE][BINNUM] = {0.0f}; int inc = 0; for(int i = 0; i < CELLBLOCKSIZE;i++) { for(int j=0; j<CELLBLOCKSIZE;j++) { int indX = leftupperX + j*CELLSIZE; int indY = leftupperY + i*CELLSIZE; for(int p=0;p<CELLSIZE;p++) { for(int q=0;q<CELLSIZE;q++) { //int currendindX = indX+p; //int currendindY = indY+q; int binind = Tag[indX+q + width*(indY+p)]-1; tmp[inc][binind] += Gr[indX+q + width*(indY+p)]; } } inc++; } } float norm = 0.0f; int nonempty = 0; for(int p=0;p<CELLBLOCKSIZE*CELLBLOCKSIZE;p++) { for(int q=0;q<BINNUM;q++) { norm+= (tmp[p][q]*tmp[p][q]); if(tmp[p][q]>0) nonempty ++; } } norm = sqrt(norm); norm +=1e-5; //if(norm <= 0.0f) // norm = 1e-5; for(int p=0;p<CELLBLOCKSIZE*CELLBLOCKSIZE;p++) { for(int q=0;q<BINNUM;q++) { //outputHOGFeature[ (tx+ty*xStepNum)*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM +(q+p*BINNUM)] = tmp[p][q]/norm; outputHOGFeature[ (ty+tx*yStepNum)*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM +(q+p*BINNUM)] = tmp[p][q]/norm; } } //outputPosition[(tx + ty*xStepNum)*2] = x; //outputPosition[(tx + ty*xStepNum)*2 + 1] = y; outputPosition[(ty+tx*yStepNum)*2] = x; outputPosition[(ty+tx*yStepNum)*2 + 1] = y; } } __global__ void computeDistance(float* output,float* hog,float* hogTar,unsigned int width,unsigned int height) { /*__shared__ float tmp[2][81]; unsigned int bx = blockIdx.x * 1; unsigned int by = blockIdx.y * 1; unsigned int tx = threadIdx.x * 1; if(tx < 81) { tmp[0][tx] = hog[tx + bx*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM]; tmp[1][tx] = hogTar[tx + by*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM]; } __syncthreads(); float perDistance=0.0; float tmp1 = 0.0; float tmp2 = 0.0; for(int i=0;i<81;i++) { tmp1 = tmp[0][i]; tmp2 = tmp[1][i]; perDistance += ((tmp1-tmp2)*(tmp1-tmp2))/(tmp1 + tmp2 + 1e-5); } output[bx + by*width] = 0.5 * perDistance;*/ unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; if(tx < width && ty < height) { float perDistance = 0.0; //float tmp[CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM]= {0.0f}; //float tmpTar[CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM]= {0.0f}; float tmp = 0.0; float tmpTar = 0.0; //int i=0; for(int i =0;i<CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM;i++) { tmp = hog[i + tx*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM]; tmpTar = hogTar[i + ty*CELLBLOCKSIZE*CELLBLOCKSIZE*BINNUM]; //float tmp = hog[i*(width) + tx]; //float tmpTar = hogTar[i*height + ty]; perDistance = perDistance + ((tmp-tmpTar)*(tmp-tmpTar))/(tmp + tmpTar + 1e-5); //perDistance=1; } output[tx + ty*width] = 0.5 * perDistance; //output[ty + tx*height] = 0.5 * perDistance; } } __global__ void vote(float *OutputVoteMatrix,float *HOGDistance, float *weight,int *offset,int *samplepixel, unsigned int width,unsigned int height,unsigned int voteWidth,unsigned int voteHeight) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; if(tx < width && ty < height) { //if (HOGDistance[tx + ty*width] <= threshold[ty]) if (HOGDistance[tx + ty*width] <= 2)// { int x = samplepixel[ty*2]; int y = samplepixel[ty*2+1]; int offsetx = offset[tx*2]; int offsety = offset[tx*2+1]; int centerx = x + offsetx; int centery = y + offsety; if(centerx >= 0 && centery >= 0 && centerx < voteWidth && centery< voteHeight) { OutputVoteMatrix[centerx + centery*voteWidth] += 1/(HOGDistance[tx + ty*width]+1e-5 + 1) * weight[tx]; } } } } __global__ void genGaussianFilter(float* output) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; __shared__ float gaussianKernel[KERNEL_WIDTH][KERNEL_WIDTH]; float sigma = 5.0; float gaussianSum = 0.0; gaussianKernel[ty][tx] = powf(2.71828,-((tx-15)*(tx-15)/(2*sigma*sigma) + (ty-15)*(ty-15)/(2*sigma*sigma))); __syncthreads(); for(int i=0;i<KERNEL_WIDTH;i++) for(int j=0;j<KERNEL_WIDTH;j++) { gaussianSum += gaussianKernel[i][j]; } output[ty*KERNEL_WIDTH + tx] = gaussianKernel[ty][tx]/gaussianSum; } __global__ void gaussianFilter(float* output,float* input, float* gaussianKernel,unsigned int width,unsigned int height) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; if(tx < width && ty < height) { float accum = 0.0; float offset = 15; for (int i = 0; i < KERNEL_WIDTH; ++i) { for (int j = 0; j < KERNEL_WIDTH; ++j) { int ind = tx+j-offset + (ty+i-offset) * width; if(ind > 0 && ind < width*height) accum += gaussianKernel[j + KERNEL_WIDTH*i] * input[ind]; } output[tx + ty*width] = accum; } } } __global__ void detectMaximal(float* output,float* input,unsigned int width,unsigned int height,float maximal) { unsigned int tx = blockIdx.x*blockDim.x + threadIdx.x; unsigned int ty = blockIdx.y*blockDim.y + threadIdx.y; if(tx < width-1 && ty < height-1 && tx >0 && ty>0) { if (input[tx + ty*width] > 0.8*maximal && input[tx + ty*width] > input[tx-1 + (ty-1)*width] && input[tx + ty*width] > input[tx + (ty-1)*width] && input[tx + ty*width] > input[tx+1 + (ty-1)*width] && input[tx + ty*width] > input[tx-1 + (ty)*width] && input[tx + ty*width] > input[tx+1 + (ty)*width] && input[tx + ty*width] > input[tx-1 + (ty+1)*width] && input[tx + ty*width] > input[tx + (ty+1)*width] && input[tx + ty*width] > input[tx+1 + (ty+1)*width]) { output[tx + ty*width] = (tx+1) + (ty)*width; } } } __global__ void computeThreshold(float* output,float* input) { unsigned int tx = threadIdx.x; unsigned int bx = blockIdx.x; __shared__ float partialMin[500]; partialMin[tx] = input[tx + bx*blockDim.x]; unsigned int stride = 1; for (stride = 1; stride < blockDim.x; stride *= 2) { __syncthreads(); if (tx % (2 * stride) == 0 && tx + stride < blockDim.x) partialMin[tx] = partialMin[tx] < partialMin[tx + stride] ? partialMin[tx]:partialMin[tx + stride]; } /*unsigned int stridenew = stride/2; float kmax = partialMin[0]; for(int i=0;i<blockDim.x;i+=stridenew) { if(kmax < partialMin[i]) kmax = partialMin[i]; }*/ output[0] = partialMin[0]; } extern "C" float* computeHOG(float *img,float* imgTar,unsigned int width,unsigned int height,unsigned int widthTar,unsigned int heightTar) { //For performance analysis hipEvent_t start, stop0,stop1, stop2, stop3, stop4, stop5, stop6,stop7,stop8,stop9,stop10,stop11,stop12,stop13,stop14; hipEventCreate(&start); hipEventCreate(&stop0); hipEventCreate(&stop1); hipEventCreate(&stop2); hipEventCreate(&stop3); hipEventCreate(&stop4); hipEventCreate(&stop5); hipEventCreate(&stop6); hipEventCreate(&stop7); hipEventCreate(&stop8); hipEventCreate(&stop9); hipEventCreate(&stop10); hipEventCreate(&stop11); hipEventCreate(&stop12); hipEventCreate(&stop13); hipEventCreate(&stop14); hipEventRecord(start, 0); int xStepNum = floor((width-CELLSIZE*CELLBLOCKSIZE)/float(STEPSIZE)); int yStepNum = floor((height-CELLSIZE*CELLBLOCKSIZE)/float(STEPSIZE)); int xStepNumTar = floor((widthTar-CELLSIZE*CELLBLOCKSIZE)/float(STEPSIZE)); int yStepNumTar = floor((heightTar-CELLSIZE*CELLBLOCKSIZE)/float(STEPSIZE)); float *Gr; hipMalloc((void**)&Gr , width*height*sizeof(float)); int *Tag; hipMalloc((void**)&Tag , width*height*sizeof(int)); float *HOGFeature; hipMalloc((void**)&HOGFeature , xStepNum*yStepNum*BINNUM*CELLBLOCKSIZE*CELLBLOCKSIZE*sizeof(float)); hipMemset(HOGFeature,0,xStepNum*yStepNum*BINNUM*CELLBLOCKSIZE*CELLBLOCKSIZE*sizeof(float)); int *centerPosition; hipMalloc((void**)&centerPosition , 2*xStepNum*yStepNum*sizeof(int)); int *offset; hipMalloc((void**)&offset , 2*xStepNum*yStepNum*sizeof(int)); float *weight; hipMalloc((void**)&weight , xStepNum*yStepNum*sizeof(float)); float *GrTar; hipMalloc((void**)&GrTar , widthTar*heightTar*sizeof(float)); int *TagTar; hipMalloc((void**)&TagTar , widthTar*heightTar*sizeof(int)); int *samplepixel; hipMalloc((void**)&samplepixel , 2*xStepNum*yStepNum*sizeof(int)); float *HOGFeatureTar; hipMalloc((void**)&HOGFeatureTar , xStepNumTar*yStepNumTar*BINNUM*CELLBLOCKSIZE*CELLBLOCKSIZE*sizeof(float)); hipMemset(HOGFeatureTar,0,xStepNumTar*yStepNumTar*BINNUM*CELLBLOCKSIZE*CELLBLOCKSIZE*sizeof(float)); float *HOGDistance; hipMalloc((void**)&HOGDistance , xStepNumTar*yStepNumTar*xStepNum*yStepNum*sizeof(float)); hipMemset(HOGDistance,0,xStepNumTar*yStepNumTar*xStepNum*yStepNum*sizeof(float)); float *votematrix; hipMalloc((void**)&votematrix , widthTar*heightTar* sizeof(float)); hipMemset(votematrix,0,widthTar*heightTar*sizeof(float)); float *gaussianKernel_dev; hipMalloc((void**)&gaussianKernel_dev , KERNEL_WIDTH*KERNEL_WIDTH* sizeof(float)); hipMemset(gaussianKernel_dev,0,KERNEL_WIDTH*KERNEL_WIDTH*sizeof(float)); float *votematrix_smoothed; hipMalloc((void**)&votematrix_smoothed , widthTar*heightTar* sizeof(float)); hipMemset(votematrix_smoothed,0,widthTar*heightTar*sizeof(float)); float *vote_maxposition; hipMalloc((void**)&vote_maxposition , widthTar*heightTar* sizeof(float)); hipMemset(vote_maxposition,0,widthTar*heightTar*sizeof(float)); //******************** hipEventRecord(stop0, 0); hipEventSynchronize(stop0); float elapsedTime0; hipEventElapsedTime(&elapsedTime0, start, stop0); printf("Initialization :%f ms\n", elapsedTime0); //******************** dim3 dimBlock(16,16,1); dim3 dimGrid(width/dimBlock.x + 1, height/dimBlock.y + 1,1); computeGradient<<<dimGrid,dimBlock>>>(Gr,Tag,img,width,height); //******************** hipEventRecord(stop1, 0); hipEventSynchronize(stop1); float elapsedTime1; hipEventElapsedTime(&elapsedTime1, start, stop1); printf("Gradient Computation for Ref Image :%f ms\n", elapsedTime1); //******************** dimBlock = dim3(16,16,1); dimGrid = dim3(xStepNum/dimBlock.x + 1, yStepNum/dimBlock.y + 1,1); computeBinHOG<<<dimGrid,dimBlock>>>(HOGFeature,weight,centerPosition,offset,Gr,Tag,xStepNum,yStepNum,width,height); //******************** hipEventRecord(stop2, 0); hipEventSynchronize(stop2); float elapsedTime2; hipEventElapsedTime(&elapsedTime2, start, stop2); printf("HOG Computation for Ref Image :%f ms\n", elapsedTime2); //******************** dimBlock = dim3(16,16,1); dimGrid = dim3(widthTar/dimBlock.x + 1, heightTar/dimBlock.y + 1,1); computeGradient<<<dimGrid,dimBlock>>>(GrTar,TagTar,imgTar,widthTar,heightTar); //******************** hipEventRecord(stop3, 0); hipEventSynchronize(stop3); float elapsedTime3; hipEventElapsedTime(&elapsedTime3, start, stop3); printf("Gradient Computation for Target Image :%f ms\n", elapsedTime3); //******************** dimBlock = dim3(16,16,1); dimGrid = dim3(xStepNumTar/dimBlock.x + 1, yStepNumTar/dimBlock.y + 1,1); computeBinHOGTar<<<dimGrid,dimBlock>>>(HOGFeatureTar,samplepixel,GrTar,TagTar,xStepNumTar,yStepNumTar,widthTar,heightTar); //******************** hipEventRecord(stop4, 0); hipEventSynchronize(stop4); float elapsedTime4; hipEventElapsedTime(&elapsedTime4, start, stop4); printf("HOG Computation for Target Image :%f ms\n", elapsedTime4); //******************** dimBlock = dim3(16,16,1); dimGrid = dim3((xStepNum*yStepNum)/dimBlock.x + 1, (xStepNumTar*yStepNumTar)/dimBlock.y + 1,1); computeDistance<<<dimGrid,dimBlock>>>(HOGDistance,HOGFeature,HOGFeatureTar,xStepNum*yStepNum,xStepNumTar*yStepNumTar); //******************** hipEventRecord(stop5, 0); hipEventSynchronize(stop5); float elapsedTime5; hipEventElapsedTime(&elapsedTime5, start, stop5); printf("Distance Matrix Computation:%f ms\n", elapsedTime5); //******************** //float *HOGDistanceSorted = new float[xStepNumTar*yStepNumTar*xStepNum*yStepNum]; //cudaMemcpy(HOGDistanceSorted,HOGDistance, xStepNumTar*yStepNumTar*xStepNum*yStepNum*sizeof(float), cudaMemcpyDeviceToHost); /*float *threshold_host = new float[xStepNumTar*yStepNumTar]; for(int i=0;i<xStepNumTar*yStepNumTar;i++) { thrust::sort(HOGDistanceSorted + i*(xStepNum*yStepNum), HOGDistanceSorted +i*(xStepNum*yStepNum) +xStepNum*yStepNum); threshold_host[i] = HOGDistanceSorted[i*(xStepNum*yStepNum) + K - 1]; }*/ /*float *threshold_dev; cudaMalloc((void**)&threshold_dev , xStepNumTar*yStepNumTar* sizeof(float)); cudaMemcpy(threshold_dev,threshold_host, xStepNumTar*yStepNumTar*sizeof(float), cudaMemcpyHostToDevice);*/ //cudaMemcpy(threshold_dev,threshold_host, xStepNumTar*yStepNumTar*sizeof(float), cudaMemcpyHostToDevice); /*dimBlock = dim3(xStepNum*yStepNum,1,1); dimGrid = dim3(xStepNumTar*yStepNumTar,1,1); float *threshold_dev; cudaMalloc((void**)&threshold_dev , xStepNumTar*yStepNumTar* sizeof(float)); cudaMemset(threshold_dev,0,xStepNumTar*yStepNumTar*sizeof(float)); computeThreshold<<<dimGrid,dimBlock>>>(threshold_dev,HOGDistance);*/ dimBlock = dim3(16,16,1); dimGrid = dim3((xStepNum*yStepNum)/dimBlock.x + 1, (xStepNumTar*yStepNumTar)/dimBlock.y + 1,1); //vote<<<dimGrid,dimBlock>>>(votematrix,threshold_dev,HOGDistance,weight,offset,samplepixel,xStepNum*yStepNum,xStepNumTar*yStepNumTar,widthTar,heightTar); vote<<<dimGrid,dimBlock>>>(votematrix,HOGDistance,weight,offset,samplepixel,xStepNum*yStepNum,xStepNumTar*yStepNumTar,widthTar,heightTar); //******************** hipEventRecord(stop6, 0); hipEventSynchronize(stop6); float elapsedTime6; hipEventElapsedTime(&elapsedTime6, start, stop6); printf("Threshold and Voting:%f ms\n", elapsedTime6); //******************** dimBlock = dim3(KERNEL_WIDTH,KERNEL_WIDTH,1); dimGrid = dim3(1,1,1); genGaussianFilter<<<dimGrid,dimBlock>>>(gaussianKernel_dev); //******************** hipEventRecord(stop7, 0); hipEventSynchronize(stop7); float elapsedTime7; hipEventElapsedTime(&elapsedTime7, start, stop7); printf("Gaussian Kernel Generation:%f ms\n", elapsedTime7); //******************** dimBlock = dim3(16,16,1); dimGrid = dim3(widthTar/dimBlock.x + 1, heightTar/dimBlock.y + 1,1); gaussianFilter<<<dimGrid,dimBlock>>>(votematrix_smoothed,votematrix, gaussianKernel_dev,widthTar,heightTar); //******************** hipEventRecord(stop8, 0); hipEventSynchronize(stop8); float elapsedTime8; hipEventElapsedTime(&elapsedTime8, start, stop8); printf("Gaussian Blur:%f ms\n", elapsedTime8); //******************** float *votematrix_smoothed_sorted = new float[widthTar*heightTar]; hipMemcpy(votematrix_smoothed_sorted,votematrix_smoothed, widthTar*heightTar*sizeof(float), hipMemcpyDeviceToHost); thrust::sort(votematrix_smoothed_sorted, votematrix_smoothed_sorted +widthTar*heightTar); float maxvalue = votematrix_smoothed_sorted[widthTar*heightTar-1]; //******************** hipEventRecord(stop9, 0); hipEventSynchronize(stop9); float elapsedTime9; hipEventElapsedTime(&elapsedTime9, start, stop9); printf("Maximal Voting Value Computation:%f ms\n", elapsedTime9); //******************** dimBlock = dim3(16,16,1); dimGrid = dim3(widthTar/dimBlock.x + 1, heightTar/dimBlock.y + 1,1); detectMaximal<<<dimGrid,dimBlock>>>(vote_maxposition,votematrix_smoothed,widthTar,heightTar,maxvalue); //******************** hipEventRecord(stop10, 0); hipEventSynchronize(stop10); float elapsedTime10; hipEventElapsedTime(&elapsedTime10, start, stop10); printf("Local Maximal Detection:%f ms\n", elapsedTime10); //******************** float *vote_maxposition_sorted = new float[widthTar*heightTar]; hipMemcpy(vote_maxposition_sorted,vote_maxposition, widthTar*heightTar*sizeof(float), hipMemcpyDeviceToHost); thrust::sort(vote_maxposition_sorted, vote_maxposition_sorted +widthTar*heightTar); int centernum = 0; for(int i=widthTar*heightTar-1;i>=0;i--) { if(vote_maxposition_sorted[i] > 0) centernum++; else break; } //******************** hipEventRecord(stop11, 0); hipEventSynchronize(stop11); float elapsedTime11; hipEventElapsedTime(&elapsedTime11, start, stop11); printf("Transfering Detected Logo Position to HOST:%f ms\n", elapsedTime11); //******************** //float *ret = new float[centernum]; //cudaMemcpy(ret,vote_maxposition_sorted + widthTar*heightTar - centernum, centernum*sizeof(float), cudaMemcpyHostToHost); float *ret = new float[centernum+1]; ret[0] = float(centernum); hipMemcpy(ret+1,vote_maxposition_sorted + widthTar*heightTar - centernum, centernum*sizeof(float), hipMemcpyHostToHost); //float *ret = new float[widthTar*heightTar]; //cudaMemcpy(ret,votematrix_smoothed, widthTar*heightTar*sizeof(float), cudaMemcpyDeviceToHost); return ret; }
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
#include <cuda_runtime.h> #include <stdio.h> #include <string.h> #include <stdlib.h> #include <math.h> #include <time.h> #define ELEMENT_MIN 0 #define ELEMENT_MAX 10 #define BLOCK_SIZE 16 #define TILE_SIZE 16 #define ZERO 1.e-6 int UI(int argc, char* argv[], int* jkl); float randGenerate(int min, int max); void initMatrix(float* matrix, int size, int min, int max); void matrixMulCPU(float* A, float* B, float* C, int j, int k, int l); __global__ void matrixMulGPU(float* d_A, float* d_B, float* d_C, int j, int k, int l); void showMatrix(float* matrix, int row, int col); void resultCheck(float* result_CPU, float* result_GPU, int size); int main(int argc, char* argv[]){ // reset rand seed srand((unsigned)time(NULL)); clock_t start, finish; int total_time; // Go through UI first. // In UI section, only command with valid param can go to the next step. int UIStatus; int jkl[3]; UIStatus = UI(argc, argv, jkl); if (UIStatus != 0) { printf("\nApplication terminates."); return 0; } printf("\nContinuing with j=%d, k=%d, l=%d", jkl[0], jkl[1], jkl[2]); // UI section ends // Initialize these two matrix A, B with random float type const int j = jkl[0]; const int k = jkl[1]; const int l = jkl[2]; float* A = (float*)malloc(j * k * sizeof(float)); float* B = (float*)malloc(k * l * sizeof(float)); initMatrix(A, (j * k), ELEMENT_MIN, ELEMENT_MAX); initMatrix(B, (k * l), ELEMENT_MIN, ELEMENT_MAX); showMatrix(A, j, k); showMatrix(B, k, l); printf("\nDone initializing matrix A with size %d*%d and matrix B with %d*%d", j, k, k, l); // Initialzing ends // CPU code for calculating the matrix multiplication // Use this result to varify the kernel result later // Only do this step when j, k, l is below 5000 float* C_GPU = (float*)malloc(j * l * sizeof(float)); float* C_CPU = (float*)malloc(j * l * sizeof(float)); if (k < 5000 && j < 5000 && l < 5000) { start = clock(); matrixMulCPU(A, B, C_CPU, j, k, l); finish = clock(); printf("\nDone matrix multiplication with CPU"); total_time = (int)(finish - start); printf("\n%d microseconds used.\n", total_time); showMatrix(C_CPU, j, l); } // Matrix multipication with CPU ends // Allocate device memory and copy data from host to device float *d_A, *d_B, *d_C; cudaMalloc((float**)&d_A, j * k * sizeof(float)); cudaMalloc((float**)&d_B, k * l * sizeof(float)); cudaMalloc((float**)&d_C, j * l * sizeof(float)); printf("\nDone allocating space in device."); cudaMemcpy(d_A, A, j * k * sizeof(float), cudaMemcpyHostToDevice); cudaMemcpy(d_B, B, k * l * sizeof(float), cudaMemcpyHostToDevice); printf("\nDone copying memory from host to device"); // Done allocating and transfering // Initialize thread block and kernel grid dimensions dim3 threads(BLOCK_SIZE, BLOCK_SIZE); dim3 grid((int)ceil(1.0 * l / threads.x), (int)ceil(1.0 * j / threads.y)); printf("\nDone initializing block dimention and grid dimention."); // Done initializing thread block and kernel grid dimensions // launch CUDA device kernel start = clock(); matrixMulGPU<<< grid, threads>>>(d_A, d_B, d_C, j, k, l); // Done CUDA device kernel // Copy results from device to host and free device memory cudaMemcpy(C_GPU, d_C, j * l * sizeof(float), cudaMemcpyDeviceToHost); finish = clock(); printf("\nDone matrix multiplication with GPU."); total_time = (int)(finish - start); printf("\n%d microseconds used.\n", total_time); showMatrix(C_GPU, j, l); cudaFree(d_A); cudaFree(d_B); cudaFree(d_C); // Done copying results and freeing device memory // Check the result of the Calculated Matrix if (k < 5000 && j < 5000 && l < 5000) resultCheck(C_CPU, C_GPU, j * l); // Done result checking. return 0; } // UI for main function // return 0 means everything's fine, just continue; // return 1 means there's invalid input or '--help', terminate running. int UI(int argc, char* argv[], int* jkl) { // UI for the exe file // while input with ? or -h or --help; tell that what we need as params // while input with [1,2] || [4, +inf) parameters; tell that we need 3 params // while input with 3 paramerters; print the size of two input matrix; check if all params are valid; if (argc == 2 && (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-h") == 0 || strcmp(argv[1], "?") == 0)) { printf("\nWe need 3 parameters j,k,l which represents the two matrix's dimention.\n"); printf("The first one with size j*k, and the second one's is k*l.\n"); printf("j,k,l must be int type and greater than 0.\n"); return 1; } if (argc <= 3 || argc >= 5) { printf("\nError.\nWe need 3 parameters as input.\nYou can use '--help' for more details.\n"); return 1; } if (argc == 4) { int n; printf("\nNumber Of Arguments Passed: %d", argc); printf("\n----Following Are The Command Line Arguments Passed----"); for (int counter = 1; counter < argc; counter++) { n = atoi(argv[counter]); if (n > 0) { printf("\nargv[%d]: %d", counter, n); jkl[counter - 1] = n; } else { printf("\nargv[%d] is invalid", counter); return 1; } } } return 0; } // random float generator within min and max float randGenerate(int min, int max) { float res; int r = rand(); res = 1.0 * rand() / RAND_MAX * (max-min) + min; return res; } // initialize matrix with float type void initMatrix(float* matrix, int size, int min, int max) { for (int i = 0; i < size; ++i) matrix[i] = randGenerate(min, max); return; } // matrix multiplication with CPU in the most stupid algo // Algo Complexity: O((2k-1)*j*l) void matrixMulCPU(float* A, float* B, float* C, int j, int k, int l) { for (int row = 0; row < j; ++row) { for (int col = 0; col < l; ++col) { float eleSum = 0.0f; for (int i = 0; i < k; ++i) { eleSum += A[row * k + i] * B[i * l + col]; } C[row * l + col] = eleSum; } } } // matrix multiplication with GPU device // using tiling algorithm, take use of the shared memory to higher the compute-to-global-memory-access ratio __global__ void matrixMulGPU(float* d_A, float* d_B, float* d_C, int j, int k, int l) { __shared__ float Ads[TILE_SIZE][TILE_SIZE]; __shared__ float Bds[TILE_SIZE][TILE_SIZE]; int bx = blockIdx.x; int by = blockIdx.y; int tx = threadIdx.x; int ty = threadIdx.y; int row = by * blockDim.y + ty; int col = bx * blockDim.x + tx; float CEleValue = 0.0f; for (int ph = 0; ph * TILE_SIZE < k; ++ph) { if (row < j && ph * TILE_SIZE + tx < k) Ads[ty][tx] = d_A[row * k + ph * TILE_SIZE + tx]; else Ads[ty][tx] = 0; if (col < l && ph * TILE_SIZE + ty < k) Bds[ty][tx] = d_B[(ph * TILE_SIZE + ty) * l + col]; else Bds[ty][tx] = 0; __syncthreads(); for (int i = 0; i < TILE_SIZE; ++i) { CEleValue += Ads[ty][i] * Bds[i][tx]; } __syncthreads(); } if (row < j && col < l) d_C[row * l + col] = CEleValue; } // show the data in the command prompt. // this function is used for configuration // please don't use it when j,k,l is too large void showMatrix(float* matrix, int row, int col) { if (row > 5 || col > 5) return; printf("\n"); for (int i = 0; i < row; i++) { for (int j = 0; j < col; j++) { printf("%f\t", matrix[i*col + j]); } printf("\n"); } return; } // check if two matrix is the same // 0 is defined at the very begining void resultCheck(float* result_CPU, float* result_GPU, int size) { // using relative error // assuming CPU result is the true value for (int i = 0; i < size; ++i) { if ((result_CPU[i] - result_GPU[i])/ result_CPU[i] > ZERO || (result_GPU[i] - result_CPU[i])/result_CPU[i] < -ZERO) { printf("\nResult check: Error!!!! Didn't pass."); return; } } printf("\nResult check: ---PASS---."); return; }
code for sm_80 Function : _Z12matrixMulGPUPfS_S_iii .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R3, SR_CTAID.X ; /* 0x0000000000037919 */ /* 0x000e220000002500 */ /*0020*/ MOV R4, c[0x0][0x17c] ; /* 0x00005f0000047a02 */ /* 0x000fe20000000f00 */ /*0030*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*0040*/ HFMA2.MMA R9, -RZ, RZ, 0, 0 ; /* 0x00000000ff097435 */ /* 0x000fe200000001ff */ /*0050*/ S2R R16, SR_TID.X ; /* 0x0000000000107919 */ /* 0x000e220000002100 */ /*0060*/ ISETP.GE.AND P1, PT, R4, 0x1, PT ; /* 0x000000010400780c */ /* 0x000fc60003f26270 */ /*0070*/ S2R R5, SR_CTAID.Y ; /* 0x0000000000057919 */ /* 0x000e680000002600 */ /*0080*/ S2R R2, SR_TID.Y ; /* 0x0000000000027919 */ /* 0x000e620000002200 */ /*0090*/ IMAD R3, R3, c[0x0][0x0], R16 ; /* 0x0000000003037a24 */ /* 0x001fca00078e0210 */ /*00a0*/ ISETP.GE.AND P0, PT, R3, c[0x0][0x180], PT ; /* 0x0000600003007a0c */ /* 0x000fe20003f06270 */ /*00b0*/ IMAD R0, R5, c[0x0][0x4], R2 ; /* 0x0000010005007a24 */ /* 0x002fca00078e0202 */ /*00c0*/ ISETP.GE.OR P0, PT, R0, c[0x0][0x178], P0 ; /* 0x00005e0000007a0c */ /* 0x000fe20000706670 */ /*00d0*/ @!P1 BRA 0x4f0 ; /* 0x0000041000009947 */ /* 0x000fd80003800000 */ /*00e0*/ SHF.L.U32 R17, R2, 0x6, RZ ; /* 0x0000000602117819 */ /* 0x000fe400000006ff */ /*00f0*/ MOV R5, RZ ; /* 0x000000ff00057202 */ /* 0x000fe40000000f00 */ /*0100*/ MOV R18, RZ ; /* 0x000000ff00127202 */ /* 0x000fe40000000f00 */ /*0110*/ MOV R9, RZ ; /* 0x000000ff00097202 */ /* 0x000fe40000000f00 */ /*0120*/ LEA R19, R16, R17, 0x2 ; /* 0x0000001110137211 */ /* 0x000fe400078e10ff */ /*0130*/ IADD3 R7, R16, R5.reuse, RZ ; /* 0x0000000510077210 */ /* 0x080fe20007ffe0ff */ /*0140*/ HFMA2.MMA R8, -RZ, RZ, 0, 0 ; /* 0x00000000ff087435 */ /* 0x000fe200000001ff */ /*0150*/ IADD3 R10, R2, R5, RZ ; /* 0x00000005020a7210 */ /* 0x000fc40007ffe0ff */ /*0160*/ ISETP.GE.AND P2, PT, R7, c[0x0][0x17c], PT ; /* 0x00005f0007007a0c */ /* 0x000fe40003f46270 */ /*0170*/ ISETP.GE.AND P1, PT, R10, c[0x0][0x17c], PT ; /* 0x00005f000a007a0c */ /* 0x000fe40003f26270 */ /*0180*/ ISETP.GE.OR P2, PT, R0, c[0x0][0x178], P2 ; /* 0x00005e0000007a0c */ /* 0x000fe40001746670 */ /*0190*/ ISETP.GE.OR P1, PT, R3, c[0x0][0x180], P1 ; /* 0x0000600003007a0c */ /* 0x000fe40000f26670 */ /*01a0*/ MOV R20, RZ ; /* 0x000000ff00147202 */ /* 0x000fd20000000f00 */ /*01b0*/ @!P2 MOV R5, 0x4 ; /* 0x000000040005a802 */ /* 0x000fe20000000f00 */ /*01c0*/ @!P2 IMAD R4, R0, c[0x0][0x17c], R7 ; /* 0x00005f000004aa24 */ /* 0x000fe200078e0207 */ /*01d0*/ @!P1 MOV R11, 0x4 ; /* 0x00000004000b9802 */ /* 0x000fe20000000f00 */ /*01e0*/ @!P1 IMAD R10, R10, c[0x0][0x180], R3 ; /* 0x000060000a0a9a24 */ /* 0x000fe400078e0203 */ /*01f0*/ @!P2 IMAD.WIDE R4, R4, R5, c[0x0][0x160] ; /* 0x000058000404a625 */ /* 0x000fc800078e0205 */ /*0200*/ @!P1 IMAD.WIDE R10, R10, R11, c[0x0][0x168] ; /* 0x00005a000a0a9625 */ /* 0x000fe200078e020b */ /*0210*/ @!P2 LDG.E R8, [R4.64] ; /* 0x000000040408a981 */ /* 0x000ea8000c1e1900 */ /*0220*/ @!P1 LDG.E R20, [R10.64] ; /* 0x000000040a149981 */ /* 0x000ee2000c1e1900 */ /*0230*/ IADD3 R18, R18, 0x1, RZ ; /* 0x0000000112127810 */ /* 0x000fc60007ffe0ff */ /*0240*/ STS [R19], R8 ; /* 0x0000000813007388 */ /* 0x004fe80000000800 */ /*0250*/ STS [R19+0x400], R20 ; /* 0x0004001413007388 */ /* 0x008fe80000000800 */ /*0260*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0270*/ LDS R23, [R16.X4+0x400] ; /* 0x0004000010177984 */ /* 0x000fe80000004800 */ /*0280*/ LDS.128 R12, [R17] ; /* 0x00000000110c7984 */ /* 0x000e280000000c00 */ /*0290*/ LDS R24, [R16.X4+0x440] ; /* 0x0004400010187984 */ /* 0x000e680000004800 */ /*02a0*/ LDS R25, [R16.X4+0x480] ; /* 0x0004800010197984 */ /* 0x000ea80000004800 */ /*02b0*/ LDS R26, [R16.X4+0x4c0] ; /* 0x0004c000101a7984 */ /* 0x000ee80000004800 */ /*02c0*/ LDS R27, [R16.X4+0x500] ; /* 0x00050000101b7984 */ /* 0x000fe80000004800 */ /*02d0*/ LDS.128 R4, [R17+0x10] ; /* 0x0000100011047984 */ /* 0x000f280000000c00 */ /*02e0*/ LDS R22, [R16.X4+0x540] ; /* 0x0005400010167984 */ /* 0x000f680000004800 */ /*02f0*/ LDS R21, [R16.X4+0x580] ; /* 0x0005800010157984 */ /* 0x000f680000004800 */ /*0300*/ LDS R20, [R16.X4+0x5c0] ; /* 0x0005c00010147984 */ /* 0x000f620000004800 */ /*0310*/ FFMA R12, R23, R12, R9 ; /* 0x0000000c170c7223 */ /* 0x001fc60000000009 */ /*0320*/ LDS R23, [R16.X4+0x600] ; /* 0x0006000010177984 */ /* 0x000fe20000004800 */ /*0330*/ FFMA R12, R24, R13, R12 ; /* 0x0000000d180c7223 */ /* 0x002fc6000000000c */ /*0340*/ LDS.128 R8, [R17+0x20] ; /* 0x0000200011087984 */ /* 0x000e220000000c00 */ /*0350*/ FFMA R12, R25, R14, R12 ; /* 0x0000000e190c7223 */ /* 0x004fc6000000000c */ /*0360*/ LDS R24, [R16.X4+0x640] ; /* 0x0006400010187984 */ /* 0x000e620000004800 */ /*0370*/ FFMA R12, R26, R15, R12 ; /* 0x0000000f1a0c7223 */ /* 0x008fc6000000000c */ /*0380*/ LDS R25, [R16.X4+0x680] ; /* 0x0006800010197984 */ /* 0x000ea20000004800 */ /*0390*/ FFMA R12, R27, R4, R12 ; /* 0x000000041b0c7223 */ /* 0x010fc6000000000c */ /*03a0*/ LDS R4, [R16.X4+0x6c0] ; /* 0x0006c00010047984 */ /* 0x000ee20000004800 */ /*03b0*/ FFMA R22, R22, R5, R12 ; /* 0x0000000516167223 */ /* 0x020fc6000000000c */ /*03c0*/ LDS R5, [R16.X4+0x700] ; /* 0x0007000010057984 */ /* 0x000fe20000004800 */ /*03d0*/ FFMA R26, R21, R6, R22 ; /* 0x00000006151a7223 */ /* 0x000fc60000000016 */ /*03e0*/ LDS.128 R12, [R17+0x30] ; /* 0x00003000110c7984 */ /* 0x000f220000000c00 */ /*03f0*/ FFMA R7, R20, R7, R26 ; /* 0x0000000714077223 */ /* 0x000fc6000000001a */ /*0400*/ LDS R22, [R16.X4+0x740] ; /* 0x0007400010167984 */ /* 0x000f680000004800 */ /*0410*/ LDS R21, [R16.X4+0x780] ; /* 0x0007800010157984 */ /* 0x000f680000004800 */ /*0420*/ LDS R6, [R16.X4+0x7c0] ; /* 0x0007c00010067984 */ /* 0x000f620000004800 */ /*0430*/ FFMA R7, R23, R8, R7 ; /* 0x0000000817077223 */ /* 0x001fc80000000007 */ /*0440*/ FFMA R7, R24, R9, R7 ; /* 0x0000000918077223 */ /* 0x002fc80000000007 */ /*0450*/ FFMA R7, R25, R10, R7 ; /* 0x0000000a19077223 */ /* 0x004fc80000000007 */ /*0460*/ FFMA R4, R4, R11, R7 ; /* 0x0000000b04047223 */ /* 0x008fc80000000007 */ /*0470*/ FFMA R4, R5, R12, R4 ; /* 0x0000000c05047223 */ /* 0x010fe20000000004 */ /*0480*/ SHF.L.U32 R5, R18, 0x4, RZ ; /* 0x0000000412057819 */ /* 0x000fc600000006ff */ /*0490*/ FFMA R4, R22, R13, R4 ; /* 0x0000000d16047223 */ /* 0x020fe20000000004 */ /*04a0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fe20000010000 */ /*04b0*/ ISETP.GE.AND P1, PT, R5, c[0x0][0x17c], PT ; /* 0x00005f0005007a0c */ /* 0x000fe40003f26270 */ /*04c0*/ FFMA R4, R21, R14, R4 ; /* 0x0000000e15047223 */ /* 0x000fc80000000004 */ /*04d0*/ FFMA R9, R6, R15, R4 ; /* 0x0000000f06097223 */ /* 0x000fce0000000004 */ /*04e0*/ @!P1 BRA 0x130 ; /* 0xfffffc4000009947 */ /* 0x000fea000383ffff */ /*04f0*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0500*/ HFMA2.MMA R2, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff027435 */ /* 0x000fe200000001ff */ /*0510*/ IMAD R3, R0, c[0x0][0x180], R3 ; /* 0x0000600000037a24 */ /* 0x000fd200078e0203 */ /*0520*/ IMAD.WIDE R2, R3, R2, c[0x0][0x170] ; /* 0x00005c0003027625 */ /* 0x000fca00078e0202 */ /*0530*/ STG.E [R2.64], R9 ; /* 0x0000000902007986 */ /* 0x000fe2000c101904 */ /*0540*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0550*/ BRA 0x550; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0560*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0570*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0580*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0590*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include <cuda_runtime.h> #include <stdio.h> #include <string.h> #include <stdlib.h> #include <math.h> #include <time.h> #define ELEMENT_MIN 0 #define ELEMENT_MAX 10 #define BLOCK_SIZE 16 #define TILE_SIZE 16 #define ZERO 1.e-6 int UI(int argc, char* argv[], int* jkl); float randGenerate(int min, int max); void initMatrix(float* matrix, int size, int min, int max); void matrixMulCPU(float* A, float* B, float* C, int j, int k, int l); __global__ void matrixMulGPU(float* d_A, float* d_B, float* d_C, int j, int k, int l); void showMatrix(float* matrix, int row, int col); void resultCheck(float* result_CPU, float* result_GPU, int size); int main(int argc, char* argv[]){ // reset rand seed srand((unsigned)time(NULL)); clock_t start, finish; int total_time; // Go through UI first. // In UI section, only command with valid param can go to the next step. int UIStatus; int jkl[3]; UIStatus = UI(argc, argv, jkl); if (UIStatus != 0) { printf("\nApplication terminates."); return 0; } printf("\nContinuing with j=%d, k=%d, l=%d", jkl[0], jkl[1], jkl[2]); // UI section ends // Initialize these two matrix A, B with random float type const int j = jkl[0]; const int k = jkl[1]; const int l = jkl[2]; float* A = (float*)malloc(j * k * sizeof(float)); float* B = (float*)malloc(k * l * sizeof(float)); initMatrix(A, (j * k), ELEMENT_MIN, ELEMENT_MAX); initMatrix(B, (k * l), ELEMENT_MIN, ELEMENT_MAX); showMatrix(A, j, k); showMatrix(B, k, l); printf("\nDone initializing matrix A with size %d*%d and matrix B with %d*%d", j, k, k, l); // Initialzing ends // CPU code for calculating the matrix multiplication // Use this result to varify the kernel result later // Only do this step when j, k, l is below 5000 float* C_GPU = (float*)malloc(j * l * sizeof(float)); float* C_CPU = (float*)malloc(j * l * sizeof(float)); if (k < 5000 && j < 5000 && l < 5000) { start = clock(); matrixMulCPU(A, B, C_CPU, j, k, l); finish = clock(); printf("\nDone matrix multiplication with CPU"); total_time = (int)(finish - start); printf("\n%d microseconds used.\n", total_time); showMatrix(C_CPU, j, l); } // Matrix multipication with CPU ends // Allocate device memory and copy data from host to device float *d_A, *d_B, *d_C; cudaMalloc((float**)&d_A, j * k * sizeof(float)); cudaMalloc((float**)&d_B, k * l * sizeof(float)); cudaMalloc((float**)&d_C, j * l * sizeof(float)); printf("\nDone allocating space in device."); cudaMemcpy(d_A, A, j * k * sizeof(float), cudaMemcpyHostToDevice); cudaMemcpy(d_B, B, k * l * sizeof(float), cudaMemcpyHostToDevice); printf("\nDone copying memory from host to device"); // Done allocating and transfering // Initialize thread block and kernel grid dimensions dim3 threads(BLOCK_SIZE, BLOCK_SIZE); dim3 grid((int)ceil(1.0 * l / threads.x), (int)ceil(1.0 * j / threads.y)); printf("\nDone initializing block dimention and grid dimention."); // Done initializing thread block and kernel grid dimensions // launch CUDA device kernel start = clock(); matrixMulGPU<<< grid, threads>>>(d_A, d_B, d_C, j, k, l); // Done CUDA device kernel // Copy results from device to host and free device memory cudaMemcpy(C_GPU, d_C, j * l * sizeof(float), cudaMemcpyDeviceToHost); finish = clock(); printf("\nDone matrix multiplication with GPU."); total_time = (int)(finish - start); printf("\n%d microseconds used.\n", total_time); showMatrix(C_GPU, j, l); cudaFree(d_A); cudaFree(d_B); cudaFree(d_C); // Done copying results and freeing device memory // Check the result of the Calculated Matrix if (k < 5000 && j < 5000 && l < 5000) resultCheck(C_CPU, C_GPU, j * l); // Done result checking. return 0; } // UI for main function // return 0 means everything's fine, just continue; // return 1 means there's invalid input or '--help', terminate running. int UI(int argc, char* argv[], int* jkl) { // UI for the exe file // while input with ? or -h or --help; tell that what we need as params // while input with [1,2] || [4, +inf) parameters; tell that we need 3 params // while input with 3 paramerters; print the size of two input matrix; check if all params are valid; if (argc == 2 && (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-h") == 0 || strcmp(argv[1], "?") == 0)) { printf("\nWe need 3 parameters j,k,l which represents the two matrix's dimention.\n"); printf("The first one with size j*k, and the second one's is k*l.\n"); printf("j,k,l must be int type and greater than 0.\n"); return 1; } if (argc <= 3 || argc >= 5) { printf("\nError.\nWe need 3 parameters as input.\nYou can use '--help' for more details.\n"); return 1; } if (argc == 4) { int n; printf("\nNumber Of Arguments Passed: %d", argc); printf("\n----Following Are The Command Line Arguments Passed----"); for (int counter = 1; counter < argc; counter++) { n = atoi(argv[counter]); if (n > 0) { printf("\nargv[%d]: %d", counter, n); jkl[counter - 1] = n; } else { printf("\nargv[%d] is invalid", counter); return 1; } } } return 0; } // random float generator within min and max float randGenerate(int min, int max) { float res; int r = rand(); res = 1.0 * rand() / RAND_MAX * (max-min) + min; return res; } // initialize matrix with float type void initMatrix(float* matrix, int size, int min, int max) { for (int i = 0; i < size; ++i) matrix[i] = randGenerate(min, max); return; } // matrix multiplication with CPU in the most stupid algo // Algo Complexity: O((2k-1)*j*l) void matrixMulCPU(float* A, float* B, float* C, int j, int k, int l) { for (int row = 0; row < j; ++row) { for (int col = 0; col < l; ++col) { float eleSum = 0.0f; for (int i = 0; i < k; ++i) { eleSum += A[row * k + i] * B[i * l + col]; } C[row * l + col] = eleSum; } } } // matrix multiplication with GPU device // using tiling algorithm, take use of the shared memory to higher the compute-to-global-memory-access ratio __global__ void matrixMulGPU(float* d_A, float* d_B, float* d_C, int j, int k, int l) { __shared__ float Ads[TILE_SIZE][TILE_SIZE]; __shared__ float Bds[TILE_SIZE][TILE_SIZE]; int bx = blockIdx.x; int by = blockIdx.y; int tx = threadIdx.x; int ty = threadIdx.y; int row = by * blockDim.y + ty; int col = bx * blockDim.x + tx; float CEleValue = 0.0f; for (int ph = 0; ph * TILE_SIZE < k; ++ph) { if (row < j && ph * TILE_SIZE + tx < k) Ads[ty][tx] = d_A[row * k + ph * TILE_SIZE + tx]; else Ads[ty][tx] = 0; if (col < l && ph * TILE_SIZE + ty < k) Bds[ty][tx] = d_B[(ph * TILE_SIZE + ty) * l + col]; else Bds[ty][tx] = 0; __syncthreads(); for (int i = 0; i < TILE_SIZE; ++i) { CEleValue += Ads[ty][i] * Bds[i][tx]; } __syncthreads(); } if (row < j && col < l) d_C[row * l + col] = CEleValue; } // show the data in the command prompt. // this function is used for configuration // please don't use it when j,k,l is too large void showMatrix(float* matrix, int row, int col) { if (row > 5 || col > 5) return; printf("\n"); for (int i = 0; i < row; i++) { for (int j = 0; j < col; j++) { printf("%f\t", matrix[i*col + j]); } printf("\n"); } return; } // check if two matrix is the same // 0 is defined at the very begining void resultCheck(float* result_CPU, float* result_GPU, int size) { // using relative error // assuming CPU result is the true value for (int i = 0; i < size; ++i) { if ((result_CPU[i] - result_GPU[i])/ result_CPU[i] > ZERO || (result_GPU[i] - result_CPU[i])/result_CPU[i] < -ZERO) { printf("\nResult check: Error!!!! Didn't pass."); return; } } printf("\nResult check: ---PASS---."); return; }
.file "tmpxft_0007abcf_00000000-6_hw1.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2066: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2066: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "--help" .LC1: .string "-h" .LC2: .string "?" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC3: .string "\nWe need 3 parameters j,k,l which represents the two matrix's dimention.\n" .align 8 .LC4: .string "The first one with size j*k, and the second one's is k*l.\n" .align 8 .LC5: .string "j,k,l must be int type and greater than 0.\n" .align 8 .LC6: .string "\nError.\nWe need 3 parameters as input.\nYou can use '--help' for more details.\n" .align 8 .LC7: .string "\nNumber Of Arguments Passed: %d" .align 8 .LC8: .string "\n----Following Are The Command Line Arguments Passed----" .section .rodata.str1.1 .LC9: .string "\nargv[%d]: %d" .LC10: .string "\nargv[%d] is invalid" .text .globl _Z2UIiPPcPi .type _Z2UIiPPcPi, @function _Z2UIiPPcPi: .LFB2058: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $8, %rsp .cfi_def_cfa_offset 64 movq %rsi, %r13 cmpl $2, %edi je .L13 movq %rdx, %r14 cmpl $4, %edi jne .L6 movl $4, %edx leaq .LC7(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT leaq .LC8(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $1, %ebp leaq .LC9(%rip), %r15 .L10: movq 0(%r13,%rbp,8), %rdi movl $10, %edx movl $0, %esi call __isoc23_strtol@PLT movq %rax, %rbx movl %eax, %ecx testl %eax, %eax jle .L9 movl %ebp, %edx movq %r15, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl %ebx, -4(%r14,%rbp,4) addq $1, %rbp cmpq $4, %rbp jne .L10 movl $0, %eax jmp .L3 .L13: movq 8(%rsi), %rbx leaq .LC0(%rip), %rsi movq %rbx, %rdi call strcmp@PLT testl %eax, %eax je .L5 leaq .LC1(%rip), %rsi movq %rbx, %rdi call strcmp@PLT testl %eax, %eax je .L5 leaq .LC2(%rip), %rsi movq %rbx, %rdi call strcmp@PLT testl %eax, %eax je .L5 .L6: leaq .LC6(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $1, %eax .L3: addq $8, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L5: .cfi_restore_state leaq .LC3(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT leaq .LC4(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT leaq .LC5(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $1, %eax jmp .L3 .L9: movl %ebp, %edx leaq .LC10(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $1, %eax jmp .L3 .cfi_endproc .LFE2058: .size _Z2UIiPPcPi, .-_Z2UIiPPcPi .globl _Z12randGenerateii .type _Z12randGenerateii, @function _Z12randGenerateii: .LFB2059: .cfi_startproc endbr64 pushq %rbp .cfi_def_cfa_offset 16 .cfi_offset 6, -16 pushq %rbx .cfi_def_cfa_offset 24 .cfi_offset 3, -24 subq $8, %rsp .cfi_def_cfa_offset 32 movl %edi, %ebp movl %esi, %ebx call rand@PLT call rand@PLT pxor %xmm0, %xmm0 cvtsi2sdl %eax, %xmm0 divsd .LC11(%rip), %xmm0 subl %ebp, %ebx pxor %xmm1, %xmm1 cvtsi2sdl %ebx, %xmm1 mulsd %xmm1, %xmm0 pxor %xmm1, %xmm1 cvtsi2sdl %ebp, %xmm1 addsd %xmm1, %xmm0 cvtsd2ss %xmm0, %xmm0 addq $8, %rsp .cfi_def_cfa_offset 24 popq %rbx .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2059: .size _Z12randGenerateii, .-_Z12randGenerateii .globl _Z10initMatrixPfiii .type _Z10initMatrixPfiii, @function _Z10initMatrixPfiii: .LFB2060: .cfi_startproc endbr64 testl %esi, %esi jle .L21 pushq %r13 .cfi_def_cfa_offset 16 .cfi_offset 13, -16 pushq %r12 .cfi_def_cfa_offset 24 .cfi_offset 12, -24 pushq %rbp .cfi_def_cfa_offset 32 .cfi_offset 6, -32 pushq %rbx .cfi_def_cfa_offset 40 .cfi_offset 3, -40 subq $8, %rsp .cfi_def_cfa_offset 48 movl %edx, %ebp movl %ecx, %r12d movq %rdi, %rbx movslq %esi, %rsi leaq (%rdi,%rsi,4), %r13 .L18: movl %r12d, %esi movl %ebp, %edi call _Z12randGenerateii movss %xmm0, (%rbx) addq $4, %rbx cmpq %r13, %rbx jne .L18 addq $8, %rsp .cfi_def_cfa_offset 40 popq %rbx .cfi_def_cfa_offset 32 popq %rbp .cfi_def_cfa_offset 24 popq %r12 .cfi_def_cfa_offset 16 popq %r13 .cfi_def_cfa_offset 8 ret .L21: .cfi_restore 3 .cfi_restore 6 .cfi_restore 12 .cfi_restore 13 ret .cfi_endproc .LFE2060: .size _Z10initMatrixPfiii, .-_Z10initMatrixPfiii .globl _Z12matrixMulCPUPfS_S_iii .type _Z12matrixMulCPUPfS_S_iii, @function _Z12matrixMulCPUPfS_S_iii: .LFB2061: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 movq %rsi, -16(%rsp) movq %rdx, -8(%rsp) movl %ecx, -20(%rsp) testl %ecx, %ecx jle .L24 movq %rdi, %rbx movl %r8d, %r10d movl %r9d, %r14d movslq %r9d, %rbp leaq 0(,%rbp,4), %rsi movl $0, %r13d movl $0, %r12d movl $0, %edx movslq %r8d, %r15 movq %r15, %rcx jmp .L26 .L27: movss (%rax), %xmm0 mulss (%rdx), %xmm0 addss %xmm0, %xmm1 addq $4, %rax addq %rsi, %rdx cmpq %rdi, %rax jne .L27 .L29: movss %xmm1, (%r11,%r8,4) addq $1, %r8 addq $4, %r9 cmpq %r8, %rbp je .L34 .L30: movq %r9, %rdx movq %r15, %rax pxor %xmm1, %xmm1 testl %r10d, %r10d jg .L27 jmp .L29 .L34: movl -24(%rsp), %edx .L28: addl $1, %edx addl %r14d, %r12d addl %r10d, %r13d cmpl %edx, -20(%rsp) je .L24 .L26: testl %r14d, %r14d jle .L28 movq -16(%rsp), %r9 movslq %r13d, %rax leaq (%rbx,%rax,4), %r15 addq %rcx, %rax leaq (%rbx,%rax,4), %rdi movslq %r12d, %rax movq -8(%rsp), %r11 leaq (%r11,%rax,4), %r11 movl $0, %r8d movl %edx, -24(%rsp) jmp .L30 .L24: popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2061: .size _Z12matrixMulCPUPfS_S_iii, .-_Z12matrixMulCPUPfS_S_iii .section .rodata.str1.1 .LC13: .string "\n" .LC14: .string "%f\t" .text .globl _Z10showMatrixPfii .type _Z10showMatrixPfii, @function _Z10showMatrixPfii: .LFB2062: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $40, %rsp .cfi_def_cfa_offset 96 movq %rdi, 16(%rsp) movl %esi, 12(%rsp) cmpl $5, %esi jg .L37 movl %esi, %ebx movl %edx, %r14d cmpl $5, %edx jg .L37 leaq .LC13(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT testl %ebx, %ebx jle .L37 movl $0, %r15d movl $0, %r13d movslq %r14d, %rax movq %rax, 24(%rsp) leaq .LC14(%rip), %r12 jmp .L39 .L41: movslq %r15d, %rax movq 16(%rsp), %rcx leaq (%rcx,%rax,4), %rbx movq 24(%rsp), %rdx addq %rdx, %rax leaq (%rcx,%rax,4), %rbp .L40: pxor %xmm0, %xmm0 cvtss2sd (%rbx), %xmm0 movq %r12, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT addq $4, %rbx cmpq %rbp, %rbx jne .L40 .L42: leaq .LC13(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT addl $1, %r13d addl %r14d, %r15d cmpl %r13d, 12(%rsp) je .L37 .L39: testl %r14d, %r14d jg .L41 jmp .L42 .L37: addq $40, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2062: .size _Z10showMatrixPfii, .-_Z10showMatrixPfii .section .rodata.str1.8 .align 8 .LC17: .string "\nResult check: Error!!!! Didn't pass." .section .rodata.str1.1 .LC18: .string "\nResult check: ---PASS---." .text .globl _Z11resultCheckPfS_i .type _Z11resultCheckPfS_i, @function _Z11resultCheckPfS_i: .LFB2063: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 testl %edx, %edx jle .L46 movslq %edx, %rdx salq $2, %rdx movl $0, %eax movsd .LC15(%rip), %xmm3 movsd .LC16(%rip), %xmm4 .L51: movss (%rdi,%rax), %xmm2 movss (%rsi,%rax), %xmm0 movaps %xmm2, %xmm1 subss %xmm0, %xmm1 divss %xmm2, %xmm1 cvtss2sd %xmm1, %xmm1 comisd %xmm3, %xmm1 ja .L47 subss %xmm2, %xmm0 divss %xmm2, %xmm0 cvtss2sd %xmm0, %xmm0 comisd %xmm0, %xmm4 ja .L47 addq $4, %rax cmpq %rdx, %rax jne .L51 .L46: leaq .LC18(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT jmp .L45 .L47: leaq .LC17(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT .L45: addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2063: .size _Z11resultCheckPfS_i, .-_Z11resultCheckPfS_i .globl _Z39__device_stub__Z12matrixMulGPUPfS_S_iiiPfS_S_iii .type _Z39__device_stub__Z12matrixMulGPUPfS_S_iiiPfS_S_iii, @function _Z39__device_stub__Z12matrixMulGPUPfS_S_iiiPfS_S_iii: .LFB2088: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movq %rdi, 40(%rsp) movq %rsi, 32(%rsp) movq %rdx, 24(%rsp) movl %ecx, 20(%rsp) movl %r8d, 16(%rsp) movl %r9d, 12(%rsp) movq %fs:40, %rax movq %rax, 168(%rsp) xorl %eax, %eax leaq 40(%rsp), %rax movq %rax, 112(%rsp) leaq 32(%rsp), %rax movq %rax, 120(%rsp) leaq 24(%rsp), %rax movq %rax, 128(%rsp) leaq 20(%rsp), %rax movq %rax, 136(%rsp) leaq 16(%rsp), %rax movq %rax, 144(%rsp) leaq 12(%rsp), %rax movq %rax, 152(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) movl $1, 72(%rsp) movl $1, 76(%rsp) movl $1, 80(%rsp) movl $1, 84(%rsp) leaq 56(%rsp), %rcx leaq 48(%rsp), %rdx leaq 76(%rsp), %rsi leaq 64(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L59 .L55: movq 168(%rsp), %rax subq %fs:40, %rax jne .L60 addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L59: .cfi_restore_state pushq 56(%rsp) .cfi_def_cfa_offset 200 pushq 56(%rsp) .cfi_def_cfa_offset 208 leaq 128(%rsp), %r9 movq 92(%rsp), %rcx movl 100(%rsp), %r8d movq 80(%rsp), %rsi movl 88(%rsp), %edx leaq _Z12matrixMulGPUPfS_S_iii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 192 jmp .L55 .L60: call __stack_chk_fail@PLT .cfi_endproc .LFE2088: .size _Z39__device_stub__Z12matrixMulGPUPfS_S_iiiPfS_S_iii, .-_Z39__device_stub__Z12matrixMulGPUPfS_S_iiiPfS_S_iii .globl _Z12matrixMulGPUPfS_S_iii .type _Z12matrixMulGPUPfS_S_iii, @function _Z12matrixMulGPUPfS_S_iii: .LFB2089: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z39__device_stub__Z12matrixMulGPUPfS_S_iiiPfS_S_iii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2089: .size _Z12matrixMulGPUPfS_S_iii, .-_Z12matrixMulGPUPfS_S_iii .section .rodata.str1.1 .LC19: .string "\nApplication terminates." .section .rodata.str1.8 .align 8 .LC20: .string "\nContinuing with j=%d, k=%d, l=%d" .align 8 .LC21: .string "\nDone initializing matrix A with size %d*%d and matrix B with %d*%d" .align 8 .LC22: .string "\nDone matrix multiplication with CPU" .section .rodata.str1.1 .LC23: .string "\n%d microseconds used.\n" .section .rodata.str1.8 .align 8 .LC24: .string "\nDone allocating space in device." .align 8 .LC25: .string "\nDone copying memory from host to device" .align 8 .LC27: .string "\nDone initializing block dimention and grid dimention." .align 8 .LC28: .string "\nDone matrix multiplication with GPU." .text .globl main .type main, @function main: .LFB2057: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $152, %rsp .cfi_def_cfa_offset 208 movl %edi, %ebx movq %rsi, %rbp movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax movl $0, %edi call time@PLT movl %eax, %edi call srand@PLT leaq 124(%rsp), %rdx movq %rbp, %rsi movl %ebx, %edi call _Z2UIiPPcPi testl %eax, %eax jne .L70 movl 132(%rsp), %ebp movl 128(%rsp), %r12d movl 124(%rsp), %ebx movl %ebp, %r8d movl %r12d, %ecx movl %ebx, %edx leaq .LC20(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl %r12d, %r15d imull %ebx, %r15d movslq %r15d, %r13 salq $2, %r13 movq %r13, %rdi call malloc@PLT movq %rax, (%rsp) movl %ebp, %r14d imull %r12d, %r14d movslq %r14d, %rax leaq 0(,%rax,4), %rcx movq %rcx, 16(%rsp) movq %rcx, %rdi call malloc@PLT movq %rax, 8(%rsp) movl $10, %ecx movl $0, %edx movl %r15d, %esi movq (%rsp), %rdi call _Z10initMatrixPfiii movl $10, %ecx movl $0, %edx movl %r14d, %esi movq 8(%rsp), %r15 movq %r15, %rdi call _Z10initMatrixPfiii movl %r12d, %edx movl %ebx, %esi movq (%rsp), %rdi call _Z10showMatrixPfii movl %ebp, %edx movl %r12d, %esi movq %r15, %rdi call _Z10showMatrixPfii movl %ebp, %r9d movl %r12d, %r8d movl %r12d, %ecx movl %ebx, %edx leaq .LC21(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl %ebp, %r15d imull %ebx, %r15d movslq %r15d, %r14 salq $2, %r14 movq %r14, %rdi call malloc@PLT movq %rax, 24(%rsp) movq %r14, %rdi call malloc@PLT movq %rax, 40(%rsp) cmpl $4999, %r12d setle %al cmpl $4999, %ebx setle %dl andl %edx, %eax cmpl $4999, %ebp setle %dl andb %dl, %al movb %al, 39(%rsp) jne .L71 .L66: leaq 72(%rsp), %rdi movq %r13, %rsi call cudaMalloc@PLT leaq 80(%rsp), %rdi movq 16(%rsp), %rsi call cudaMalloc@PLT leaq 88(%rsp), %rdi movq %r14, %rsi call cudaMalloc@PLT leaq .LC24(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $1, %ecx movq %r13, %rdx movq (%rsp), %rsi movq 72(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movq 16(%rsp), %rdx movq 8(%rsp), %rsi movq 80(%rsp), %rdi call cudaMemcpy@PLT leaq .LC25(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $1, 108(%rsp) pxor %xmm0, %xmm0 cvtsi2sdl %ebx, %xmm0 mulsd .LC26(%rip), %xmm0 call ceil@PLT movsd %xmm0, (%rsp) pxor %xmm0, %xmm0 cvtsi2sdl %ebp, %xmm0 mulsd .LC26(%rip), %xmm0 call ceil@PLT cvttsd2sil %xmm0, %eax movl %eax, 112(%rsp) cvttsd2sil (%rsp), %eax movl %eax, 116(%rsp) movl $1, 120(%rsp) leaq .LC27(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT call clock@PLT movq %rax, %r13 movl $16, 100(%rsp) movl $16, 104(%rsp) movl 108(%rsp), %ecx movl $0, %r9d movl $0, %r8d movq 100(%rsp), %rdx movq 112(%rsp), %rdi movl 120(%rsp), %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L72 .L67: movl $2, %ecx movq %r14, %rdx movq 88(%rsp), %rsi movq 24(%rsp), %r14 movq %r14, %rdi call cudaMemcpy@PLT call clock@PLT movq %rax, %r12 leaq .LC28(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl %r12d, %edx subl %r13d, %edx leaq .LC23(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl %ebp, %edx movl %ebx, %esi movq %r14, %rdi call _Z10showMatrixPfii movq 72(%rsp), %rdi call cudaFree@PLT movq 80(%rsp), %rdi call cudaFree@PLT movq 88(%rsp), %rdi call cudaFree@PLT cmpb $0, 39(%rsp) jne .L73 .L65: movq 136(%rsp), %rax subq %fs:40, %rax jne .L74 movl $0, %eax addq $152, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L70: .cfi_restore_state leaq .LC19(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT jmp .L65 .L71: call clock@PLT movq %rax, 48(%rsp) movl %ebp, %r9d movl %r12d, %r8d movl %ebx, %ecx movq 40(%rsp), %rdx movq 8(%rsp), %rsi movq (%rsp), %rdi call _Z12matrixMulCPUPfS_S_iii call clock@PLT movq %rax, 56(%rsp) leaq .LC22(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 56(%rsp), %ecx movl 48(%rsp), %eax subl %eax, %ecx movl %ecx, %edx leaq .LC23(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl %ebp, %edx movl %ebx, %esi movq 40(%rsp), %rdi call _Z10showMatrixPfii jmp .L66 .L72: movl %ebp, %r9d movl %r12d, %r8d movl %ebx, %ecx movq 88(%rsp), %rdx movq 80(%rsp), %rsi movq 72(%rsp), %rdi call _Z39__device_stub__Z12matrixMulGPUPfS_S_iiiPfS_S_iii jmp .L67 .L73: movl %r15d, %edx movq 24(%rsp), %rsi movq 40(%rsp), %rdi call _Z11resultCheckPfS_i jmp .L65 .L74: call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size main, .-main .section .rodata.str1.1 .LC29: .string "_Z12matrixMulGPUPfS_S_iii" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2091: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC29(%rip), %rdx movq %rdx, %rcx leaq _Z12matrixMulGPUPfS_S_iii(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2091: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .section .rodata.cst8,"aM",@progbits,8 .align 8 .LC11: .long -4194304 .long 1105199103 .align 8 .LC15: .long -1598689907 .long 1051772663 .align 8 .LC16: .long -1598689907 .long -1095710985 .align 8 .LC26: .long 0 .long 1068498944 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include <cuda_runtime.h> #include <stdio.h> #include <string.h> #include <stdlib.h> #include <math.h> #include <time.h> #define ELEMENT_MIN 0 #define ELEMENT_MAX 10 #define BLOCK_SIZE 16 #define TILE_SIZE 16 #define ZERO 1.e-6 int UI(int argc, char* argv[], int* jkl); float randGenerate(int min, int max); void initMatrix(float* matrix, int size, int min, int max); void matrixMulCPU(float* A, float* B, float* C, int j, int k, int l); __global__ void matrixMulGPU(float* d_A, float* d_B, float* d_C, int j, int k, int l); void showMatrix(float* matrix, int row, int col); void resultCheck(float* result_CPU, float* result_GPU, int size); int main(int argc, char* argv[]){ // reset rand seed srand((unsigned)time(NULL)); clock_t start, finish; int total_time; // Go through UI first. // In UI section, only command with valid param can go to the next step. int UIStatus; int jkl[3]; UIStatus = UI(argc, argv, jkl); if (UIStatus != 0) { printf("\nApplication terminates."); return 0; } printf("\nContinuing with j=%d, k=%d, l=%d", jkl[0], jkl[1], jkl[2]); // UI section ends // Initialize these two matrix A, B with random float type const int j = jkl[0]; const int k = jkl[1]; const int l = jkl[2]; float* A = (float*)malloc(j * k * sizeof(float)); float* B = (float*)malloc(k * l * sizeof(float)); initMatrix(A, (j * k), ELEMENT_MIN, ELEMENT_MAX); initMatrix(B, (k * l), ELEMENT_MIN, ELEMENT_MAX); showMatrix(A, j, k); showMatrix(B, k, l); printf("\nDone initializing matrix A with size %d*%d and matrix B with %d*%d", j, k, k, l); // Initialzing ends // CPU code for calculating the matrix multiplication // Use this result to varify the kernel result later // Only do this step when j, k, l is below 5000 float* C_GPU = (float*)malloc(j * l * sizeof(float)); float* C_CPU = (float*)malloc(j * l * sizeof(float)); if (k < 5000 && j < 5000 && l < 5000) { start = clock(); matrixMulCPU(A, B, C_CPU, j, k, l); finish = clock(); printf("\nDone matrix multiplication with CPU"); total_time = (int)(finish - start); printf("\n%d microseconds used.\n", total_time); showMatrix(C_CPU, j, l); } // Matrix multipication with CPU ends // Allocate device memory and copy data from host to device float *d_A, *d_B, *d_C; cudaMalloc((float**)&d_A, j * k * sizeof(float)); cudaMalloc((float**)&d_B, k * l * sizeof(float)); cudaMalloc((float**)&d_C, j * l * sizeof(float)); printf("\nDone allocating space in device."); cudaMemcpy(d_A, A, j * k * sizeof(float), cudaMemcpyHostToDevice); cudaMemcpy(d_B, B, k * l * sizeof(float), cudaMemcpyHostToDevice); printf("\nDone copying memory from host to device"); // Done allocating and transfering // Initialize thread block and kernel grid dimensions dim3 threads(BLOCK_SIZE, BLOCK_SIZE); dim3 grid((int)ceil(1.0 * l / threads.x), (int)ceil(1.0 * j / threads.y)); printf("\nDone initializing block dimention and grid dimention."); // Done initializing thread block and kernel grid dimensions // launch CUDA device kernel start = clock(); matrixMulGPU<<< grid, threads>>>(d_A, d_B, d_C, j, k, l); // Done CUDA device kernel // Copy results from device to host and free device memory cudaMemcpy(C_GPU, d_C, j * l * sizeof(float), cudaMemcpyDeviceToHost); finish = clock(); printf("\nDone matrix multiplication with GPU."); total_time = (int)(finish - start); printf("\n%d microseconds used.\n", total_time); showMatrix(C_GPU, j, l); cudaFree(d_A); cudaFree(d_B); cudaFree(d_C); // Done copying results and freeing device memory // Check the result of the Calculated Matrix if (k < 5000 && j < 5000 && l < 5000) resultCheck(C_CPU, C_GPU, j * l); // Done result checking. return 0; } // UI for main function // return 0 means everything's fine, just continue; // return 1 means there's invalid input or '--help', terminate running. int UI(int argc, char* argv[], int* jkl) { // UI for the exe file // while input with ? or -h or --help; tell that what we need as params // while input with [1,2] || [4, +inf) parameters; tell that we need 3 params // while input with 3 paramerters; print the size of two input matrix; check if all params are valid; if (argc == 2 && (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-h") == 0 || strcmp(argv[1], "?") == 0)) { printf("\nWe need 3 parameters j,k,l which represents the two matrix's dimention.\n"); printf("The first one with size j*k, and the second one's is k*l.\n"); printf("j,k,l must be int type and greater than 0.\n"); return 1; } if (argc <= 3 || argc >= 5) { printf("\nError.\nWe need 3 parameters as input.\nYou can use '--help' for more details.\n"); return 1; } if (argc == 4) { int n; printf("\nNumber Of Arguments Passed: %d", argc); printf("\n----Following Are The Command Line Arguments Passed----"); for (int counter = 1; counter < argc; counter++) { n = atoi(argv[counter]); if (n > 0) { printf("\nargv[%d]: %d", counter, n); jkl[counter - 1] = n; } else { printf("\nargv[%d] is invalid", counter); return 1; } } } return 0; } // random float generator within min and max float randGenerate(int min, int max) { float res; int r = rand(); res = 1.0 * rand() / RAND_MAX * (max-min) + min; return res; } // initialize matrix with float type void initMatrix(float* matrix, int size, int min, int max) { for (int i = 0; i < size; ++i) matrix[i] = randGenerate(min, max); return; } // matrix multiplication with CPU in the most stupid algo // Algo Complexity: O((2k-1)*j*l) void matrixMulCPU(float* A, float* B, float* C, int j, int k, int l) { for (int row = 0; row < j; ++row) { for (int col = 0; col < l; ++col) { float eleSum = 0.0f; for (int i = 0; i < k; ++i) { eleSum += A[row * k + i] * B[i * l + col]; } C[row * l + col] = eleSum; } } } // matrix multiplication with GPU device // using tiling algorithm, take use of the shared memory to higher the compute-to-global-memory-access ratio __global__ void matrixMulGPU(float* d_A, float* d_B, float* d_C, int j, int k, int l) { __shared__ float Ads[TILE_SIZE][TILE_SIZE]; __shared__ float Bds[TILE_SIZE][TILE_SIZE]; int bx = blockIdx.x; int by = blockIdx.y; int tx = threadIdx.x; int ty = threadIdx.y; int row = by * blockDim.y + ty; int col = bx * blockDim.x + tx; float CEleValue = 0.0f; for (int ph = 0; ph * TILE_SIZE < k; ++ph) { if (row < j && ph * TILE_SIZE + tx < k) Ads[ty][tx] = d_A[row * k + ph * TILE_SIZE + tx]; else Ads[ty][tx] = 0; if (col < l && ph * TILE_SIZE + ty < k) Bds[ty][tx] = d_B[(ph * TILE_SIZE + ty) * l + col]; else Bds[ty][tx] = 0; __syncthreads(); for (int i = 0; i < TILE_SIZE; ++i) { CEleValue += Ads[ty][i] * Bds[i][tx]; } __syncthreads(); } if (row < j && col < l) d_C[row * l + col] = CEleValue; } // show the data in the command prompt. // this function is used for configuration // please don't use it when j,k,l is too large void showMatrix(float* matrix, int row, int col) { if (row > 5 || col > 5) return; printf("\n"); for (int i = 0; i < row; i++) { for (int j = 0; j < col; j++) { printf("%f\t", matrix[i*col + j]); } printf("\n"); } return; } // check if two matrix is the same // 0 is defined at the very begining void resultCheck(float* result_CPU, float* result_GPU, int size) { // using relative error // assuming CPU result is the true value for (int i = 0; i < size; ++i) { if ((result_CPU[i] - result_GPU[i])/ result_CPU[i] > ZERO || (result_GPU[i] - result_CPU[i])/result_CPU[i] < -ZERO) { printf("\nResult check: Error!!!! Didn't pass."); return; } } printf("\nResult check: ---PASS---."); return; }
#include <hip/hip_runtime.h> #include <stdio.h> #include <string.h> #include <stdlib.h> #include <math.h> #include <time.h> #define ELEMENT_MIN 0 #define ELEMENT_MAX 10 #define BLOCK_SIZE 16 #define TILE_SIZE 16 #define ZERO 1.e-6 int UI(int argc, char* argv[], int* jkl); float randGenerate(int min, int max); void initMatrix(float* matrix, int size, int min, int max); void matrixMulCPU(float* A, float* B, float* C, int j, int k, int l); __global__ void matrixMulGPU(float* d_A, float* d_B, float* d_C, int j, int k, int l); void showMatrix(float* matrix, int row, int col); void resultCheck(float* result_CPU, float* result_GPU, int size); int main(int argc, char* argv[]){ // reset rand seed srand((unsigned)time(NULL)); clock_t start, finish; int total_time; // Go through UI first. // In UI section, only command with valid param can go to the next step. int UIStatus; int jkl[3]; UIStatus = UI(argc, argv, jkl); if (UIStatus != 0) { printf("\nApplication terminates."); return 0; } printf("\nContinuing with j=%d, k=%d, l=%d", jkl[0], jkl[1], jkl[2]); // UI section ends // Initialize these two matrix A, B with random float type const int j = jkl[0]; const int k = jkl[1]; const int l = jkl[2]; float* A = (float*)malloc(j * k * sizeof(float)); float* B = (float*)malloc(k * l * sizeof(float)); initMatrix(A, (j * k), ELEMENT_MIN, ELEMENT_MAX); initMatrix(B, (k * l), ELEMENT_MIN, ELEMENT_MAX); showMatrix(A, j, k); showMatrix(B, k, l); printf("\nDone initializing matrix A with size %d*%d and matrix B with %d*%d", j, k, k, l); // Initialzing ends // CPU code for calculating the matrix multiplication // Use this result to varify the kernel result later // Only do this step when j, k, l is below 5000 float* C_GPU = (float*)malloc(j * l * sizeof(float)); float* C_CPU = (float*)malloc(j * l * sizeof(float)); if (k < 5000 && j < 5000 && l < 5000) { start = clock(); matrixMulCPU(A, B, C_CPU, j, k, l); finish = clock(); printf("\nDone matrix multiplication with CPU"); total_time = (int)(finish - start); printf("\n%d microseconds used.\n", total_time); showMatrix(C_CPU, j, l); } // Matrix multipication with CPU ends // Allocate device memory and copy data from host to device float *d_A, *d_B, *d_C; hipMalloc((float**)&d_A, j * k * sizeof(float)); hipMalloc((float**)&d_B, k * l * sizeof(float)); hipMalloc((float**)&d_C, j * l * sizeof(float)); printf("\nDone allocating space in device."); hipMemcpy(d_A, A, j * k * sizeof(float), hipMemcpyHostToDevice); hipMemcpy(d_B, B, k * l * sizeof(float), hipMemcpyHostToDevice); printf("\nDone copying memory from host to device"); // Done allocating and transfering // Initialize thread block and kernel grid dimensions dim3 threads(BLOCK_SIZE, BLOCK_SIZE); dim3 grid((int)ceil(1.0 * l / threads.x), (int)ceil(1.0 * j / threads.y)); printf("\nDone initializing block dimention and grid dimention."); // Done initializing thread block and kernel grid dimensions // launch CUDA device kernel start = clock(); matrixMulGPU<<< grid, threads>>>(d_A, d_B, d_C, j, k, l); // Done CUDA device kernel // Copy results from device to host and free device memory hipMemcpy(C_GPU, d_C, j * l * sizeof(float), hipMemcpyDeviceToHost); finish = clock(); printf("\nDone matrix multiplication with GPU."); total_time = (int)(finish - start); printf("\n%d microseconds used.\n", total_time); showMatrix(C_GPU, j, l); hipFree(d_A); hipFree(d_B); hipFree(d_C); // Done copying results and freeing device memory // Check the result of the Calculated Matrix if (k < 5000 && j < 5000 && l < 5000) resultCheck(C_CPU, C_GPU, j * l); // Done result checking. return 0; } // UI for main function // return 0 means everything's fine, just continue; // return 1 means there's invalid input or '--help', terminate running. int UI(int argc, char* argv[], int* jkl) { // UI for the exe file // while input with ? or -h or --help; tell that what we need as params // while input with [1,2] || [4, +inf) parameters; tell that we need 3 params // while input with 3 paramerters; print the size of two input matrix; check if all params are valid; if (argc == 2 && (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-h") == 0 || strcmp(argv[1], "?") == 0)) { printf("\nWe need 3 parameters j,k,l which represents the two matrix's dimention.\n"); printf("The first one with size j*k, and the second one's is k*l.\n"); printf("j,k,l must be int type and greater than 0.\n"); return 1; } if (argc <= 3 || argc >= 5) { printf("\nError.\nWe need 3 parameters as input.\nYou can use '--help' for more details.\n"); return 1; } if (argc == 4) { int n; printf("\nNumber Of Arguments Passed: %d", argc); printf("\n----Following Are The Command Line Arguments Passed----"); for (int counter = 1; counter < argc; counter++) { n = atoi(argv[counter]); if (n > 0) { printf("\nargv[%d]: %d", counter, n); jkl[counter - 1] = n; } else { printf("\nargv[%d] is invalid", counter); return 1; } } } return 0; } // random float generator within min and max float randGenerate(int min, int max) { float res; int r = rand(); res = 1.0 * rand() / RAND_MAX * (max-min) + min; return res; } // initialize matrix with float type void initMatrix(float* matrix, int size, int min, int max) { for (int i = 0; i < size; ++i) matrix[i] = randGenerate(min, max); return; } // matrix multiplication with CPU in the most stupid algo // Algo Complexity: O((2k-1)*j*l) void matrixMulCPU(float* A, float* B, float* C, int j, int k, int l) { for (int row = 0; row < j; ++row) { for (int col = 0; col < l; ++col) { float eleSum = 0.0f; for (int i = 0; i < k; ++i) { eleSum += A[row * k + i] * B[i * l + col]; } C[row * l + col] = eleSum; } } } // matrix multiplication with GPU device // using tiling algorithm, take use of the shared memory to higher the compute-to-global-memory-access ratio __global__ void matrixMulGPU(float* d_A, float* d_B, float* d_C, int j, int k, int l) { __shared__ float Ads[TILE_SIZE][TILE_SIZE]; __shared__ float Bds[TILE_SIZE][TILE_SIZE]; int bx = blockIdx.x; int by = blockIdx.y; int tx = threadIdx.x; int ty = threadIdx.y; int row = by * blockDim.y + ty; int col = bx * blockDim.x + tx; float CEleValue = 0.0f; for (int ph = 0; ph * TILE_SIZE < k; ++ph) { if (row < j && ph * TILE_SIZE + tx < k) Ads[ty][tx] = d_A[row * k + ph * TILE_SIZE + tx]; else Ads[ty][tx] = 0; if (col < l && ph * TILE_SIZE + ty < k) Bds[ty][tx] = d_B[(ph * TILE_SIZE + ty) * l + col]; else Bds[ty][tx] = 0; __syncthreads(); for (int i = 0; i < TILE_SIZE; ++i) { CEleValue += Ads[ty][i] * Bds[i][tx]; } __syncthreads(); } if (row < j && col < l) d_C[row * l + col] = CEleValue; } // show the data in the command prompt. // this function is used for configuration // please don't use it when j,k,l is too large void showMatrix(float* matrix, int row, int col) { if (row > 5 || col > 5) return; printf("\n"); for (int i = 0; i < row; i++) { for (int j = 0; j < col; j++) { printf("%f\t", matrix[i*col + j]); } printf("\n"); } return; } // check if two matrix is the same // 0 is defined at the very begining void resultCheck(float* result_CPU, float* result_GPU, int size) { // using relative error // assuming CPU result is the true value for (int i = 0; i < size; ++i) { if ((result_CPU[i] - result_GPU[i])/ result_CPU[i] > ZERO || (result_GPU[i] - result_CPU[i])/result_CPU[i] < -ZERO) { printf("\nResult check: Error!!!! Didn't pass."); return; } } printf("\nResult check: ---PASS---."); return; }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include <stdio.h> #include <string.h> #include <stdlib.h> #include <math.h> #include <time.h> #define ELEMENT_MIN 0 #define ELEMENT_MAX 10 #define BLOCK_SIZE 16 #define TILE_SIZE 16 #define ZERO 1.e-6 int UI(int argc, char* argv[], int* jkl); float randGenerate(int min, int max); void initMatrix(float* matrix, int size, int min, int max); void matrixMulCPU(float* A, float* B, float* C, int j, int k, int l); __global__ void matrixMulGPU(float* d_A, float* d_B, float* d_C, int j, int k, int l); void showMatrix(float* matrix, int row, int col); void resultCheck(float* result_CPU, float* result_GPU, int size); int main(int argc, char* argv[]){ // reset rand seed srand((unsigned)time(NULL)); clock_t start, finish; int total_time; // Go through UI first. // In UI section, only command with valid param can go to the next step. int UIStatus; int jkl[3]; UIStatus = UI(argc, argv, jkl); if (UIStatus != 0) { printf("\nApplication terminates."); return 0; } printf("\nContinuing with j=%d, k=%d, l=%d", jkl[0], jkl[1], jkl[2]); // UI section ends // Initialize these two matrix A, B with random float type const int j = jkl[0]; const int k = jkl[1]; const int l = jkl[2]; float* A = (float*)malloc(j * k * sizeof(float)); float* B = (float*)malloc(k * l * sizeof(float)); initMatrix(A, (j * k), ELEMENT_MIN, ELEMENT_MAX); initMatrix(B, (k * l), ELEMENT_MIN, ELEMENT_MAX); showMatrix(A, j, k); showMatrix(B, k, l); printf("\nDone initializing matrix A with size %d*%d and matrix B with %d*%d", j, k, k, l); // Initialzing ends // CPU code for calculating the matrix multiplication // Use this result to varify the kernel result later // Only do this step when j, k, l is below 5000 float* C_GPU = (float*)malloc(j * l * sizeof(float)); float* C_CPU = (float*)malloc(j * l * sizeof(float)); if (k < 5000 && j < 5000 && l < 5000) { start = clock(); matrixMulCPU(A, B, C_CPU, j, k, l); finish = clock(); printf("\nDone matrix multiplication with CPU"); total_time = (int)(finish - start); printf("\n%d microseconds used.\n", total_time); showMatrix(C_CPU, j, l); } // Matrix multipication with CPU ends // Allocate device memory and copy data from host to device float *d_A, *d_B, *d_C; hipMalloc((float**)&d_A, j * k * sizeof(float)); hipMalloc((float**)&d_B, k * l * sizeof(float)); hipMalloc((float**)&d_C, j * l * sizeof(float)); printf("\nDone allocating space in device."); hipMemcpy(d_A, A, j * k * sizeof(float), hipMemcpyHostToDevice); hipMemcpy(d_B, B, k * l * sizeof(float), hipMemcpyHostToDevice); printf("\nDone copying memory from host to device"); // Done allocating and transfering // Initialize thread block and kernel grid dimensions dim3 threads(BLOCK_SIZE, BLOCK_SIZE); dim3 grid((int)ceil(1.0 * l / threads.x), (int)ceil(1.0 * j / threads.y)); printf("\nDone initializing block dimention and grid dimention."); // Done initializing thread block and kernel grid dimensions // launch CUDA device kernel start = clock(); matrixMulGPU<<< grid, threads>>>(d_A, d_B, d_C, j, k, l); // Done CUDA device kernel // Copy results from device to host and free device memory hipMemcpy(C_GPU, d_C, j * l * sizeof(float), hipMemcpyDeviceToHost); finish = clock(); printf("\nDone matrix multiplication with GPU."); total_time = (int)(finish - start); printf("\n%d microseconds used.\n", total_time); showMatrix(C_GPU, j, l); hipFree(d_A); hipFree(d_B); hipFree(d_C); // Done copying results and freeing device memory // Check the result of the Calculated Matrix if (k < 5000 && j < 5000 && l < 5000) resultCheck(C_CPU, C_GPU, j * l); // Done result checking. return 0; } // UI for main function // return 0 means everything's fine, just continue; // return 1 means there's invalid input or '--help', terminate running. int UI(int argc, char* argv[], int* jkl) { // UI for the exe file // while input with ? or -h or --help; tell that what we need as params // while input with [1,2] || [4, +inf) parameters; tell that we need 3 params // while input with 3 paramerters; print the size of two input matrix; check if all params are valid; if (argc == 2 && (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-h") == 0 || strcmp(argv[1], "?") == 0)) { printf("\nWe need 3 parameters j,k,l which represents the two matrix's dimention.\n"); printf("The first one with size j*k, and the second one's is k*l.\n"); printf("j,k,l must be int type and greater than 0.\n"); return 1; } if (argc <= 3 || argc >= 5) { printf("\nError.\nWe need 3 parameters as input.\nYou can use '--help' for more details.\n"); return 1; } if (argc == 4) { int n; printf("\nNumber Of Arguments Passed: %d", argc); printf("\n----Following Are The Command Line Arguments Passed----"); for (int counter = 1; counter < argc; counter++) { n = atoi(argv[counter]); if (n > 0) { printf("\nargv[%d]: %d", counter, n); jkl[counter - 1] = n; } else { printf("\nargv[%d] is invalid", counter); return 1; } } } return 0; } // random float generator within min and max float randGenerate(int min, int max) { float res; int r = rand(); res = 1.0 * rand() / RAND_MAX * (max-min) + min; return res; } // initialize matrix with float type void initMatrix(float* matrix, int size, int min, int max) { for (int i = 0; i < size; ++i) matrix[i] = randGenerate(min, max); return; } // matrix multiplication with CPU in the most stupid algo // Algo Complexity: O((2k-1)*j*l) void matrixMulCPU(float* A, float* B, float* C, int j, int k, int l) { for (int row = 0; row < j; ++row) { for (int col = 0; col < l; ++col) { float eleSum = 0.0f; for (int i = 0; i < k; ++i) { eleSum += A[row * k + i] * B[i * l + col]; } C[row * l + col] = eleSum; } } } // matrix multiplication with GPU device // using tiling algorithm, take use of the shared memory to higher the compute-to-global-memory-access ratio __global__ void matrixMulGPU(float* d_A, float* d_B, float* d_C, int j, int k, int l) { __shared__ float Ads[TILE_SIZE][TILE_SIZE]; __shared__ float Bds[TILE_SIZE][TILE_SIZE]; int bx = blockIdx.x; int by = blockIdx.y; int tx = threadIdx.x; int ty = threadIdx.y; int row = by * blockDim.y + ty; int col = bx * blockDim.x + tx; float CEleValue = 0.0f; for (int ph = 0; ph * TILE_SIZE < k; ++ph) { if (row < j && ph * TILE_SIZE + tx < k) Ads[ty][tx] = d_A[row * k + ph * TILE_SIZE + tx]; else Ads[ty][tx] = 0; if (col < l && ph * TILE_SIZE + ty < k) Bds[ty][tx] = d_B[(ph * TILE_SIZE + ty) * l + col]; else Bds[ty][tx] = 0; __syncthreads(); for (int i = 0; i < TILE_SIZE; ++i) { CEleValue += Ads[ty][i] * Bds[i][tx]; } __syncthreads(); } if (row < j && col < l) d_C[row * l + col] = CEleValue; } // show the data in the command prompt. // this function is used for configuration // please don't use it when j,k,l is too large void showMatrix(float* matrix, int row, int col) { if (row > 5 || col > 5) return; printf("\n"); for (int i = 0; i < row; i++) { for (int j = 0; j < col; j++) { printf("%f\t", matrix[i*col + j]); } printf("\n"); } return; } // check if two matrix is the same // 0 is defined at the very begining void resultCheck(float* result_CPU, float* result_GPU, int size) { // using relative error // assuming CPU result is the true value for (int i = 0; i < size; ++i) { if ((result_CPU[i] - result_GPU[i])/ result_CPU[i] > ZERO || (result_GPU[i] - result_CPU[i])/result_CPU[i] < -ZERO) { printf("\nResult check: Error!!!! Didn't pass."); return; } } printf("\nResult check: ---PASS---."); return; }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z12matrixMulGPUPfS_S_iii .globl _Z12matrixMulGPUPfS_S_iii .p2align 8 .type _Z12matrixMulGPUPfS_S_iii,@function _Z12matrixMulGPUPfS_S_iii: s_clause 0x2 s_load_b32 s2, s[0:1], 0x34 s_load_b64 s[8:9], s[0:1], 0x18 s_load_b32 s3, s[0:1], 0x20 v_bfe_u32 v3, v0, 10, 10 v_and_b32_e32 v4, 0x3ff, v0 s_waitcnt lgkmcnt(0) s_lshr_b32 s4, s2, 16 s_and_b32 s2, s2, 0xffff s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_2) v_mad_u64_u32 v[0:1], null, s15, s4, v[3:4] v_mad_u64_u32 v[1:2], null, s14, s2, v[4:5] s_cmp_lt_i32 s9, 1 v_cmp_gt_i32_e64 s2, s8, v0 s_delay_alu instid0(VALU_DEP_2) v_cmp_gt_i32_e32 vcc_lo, s3, v1 s_cbranch_scc1 .LBB0_15 s_load_b128 s[4:7], s[0:1], 0x0 v_lshlrev_b32_e32 v2, 2, v4 v_dual_mov_b32 v10, 0 :: v_dual_lshlrev_b32 v7, 6, v3 v_mad_u64_u32 v[5:6], null, v0, s9, v[4:5] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_nc_u32_e32 v8, 0x400, v2 v_dual_mov_b32 v2, 0 :: v_dual_add_nc_u32 v9, v7, v2 v_cmp_le_i32_e64 s10, s3, v1 s_mov_b32 s13, 0 s_delay_alu instid0(VALU_DEP_3) v_add_nc_u32_e32 v6, v8, v7 s_xor_b32 s11, s2, -1 s_mov_b32 s12, 0 .LBB0_2: v_add_nc_u32_e32 v12, s13, v4 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cmp_le_i32_e64 s2, s9, v12 s_or_b32 s2, s11, s2 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_and_saveexec_b32 s14, s2 s_xor_b32 s2, exec_lo, s14 s_cbranch_execz .LBB0_4 ds_store_b32 v9, v10 .LBB0_4: s_and_not1_saveexec_b32 s14, s2 s_cbranch_execz .LBB0_6 v_add_nc_u32_e32 v12, s13, v5 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v13, 31, v12 v_lshlrev_b64 v[12:13], 2, v[12:13] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s2, s4, v12 v_add_co_ci_u32_e64 v13, s2, s5, v13, s2 global_load_b32 v12, v[12:13], off s_waitcnt vmcnt(0) ds_store_b32 v9, v12 .LBB0_6: s_or_b32 exec_lo, exec_lo, s14 s_mov_b32 s14, 0 s_mov_b32 s2, s10 s_and_saveexec_b32 s15, vcc_lo v_add_nc_u32_e32 v11, s13, v3 s_and_not1_b32 s13, s10, exec_lo s_mov_b32 s14, exec_lo s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cmp_le_i32_e64 s2, s9, v11 s_and_b32 s2, s2, exec_lo s_delay_alu instid0(SALU_CYCLE_1) s_or_b32 s2, s13, s2 s_or_b32 exec_lo, exec_lo, s15 s_and_saveexec_b32 s13, s2 s_delay_alu instid0(SALU_CYCLE_1) s_xor_b32 s2, exec_lo, s13 s_cbranch_execz .LBB0_10 s_and_not1_b32 s14, s14, exec_lo ds_store_b32 v6, v10 .LBB0_10: s_or_b32 exec_lo, exec_lo, s2 s_and_saveexec_b32 s13, s14 s_cbranch_execz .LBB0_12 v_mad_u64_u32 v[12:13], null, v11, s3, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v13, 31, v12 v_lshlrev_b64 v[12:13], 2, v[12:13] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s2, s6, v12 v_add_co_ci_u32_e64 v13, s2, s7, v13, s2 global_load_b32 v12, v[12:13], off s_waitcnt vmcnt(0) ds_store_b32 v6, v12 .LBB0_12: s_or_b32 exec_lo, exec_lo, s13 v_mov_b32_e32 v12, v8 s_mov_b32 s2, 0 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv .LBB0_13: v_add_nc_u32_e32 v13, s2, v7 s_add_i32 s2, s2, 4 ds_load_b32 v14, v12 ds_load_b32 v13, v13 v_add_nc_u32_e32 v12, 64, v12 s_cmp_eq_u32 s2, 64 s_waitcnt lgkmcnt(0) v_fmac_f32_e32 v2, v13, v14 s_cbranch_scc0 .LBB0_13 s_add_i32 s12, s12, 1 s_delay_alu instid0(SALU_CYCLE_1) s_lshl_b32 s13, s12, 4 s_barrier s_cmp_ge_i32 s13, s9 buffer_gl0_inv s_cbranch_scc0 .LBB0_2 s_branch .LBB0_16 .LBB0_15: v_mov_b32_e32 v2, 0 .LBB0_16: v_cmp_gt_i32_e32 vcc_lo, s8, v0 v_cmp_gt_i32_e64 s2, s3, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_and_b32 s2, vcc_lo, s2 s_and_saveexec_b32 s4, s2 s_cbranch_execz .LBB0_18 s_load_b64 s[0:1], s[0:1], 0x10 v_mad_u64_u32 v[3:4], null, v0, s3, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v4, 31, v3 v_lshlrev_b64 v[0:1], 2, v[3:4] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v0, vcc_lo, s0, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo global_store_b32 v[0:1], v2, off .LBB0_18: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z12matrixMulGPUPfS_S_iii .amdhsa_group_segment_fixed_size 2048 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 296 .amdhsa_user_sgpr_count 14 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 1 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 1 .amdhsa_next_free_vgpr 15 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z12matrixMulGPUPfS_S_iii, .Lfunc_end0-_Z12matrixMulGPUPfS_S_iii .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 8 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .offset: 24 .size: 4 .value_kind: by_value - .offset: 28 .size: 4 .value_kind: by_value - .offset: 32 .size: 4 .value_kind: by_value - .offset: 40 .size: 4 .value_kind: hidden_block_count_x - .offset: 44 .size: 4 .value_kind: hidden_block_count_y - .offset: 48 .size: 4 .value_kind: hidden_block_count_z - .offset: 52 .size: 2 .value_kind: hidden_group_size_x - .offset: 54 .size: 2 .value_kind: hidden_group_size_y - .offset: 56 .size: 2 .value_kind: hidden_group_size_z - .offset: 58 .size: 2 .value_kind: hidden_remainder_x - .offset: 60 .size: 2 .value_kind: hidden_remainder_y - .offset: 62 .size: 2 .value_kind: hidden_remainder_z - .offset: 80 .size: 8 .value_kind: hidden_global_offset_x - .offset: 88 .size: 8 .value_kind: hidden_global_offset_y - .offset: 96 .size: 8 .value_kind: hidden_global_offset_z - .offset: 104 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 2048 .kernarg_segment_align: 8 .kernarg_segment_size: 296 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z12matrixMulGPUPfS_S_iii .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z12matrixMulGPUPfS_S_iii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 15 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <hip/hip_runtime.h> #include <stdio.h> #include <string.h> #include <stdlib.h> #include <math.h> #include <time.h> #define ELEMENT_MIN 0 #define ELEMENT_MAX 10 #define BLOCK_SIZE 16 #define TILE_SIZE 16 #define ZERO 1.e-6 int UI(int argc, char* argv[], int* jkl); float randGenerate(int min, int max); void initMatrix(float* matrix, int size, int min, int max); void matrixMulCPU(float* A, float* B, float* C, int j, int k, int l); __global__ void matrixMulGPU(float* d_A, float* d_B, float* d_C, int j, int k, int l); void showMatrix(float* matrix, int row, int col); void resultCheck(float* result_CPU, float* result_GPU, int size); int main(int argc, char* argv[]){ // reset rand seed srand((unsigned)time(NULL)); clock_t start, finish; int total_time; // Go through UI first. // In UI section, only command with valid param can go to the next step. int UIStatus; int jkl[3]; UIStatus = UI(argc, argv, jkl); if (UIStatus != 0) { printf("\nApplication terminates."); return 0; } printf("\nContinuing with j=%d, k=%d, l=%d", jkl[0], jkl[1], jkl[2]); // UI section ends // Initialize these two matrix A, B with random float type const int j = jkl[0]; const int k = jkl[1]; const int l = jkl[2]; float* A = (float*)malloc(j * k * sizeof(float)); float* B = (float*)malloc(k * l * sizeof(float)); initMatrix(A, (j * k), ELEMENT_MIN, ELEMENT_MAX); initMatrix(B, (k * l), ELEMENT_MIN, ELEMENT_MAX); showMatrix(A, j, k); showMatrix(B, k, l); printf("\nDone initializing matrix A with size %d*%d and matrix B with %d*%d", j, k, k, l); // Initialzing ends // CPU code for calculating the matrix multiplication // Use this result to varify the kernel result later // Only do this step when j, k, l is below 5000 float* C_GPU = (float*)malloc(j * l * sizeof(float)); float* C_CPU = (float*)malloc(j * l * sizeof(float)); if (k < 5000 && j < 5000 && l < 5000) { start = clock(); matrixMulCPU(A, B, C_CPU, j, k, l); finish = clock(); printf("\nDone matrix multiplication with CPU"); total_time = (int)(finish - start); printf("\n%d microseconds used.\n", total_time); showMatrix(C_CPU, j, l); } // Matrix multipication with CPU ends // Allocate device memory and copy data from host to device float *d_A, *d_B, *d_C; hipMalloc((float**)&d_A, j * k * sizeof(float)); hipMalloc((float**)&d_B, k * l * sizeof(float)); hipMalloc((float**)&d_C, j * l * sizeof(float)); printf("\nDone allocating space in device."); hipMemcpy(d_A, A, j * k * sizeof(float), hipMemcpyHostToDevice); hipMemcpy(d_B, B, k * l * sizeof(float), hipMemcpyHostToDevice); printf("\nDone copying memory from host to device"); // Done allocating and transfering // Initialize thread block and kernel grid dimensions dim3 threads(BLOCK_SIZE, BLOCK_SIZE); dim3 grid((int)ceil(1.0 * l / threads.x), (int)ceil(1.0 * j / threads.y)); printf("\nDone initializing block dimention and grid dimention."); // Done initializing thread block and kernel grid dimensions // launch CUDA device kernel start = clock(); matrixMulGPU<<< grid, threads>>>(d_A, d_B, d_C, j, k, l); // Done CUDA device kernel // Copy results from device to host and free device memory hipMemcpy(C_GPU, d_C, j * l * sizeof(float), hipMemcpyDeviceToHost); finish = clock(); printf("\nDone matrix multiplication with GPU."); total_time = (int)(finish - start); printf("\n%d microseconds used.\n", total_time); showMatrix(C_GPU, j, l); hipFree(d_A); hipFree(d_B); hipFree(d_C); // Done copying results and freeing device memory // Check the result of the Calculated Matrix if (k < 5000 && j < 5000 && l < 5000) resultCheck(C_CPU, C_GPU, j * l); // Done result checking. return 0; } // UI for main function // return 0 means everything's fine, just continue; // return 1 means there's invalid input or '--help', terminate running. int UI(int argc, char* argv[], int* jkl) { // UI for the exe file // while input with ? or -h or --help; tell that what we need as params // while input with [1,2] || [4, +inf) parameters; tell that we need 3 params // while input with 3 paramerters; print the size of two input matrix; check if all params are valid; if (argc == 2 && (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-h") == 0 || strcmp(argv[1], "?") == 0)) { printf("\nWe need 3 parameters j,k,l which represents the two matrix's dimention.\n"); printf("The first one with size j*k, and the second one's is k*l.\n"); printf("j,k,l must be int type and greater than 0.\n"); return 1; } if (argc <= 3 || argc >= 5) { printf("\nError.\nWe need 3 parameters as input.\nYou can use '--help' for more details.\n"); return 1; } if (argc == 4) { int n; printf("\nNumber Of Arguments Passed: %d", argc); printf("\n----Following Are The Command Line Arguments Passed----"); for (int counter = 1; counter < argc; counter++) { n = atoi(argv[counter]); if (n > 0) { printf("\nargv[%d]: %d", counter, n); jkl[counter - 1] = n; } else { printf("\nargv[%d] is invalid", counter); return 1; } } } return 0; } // random float generator within min and max float randGenerate(int min, int max) { float res; int r = rand(); res = 1.0 * rand() / RAND_MAX * (max-min) + min; return res; } // initialize matrix with float type void initMatrix(float* matrix, int size, int min, int max) { for (int i = 0; i < size; ++i) matrix[i] = randGenerate(min, max); return; } // matrix multiplication with CPU in the most stupid algo // Algo Complexity: O((2k-1)*j*l) void matrixMulCPU(float* A, float* B, float* C, int j, int k, int l) { for (int row = 0; row < j; ++row) { for (int col = 0; col < l; ++col) { float eleSum = 0.0f; for (int i = 0; i < k; ++i) { eleSum += A[row * k + i] * B[i * l + col]; } C[row * l + col] = eleSum; } } } // matrix multiplication with GPU device // using tiling algorithm, take use of the shared memory to higher the compute-to-global-memory-access ratio __global__ void matrixMulGPU(float* d_A, float* d_B, float* d_C, int j, int k, int l) { __shared__ float Ads[TILE_SIZE][TILE_SIZE]; __shared__ float Bds[TILE_SIZE][TILE_SIZE]; int bx = blockIdx.x; int by = blockIdx.y; int tx = threadIdx.x; int ty = threadIdx.y; int row = by * blockDim.y + ty; int col = bx * blockDim.x + tx; float CEleValue = 0.0f; for (int ph = 0; ph * TILE_SIZE < k; ++ph) { if (row < j && ph * TILE_SIZE + tx < k) Ads[ty][tx] = d_A[row * k + ph * TILE_SIZE + tx]; else Ads[ty][tx] = 0; if (col < l && ph * TILE_SIZE + ty < k) Bds[ty][tx] = d_B[(ph * TILE_SIZE + ty) * l + col]; else Bds[ty][tx] = 0; __syncthreads(); for (int i = 0; i < TILE_SIZE; ++i) { CEleValue += Ads[ty][i] * Bds[i][tx]; } __syncthreads(); } if (row < j && col < l) d_C[row * l + col] = CEleValue; } // show the data in the command prompt. // this function is used for configuration // please don't use it when j,k,l is too large void showMatrix(float* matrix, int row, int col) { if (row > 5 || col > 5) return; printf("\n"); for (int i = 0; i < row; i++) { for (int j = 0; j < col; j++) { printf("%f\t", matrix[i*col + j]); } printf("\n"); } return; } // check if two matrix is the same // 0 is defined at the very begining void resultCheck(float* result_CPU, float* result_GPU, int size) { // using relative error // assuming CPU result is the true value for (int i = 0; i < size; ++i) { if ((result_CPU[i] - result_GPU[i])/ result_CPU[i] > ZERO || (result_GPU[i] - result_CPU[i])/result_CPU[i] < -ZERO) { printf("\nResult check: Error!!!! Didn't pass."); return; } } printf("\nResult check: ---PASS---."); return; }
.text .file "hw1.hip" .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 # -- Begin function main .LCPI0_0: .quad 0x41dfffffffc00000 # double 2147483647 .LCPI0_1: .quad 0x4024000000000000 # double 10 .LCPI0_2: .quad 0x3fb0000000000000 # double 0.0625 .LCPI0_3: .quad 0x3eb0c6f7a0b5ed8d # double 9.9999999999999995E-7 .LCPI0_4: .quad 0xbeb0c6f7a0b5ed8d # double -9.9999999999999995E-7 .LCPI0_5: .quad 0x0000000000000000 # double 0 .text .globl main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 subq $264, %rsp # imm = 0x108 .cfi_def_cfa_offset 320 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movq %rsi, %rbx movl %edi, %ebp xorl %edi, %edi callq time movl %eax, %edi callq srand leaq 100(%rsp), %rdx movl %ebp, %edi movq %rbx, %rsi callq _Z2UIiPPcPi testl %eax, %eax je .LBB0_4 # %bb.1: movl $.L.str, %edi .LBB0_2: # %_Z11resultCheckPfS_i.exit.sink.split xorl %eax, %eax callq printf .LBB0_3: xorl %eax, %eax addq $264, %rsp # imm = 0x108 .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .LBB0_4: .cfi_def_cfa_offset 320 movl 100(%rsp), %r12d movl 104(%rsp), %r15d movl 108(%rsp), %r14d movl $.L.str.1, %edi movl %r12d, %esi movl %r15d, %edx movl %r14d, %ecx xorl %eax, %eax callq printf movl %r15d, %ebx movq %r12, 32(%rsp) # 8-byte Spill imull %r12d, %ebx movslq %ebx, %r12 leaq (,%r12,4), %rdi movq %rdi, 128(%rsp) # 8-byte Spill callq malloc movq %rax, %rbp movl %r14d, %r13d imull %r15d, %r13d movslq %r13d, %rdi shlq $2, %rdi movq %rdi, 120(%rsp) # 8-byte Spill callq malloc movq %rax, 24(%rsp) # 8-byte Spill testl %r12d, %r12d movq %rbp, %r12 jle .LBB0_7 # %bb.5: # %.lr.ph.i movl %ebx, %ebx xorl %ebp, %ebp .p2align 4, 0x90 .LBB0_6: # =>This Inner Loop Header: Depth=1 callq rand callq rand xorps %xmm0, %xmm0 cvtsi2sd %eax, %xmm0 divsd .LCPI0_0(%rip), %xmm0 mulsd .LCPI0_1(%rip), %xmm0 addsd .LCPI0_5(%rip), %xmm0 cvtsd2ss %xmm0, %xmm0 movss %xmm0, (%r12,%rbp,4) incq %rbp cmpq %rbp, %rbx jne .LBB0_6 .LBB0_7: # %_Z10initMatrixPfiii.exit movq %r12, 40(%rsp) # 8-byte Spill testl %r13d, %r13d movq 24(%rsp), %r12 # 8-byte Reload jle .LBB0_10 # %bb.8: # %.lr.ph.i91 movl %r13d, %ebx xorl %r13d, %r13d .p2align 4, 0x90 .LBB0_9: # =>This Inner Loop Header: Depth=1 callq rand callq rand xorps %xmm0, %xmm0 cvtsi2sd %eax, %xmm0 divsd .LCPI0_0(%rip), %xmm0 mulsd .LCPI0_1(%rip), %xmm0 addsd .LCPI0_5(%rip), %xmm0 cvtsd2ss %xmm0, %xmm0 movss %xmm0, (%r12,%r13,4) incq %r13 cmpq %r13, %rbx jne .LBB0_9 .LBB0_10: # %_Z10initMatrixPfiii.exit96 movq 32(%rsp), %rbx # 8-byte Reload cmpl $5, %ebx jg .LBB0_18 # %bb.11: # %_Z10initMatrixPfiii.exit96 cmpl $5, %r15d jg .LBB0_18 # %bb.12: movl $10, %edi callq putchar@PLT testl %ebx, %ebx jle .LBB0_18 # %bb.13: # %.preheader.lr.ph.i xorl %r13d, %r13d xorl %ebx, %ebx jmp .LBB0_15 .p2align 4, 0x90 .LBB0_14: # %._crit_edge.i # in Loop: Header=BB0_15 Depth=1 movl $10, %edi callq putchar@PLT incq %rbx addl %r15d, %r13d cmpq 32(%rsp), %rbx # 8-byte Folded Reload je .LBB0_18 .LBB0_15: # %.preheader.i # =>This Loop Header: Depth=1 # Child Loop BB0_17 Depth 2 testl %r15d, %r15d jle .LBB0_14 # %bb.16: # %.lr.ph.i97 # in Loop: Header=BB0_15 Depth=1 movl %r13d, %eax movq 40(%rsp), %rcx # 8-byte Reload leaq (%rcx,%rax,4), %rbp xorl %r12d, %r12d .p2align 4, 0x90 .LBB0_17: # Parent Loop BB0_15 Depth=1 # => This Inner Loop Header: Depth=2 movss (%rbp,%r12,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str.21, %edi movb $1, %al callq printf incq %r12 cmpq %r12, %r15 jne .LBB0_17 jmp .LBB0_14 .LBB0_18: # %_Z10showMatrixPfii.exit cmpl $5, %r15d jg .LBB0_26 # %bb.19: # %_Z10showMatrixPfii.exit cmpl $5, %r14d jg .LBB0_26 # %bb.20: movl $10, %edi callq putchar@PLT testl %r15d, %r15d jle .LBB0_26 # %bb.21: # %.preheader.lr.ph.i104 xorl %r13d, %r13d xorl %ebx, %ebx jmp .LBB0_23 .p2align 4, 0x90 .LBB0_22: # %._crit_edge.i108 # in Loop: Header=BB0_23 Depth=1 movl $10, %edi callq putchar@PLT incq %rbx addl %r14d, %r13d cmpq %r15, %rbx je .LBB0_26 .LBB0_23: # %.preheader.i106 # =>This Loop Header: Depth=1 # Child Loop BB0_25 Depth 2 testl %r14d, %r14d jle .LBB0_22 # %bb.24: # %.lr.ph.i112 # in Loop: Header=BB0_23 Depth=1 movl %r13d, %eax movq 24(%rsp), %rcx # 8-byte Reload leaq (%rcx,%rax,4), %rbp xorl %r12d, %r12d .p2align 4, 0x90 .LBB0_25: # Parent Loop BB0_23 Depth=1 # => This Inner Loop Header: Depth=2 movss (%rbp,%r12,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str.21, %edi movb $1, %al callq printf incq %r12 cmpq %r12, %r14 jne .LBB0_25 jmp .LBB0_22 .LBB0_26: # %_Z10showMatrixPfii.exit117 movl $.L.str.2, %edi movq 32(%rsp), %rbp # 8-byte Reload movl %ebp, %esi movl %r15d, %edx movl %r15d, %ecx movl %r14d, %r8d xorl %eax, %eax callq printf movl %r14d, %eax imull %ebp, %eax movl %eax, 4(%rsp) # 4-byte Spill movslq %eax, %rbx shlq $2, %rbx movq %rbx, %rdi callq malloc movq %rax, 16(%rsp) # 8-byte Spill movq %rbx, 88(%rsp) # 8-byte Spill movq %rbx, %rdi callq malloc movq %rax, 8(%rsp) # 8-byte Spill cmpl $5000, %r15d # imm = 0x1388 setl %al cmpl $5000, %ebp # imm = 0x1388 setl %cl andb %al, %cl cmpl $5000, %r14d # imm = 0x1388 setl %al andb %cl, %al movb %al, 3(%rsp) # 1-byte Spill cmpb $1, %al jne .LBB0_44 # %bb.27: callq clock movq %rax, 112(%rsp) # 8-byte Spill testl %ebp, %ebp movq 8(%rsp), %r12 # 8-byte Reload jle .LBB0_36 # %bb.28: # %.preheader27.lr.ph.i movslq %r14d, %rax leaq (,%rax,4), %rcx xorl %edx, %edx xorl %esi, %esi jmp .LBB0_30 .p2align 4, 0x90 .LBB0_29: # %._crit_edge31.i # in Loop: Header=BB0_30 Depth=1 incq %rsi addl %r15d, %edx cmpq %rbp, %rsi je .LBB0_36 .LBB0_30: # %.preheader27.i # =>This Loop Header: Depth=1 # Child Loop BB0_33 Depth 2 # Child Loop BB0_35 Depth 3 testl %r14d, %r14d jle .LBB0_29 # %bb.31: # %.preheader.lr.ph.i118 # in Loop: Header=BB0_30 Depth=1 movl %edx, %edi movq 40(%rsp), %r8 # 8-byte Reload leaq (%r8,%rdi,4), %rdi movq %rsi, %r8 imulq %rax, %r8 leaq (%r12,%r8,4), %r8 movq 24(%rsp), %r9 # 8-byte Reload xorl %r10d, %r10d jmp .LBB0_33 .p2align 4, 0x90 .LBB0_32: # %._crit_edge.i120 # in Loop: Header=BB0_33 Depth=2 movss %xmm0, (%r8,%r10,4) incq %r10 addq $4, %r9 cmpq %r14, %r10 je .LBB0_29 .LBB0_33: # %.preheader.i119 # Parent Loop BB0_30 Depth=1 # => This Loop Header: Depth=2 # Child Loop BB0_35 Depth 3 xorps %xmm0, %xmm0 testl %r15d, %r15d jle .LBB0_32 # %bb.34: # %.lr.ph.preheader.i # in Loop: Header=BB0_33 Depth=2 movq %r9, %r11 xorl %ebx, %ebx .p2align 4, 0x90 .LBB0_35: # %.lr.ph.i122 # Parent Loop BB0_30 Depth=1 # Parent Loop BB0_33 Depth=2 # => This Inner Loop Header: Depth=3 movss (%rdi,%rbx,4), %xmm1 # xmm1 = mem[0],zero,zero,zero mulss (%r11), %xmm1 addss %xmm1, %xmm0 incq %rbx addq %rcx, %r11 cmpq %rbx, %r15 jne .LBB0_35 jmp .LBB0_32 .LBB0_36: # %_Z12matrixMulCPUPfS_S_iii.exit callq clock movq %rax, %r13 movl $.L.str.3, %edi xorl %eax, %eax callq printf subl 112(%rsp), %r13d # 4-byte Folded Reload movl $.L.str.4, %edi movl %r13d, %esi xorl %eax, %eax callq printf cmpl $5, %ebp jg .LBB0_44 # %bb.37: # %_Z12matrixMulCPUPfS_S_iii.exit cmpl $5, %r14d jg .LBB0_44 # %bb.38: movl $10, %edi callq putchar@PLT testl %ebp, %ebp jle .LBB0_44 # %bb.39: # %.preheader.lr.ph.i128 xorl %r13d, %r13d xorl %ebx, %ebx jmp .LBB0_41 .p2align 4, 0x90 .LBB0_40: # %._crit_edge.i132 # in Loop: Header=BB0_41 Depth=1 movl $10, %edi callq putchar@PLT incq %rbx addl %r14d, %r13d movq 32(%rsp), %rbp # 8-byte Reload cmpq %rbp, %rbx je .LBB0_44 .LBB0_41: # %.preheader.i130 # =>This Loop Header: Depth=1 # Child Loop BB0_43 Depth 2 testl %r14d, %r14d jle .LBB0_40 # %bb.42: # %.lr.ph.i136 # in Loop: Header=BB0_41 Depth=1 movl %r13d, %eax movq 8(%rsp), %rcx # 8-byte Reload leaq (%rcx,%rax,4), %rbp xorl %r12d, %r12d .p2align 4, 0x90 .LBB0_43: # Parent Loop BB0_41 Depth=1 # => This Inner Loop Header: Depth=2 movss (%rbp,%r12,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str.21, %edi movb $1, %al callq printf incq %r12 cmpq %r12, %r14 jne .LBB0_43 jmp .LBB0_40 .LBB0_44: # %_Z10showMatrixPfii.exit141 leaq 64(%rsp), %rdi movq 128(%rsp), %rbx # 8-byte Reload movq %rbx, %rsi callq hipMalloc leaq 56(%rsp), %rdi movq 120(%rsp), %r12 # 8-byte Reload movq %r12, %rsi callq hipMalloc leaq 48(%rsp), %rdi movq 88(%rsp), %rsi # 8-byte Reload callq hipMalloc movl $.L.str.5, %edi xorl %eax, %eax callq printf movq 64(%rsp), %rdi movq 40(%rsp), %rsi # 8-byte Reload movq %rbx, %rdx movl $1, %ecx callq hipMemcpy movq 56(%rsp), %rdi movq 24(%rsp), %rsi # 8-byte Reload movq %r12, %rdx movl $1, %ecx callq hipMemcpy movl $.L.str.6, %edi xorl %eax, %eax callq printf xorps %xmm0, %xmm0 cvtsi2sd %r14d, %xmm0 mulsd .LCPI0_2(%rip), %xmm0 callq ceil@PLT cvttsd2si %xmm0, %ebx xorps %xmm0, %xmm0 cvtsi2sd %ebp, %xmm0 mulsd .LCPI0_2(%rip), %xmm0 callq ceil@PLT cvttsd2si %xmm0, %r13d shlq $32, %r13 orq %rbx, %r13 movl $.L.str.7, %edi xorl %eax, %eax callq printf callq clock movq %rax, %r12 movabsq $68719476752, %rdx # imm = 0x1000000010 movq %r13, %rdi movl $1, %esi movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB0_46 # %bb.45: movq 64(%rsp), %rax movq 56(%rsp), %rcx movq 48(%rsp), %rdx movq %rax, 200(%rsp) movq %rcx, 192(%rsp) movq %rdx, 184(%rsp) movl %ebp, 84(%rsp) movl %r15d, 80(%rsp) movl %r14d, 76(%rsp) leaq 200(%rsp), %rax movq %rax, 208(%rsp) leaq 192(%rsp), %rax movq %rax, 216(%rsp) leaq 184(%rsp), %rax movq %rax, 224(%rsp) leaq 84(%rsp), %rax movq %rax, 232(%rsp) leaq 80(%rsp), %rax movq %rax, 240(%rsp) leaq 76(%rsp), %rax movq %rax, 248(%rsp) leaq 168(%rsp), %rdi leaq 152(%rsp), %rsi leaq 144(%rsp), %rdx leaq 136(%rsp), %rcx callq __hipPopCallConfiguration movq 168(%rsp), %rsi movl 176(%rsp), %edx movq 152(%rsp), %rcx movl 160(%rsp), %r8d leaq 208(%rsp), %r9 movl $_Z12matrixMulGPUPfS_S_iii, %edi pushq 136(%rsp) .cfi_adjust_cfa_offset 8 pushq 152(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB0_46: movq 48(%rsp), %rsi movq 16(%rsp), %rdi # 8-byte Reload movq 88(%rsp), %rdx # 8-byte Reload movl $2, %ecx callq hipMemcpy callq clock movq %rax, %r15 movl $.L.str.8, %edi xorl %eax, %eax callq printf subl %r12d, %r15d movl $.L.str.4, %edi movl %r15d, %esi xorl %eax, %eax callq printf cmpl $5, %ebp jg .LBB0_54 # %bb.47: cmpl $5, %r14d jg .LBB0_54 # %bb.48: movl $10, %edi callq putchar@PLT testl %ebp, %ebp jle .LBB0_54 # %bb.49: # %.preheader.lr.ph.i144 xorl %r15d, %r15d xorl %ebx, %ebx jmp .LBB0_51 .p2align 4, 0x90 .LBB0_50: # %._crit_edge.i148 # in Loop: Header=BB0_51 Depth=1 movl $10, %edi callq putchar@PLT incq %rbx addl %r14d, %r15d cmpq %rbp, %rbx je .LBB0_54 .LBB0_51: # %.preheader.i146 # =>This Loop Header: Depth=1 # Child Loop BB0_53 Depth 2 testl %r14d, %r14d jle .LBB0_50 # %bb.52: # %.lr.ph.i152 # in Loop: Header=BB0_51 Depth=1 movl %r15d, %eax movq 16(%rsp), %rcx # 8-byte Reload leaq (%rcx,%rax,4), %r12 xorl %r13d, %r13d .p2align 4, 0x90 .LBB0_53: # Parent Loop BB0_51 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r12,%r13,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str.21, %edi movb $1, %al callq printf incq %r13 cmpq %r13, %r14 jne .LBB0_53 jmp .LBB0_50 .LBB0_54: # %_Z10showMatrixPfii.exit157 movq 64(%rsp), %rdi callq hipFree movq 56(%rsp), %rdi callq hipFree movq 48(%rsp), %rdi callq hipFree cmpb $0, 3(%rsp) # 1-byte Folded Reload je .LBB0_3 # %bb.55: cmpl $0, 4(%rsp) # 4-byte Folded Reload jle .LBB0_64 # %bb.56: # %.lr.ph.preheader.i158 movq 8(%rsp), %rax # 8-byte Reload movss (%rax), %xmm0 # xmm0 = mem[0],zero,zero,zero movq 16(%rsp), %rax # 8-byte Reload movss (%rax), %xmm1 # xmm1 = mem[0],zero,zero,zero movaps %xmm0, %xmm2 subss %xmm1, %xmm2 divss %xmm0, %xmm2 cvtss2sd %xmm2, %xmm2 movl $.L.str.22, %edi ucomisd .LCPI0_3(%rip), %xmm2 ja .LBB0_2 # %bb.57: # %.lr.ph.preheader subss %xmm0, %xmm1 divss %xmm0, %xmm1 cvtss2sd %xmm1, %xmm1 movsd .LCPI0_4(%rip), %xmm0 # xmm0 = mem[0],zero ucomisd %xmm1, %xmm0 ja .LBB0_2 # %bb.58: # %.lr.ph183.preheader movl 4(%rsp), %ebx # 4-byte Reload movl $1, %eax movsd .LCPI0_3(%rip), %xmm1 # xmm1 = mem[0],zero movq 8(%rsp), %rcx # 8-byte Reload .p2align 4, 0x90 .LBB0_59: # %.lr.ph183 # =>This Inner Loop Header: Depth=1 cmpq %rax, %rbx je .LBB0_64 # %bb.60: # %.lr.ph.i160 # in Loop: Header=BB0_59 Depth=1 movq %rax, %r14 movss (%rcx,%rax,4), %xmm2 # xmm2 = mem[0],zero,zero,zero movq 16(%rsp), %rax # 8-byte Reload movss (%rax,%r14,4), %xmm3 # xmm3 = mem[0],zero,zero,zero movaps %xmm2, %xmm4 subss %xmm3, %xmm4 divss %xmm2, %xmm4 cvtss2sd %xmm4, %xmm4 ucomisd %xmm1, %xmm4 ja .LBB0_62 # %bb.61: # %.lr.ph # in Loop: Header=BB0_59 Depth=1 subss %xmm2, %xmm3 divss %xmm2, %xmm3 xorps %xmm2, %xmm2 cvtss2sd %xmm3, %xmm2 leaq 1(%r14), %rax ucomisd %xmm2, %xmm0 jbe .LBB0_59 .LBB0_62: # %._crit_edge movl $.L.str.22, %edi xorl %eax, %eax callq printf movl $.L.str.23, %edi cmpq %rbx, %r14 jae .LBB0_2 jmp .LBB0_3 .LBB0_64: movl $.L.str.23, %edi jmp .LBB0_2 .Lfunc_end0: .size main, .Lfunc_end0-main .cfi_endproc # -- End function .globl _Z2UIiPPcPi # -- Begin function _Z2UIiPPcPi .p2align 4, 0x90 .type _Z2UIiPPcPi,@function _Z2UIiPPcPi: # @_Z2UIiPPcPi .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r12 .cfi_def_cfa_offset 40 pushq %rbx .cfi_def_cfa_offset 48 .cfi_offset %rbx, -48 .cfi_offset %r12, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movq %rdx, %rbx movq %rsi, %r14 cmpl $4, %edi je .LBB1_9 # %bb.1: cmpl $2, %edi jne .LBB1_7 # %bb.2: movl %edi, %ebp movq 8(%r14), %r15 movl $.L.str.9, %esi movq %r15, %rdi callq strcmp testl %eax, %eax je .LBB1_5 # %bb.3: movl $.L.str.10, %esi movq %r15, %rdi callq strcmp testl %eax, %eax je .LBB1_5 # %bb.4: movl $.L.str.11, %esi movq %r15, %rdi callq strcmp testl %eax, %eax je .LBB1_5 # %bb.6: cmpl $4, %ebp jne .LBB1_7 .LBB1_9: movl $.L.str.16, %edi movl $4, %esi xorl %eax, %eax callq printf movl $.L.str.17, %edi xorl %eax, %eax callq printf movq 8(%r14), %rdi xorl %esi, %esi movl $10, %edx callq __isoc23_strtol movq %rax, %r12 testl %r12d, %r12d jle .LBB1_10 # %bb.11: # %.lr.ph.preheader xorl %r15d, %r15d .p2align 4, 0x90 .LBB1_13: # %.lr.ph # =>This Inner Loop Header: Depth=1 leal 1(%r15), %esi movl $.L.str.18, %edi movl %r12d, %edx xorl %eax, %eax callq printf movl %r12d, (%rbx,%r15,4) cmpq $2, %r15 je .LBB1_14 # %bb.12: # in Loop: Header=BB1_13 Depth=1 movq 16(%r14,%r15,8), %rdi xorl %esi, %esi movl $10, %edx callq __isoc23_strtol movq %rax, %r12 incq %r15 testl %r12d, %r12d jg .LBB1_13 # %bb.15: # %._crit_edge cmpq $3, %r15 setb %bl incl %r15d jmp .LBB1_16 .LBB1_7: movl $.Lstr, %edi jmp .LBB1_8 .LBB1_5: movl $.Lstr.1, %edi callq puts@PLT movl $.Lstr.2, %edi callq puts@PLT movl $.Lstr.3, %edi .LBB1_8: callq puts@PLT movl $1, %eax jmp .LBB1_18 .LBB1_10: movb $1, %bl movl $1, %r15d .LBB1_16: movl $.L.str.19, %edi movl %r15d, %esi xorl %eax, %eax callq printf jmp .LBB1_17 .LBB1_14: # %.loopexit.loopexit incq %r15 cmpq $3, %r15 setb %bl .LBB1_17: # %.loopexit movzbl %bl, %eax .LBB1_18: popq %rbx .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end1: .size _Z2UIiPPcPi, .Lfunc_end1-_Z2UIiPPcPi .cfi_endproc # -- End function .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 # -- Begin function _Z10initMatrixPfiii .LCPI2_0: .quad 0x41dfffffffc00000 # double 2147483647 .text .globl _Z10initMatrixPfiii .p2align 4, 0x90 .type _Z10initMatrixPfiii,@function _Z10initMatrixPfiii: # @_Z10initMatrixPfiii .cfi_startproc # %bb.0: testl %esi, %esi jle .LBB2_4 # %bb.1: # %.lr.ph pushq %r15 .cfi_def_cfa_offset 16 pushq %r14 .cfi_def_cfa_offset 24 pushq %rbx .cfi_def_cfa_offset 32 subq $16, %rsp .cfi_def_cfa_offset 48 .cfi_offset %rbx, -32 .cfi_offset %r14, -24 .cfi_offset %r15, -16 movq %rdi, %rbx subl %edx, %ecx cvtsi2sd %ecx, %xmm0 movsd %xmm0, 8(%rsp) # 8-byte Spill xorps %xmm0, %xmm0 cvtsi2sd %edx, %xmm0 movsd %xmm0, (%rsp) # 8-byte Spill movl %esi, %r14d xorl %r15d, %r15d .p2align 4, 0x90 .LBB2_2: # =>This Inner Loop Header: Depth=1 callq rand callq rand xorps %xmm0, %xmm0 cvtsi2sd %eax, %xmm0 divsd .LCPI2_0(%rip), %xmm0 mulsd 8(%rsp), %xmm0 # 8-byte Folded Reload addsd (%rsp), %xmm0 # 8-byte Folded Reload cvtsd2ss %xmm0, %xmm0 movss %xmm0, (%rbx,%r15,4) incq %r15 cmpq %r15, %r14 jne .LBB2_2 # %bb.3: addq $16, %rsp .cfi_def_cfa_offset 32 popq %rbx .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 .cfi_restore %rbx .cfi_restore %r14 .cfi_restore %r15 .LBB2_4: # %._crit_edge retq .Lfunc_end2: .size _Z10initMatrixPfiii, .Lfunc_end2-_Z10initMatrixPfiii .cfi_endproc # -- End function .globl _Z10showMatrixPfii # -- Begin function _Z10showMatrixPfii .p2align 4, 0x90 .type _Z10showMatrixPfii,@function _Z10showMatrixPfii: # @_Z10showMatrixPfii .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 subq $24, %rsp .cfi_def_cfa_offset 80 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movq %rdi, 8(%rsp) # 8-byte Spill cmpl $5, %esi jg .LBB3_8 # %bb.1: movl %edx, %ebx cmpl $5, %edx jg .LBB3_8 # %bb.2: movl %esi, %ebp movl $10, %edi callq putchar@PLT testl %ebp, %ebp jle .LBB3_8 # %bb.3: # %.preheader.lr.ph movl %ebp, %eax movq %rax, 16(%rsp) # 8-byte Spill movl %ebx, %r12d xorl %r13d, %r13d xorl %ebp, %ebp jmp .LBB3_4 .p2align 4, 0x90 .LBB3_7: # %._crit_edge # in Loop: Header=BB3_4 Depth=1 movl $10, %edi callq putchar@PLT incq %rbp addl %ebx, %r13d cmpq 16(%rsp), %rbp # 8-byte Folded Reload je .LBB3_8 .LBB3_4: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB3_6 Depth 2 testl %ebx, %ebx jle .LBB3_7 # %bb.5: # %.lr.ph # in Loop: Header=BB3_4 Depth=1 movl %r13d, %eax movq 8(%rsp), %rcx # 8-byte Reload leaq (%rcx,%rax,4), %r14 xorl %r15d, %r15d .p2align 4, 0x90 .LBB3_6: # Parent Loop BB3_4 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r14,%r15,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str.21, %edi movb $1, %al callq printf incq %r15 cmpq %r15, %r12 jne .LBB3_6 jmp .LBB3_7 .LBB3_8: # %.loopexit addq $24, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end3: .size _Z10showMatrixPfii, .Lfunc_end3-_Z10showMatrixPfii .cfi_endproc # -- End function .globl _Z12matrixMulCPUPfS_S_iii # -- Begin function _Z12matrixMulCPUPfS_S_iii .p2align 4, 0x90 .type _Z12matrixMulCPUPfS_S_iii,@function _Z12matrixMulCPUPfS_S_iii: # @_Z12matrixMulCPUPfS_S_iii .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movq %rdx, -8(%rsp) # 8-byte Spill movq %rsi, -16(%rsp) # 8-byte Spill movq %rdi, -24(%rsp) # 8-byte Spill testl %ecx, %ecx jle .LBB4_9 # %bb.1: # %.preheader27.lr.ph movslq %r9d, %rdi movl %ecx, %ecx movl %edi, %r10d movl %r8d, %r11d leaq (,%rdi,4), %rbx xorl %r14d, %r14d xorl %r15d, %r15d jmp .LBB4_2 .p2align 4, 0x90 .LBB4_8: # %._crit_edge31 # in Loop: Header=BB4_2 Depth=1 incq %r15 addl %r8d, %r14d cmpq %rcx, %r15 je .LBB4_9 .LBB4_2: # %.preheader27 # =>This Loop Header: Depth=1 # Child Loop BB4_4 Depth 2 # Child Loop BB4_6 Depth 3 testl %r9d, %r9d jle .LBB4_8 # %bb.3: # %.preheader.lr.ph # in Loop: Header=BB4_2 Depth=1 movl %r14d, %eax movq -24(%rsp), %rdx # 8-byte Reload leaq (%rdx,%rax,4), %r12 movq %r15, %rax imulq %rdi, %rax movq -8(%rsp), %rdx # 8-byte Reload leaq (%rdx,%rax,4), %r13 movq -16(%rsp), %rsi # 8-byte Reload xorl %eax, %eax jmp .LBB4_4 .p2align 4, 0x90 .LBB4_7: # %._crit_edge # in Loop: Header=BB4_4 Depth=2 movss %xmm0, (%r13,%rax,4) incq %rax addq $4, %rsi cmpq %r10, %rax je .LBB4_8 .LBB4_4: # %.preheader # Parent Loop BB4_2 Depth=1 # => This Loop Header: Depth=2 # Child Loop BB4_6 Depth 3 xorps %xmm0, %xmm0 testl %r8d, %r8d jle .LBB4_7 # %bb.5: # %.lr.ph.preheader # in Loop: Header=BB4_4 Depth=2 movq %rsi, %rbp xorl %edx, %edx .p2align 4, 0x90 .LBB4_6: # %.lr.ph # Parent Loop BB4_2 Depth=1 # Parent Loop BB4_4 Depth=2 # => This Inner Loop Header: Depth=3 movss (%r12,%rdx,4), %xmm1 # xmm1 = mem[0],zero,zero,zero mulss (%rbp), %xmm1 addss %xmm1, %xmm0 incq %rdx addq %rbx, %rbp cmpq %rdx, %r11 jne .LBB4_6 jmp .LBB4_7 .LBB4_9: # %._crit_edge33 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end4: .size _Z12matrixMulCPUPfS_S_iii, .Lfunc_end4-_Z12matrixMulCPUPfS_S_iii .cfi_endproc # -- End function .globl _Z27__device_stub__matrixMulGPUPfS_S_iii # -- Begin function _Z27__device_stub__matrixMulGPUPfS_S_iii .p2align 4, 0x90 .type _Z27__device_stub__matrixMulGPUPfS_S_iii,@function _Z27__device_stub__matrixMulGPUPfS_S_iii: # @_Z27__device_stub__matrixMulGPUPfS_S_iii .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 88(%rsp) movq %rsi, 80(%rsp) movq %rdx, 72(%rsp) movl %ecx, 20(%rsp) movl %r8d, 16(%rsp) movl %r9d, 12(%rsp) leaq 88(%rsp), %rax movq %rax, 96(%rsp) leaq 80(%rsp), %rax movq %rax, 104(%rsp) leaq 72(%rsp), %rax movq %rax, 112(%rsp) leaq 20(%rsp), %rax movq %rax, 120(%rsp) leaq 16(%rsp), %rax movq %rax, 128(%rsp) leaq 12(%rsp), %rax movq %rax, 136(%rsp) leaq 56(%rsp), %rdi leaq 40(%rsp), %rsi leaq 32(%rsp), %rdx leaq 24(%rsp), %rcx callq __hipPopCallConfiguration movq 56(%rsp), %rsi movl 64(%rsp), %edx movq 40(%rsp), %rcx movl 48(%rsp), %r8d leaq 96(%rsp), %r9 movl $_Z12matrixMulGPUPfS_S_iii, %edi pushq 24(%rsp) .cfi_adjust_cfa_offset 8 pushq 40(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $168, %rsp .cfi_adjust_cfa_offset -168 retq .Lfunc_end5: .size _Z27__device_stub__matrixMulGPUPfS_S_iii, .Lfunc_end5-_Z27__device_stub__matrixMulGPUPfS_S_iii .cfi_endproc # -- End function .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 # -- Begin function _Z11resultCheckPfS_i .LCPI6_0: .quad 0x3eb0c6f7a0b5ed8d # double 9.9999999999999995E-7 .LCPI6_1: .quad 0xbeb0c6f7a0b5ed8d # double -9.9999999999999995E-7 .text .globl _Z11resultCheckPfS_i .p2align 4, 0x90 .type _Z11resultCheckPfS_i,@function _Z11resultCheckPfS_i: # @_Z11resultCheckPfS_i .cfi_startproc # %bb.0: pushq %rbx .cfi_def_cfa_offset 16 .cfi_offset %rbx, -16 testl %edx, %edx setg %bl jle .LBB6_8 # %bb.1: # %.lr.ph.preheader movss (%rdi), %xmm0 # xmm0 = mem[0],zero,zero,zero movss (%rsi), %xmm2 # xmm2 = mem[0],zero,zero,zero movaps %xmm0, %xmm1 subss %xmm2, %xmm1 divss %xmm0, %xmm1 cvtss2sd %xmm1, %xmm1 ucomisd .LCPI6_0(%rip), %xmm1 ja .LBB6_6 # %bb.2: # %.lr.ph24.preheader movl %edx, %eax movl $1, %ecx movsd .LCPI6_1(%rip), %xmm1 # xmm1 = mem[0],zero movsd .LCPI6_0(%rip), %xmm3 # xmm3 = mem[0],zero .p2align 4, 0x90 .LBB6_5: # %.lr.ph24 # =>This Inner Loop Header: Depth=1 subss %xmm0, %xmm2 divss %xmm0, %xmm2 xorps %xmm0, %xmm0 cvtss2sd %xmm2, %xmm0 ucomisd %xmm0, %xmm1 ja .LBB6_6 # %bb.3: # in Loop: Header=BB6_5 Depth=1 cmpq %rcx, %rax je .LBB6_8 # %bb.4: # %.lr.ph # in Loop: Header=BB6_5 Depth=1 seta %bl movss (%rdi,%rcx,4), %xmm0 # xmm0 = mem[0],zero,zero,zero movss (%rsi,%rcx,4), %xmm2 # xmm2 = mem[0],zero,zero,zero movaps %xmm0, %xmm4 subss %xmm2, %xmm4 divss %xmm0, %xmm4 cvtss2sd %xmm4, %xmm4 incq %rcx ucomisd %xmm3, %xmm4 jbe .LBB6_5 .LBB6_6: # %._crit_edge movl $.L.str.22, %edi xorl %eax, %eax callq printf testb $1, %bl je .LBB6_8 # %bb.7: popq %rbx .cfi_def_cfa_offset 8 retq .LBB6_8: # %.critedge .cfi_def_cfa_offset 16 movl $.L.str.23, %edi xorl %eax, %eax popq %rbx .cfi_def_cfa_offset 8 jmp printf # TAILCALL .Lfunc_end6: .size _Z11resultCheckPfS_i, .Lfunc_end6-_Z11resultCheckPfS_i .cfi_endproc # -- End function .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 # -- Begin function _Z12randGenerateii .LCPI7_0: .quad 0x41dfffffffc00000 # double 2147483647 .text .globl _Z12randGenerateii .p2align 4, 0x90 .type _Z12randGenerateii,@function _Z12randGenerateii: # @_Z12randGenerateii .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %rbx .cfi_def_cfa_offset 24 pushq %rax .cfi_def_cfa_offset 32 .cfi_offset %rbx, -24 .cfi_offset %rbp, -16 movl %esi, %ebx movl %edi, %ebp callq rand callq rand cvtsi2sd %eax, %xmm0 divsd .LCPI7_0(%rip), %xmm0 subl %ebp, %ebx cvtsi2sd %ebx, %xmm1 mulsd %xmm0, %xmm1 xorps %xmm0, %xmm0 cvtsi2sd %ebp, %xmm0 addsd %xmm1, %xmm0 cvtsd2ss %xmm0, %xmm0 addq $8, %rsp .cfi_def_cfa_offset 24 popq %rbx .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end7: .size _Z12randGenerateii, .Lfunc_end7-_Z12randGenerateii .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB8_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB8_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z12matrixMulGPUPfS_S_iii, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end8: .size __hip_module_ctor, .Lfunc_end8-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB9_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB9_2: retq .Lfunc_end9: .size __hip_module_dtor, .Lfunc_end9-__hip_module_dtor .cfi_endproc # -- End function .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "\nApplication terminates." .size .L.str, 25 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "\nContinuing with j=%d, k=%d, l=%d" .size .L.str.1, 34 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "\nDone initializing matrix A with size %d*%d and matrix B with %d*%d" .size .L.str.2, 68 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "\nDone matrix multiplication with CPU" .size .L.str.3, 37 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "\n%d microseconds used.\n" .size .L.str.4, 24 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "\nDone allocating space in device." .size .L.str.5, 34 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz "\nDone copying memory from host to device" .size .L.str.6, 41 .type .L.str.7,@object # @.str.7 .L.str.7: .asciz "\nDone initializing block dimention and grid dimention." .size .L.str.7, 55 .type _Z12matrixMulGPUPfS_S_iii,@object # @_Z12matrixMulGPUPfS_S_iii .section .rodata,"a",@progbits .globl _Z12matrixMulGPUPfS_S_iii .p2align 3, 0x0 _Z12matrixMulGPUPfS_S_iii: .quad _Z27__device_stub__matrixMulGPUPfS_S_iii .size _Z12matrixMulGPUPfS_S_iii, 8 .type .L.str.8,@object # @.str.8 .section .rodata.str1.1,"aMS",@progbits,1 .L.str.8: .asciz "\nDone matrix multiplication with GPU." .size .L.str.8, 38 .type .L.str.9,@object # @.str.9 .L.str.9: .asciz "--help" .size .L.str.9, 7 .type .L.str.10,@object # @.str.10 .L.str.10: .asciz "-h" .size .L.str.10, 3 .type .L.str.11,@object # @.str.11 .L.str.11: .asciz "?" .size .L.str.11, 2 .type .L.str.16,@object # @.str.16 .L.str.16: .asciz "\nNumber Of Arguments Passed: %d" .size .L.str.16, 32 .type .L.str.17,@object # @.str.17 .L.str.17: .asciz "\n----Following Are The Command Line Arguments Passed----" .size .L.str.17, 57 .type .L.str.18,@object # @.str.18 .L.str.18: .asciz "\nargv[%d]: %d" .size .L.str.18, 14 .type .L.str.19,@object # @.str.19 .L.str.19: .asciz "\nargv[%d] is invalid" .size .L.str.19, 21 .type .L.str.21,@object # @.str.21 .L.str.21: .asciz "%f\t" .size .L.str.21, 4 .type .L.str.22,@object # @.str.22 .L.str.22: .asciz "\nResult check: Error!!!! Didn't pass." .size .L.str.22, 38 .type .L.str.23,@object # @.str.23 .L.str.23: .asciz "\nResult check: ---PASS---." .size .L.str.23, 27 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z12matrixMulGPUPfS_S_iii" .size .L__unnamed_1, 26 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .type .Lstr,@object # @str .section .rodata.str1.1,"aMS",@progbits,1 .Lstr: .asciz "\nError.\nWe need 3 parameters as input.\nYou can use '--help' for more details." .size .Lstr, 78 .type .Lstr.1,@object # @str.1 .Lstr.1: .asciz "\nWe need 3 parameters j,k,l which represents the two matrix's dimention." .size .Lstr.1, 73 .type .Lstr.2,@object # @str.2 .Lstr.2: .asciz "The first one with size j*k, and the second one's is k*l." .size .Lstr.2, 58 .type .Lstr.3,@object # @str.3 .Lstr.3: .asciz "j,k,l must be int type and greater than 0." .size .Lstr.3, 43 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z27__device_stub__matrixMulGPUPfS_S_iii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z12matrixMulGPUPfS_S_iii .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA device assembly to AMD device assembly.
code for sm_80 Function : _Z12matrixMulGPUPfS_S_iii .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R3, SR_CTAID.X ; /* 0x0000000000037919 */ /* 0x000e220000002500 */ /*0020*/ MOV R4, c[0x0][0x17c] ; /* 0x00005f0000047a02 */ /* 0x000fe20000000f00 */ /*0030*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*0040*/ HFMA2.MMA R9, -RZ, RZ, 0, 0 ; /* 0x00000000ff097435 */ /* 0x000fe200000001ff */ /*0050*/ S2R R16, SR_TID.X ; /* 0x0000000000107919 */ /* 0x000e220000002100 */ /*0060*/ ISETP.GE.AND P1, PT, R4, 0x1, PT ; /* 0x000000010400780c */ /* 0x000fc60003f26270 */ /*0070*/ S2R R5, SR_CTAID.Y ; /* 0x0000000000057919 */ /* 0x000e680000002600 */ /*0080*/ S2R R2, SR_TID.Y ; /* 0x0000000000027919 */ /* 0x000e620000002200 */ /*0090*/ IMAD R3, R3, c[0x0][0x0], R16 ; /* 0x0000000003037a24 */ /* 0x001fca00078e0210 */ /*00a0*/ ISETP.GE.AND P0, PT, R3, c[0x0][0x180], PT ; /* 0x0000600003007a0c */ /* 0x000fe20003f06270 */ /*00b0*/ IMAD R0, R5, c[0x0][0x4], R2 ; /* 0x0000010005007a24 */ /* 0x002fca00078e0202 */ /*00c0*/ ISETP.GE.OR P0, PT, R0, c[0x0][0x178], P0 ; /* 0x00005e0000007a0c */ /* 0x000fe20000706670 */ /*00d0*/ @!P1 BRA 0x4f0 ; /* 0x0000041000009947 */ /* 0x000fd80003800000 */ /*00e0*/ SHF.L.U32 R17, R2, 0x6, RZ ; /* 0x0000000602117819 */ /* 0x000fe400000006ff */ /*00f0*/ MOV R5, RZ ; /* 0x000000ff00057202 */ /* 0x000fe40000000f00 */ /*0100*/ MOV R18, RZ ; /* 0x000000ff00127202 */ /* 0x000fe40000000f00 */ /*0110*/ MOV R9, RZ ; /* 0x000000ff00097202 */ /* 0x000fe40000000f00 */ /*0120*/ LEA R19, R16, R17, 0x2 ; /* 0x0000001110137211 */ /* 0x000fe400078e10ff */ /*0130*/ IADD3 R7, R16, R5.reuse, RZ ; /* 0x0000000510077210 */ /* 0x080fe20007ffe0ff */ /*0140*/ HFMA2.MMA R8, -RZ, RZ, 0, 0 ; /* 0x00000000ff087435 */ /* 0x000fe200000001ff */ /*0150*/ IADD3 R10, R2, R5, RZ ; /* 0x00000005020a7210 */ /* 0x000fc40007ffe0ff */ /*0160*/ ISETP.GE.AND P2, PT, R7, c[0x0][0x17c], PT ; /* 0x00005f0007007a0c */ /* 0x000fe40003f46270 */ /*0170*/ ISETP.GE.AND P1, PT, R10, c[0x0][0x17c], PT ; /* 0x00005f000a007a0c */ /* 0x000fe40003f26270 */ /*0180*/ ISETP.GE.OR P2, PT, R0, c[0x0][0x178], P2 ; /* 0x00005e0000007a0c */ /* 0x000fe40001746670 */ /*0190*/ ISETP.GE.OR P1, PT, R3, c[0x0][0x180], P1 ; /* 0x0000600003007a0c */ /* 0x000fe40000f26670 */ /*01a0*/ MOV R20, RZ ; /* 0x000000ff00147202 */ /* 0x000fd20000000f00 */ /*01b0*/ @!P2 MOV R5, 0x4 ; /* 0x000000040005a802 */ /* 0x000fe20000000f00 */ /*01c0*/ @!P2 IMAD R4, R0, c[0x0][0x17c], R7 ; /* 0x00005f000004aa24 */ /* 0x000fe200078e0207 */ /*01d0*/ @!P1 MOV R11, 0x4 ; /* 0x00000004000b9802 */ /* 0x000fe20000000f00 */ /*01e0*/ @!P1 IMAD R10, R10, c[0x0][0x180], R3 ; /* 0x000060000a0a9a24 */ /* 0x000fe400078e0203 */ /*01f0*/ @!P2 IMAD.WIDE R4, R4, R5, c[0x0][0x160] ; /* 0x000058000404a625 */ /* 0x000fc800078e0205 */ /*0200*/ @!P1 IMAD.WIDE R10, R10, R11, c[0x0][0x168] ; /* 0x00005a000a0a9625 */ /* 0x000fe200078e020b */ /*0210*/ @!P2 LDG.E R8, [R4.64] ; /* 0x000000040408a981 */ /* 0x000ea8000c1e1900 */ /*0220*/ @!P1 LDG.E R20, [R10.64] ; /* 0x000000040a149981 */ /* 0x000ee2000c1e1900 */ /*0230*/ IADD3 R18, R18, 0x1, RZ ; /* 0x0000000112127810 */ /* 0x000fc60007ffe0ff */ /*0240*/ STS [R19], R8 ; /* 0x0000000813007388 */ /* 0x004fe80000000800 */ /*0250*/ STS [R19+0x400], R20 ; /* 0x0004001413007388 */ /* 0x008fe80000000800 */ /*0260*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0270*/ LDS R23, [R16.X4+0x400] ; /* 0x0004000010177984 */ /* 0x000fe80000004800 */ /*0280*/ LDS.128 R12, [R17] ; /* 0x00000000110c7984 */ /* 0x000e280000000c00 */ /*0290*/ LDS R24, [R16.X4+0x440] ; /* 0x0004400010187984 */ /* 0x000e680000004800 */ /*02a0*/ LDS R25, [R16.X4+0x480] ; /* 0x0004800010197984 */ /* 0x000ea80000004800 */ /*02b0*/ LDS R26, [R16.X4+0x4c0] ; /* 0x0004c000101a7984 */ /* 0x000ee80000004800 */ /*02c0*/ LDS R27, [R16.X4+0x500] ; /* 0x00050000101b7984 */ /* 0x000fe80000004800 */ /*02d0*/ LDS.128 R4, [R17+0x10] ; /* 0x0000100011047984 */ /* 0x000f280000000c00 */ /*02e0*/ LDS R22, [R16.X4+0x540] ; /* 0x0005400010167984 */ /* 0x000f680000004800 */ /*02f0*/ LDS R21, [R16.X4+0x580] ; /* 0x0005800010157984 */ /* 0x000f680000004800 */ /*0300*/ LDS R20, [R16.X4+0x5c0] ; /* 0x0005c00010147984 */ /* 0x000f620000004800 */ /*0310*/ FFMA R12, R23, R12, R9 ; /* 0x0000000c170c7223 */ /* 0x001fc60000000009 */ /*0320*/ LDS R23, [R16.X4+0x600] ; /* 0x0006000010177984 */ /* 0x000fe20000004800 */ /*0330*/ FFMA R12, R24, R13, R12 ; /* 0x0000000d180c7223 */ /* 0x002fc6000000000c */ /*0340*/ LDS.128 R8, [R17+0x20] ; /* 0x0000200011087984 */ /* 0x000e220000000c00 */ /*0350*/ FFMA R12, R25, R14, R12 ; /* 0x0000000e190c7223 */ /* 0x004fc6000000000c */ /*0360*/ LDS R24, [R16.X4+0x640] ; /* 0x0006400010187984 */ /* 0x000e620000004800 */ /*0370*/ FFMA R12, R26, R15, R12 ; /* 0x0000000f1a0c7223 */ /* 0x008fc6000000000c */ /*0380*/ LDS R25, [R16.X4+0x680] ; /* 0x0006800010197984 */ /* 0x000ea20000004800 */ /*0390*/ FFMA R12, R27, R4, R12 ; /* 0x000000041b0c7223 */ /* 0x010fc6000000000c */ /*03a0*/ LDS R4, [R16.X4+0x6c0] ; /* 0x0006c00010047984 */ /* 0x000ee20000004800 */ /*03b0*/ FFMA R22, R22, R5, R12 ; /* 0x0000000516167223 */ /* 0x020fc6000000000c */ /*03c0*/ LDS R5, [R16.X4+0x700] ; /* 0x0007000010057984 */ /* 0x000fe20000004800 */ /*03d0*/ FFMA R26, R21, R6, R22 ; /* 0x00000006151a7223 */ /* 0x000fc60000000016 */ /*03e0*/ LDS.128 R12, [R17+0x30] ; /* 0x00003000110c7984 */ /* 0x000f220000000c00 */ /*03f0*/ FFMA R7, R20, R7, R26 ; /* 0x0000000714077223 */ /* 0x000fc6000000001a */ /*0400*/ LDS R22, [R16.X4+0x740] ; /* 0x0007400010167984 */ /* 0x000f680000004800 */ /*0410*/ LDS R21, [R16.X4+0x780] ; /* 0x0007800010157984 */ /* 0x000f680000004800 */ /*0420*/ LDS R6, [R16.X4+0x7c0] ; /* 0x0007c00010067984 */ /* 0x000f620000004800 */ /*0430*/ FFMA R7, R23, R8, R7 ; /* 0x0000000817077223 */ /* 0x001fc80000000007 */ /*0440*/ FFMA R7, R24, R9, R7 ; /* 0x0000000918077223 */ /* 0x002fc80000000007 */ /*0450*/ FFMA R7, R25, R10, R7 ; /* 0x0000000a19077223 */ /* 0x004fc80000000007 */ /*0460*/ FFMA R4, R4, R11, R7 ; /* 0x0000000b04047223 */ /* 0x008fc80000000007 */ /*0470*/ FFMA R4, R5, R12, R4 ; /* 0x0000000c05047223 */ /* 0x010fe20000000004 */ /*0480*/ SHF.L.U32 R5, R18, 0x4, RZ ; /* 0x0000000412057819 */ /* 0x000fc600000006ff */ /*0490*/ FFMA R4, R22, R13, R4 ; /* 0x0000000d16047223 */ /* 0x020fe20000000004 */ /*04a0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fe20000010000 */ /*04b0*/ ISETP.GE.AND P1, PT, R5, c[0x0][0x17c], PT ; /* 0x00005f0005007a0c */ /* 0x000fe40003f26270 */ /*04c0*/ FFMA R4, R21, R14, R4 ; /* 0x0000000e15047223 */ /* 0x000fc80000000004 */ /*04d0*/ FFMA R9, R6, R15, R4 ; /* 0x0000000f06097223 */ /* 0x000fce0000000004 */ /*04e0*/ @!P1 BRA 0x130 ; /* 0xfffffc4000009947 */ /* 0x000fea000383ffff */ /*04f0*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0500*/ HFMA2.MMA R2, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff027435 */ /* 0x000fe200000001ff */ /*0510*/ IMAD R3, R0, c[0x0][0x180], R3 ; /* 0x0000600000037a24 */ /* 0x000fd200078e0203 */ /*0520*/ IMAD.WIDE R2, R3, R2, c[0x0][0x170] ; /* 0x00005c0003027625 */ /* 0x000fca00078e0202 */ /*0530*/ STG.E [R2.64], R9 ; /* 0x0000000902007986 */ /* 0x000fe2000c101904 */ /*0540*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0550*/ BRA 0x550; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0560*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0570*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0580*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0590*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*05f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z12matrixMulGPUPfS_S_iii .globl _Z12matrixMulGPUPfS_S_iii .p2align 8 .type _Z12matrixMulGPUPfS_S_iii,@function _Z12matrixMulGPUPfS_S_iii: s_clause 0x2 s_load_b32 s2, s[0:1], 0x34 s_load_b64 s[8:9], s[0:1], 0x18 s_load_b32 s3, s[0:1], 0x20 v_bfe_u32 v3, v0, 10, 10 v_and_b32_e32 v4, 0x3ff, v0 s_waitcnt lgkmcnt(0) s_lshr_b32 s4, s2, 16 s_and_b32 s2, s2, 0xffff s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_2) v_mad_u64_u32 v[0:1], null, s15, s4, v[3:4] v_mad_u64_u32 v[1:2], null, s14, s2, v[4:5] s_cmp_lt_i32 s9, 1 v_cmp_gt_i32_e64 s2, s8, v0 s_delay_alu instid0(VALU_DEP_2) v_cmp_gt_i32_e32 vcc_lo, s3, v1 s_cbranch_scc1 .LBB0_15 s_load_b128 s[4:7], s[0:1], 0x0 v_lshlrev_b32_e32 v2, 2, v4 v_dual_mov_b32 v10, 0 :: v_dual_lshlrev_b32 v7, 6, v3 v_mad_u64_u32 v[5:6], null, v0, s9, v[4:5] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_nc_u32_e32 v8, 0x400, v2 v_dual_mov_b32 v2, 0 :: v_dual_add_nc_u32 v9, v7, v2 v_cmp_le_i32_e64 s10, s3, v1 s_mov_b32 s13, 0 s_delay_alu instid0(VALU_DEP_3) v_add_nc_u32_e32 v6, v8, v7 s_xor_b32 s11, s2, -1 s_mov_b32 s12, 0 .LBB0_2: v_add_nc_u32_e32 v12, s13, v4 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cmp_le_i32_e64 s2, s9, v12 s_or_b32 s2, s11, s2 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_and_saveexec_b32 s14, s2 s_xor_b32 s2, exec_lo, s14 s_cbranch_execz .LBB0_4 ds_store_b32 v9, v10 .LBB0_4: s_and_not1_saveexec_b32 s14, s2 s_cbranch_execz .LBB0_6 v_add_nc_u32_e32 v12, s13, v5 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v13, 31, v12 v_lshlrev_b64 v[12:13], 2, v[12:13] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s2, s4, v12 v_add_co_ci_u32_e64 v13, s2, s5, v13, s2 global_load_b32 v12, v[12:13], off s_waitcnt vmcnt(0) ds_store_b32 v9, v12 .LBB0_6: s_or_b32 exec_lo, exec_lo, s14 s_mov_b32 s14, 0 s_mov_b32 s2, s10 s_and_saveexec_b32 s15, vcc_lo v_add_nc_u32_e32 v11, s13, v3 s_and_not1_b32 s13, s10, exec_lo s_mov_b32 s14, exec_lo s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cmp_le_i32_e64 s2, s9, v11 s_and_b32 s2, s2, exec_lo s_delay_alu instid0(SALU_CYCLE_1) s_or_b32 s2, s13, s2 s_or_b32 exec_lo, exec_lo, s15 s_and_saveexec_b32 s13, s2 s_delay_alu instid0(SALU_CYCLE_1) s_xor_b32 s2, exec_lo, s13 s_cbranch_execz .LBB0_10 s_and_not1_b32 s14, s14, exec_lo ds_store_b32 v6, v10 .LBB0_10: s_or_b32 exec_lo, exec_lo, s2 s_and_saveexec_b32 s13, s14 s_cbranch_execz .LBB0_12 v_mad_u64_u32 v[12:13], null, v11, s3, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v13, 31, v12 v_lshlrev_b64 v[12:13], 2, v[12:13] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s2, s6, v12 v_add_co_ci_u32_e64 v13, s2, s7, v13, s2 global_load_b32 v12, v[12:13], off s_waitcnt vmcnt(0) ds_store_b32 v6, v12 .LBB0_12: s_or_b32 exec_lo, exec_lo, s13 v_mov_b32_e32 v12, v8 s_mov_b32 s2, 0 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv .LBB0_13: v_add_nc_u32_e32 v13, s2, v7 s_add_i32 s2, s2, 4 ds_load_b32 v14, v12 ds_load_b32 v13, v13 v_add_nc_u32_e32 v12, 64, v12 s_cmp_eq_u32 s2, 64 s_waitcnt lgkmcnt(0) v_fmac_f32_e32 v2, v13, v14 s_cbranch_scc0 .LBB0_13 s_add_i32 s12, s12, 1 s_delay_alu instid0(SALU_CYCLE_1) s_lshl_b32 s13, s12, 4 s_barrier s_cmp_ge_i32 s13, s9 buffer_gl0_inv s_cbranch_scc0 .LBB0_2 s_branch .LBB0_16 .LBB0_15: v_mov_b32_e32 v2, 0 .LBB0_16: v_cmp_gt_i32_e32 vcc_lo, s8, v0 v_cmp_gt_i32_e64 s2, s3, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_and_b32 s2, vcc_lo, s2 s_and_saveexec_b32 s4, s2 s_cbranch_execz .LBB0_18 s_load_b64 s[0:1], s[0:1], 0x10 v_mad_u64_u32 v[3:4], null, v0, s3, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v4, 31, v3 v_lshlrev_b64 v[0:1], 2, v[3:4] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v0, vcc_lo, s0, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo global_store_b32 v[0:1], v2, off .LBB0_18: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z12matrixMulGPUPfS_S_iii .amdhsa_group_segment_fixed_size 2048 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 296 .amdhsa_user_sgpr_count 14 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 1 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 1 .amdhsa_next_free_vgpr 15 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z12matrixMulGPUPfS_S_iii, .Lfunc_end0-_Z12matrixMulGPUPfS_S_iii .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 8 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .offset: 24 .size: 4 .value_kind: by_value - .offset: 28 .size: 4 .value_kind: by_value - .offset: 32 .size: 4 .value_kind: by_value - .offset: 40 .size: 4 .value_kind: hidden_block_count_x - .offset: 44 .size: 4 .value_kind: hidden_block_count_y - .offset: 48 .size: 4 .value_kind: hidden_block_count_z - .offset: 52 .size: 2 .value_kind: hidden_group_size_x - .offset: 54 .size: 2 .value_kind: hidden_group_size_y - .offset: 56 .size: 2 .value_kind: hidden_group_size_z - .offset: 58 .size: 2 .value_kind: hidden_remainder_x - .offset: 60 .size: 2 .value_kind: hidden_remainder_y - .offset: 62 .size: 2 .value_kind: hidden_remainder_z - .offset: 80 .size: 8 .value_kind: hidden_global_offset_x - .offset: 88 .size: 8 .value_kind: hidden_global_offset_y - .offset: 96 .size: 8 .value_kind: hidden_global_offset_z - .offset: 104 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 2048 .kernarg_segment_align: 8 .kernarg_segment_size: 296 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z12matrixMulGPUPfS_S_iii .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z12matrixMulGPUPfS_S_iii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 15 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
#include <stdio.h> #include <pthread.h> const int N = 1 << 27; __global__ void kernel(float *x, int n) { int tid = threadIdx.x + blockIdx.x * blockDim.x; for (int i = tid; i < n; i += blockDim.x * gridDim.x) { x[i] = sqrt(pow(3.14159,i)); } } void *thread(void *args) { int * thread_data = (int*) args; const int num_streams = 8; cudaStream_t streams[num_streams]; float *data[num_streams]; int deviceNum; cudaError_t ret = cudaGetDeviceCount(&deviceNum); ret = cudaSetDevice(*thread_data); printf("device num: %d\n", *thread_data); for (int i = 0; i < num_streams; i++) { cudaStreamCreate(&streams[i]); cudaError_t ret0 = cudaMalloc(&data[i], N * sizeof(float)); if (ret0 != cudaSuccess) { printf("allocate failed\n"); return 0; } else { printf("%d MB\n", N*sizeof(float)/1024/1024); } // launch one worker kernel per stream kernel<<<1, 64, 0, streams[i]>>>(data[i], N); // launch a dummy kernel on the default stream kernel<<<1, 1>>>(0, 0); printf("finished stream syn %d\n", i); } printf("finished all stream %d\n"); cudaDeviceReset(); printf("finished device reset %d\n"); return 0; } int main() { pthread_t threads[4]; int thread_data[4]; for(int t=0;t<4;t++){ printf("In main: creating thread %ld\n", t); thread_data[t] = t; int rc = pthread_create(&threads[t], NULL, thread, &thread_data[t]); if (rc){ printf("ERROR; return code from pthread_create() is %d\n", rc); exit(-1); } } for(int i = 0; i < 4; i++) pthread_join(threads[i], NULL); /* Last thing that main() should do */ pthread_exit(NULL); }
code for sm_80 Function : _Z6kernelPfi .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ IMAD.MOV.U32 R1, RZ, RZ, c[0x0][0x28] ; /* 0x00000a00ff017624 */ /* 0x000fe400078e00ff */ /*0010*/ S2R R8, SR_CTAID.X ; /* 0x0000000000087919 */ /* 0x000e280000002500 */ /*0020*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e240000002100 */ /*0030*/ IMAD R8, R8, c[0x0][0x0], R3 ; /* 0x0000000008087a24 */ /* 0x001fca00078e0203 */ /*0040*/ ISETP.GE.AND P0, PT, R8, c[0x0][0x168], PT ; /* 0x00005a0008007a0c */ /* 0x000fda0003f06270 */ /*0050*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0060*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe40000000a00 */ /*0070*/ I2F.F64 R2, R8 ; /* 0x0000000800027312 */ /* 0x0000620000201c00 */ /*0080*/ BSSY B0, 0xc0 ; /* 0x0000003000007945 */ /* 0x000fe20003800000 */ /*0090*/ MOV R0, 0xb0 ; /* 0x000000b000007802 */ /* 0x000fcc0000000f00 */ /*00a0*/ CALL.REL.NOINC 0x6c0 ; /* 0x0000061000007944 */ /* 0x003fea0003c00000 */ /*00b0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*00c0*/ DADD R4, R2, c[0x2][0x0] ; /* 0x0080000002047629 */ /* 0x000e620000000000 */ /*00d0*/ BSSY B0, 0x1f0 ; /* 0x0000011000007945 */ /* 0x000ff20003800000 */ /*00e0*/ LOP3.LUT R4, R5, 0x7ff00000, RZ, 0xc0, !PT ; /* 0x7ff0000005047812 */ /* 0x002fe200078ec0ff */ /*00f0*/ IMAD.MOV.U32 R5, RZ, RZ, R15 ; /* 0x000000ffff057224 */ /* 0x000fc600078e000f */ /*0100*/ ISETP.NE.AND P0, PT, R4, 0x7ff00000, PT ; /* 0x7ff000000400780c */ /* 0x000fe20003f05270 */ /*0110*/ IMAD.MOV.U32 R4, RZ, RZ, R14 ; /* 0x000000ffff047224 */ /* 0x000fd800078e000e */ /*0120*/ @P0 BRA 0x1e0 ; /* 0x000000b000000947 */ /* 0x000fea0003800000 */ /*0130*/ DSETP.GTU.AND P0, PT, |R2|, +INF , PT ; /* 0x7ff000000200742a */ /* 0x000e5c0003f0c200 */ /*0140*/ @P0 BRA 0x1d0 ; /* 0x0000008000000947 */ /* 0x002fea0003800000 */ /*0150*/ ISETP.NE.AND P0, PT, R2, RZ, PT ; /* 0x000000ff0200720c */ /* 0x000fe40003f05270 */ /*0160*/ LOP3.LUT R0, R3, 0x7fffffff, RZ, 0xc0, !PT ; /* 0x7fffffff03007812 */ /* 0x000fc800078ec0ff */ /*0170*/ ISETP.NE.OR P0, PT, R0, 0x7ff00000, P0 ; /* 0x7ff000000000780c */ /* 0x000fda0000705670 */ /*0180*/ @P0 BRA 0x1e0 ; /* 0x0000005000000947 */ /* 0x000fea0003800000 */ /*0190*/ ISETP.GE.AND P0, PT, R3, RZ, PT ; /* 0x000000ff0300720c */ /* 0x000fe20003f06270 */ /*01a0*/ IMAD.MOV.U32 R4, RZ, RZ, RZ ; /* 0x000000ffff047224 */ /* 0x000fc600078e00ff */ /*01b0*/ SEL R5, RZ, 0x7ff00000, !P0 ; /* 0x7ff00000ff057807 */ /* 0x000fe20004000000 */ /*01c0*/ BRA 0x1e0 ; /* 0x0000001000007947 */ /* 0x000fea0003800000 */ /*01d0*/ DADD R4, R2, c[0x2][0x0] ; /* 0x0080000002047629 */ /* 0x00028c0000000000 */ /*01e0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*01f0*/ ISETP.NE.AND P0, PT, R8, RZ, PT ; /* 0x000000ff0800720c */ /* 0x000fe20003f05270 */ /*0200*/ IMAD.MOV.U32 R10, RZ, RZ, 0x0 ; /* 0x00000000ff0a7424 */ /* 0x000fe200078e00ff */ /*0210*/ BSSY B0, 0x360 ; /* 0x0000014000007945 */ /* 0x000fe20003800000 */ /*0220*/ IMAD.MOV.U32 R11, RZ, RZ, 0x3fd80000 ; /* 0x3fd80000ff0b7424 */ /* 0x000fe200078e00ff */ /*0230*/ FSEL R5, R5, 1.875, P0 ; /* 0x3ff0000005057808 */ /* 0x004fe40000000000 */ /*0240*/ FSEL R4, R4, RZ, P0 ; /* 0x000000ff04047208 */ /* 0x000fe40000000000 */ /*0250*/ MUFU.RSQ64H R3, R5 ; /* 0x0000000500037308 */ /* 0x002e620000001c00 */ /*0260*/ IADD3 R2, R5, -0x3500000, RZ ; /* 0xfcb0000005027810 */ /* 0x000fc80007ffe0ff */ /*0270*/ ISETP.GE.U32.AND P0, PT, R2, 0x7ca00000, PT ; /* 0x7ca000000200780c */ /* 0x000fe40003f06070 */ /*0280*/ DMUL R6, R2, R2 ; /* 0x0000000202067228 */ /* 0x002e4c0000000000 */ /*0290*/ DFMA R6, -R6, R4, 1 ; /* 0x3ff000000606742b */ /* 0x002e4c0000000104 */ /*02a0*/ DFMA R10, R6, R10, 0.5 ; /* 0x3fe00000060a742b */ /* 0x002fc8000000000a */ /*02b0*/ DMUL R6, R2, R6 ; /* 0x0000000602067228 */ /* 0x000e4c0000000000 */ /*02c0*/ DFMA R12, R10, R6, R2 ; /* 0x000000060a0c722b */ /* 0x003e0c0000000002 */ /*02d0*/ DMUL R14, R12, R4 ; /* 0x000000040c0e7228 */ /* 0x001e080000000000 */ /*02e0*/ IADD3 R11, R13, -0x100000, RZ ; /* 0xfff000000d0b7810 */ /* 0x000fe20007ffe0ff */ /*02f0*/ IMAD.MOV.U32 R10, RZ, RZ, R12 ; /* 0x000000ffff0a7224 */ /* 0x000fe200078e000c */ /*0300*/ DFMA R16, R14, -R14, R4 ; /* 0x8000000e0e10722b */ /* 0x001e0c0000000004 */ /*0310*/ DFMA R6, R16, R10, R14 ; /* 0x0000000a1006722b */ /* 0x001062000000000e */ /*0320*/ @!P0 BRA 0x350 ; /* 0x0000002000008947 */ /* 0x000fea0003800000 */ /*0330*/ MOV R0, 0x350 ; /* 0x0000035000007802 */ /* 0x000fca0000000f00 */ /*0340*/ CALL.REL.NOINC 0x3f0 ; /* 0x000000a000007944 */ /* 0x003fea0003c00000 */ /*0350*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0360*/ F2F.F32.F64 R7, R6 ; /* 0x0000000600077310 */ /* 0x002e620000301000 */ /*0370*/ MOV R3, 0x4 ; /* 0x0000000400037802 */ /* 0x000fe20000000f00 */ /*0380*/ IMAD.MOV.U32 R5, RZ, RZ, c[0x0][0x0] ; /* 0x00000000ff057624 */ /* 0x001fc800078e00ff */ /*0390*/ IMAD.WIDE R2, R8, R3, c[0x0][0x160] ; /* 0x0000580008027625 */ /* 0x000fc800078e0203 */ /*03a0*/ IMAD R8, R5, c[0x0][0xc], R8 ; /* 0x0000030005087a24 */ /* 0x000fca00078e0208 */ /*03b0*/ ISETP.GE.AND P0, PT, R8, c[0x0][0x168], PT ; /* 0x00005a0008007a0c */ /* 0x000fe20003f06270 */ /*03c0*/ STG.E [R2.64], R7 ; /* 0x0000000702007986 */ /* 0x0021d8000c101904 */ /*03d0*/ @!P0 BRA 0x70 ; /* 0xfffffc9000008947 */ /* 0x001fea000383ffff */ /*03e0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*03f0*/ ISETP.GE.U32.AND P0, PT, R2, -0x3400000, PT ; /* 0xfcc000000200780c */ /* 0x000fe20003f06070 */ /*0400*/ BSSY B1, 0x670 ; /* 0x0000026000017945 */ /* 0x000fe20003800000 */ /*0410*/ IMAD.MOV.U32 R10, RZ, RZ, R12 ; /* 0x000000ffff0a7224 */ /* 0x000fe400078e000c */ /*0420*/ IMAD.MOV.U32 R2, RZ, RZ, R16 ; /* 0x000000ffff027224 */ /* 0x000fe400078e0010 */ /*0430*/ IMAD.MOV.U32 R3, RZ, RZ, R17 ; /* 0x000000ffff037224 */ /* 0x000fce00078e0011 */ /*0440*/ @!P0 BRA 0x4d0 ; /* 0x0000008000008947 */ /* 0x000fea0003800000 */ /*0450*/ DFMA.RM R2, R2, R10, R14 ; /* 0x0000000a0202722b */ /* 0x000e14000000400e */ /*0460*/ IADD3 R6, P0, R2, 0x1, RZ ; /* 0x0000000102067810 */ /* 0x001fca0007f1e0ff */ /*0470*/ IMAD.X R7, RZ, RZ, R3, P0 ; /* 0x000000ffff077224 */ /* 0x000fcc00000e0603 */ /*0480*/ DFMA.RP R4, -R2, R6, R4 ; /* 0x000000060204722b */ /* 0x000e0c0000008104 */ /*0490*/ DSETP.GT.AND P0, PT, R4, RZ, PT ; /* 0x000000ff0400722a */ /* 0x001e0c0003f04000 */ /*04a0*/ FSEL R2, R6, R2, P0 ; /* 0x0000000206027208 */ /* 0x001fe40000000000 */ /*04b0*/ FSEL R3, R7, R3, P0 ; /* 0x0000000307037208 */ /* 0x000fe20000000000 */ /*04c0*/ BRA 0x660 ; /* 0x0000019000007947 */ /* 0x000fea0003800000 */ /*04d0*/ DSETP.NE.AND P0, PT, R4, RZ, PT ; /* 0x000000ff0400722a */ /* 0x000e1c0003f05000 */ /*04e0*/ @!P0 BRA 0x650 ; /* 0x0000016000008947 */ /* 0x001fea0003800000 */ /*04f0*/ ISETP.GE.AND P0, PT, R5, RZ, PT ; /* 0x000000ff0500720c */ /* 0x000fda0003f06270 */ /*0500*/ @!P0 IMAD.MOV.U32 R2, RZ, RZ, 0x0 ; /* 0x00000000ff028424 */ /* 0x000fe200078e00ff */ /*0510*/ @!P0 MOV R3, 0xfff80000 ; /* 0xfff8000000038802 */ /* 0x000fe20000000f00 */ /*0520*/ @!P0 BRA 0x660 ; /* 0x0000013000008947 */ /* 0x000fea0003800000 */ /*0530*/ ISETP.GT.AND P0, PT, R5, 0x7fefffff, PT ; /* 0x7fefffff0500780c */ /* 0x000fda0003f04270 */ /*0540*/ @P0 BRA 0x650 ; /* 0x0000010000000947 */ /* 0x000fea0003800000 */ /*0550*/ DMUL R2, R4, 8.11296384146066816958e+31 ; /* 0x4690000004027828 */ /* 0x0000620000000000 */ /*0560*/ IMAD.MOV.U32 R10, RZ, RZ, 0x0 ; /* 0x00000000ff0a7424 */ /* 0x000fe400078e00ff */ /*0570*/ IMAD.MOV.U32 R4, RZ, RZ, RZ ; /* 0x000000ffff047224 */ /* 0x001fe400078e00ff */ /*0580*/ IMAD.MOV.U32 R11, RZ, RZ, 0x3fd80000 ; /* 0x3fd80000ff0b7424 */ /* 0x000fe200078e00ff */ /*0590*/ MUFU.RSQ64H R5, R3 ; /* 0x0000000300057308 */ /* 0x002e260000001c00 */ /*05a0*/ DMUL R6, R4, R4 ; /* 0x0000000404067228 */ /* 0x001e0c0000000000 */ /*05b0*/ DFMA R6, R2, -R6, 1 ; /* 0x3ff000000206742b */ /* 0x001e0c0000000806 */ /*05c0*/ DFMA R10, R6, R10, 0.5 ; /* 0x3fe00000060a742b */ /* 0x001fc8000000000a */ /*05d0*/ DMUL R6, R4, R6 ; /* 0x0000000604067228 */ /* 0x000e0c0000000000 */ /*05e0*/ DFMA R6, R10, R6, R4 ; /* 0x000000060a06722b */ /* 0x001e0c0000000004 */ /*05f0*/ DMUL R4, R2, R6 ; /* 0x0000000602047228 */ /* 0x0010480000000000 */ /*0600*/ IADD3 R7, R7, -0x100000, RZ ; /* 0xfff0000007077810 */ /* 0x001fe40007ffe0ff */ /*0610*/ DFMA R10, R4, -R4, R2 ; /* 0x80000004040a722b */ /* 0x002e0c0000000002 */ /*0620*/ DFMA R2, R6, R10, R4 ; /* 0x0000000a0602722b */ /* 0x001e140000000004 */ /*0630*/ IADD3 R3, R3, -0x3500000, RZ ; /* 0xfcb0000003037810 */ /* 0x001fe20007ffe0ff */ /*0640*/ BRA 0x660 ; /* 0x0000001000007947 */ /* 0x000fea0003800000 */ /*0650*/ DADD R2, R4, R4 ; /* 0x0000000004027229 */ /* 0x00004c0000000004 */ /*0660*/ BSYNC B1 ; /* 0x0000000000017941 */ /* 0x000fea0003800000 */ /*0670*/ IMAD.MOV.U32 R6, RZ, RZ, R2 ; /* 0x000000ffff067224 */ /* 0x002fe400078e0002 */ /*0680*/ IMAD.MOV.U32 R7, RZ, RZ, R3 ; /* 0x000000ffff077224 */ /* 0x000fe200078e0003 */ /*0690*/ MOV R3, 0x0 ; /* 0x0000000000037802 */ /* 0x000fe20000000f00 */ /*06a0*/ IMAD.MOV.U32 R2, RZ, RZ, R0 ; /* 0x000000ffff027224 */ /* 0x000fc800078e0000 */ /*06b0*/ RET.REL.NODEC R2 0x0 ; /* 0xfffff94002007950 */ /* 0x000fea0003c3ffff */ /*06c0*/ MUFU.RCP64H R11, 1.785396575927734375 ; /* 0x3ffc90fc000b7908 */ /* 0x000e220000001800 */ /*06d0*/ IMAD.MOV.U32 R4, RZ, RZ, -0x7f23cc9 ; /* 0xf80dc337ff047424 */ /* 0x000fe400078e00ff */ /*06e0*/ IMAD.MOV.U32 R5, RZ, RZ, 0x3ffc90fc ; /* 0x3ffc90fcff057424 */ /* 0x000fe400078e00ff */ /*06f0*/ IMAD.MOV.U32 R10, RZ, RZ, RZ ; /* 0x000000ffff0a7224 */ /* 0x000fe400078e00ff */ /*0700*/ IMAD.MOV.U32 R6, RZ, RZ, 0x7d2cafe2 ; /* 0x7d2cafe2ff067424 */ /* 0x000fe400078e00ff */ /*0710*/ IMAD.MOV.U32 R7, RZ, RZ, 0x3eb0f5ff ; /* 0x3eb0f5ffff077424 */ /* 0x000fe400078e00ff */ /*0720*/ IMAD.SHL.U32 R9, R3, 0x2, RZ ; /* 0x0000000203097824 */ /* 0x000fe200078e00ff */ /*0730*/ DFMA R4, R10, -R4, 1 ; /* 0x3ff000000a04742b */ /* 0x001e080000000804 */ /*0740*/ ISETP.GT.U32.AND P0, PT, R9, -0x2000001, PT ; /* 0xfdffffff0900780c */ /* 0x000fe40003f04070 */ /*0750*/ DFMA R4, R4, R4, R4 ; /* 0x000000040404722b */ /* 0x001e0c0000000004 */ /*0760*/ DFMA R10, R10, R4, R10 ; /* 0x000000040a0a722b */ /* 0x001e0c000000000a */ /*0770*/ DMUL R4, R10, c[0x2][0x8] ; /* 0x008002000a047a28 */ /* 0x001e0c0000000000 */ /*0780*/ DFMA R4, R10, c[0x2][0x8], R4 ; /* 0x008002000a047a2b */ /* 0x001e0c0000000004 */ /*0790*/ DMUL R12, R4, R4 ; /* 0x00000004040c7228 */ /* 0x001e080000000000 */ /*07a0*/ DADD R14, -R4, c[0x2][0x8] ; /* 0x00800200040e7629 */ /* 0x000e480000000100 */ /*07b0*/ DFMA R6, R12, R6, c[0x2][0x10] ; /* 0x008004000c06762b */ /* 0x001e080000000006 */ /*07c0*/ DADD R16, R14, R14 ; /* 0x000000000e107229 */ /* 0x002fc8000000000e */ /*07d0*/ DFMA R6, R12, R6, c[0x2][0x18] ; /* 0x008006000c06762b */ /* 0x001e080000000006 */ /*07e0*/ DMUL R14, R4, R4 ; /* 0x00000004040e7228 */ /* 0x000fc80000000000 */ /*07f0*/ DFMA R6, R12, R6, c[0x2][0x20] ; /* 0x008008000c06762b */ /* 0x001e080000000006 */ /*0800*/ DFMA R20, -R4, c[0x2][0x8], R16 ; /* 0x0080020004147a2b */ /* 0x000e480000000110 */ /*0810*/ DFMA R6, R12, R6, c[0x2][0x28] ; /* 0x00800a000c06762b */ /* 0x001e080000000006 */ /*0820*/ DMUL R10, R10, R20 ; /* 0x000000140a0a7228 */ /* 0x002fc80000000000 */ /*0830*/ DFMA R6, R12, R6, c[0x2][0x30] ; /* 0x00800c000c06762b */ /* 0x001e080000000006 */ /*0840*/ DFMA R16, R4, R4, -R14 ; /* 0x000000040410722b */ /* 0x000fc8000000080e */ /*0850*/ DFMA R18, R12, R6, c[0x2][0x38] ; /* 0x00800e000c12762b */ /* 0x001e0c0000000006 */ /*0860*/ DFMA R6, R12, R18, c[0x2][0x40] ; /* 0x008010000c06762b */ /* 0x001e0c0000000012 */ /*0870*/ DADD R22, -R6, c[0x2][0x40] ; /* 0x0080100006167629 */ /* 0x001e0c0000000100 */ /*0880*/ DFMA R22, R12, R18, R22 ; /* 0x000000120c16722b */ /* 0x001e080000000016 */ /*0890*/ DMUL R12, R4, R14 ; /* 0x0000000e040c7228 */ /* 0x000e480000000000 */ /*08a0*/ DADD R18, RZ, R22 ; /* 0x00000000ff127229 */ /* 0x001e080000000016 */ /*08b0*/ DFMA R20, R4, R14, -R12 ; /* 0x0000000e0414722b */ /* 0x002e48000000080c */ /*08c0*/ DADD R18, R18, c[0x2][0x48] ; /* 0x0080120012127629 */ /* 0x001e080000000000 */ /*08d0*/ DFMA R22, R10, R14, R20 ; /* 0x0000000e0a16722b */ /* 0x0023e40000000014 */ /*08e0*/ IADD3 R21, R11, 0x100000, RZ ; /* 0x001000000b157810 */ /* 0x002fe20007ffe0ff */ /*08f0*/ IMAD.MOV.U32 R20, RZ, RZ, R10 ; /* 0x000000ffff147224 */ /* 0x000fe200078e000a */ /*0900*/ DADD R14, R6, R18 ; /* 0x00000000060e7229 */ /* 0x001e0a0000000012 */ /*0910*/ DFMA R20, R4, R20, R16 ; /* 0x000000140414722b */ /* 0x000e480000000010 */ /*0920*/ DMUL R16, R14, R12 ; /* 0x0000000c0e107228 */ /* 0x001e080000000000 */ /*0930*/ DFMA R20, R4, R20, R22 ; /* 0x000000140414722b */ /* 0x002fc80000000016 */ /*0940*/ DADD R6, R6, -R14 ; /* 0x0000000006067229 */ /* 0x000e48000000080e */ /*0950*/ DFMA R22, R14, R12, -R16 ; /* 0x0000000c0e16722b */ /* 0x001e080000000810 */ /*0960*/ DADD R6, R18, R6 ; /* 0x0000000012067229 */ /* 0x0023e40000000006 */ /*0970*/ IMAD.MOV.U32 R18, RZ, RZ, 0x69ce2bdf ; /* 0x69ce2bdfff127424 */ /* 0x002fe400078e00ff */ /*0980*/ DFMA R20, R14, R20, R22 ; /* 0x000000140e14722b */ /* 0x0010620000000016 */ /*0990*/ IMAD.MOV.U32 R19, RZ, RZ, 0x3e5ade15 ; /* 0x3e5ade15ff137424 */ /* 0x000fe200078e00ff */ /*09a0*/ MOV R14, 0xfefa39ef ; /* 0xfefa39ef000e7802 */ /* 0x001fe20000000f00 */ /*09b0*/ IMAD.MOV.U32 R15, RZ, RZ, 0x3fe62e42 ; /* 0x3fe62e42ff0f7424 */ /* 0x000fc600078e00ff */ /*09c0*/ DFMA R20, R6, R12, R20 ; /* 0x0000000c0614722b */ /* 0x002e0c0000000014 */ /*09d0*/ DADD R12, R16, R20 ; /* 0x00000000100c7229 */ /* 0x001e0c0000000014 */ /*09e0*/ DADD R6, R4, R12 ; /* 0x0000000004067229 */ /* 0x001e08000000000c */ /*09f0*/ DADD R16, R16, -R12 ; /* 0x0000000010107229 */ /* 0x000e48000000080c */ /*0a00*/ DADD R4, R4, -R6 ; /* 0x0000000004047229 */ /* 0x001e080000000806 */ /*0a10*/ DADD R16, R20, R16 ; /* 0x0000000014107229 */ /* 0x002fc80000000010 */ /*0a20*/ DADD R4, R12, R4 ; /* 0x000000000c047229 */ /* 0x001e0c0000000004 */ /*0a30*/ DADD R4, R16, R4 ; /* 0x0000000010047229 */ /* 0x001e0c0000000004 */ /*0a40*/ DADD R10, R10, R4 ; /* 0x000000000a0a7229 */ /* 0x001e0c0000000004 */ /*0a50*/ DADD R12, R6, R10 ; /* 0x00000000060c7229 */ /* 0x001e0c000000000a */ /*0a60*/ DFMA R4, R14, 2, R12 ; /* 0x400000000e04782b */ /* 0x001e08000000000c */ /*0a70*/ DADD R6, R6, -R12 ; /* 0x0000000006067229 */ /* 0x000e48000000080c */ /*0a80*/ DFMA R14, R14, -2, R4 ; /* 0xc00000000e0e782b */ /* 0x001e080000000004 */ /*0a90*/ DADD R6, R10, R6 ; /* 0x000000000a067229 */ /* 0x0023e40000000006 */ /*0aa0*/ IMAD.MOV.U32 R10, RZ, RZ, 0x3b39803f ; /* 0x3b39803fff0a7424 */ /* 0x002fe400078e00ff */ /*0ab0*/ DADD R14, -R12, R14 ; /* 0x000000000c0e7229 */ /* 0x001e22000000010e */ /*0ac0*/ IMAD.MOV.U32 R11, RZ, RZ, 0x3c7abc9e ; /* 0x3c7abc9eff0b7424 */ /* 0x000fca00078e00ff */ /*0ad0*/ DADD R6, R6, -R14 ; /* 0x0000000006067229 */ /* 0x001064000000080e */ /*0ae0*/ LOP3.LUT R15, R3, 0xff0fffff, RZ, 0xc0, !PT ; /* 0xff0fffff030f7812 */ /* 0x001fe200078ec0ff */ /*0af0*/ IMAD.MOV.U32 R14, RZ, RZ, R2 ; /* 0x000000ffff0e7224 */ /* 0x000fc600078e0002 */ /*0b00*/ DFMA R6, R10, 2, R6 ; /* 0x400000000a06782b */ /* 0x002e220000000006 */ /*0b10*/ SEL R15, R15, R3, P0 ; /* 0x000000030f0f7207 */ /* 0x000fca0000000000 */ /*0b20*/ DADD R10, R4, R6 ; /* 0x00000000040a7229 */ /* 0x001e0c0000000006 */ /*0b30*/ DADD R12, R4, -R10 ; /* 0x00000000040c7229 */ /* 0x001e08000000080a */ /*0b40*/ DMUL R4, R10, R14 ; /* 0x0000000e0a047228 */ /* 0x000e480000000000 */ /*0b50*/ DADD R6, R6, R12 ; /* 0x0000000006067229 */ /* 0x0011e4000000000c */ /*0b60*/ IMAD.MOV.U32 R12, RZ, RZ, 0x652b82fe ; /* 0x652b82feff0c7424 */ /* 0x001fe200078e00ff */ /*0b70*/ MOV R13, 0x3ff71547 ; /* 0x3ff71547000d7802 */ /* 0x000fe20000000f00 */ /*0b80*/ DFMA R10, R10, R14, -R4 ; /* 0x0000000e0a0a722b */ /* 0x002e0c0000000804 */ /*0b90*/ DFMA R6, R6, R14, R10 ; /* 0x0000000e0606722b */ /* 0x001e0c000000000a */ /*0ba0*/ DADD R10, R4, R6 ; /* 0x00000000040a7229 */ /* 0x001e0c0000000006 */ /*0bb0*/ DFMA R12, R10, R12, 6.75539944105574400000e+15 ; /* 0x433800000a0c742b */ /* 0x001e08000000000c */ /*0bc0*/ FSETP.GEU.AND P0, PT, |R11|, 4.1917929649353027344, PT ; /* 0x4086232b0b00780b */ /* 0x000fe40003f0e200 */ /*0bd0*/ DADD R14, R12, -6.75539944105574400000e+15 ; /* 0xc33800000c0e7429 */ /* 0x001e0c0000000000 */ /*0be0*/ DFMA R16, R14, c[0x2][0x50], R10 ; /* 0x008014000e107a2b */ /* 0x001e0c000000000a */ /*0bf0*/ DFMA R14, R14, c[0x2][0x58], R16 ; /* 0x008016000e0e7a2b */ /* 0x001e0c0000000010 */ /*0c00*/ DFMA R16, R14, R18, c[0x2][0x60] ; /* 0x008018000e10762b */ /* 0x001e0c0000000012 */ /*0c10*/ DFMA R16, R14, R16, c[0x2][0x68] ; /* 0x00801a000e10762b */ /* 0x001e0c0000000010 */ /*0c20*/ DFMA R16, R14, R16, c[0x2][0x70] ; /* 0x00801c000e10762b */ /* 0x001e0c0000000010 */ /*0c30*/ DFMA R16, R14, R16, c[0x2][0x78] ; /* 0x00801e000e10762b */ /* 0x001e0c0000000010 */ /*0c40*/ DFMA R16, R14, R16, c[0x2][0x80] ; /* 0x008020000e10762b */ /* 0x001e0c0000000010 */ /*0c50*/ DFMA R16, R14, R16, c[0x2][0x88] ; /* 0x008022000e10762b */ /* 0x001e0c0000000010 */ /*0c60*/ DFMA R16, R14, R16, c[0x2][0x90] ; /* 0x008024000e10762b */ /* 0x001e0c0000000010 */ /*0c70*/ DFMA R16, R14, R16, c[0x2][0x98] ; /* 0x008026000e10762b */ /* 0x001e0c0000000010 */ /*0c80*/ DFMA R16, R14, R16, c[0x2][0xa0] ; /* 0x008028000e10762b */ /* 0x001e0c0000000010 */ /*0c90*/ DFMA R16, R14, R16, 1 ; /* 0x3ff000000e10742b */ /* 0x001e0c0000000010 */ /*0ca0*/ DFMA R16, R14, R16, 1 ; /* 0x3ff000000e10742b */ /* 0x001e140000000010 */ /*0cb0*/ IMAD R15, R12, 0x100000, R17 ; /* 0x001000000c0f7824 */ /* 0x001fe400078e0211 */ /*0cc0*/ IMAD.MOV.U32 R14, RZ, RZ, R16 ; /* 0x000000ffff0e7224 */ /* 0x000fe200078e0010 */ /*0cd0*/ @!P0 BRA 0xdb0 ; /* 0x000000d000008947 */ /* 0x000fea0003800000 */ /*0ce0*/ FSETP.GEU.AND P1, PT, |R11|, 4.2275390625, PT ; /* 0x408748000b00780b */ /* 0x000fe20003f2e200 */ /*0cf0*/ DADD R14, R10, +INF ; /* 0x7ff000000a0e7429 */ /* 0x000fc80000000000 */ /*0d00*/ DSETP.GEU.AND P0, PT, R10, RZ, PT ; /* 0x000000ff0a00722a */ /* 0x000e0c0003f0e000 */ /*0d10*/ FSEL R14, R14, RZ, P0 ; /* 0x000000ff0e0e7208 */ /* 0x001fe40000000000 */ /*0d20*/ @!P1 LEA.HI R9, R12, R12, RZ, 0x1 ; /* 0x0000000c0c099211 */ /* 0x000fe400078f08ff */ /*0d30*/ FSEL R15, R15, RZ, P0 ; /* 0x000000ff0f0f7208 */ /* 0x000fe40000000000 */ /*0d40*/ @!P1 SHF.R.S32.HI R9, RZ, 0x1, R9 ; /* 0x00000001ff099819 */ /* 0x000fc80000011409 */ /*0d50*/ @!P1 LEA R13, R9, R17, 0x14 ; /* 0x00000011090d9211 */ /* 0x000fe200078ea0ff */ /*0d60*/ @!P1 IMAD.IADD R12, R12, 0x1, -R9 ; /* 0x000000010c0c9824 */ /* 0x000fca00078e0a09 */ /*0d70*/ @!P1 LEA R17, R12, 0x3ff00000, 0x14 ; /* 0x3ff000000c119811 */ /* 0x000fe200078ea0ff */ /*0d80*/ @!P1 IMAD.MOV.U32 R12, RZ, RZ, R16 ; /* 0x000000ffff0c9224 */ /* 0x000fe400078e0010 */ /*0d90*/ @!P1 IMAD.MOV.U32 R16, RZ, RZ, RZ ; /* 0x000000ffff109224 */ /* 0x000fcc00078e00ff */ /*0da0*/ @!P1 DMUL R14, R12, R16 ; /* 0x000000100c0e9228 */ /* 0x0000540000000000 */ /*0db0*/ LOP3.LUT R9, R15, 0x7fffffff, RZ, 0xc0, !PT ; /* 0x7fffffff0f097812 */ /* 0x002fe200078ec0ff */ /*0dc0*/ DADD R4, R4, -R10 ; /* 0x0000000004047229 */ /* 0x000e46000000080a */ /*0dd0*/ ISETP.NE.AND P0, PT, R9, 0x7ff00000, PT ; /* 0x7ff000000900780c */ /* 0x000fc60003f05270 */ /*0de0*/ DADD R4, R6, R4 ; /* 0x0000000006047229 */ /* 0x002e620000000004 */ /*0df0*/ ISETP.EQ.AND P0, PT, R14, RZ, !P0 ; /* 0x000000ff0e00720c */ /* 0x000fda0004702270 */ /*0e00*/ @!P0 DFMA R14, R4, R14, R14 ; /* 0x0000000e040e822b */ /* 0x0022a4000000000e */ /*0e10*/ IMAD.MOV.U32 R4, RZ, RZ, R0 ; /* 0x000000ffff047224 */ /* 0x002fe400078e0000 */ /*0e20*/ IMAD.MOV.U32 R5, RZ, RZ, 0x0 ; /* 0x00000000ff057424 */ /* 0x000fc800078e00ff */ /*0e30*/ RET.REL.NODEC R4 0x0 ; /* 0xfffff1c004007950 */ /* 0x004fea0003c3ffff */ /*0e40*/ BRA 0xe40; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0e50*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0e60*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0e70*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0e80*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0e90*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ea0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0eb0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ec0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ed0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ee0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ef0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include <stdio.h> #include <pthread.h> const int N = 1 << 27; __global__ void kernel(float *x, int n) { int tid = threadIdx.x + blockIdx.x * blockDim.x; for (int i = tid; i < n; i += blockDim.x * gridDim.x) { x[i] = sqrt(pow(3.14159,i)); } } void *thread(void *args) { int * thread_data = (int*) args; const int num_streams = 8; cudaStream_t streams[num_streams]; float *data[num_streams]; int deviceNum; cudaError_t ret = cudaGetDeviceCount(&deviceNum); ret = cudaSetDevice(*thread_data); printf("device num: %d\n", *thread_data); for (int i = 0; i < num_streams; i++) { cudaStreamCreate(&streams[i]); cudaError_t ret0 = cudaMalloc(&data[i], N * sizeof(float)); if (ret0 != cudaSuccess) { printf("allocate failed\n"); return 0; } else { printf("%d MB\n", N*sizeof(float)/1024/1024); } // launch one worker kernel per stream kernel<<<1, 64, 0, streams[i]>>>(data[i], N); // launch a dummy kernel on the default stream kernel<<<1, 1>>>(0, 0); printf("finished stream syn %d\n", i); } printf("finished all stream %d\n"); cudaDeviceReset(); printf("finished device reset %d\n"); return 0; } int main() { pthread_t threads[4]; int thread_data[4]; for(int t=0;t<4;t++){ printf("In main: creating thread %ld\n", t); thread_data[t] = t; int rc = pthread_create(&threads[t], NULL, thread, &thread_data[t]); if (rc){ printf("ERROR; return code from pthread_create() is %d\n", rc); exit(-1); } } for(int i = 0; i < 4; i++) pthread_join(threads[i], NULL); /* Last thing that main() should do */ pthread_exit(NULL); }
.file "tmpxft_00173385_00000000-6_defaultstream_devices.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2071: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2071: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "In main: creating thread %ld\n" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC1: .string "ERROR; return code from pthread_create() is %d\n" .text .globl main .type main, @function main: .LFB2068: .cfi_startproc endbr64 pushq %r14 .cfi_def_cfa_offset 16 .cfi_offset 14, -16 pushq %r13 .cfi_def_cfa_offset 24 .cfi_offset 13, -24 pushq %r12 .cfi_def_cfa_offset 32 .cfi_offset 12, -32 pushq %rbp .cfi_def_cfa_offset 40 .cfi_offset 6, -40 pushq %rbx .cfi_def_cfa_offset 48 .cfi_offset 3, -48 subq $64, %rsp .cfi_def_cfa_offset 112 movq %fs:40, %rax movq %rax, 56(%rsp) xorl %eax, %eax movq %rsp, %r12 movl $0, %ebp leaq .LC0(%rip), %r14 leaq _Z6threadPv(%rip), %r13 .L5: movl %ebp, %edx movq %r14, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl %ebp, (%r12) leaq 16(%rsp), %rbx leaq (%rbx,%rbp,8), %rdi movq %r12, %rcx movq %r13, %rdx movl $0, %esi call pthread_create@PLT testl %eax, %eax jne .L12 addq $1, %rbp addq $4, %r12 cmpq $4, %rbp jne .L5 leaq 32(%rbx), %rbp .L6: movq (%rbx), %rdi movl $0, %esi call pthread_join@PLT addq $8, %rbx cmpq %rbp, %rbx jne .L6 movq 56(%rsp), %rax subq %fs:40, %rax jne .L13 movl $0, %edi call pthread_exit@PLT .L12: movl %eax, %edx leaq .LC1(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $-1, %edi call exit@PLT .L13: call __stack_chk_fail@PLT .cfi_endproc .LFE2068: .size main, .-main .globl _Z26__device_stub__Z6kernelPfiPfi .type _Z26__device_stub__Z6kernelPfiPfi, @function _Z26__device_stub__Z6kernelPfiPfi: .LFB2093: .cfi_startproc endbr64 subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 8(%rsp) movl %esi, 4(%rsp) movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax leaq 8(%rsp), %rax movq %rax, 80(%rsp) leaq 4(%rsp), %rax movq %rax, 88(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) movl $1, 40(%rsp) movl $1, 44(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) leaq 24(%rsp), %rcx leaq 16(%rsp), %rdx leaq 44(%rsp), %rsi leaq 32(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L18 .L14: movq 104(%rsp), %rax subq %fs:40, %rax jne .L19 addq $120, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L18: .cfi_restore_state pushq 24(%rsp) .cfi_def_cfa_offset 136 pushq 24(%rsp) .cfi_def_cfa_offset 144 leaq 96(%rsp), %r9 movq 60(%rsp), %rcx movl 68(%rsp), %r8d movq 48(%rsp), %rsi movl 56(%rsp), %edx leaq _Z6kernelPfi(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 128 jmp .L14 .L19: call __stack_chk_fail@PLT .cfi_endproc .LFE2093: .size _Z26__device_stub__Z6kernelPfiPfi, .-_Z26__device_stub__Z6kernelPfiPfi .globl _Z6kernelPfi .type _Z6kernelPfi, @function _Z6kernelPfi: .LFB2094: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z26__device_stub__Z6kernelPfiPfi addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2094: .size _Z6kernelPfi, .-_Z6kernelPfi .section .rodata.str1.1 .LC2: .string "device num: %d\n" .LC3: .string "allocate failed\n" .LC4: .string "%d MB\n" .LC5: .string "finished stream syn %d\n" .LC6: .string "finished all stream %d\n" .LC7: .string "finished device reset %d\n" .text .globl _Z6threadPv .type _Z6threadPv, @function _Z6threadPv: .LFB2067: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $184, %rsp .cfi_def_cfa_offset 240 movq %rdi, %rbx movq %fs:40, %rax movq %rax, 168(%rsp) xorl %eax, %eax leaq 4(%rsp), %rdi call cudaGetDeviceCount@PLT movl (%rbx), %edi call cudaSetDevice@PLT movl (%rbx), %edx leaq .LC2(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT leaq 32(%rsp), %rbp movl $0, %ebx leaq 96(%rsp), %r13 leaq .LC4(%rip), %r15 leaq .LC5(%rip), %r14 jmp .L27 .L32: leaq .LC3(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT jmp .L24 .L33: movq 0(%r13,%rbx,8), %rdi movl $134217728, %esi call _Z26__device_stub__Z6kernelPfiPfi jmp .L25 .L26: movl %r12d, %edx movq %r14, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT addq $1, %rbx addq $8, %rbp cmpq $8, %rbx je .L31 .L27: movl %ebx, %r12d movq %rbp, %rdi call cudaStreamCreate@PLT leaq 0(%r13,%rbx,8), %rdi movl $536870912, %esi call cudaMalloc@PLT testl %eax, %eax jne .L32 movl $512, %edx movq %r15, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $64, 20(%rsp) movl $1, 24(%rsp) movl $1, 28(%rsp) movl $1, 8(%rsp) movl $1, 12(%rsp) movl $1, 16(%rsp) movq 0(%rbp), %r9 movl $0, %r8d movq 20(%rsp), %rdx movl $1, %ecx movq 8(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L33 .L25: movl $1, 20(%rsp) movl $1, 24(%rsp) movl $1, 28(%rsp) movl $1, 8(%rsp) movl $1, 12(%rsp) movl $1, 16(%rsp) movl $0, %r9d movl $0, %r8d movq 20(%rsp), %rdx movl $1, %ecx movq 8(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax jne .L26 movl $0, %esi movl $0, %edi call _Z26__device_stub__Z6kernelPfiPfi jmp .L26 .L31: leaq .LC6(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT call cudaDeviceReset@PLT leaq .LC7(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT .L24: movq 168(%rsp), %rax subq %fs:40, %rax jne .L34 movl $0, %eax addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L34: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2067: .size _Z6threadPv, .-_Z6threadPv .section .rodata.str1.1 .LC8: .string "_Z6kernelPfi" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2096: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC8(%rip), %rdx movq %rdx, %rcx leaq _Z6kernelPfi(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2096: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include <stdio.h> #include <pthread.h> const int N = 1 << 27; __global__ void kernel(float *x, int n) { int tid = threadIdx.x + blockIdx.x * blockDim.x; for (int i = tid; i < n; i += blockDim.x * gridDim.x) { x[i] = sqrt(pow(3.14159,i)); } } void *thread(void *args) { int * thread_data = (int*) args; const int num_streams = 8; cudaStream_t streams[num_streams]; float *data[num_streams]; int deviceNum; cudaError_t ret = cudaGetDeviceCount(&deviceNum); ret = cudaSetDevice(*thread_data); printf("device num: %d\n", *thread_data); for (int i = 0; i < num_streams; i++) { cudaStreamCreate(&streams[i]); cudaError_t ret0 = cudaMalloc(&data[i], N * sizeof(float)); if (ret0 != cudaSuccess) { printf("allocate failed\n"); return 0; } else { printf("%d MB\n", N*sizeof(float)/1024/1024); } // launch one worker kernel per stream kernel<<<1, 64, 0, streams[i]>>>(data[i], N); // launch a dummy kernel on the default stream kernel<<<1, 1>>>(0, 0); printf("finished stream syn %d\n", i); } printf("finished all stream %d\n"); cudaDeviceReset(); printf("finished device reset %d\n"); return 0; } int main() { pthread_t threads[4]; int thread_data[4]; for(int t=0;t<4;t++){ printf("In main: creating thread %ld\n", t); thread_data[t] = t; int rc = pthread_create(&threads[t], NULL, thread, &thread_data[t]); if (rc){ printf("ERROR; return code from pthread_create() is %d\n", rc); exit(-1); } } for(int i = 0; i < 4; i++) pthread_join(threads[i], NULL); /* Last thing that main() should do */ pthread_exit(NULL); }
#include <hip/hip_runtime.h> #include <stdio.h> #include <pthread.h> const int N = 1 << 27; __global__ void kernel(float *x, int n) { int tid = threadIdx.x + blockIdx.x * blockDim.x; for (int i = tid; i < n; i += blockDim.x * gridDim.x) { x[i] = sqrt(pow(3.14159,i)); } } void *thread(void *args) { int * thread_data = (int*) args; const int num_streams = 8; hipStream_t streams[num_streams]; float *data[num_streams]; int deviceNum; hipError_t ret = hipGetDeviceCount(&deviceNum); ret = hipSetDevice(*thread_data); printf("device num: %d\n", *thread_data); for (int i = 0; i < num_streams; i++) { hipStreamCreate(&streams[i]); hipError_t ret0 = hipMalloc(&data[i], N * sizeof(float)); if (ret0 != hipSuccess) { printf("allocate failed\n"); return 0; } else { printf("%d MB\n", N*sizeof(float)/1024/1024); } // launch one worker kernel per stream kernel<<<1, 64, 0, streams[i]>>>(data[i], N); // launch a dummy kernel on the default stream kernel<<<1, 1>>>(0, 0); printf("finished stream syn %d\n", i); } printf("finished all stream %d\n"); hipDeviceReset(); printf("finished device reset %d\n"); return 0; } int main() { pthread_t threads[4]; int thread_data[4]; for(int t=0;t<4;t++){ printf("In main: creating thread %ld\n", t); thread_data[t] = t; int rc = pthread_create(&threads[t], NULL, thread, &thread_data[t]); if (rc){ printf("ERROR; return code from pthread_create() is %d\n", rc); exit(-1); } } for(int i = 0; i < 4; i++) pthread_join(threads[i], NULL); /* Last thing that main() should do */ pthread_exit(NULL); }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include <stdio.h> #include <pthread.h> const int N = 1 << 27; __global__ void kernel(float *x, int n) { int tid = threadIdx.x + blockIdx.x * blockDim.x; for (int i = tid; i < n; i += blockDim.x * gridDim.x) { x[i] = sqrt(pow(3.14159,i)); } } void *thread(void *args) { int * thread_data = (int*) args; const int num_streams = 8; hipStream_t streams[num_streams]; float *data[num_streams]; int deviceNum; hipError_t ret = hipGetDeviceCount(&deviceNum); ret = hipSetDevice(*thread_data); printf("device num: %d\n", *thread_data); for (int i = 0; i < num_streams; i++) { hipStreamCreate(&streams[i]); hipError_t ret0 = hipMalloc(&data[i], N * sizeof(float)); if (ret0 != hipSuccess) { printf("allocate failed\n"); return 0; } else { printf("%d MB\n", N*sizeof(float)/1024/1024); } // launch one worker kernel per stream kernel<<<1, 64, 0, streams[i]>>>(data[i], N); // launch a dummy kernel on the default stream kernel<<<1, 1>>>(0, 0); printf("finished stream syn %d\n", i); } printf("finished all stream %d\n"); hipDeviceReset(); printf("finished device reset %d\n"); return 0; } int main() { pthread_t threads[4]; int thread_data[4]; for(int t=0;t<4;t++){ printf("In main: creating thread %ld\n", t); thread_data[t] = t; int rc = pthread_create(&threads[t], NULL, thread, &thread_data[t]); if (rc){ printf("ERROR; return code from pthread_create() is %d\n", rc); exit(-1); } } for(int i = 0; i < 4; i++) pthread_join(threads[i], NULL); /* Last thing that main() should do */ pthread_exit(NULL); }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z6kernelPfi .globl _Z6kernelPfi .p2align 8 .type _Z6kernelPfi,@function _Z6kernelPfi: s_clause 0x1 s_load_b32 s4, s[0:1], 0x1c s_load_b32 s33, s[0:1], 0x8 s_add_u32 s2, s0, 16 s_addc_u32 s3, s1, 0 s_waitcnt lgkmcnt(0) s_and_b32 s20, s4, 0xffff s_mov_b32 s4, exec_lo v_mad_u64_u32 v[1:2], null, s15, s20, v[0:1] s_delay_alu instid0(VALU_DEP_1) v_cmpx_gt_i32_e64 s33, v1 s_cbranch_execz .LBB0_3 s_load_b32 s61, s[2:3], 0x0 s_load_b64 s[2:3], s[0:1], 0x0 v_mov_b32_e32 v0, 0x3ff00000 s_mov_b32 s60, 0 s_mov_b32 s5, 0x3fe55555 s_mov_b32 s4, 0x55555555 s_mov_b32 s7, 0x3fba6564 s_mov_b32 s6, 0x968915a9 s_mov_b32 s9, 0x3fbdee67 s_mov_b32 s8, 0x4222de17 s_mov_b32 s11, 0x3fbe25e4 s_mov_b32 s10, 0x3abe935a s_mov_b32 s13, 0x3fc110ef s_mov_b32 s12, 0x47e6c9c2 s_mov_b32 s15, 0x3fc3b13b s_mov_b32 s14, 0xcfa74449 s_mov_b32 s17, 0x3fc745d1 s_mov_b32 s16, 0x71bf3c30 s_mov_b32 s19, 0x3fcc71c7 s_mov_b32 s18, 0x1c7792ce s_waitcnt lgkmcnt(0) s_mul_i32 s61, s61, s20 s_mov_b32 s21, 0x3fd24924 s_mov_b32 s20, 0x924920da s_mov_b32 s23, 0x3fd99999 s_mov_b32 s22, 0x9999999c s_mov_b32 s25, 0x3fe62e42 s_mov_b32 s24, 0xfefa39ef s_mov_b32 s27, 0x3c7abc9e s_mov_b32 s26, 0x3b39803f s_mov_b32 s29, 0xbfe55555 s_mov_b32 s31, 0x3c8543b0 s_mov_b32 s30, 0xd5df274d s_mov_b32 s35, 0x3ff71547 s_mov_b32 s34, 0x652b82fe s_mov_b32 s37, 0xbfe62e42 s_mov_b32 s39, 0xbc7abc9e s_mov_b32 s41, 0x3e928af3 s_mov_b32 s40, 0xfca7ab0c s_mov_b32 s43, 0x3e5ade15 s_mov_b32 s42, 0x6a5dcb37 s_mov_b32 s45, 0x3ec71dee s_mov_b32 s44, 0x623fde64 s_mov_b32 s47, 0x3efa0199 s_mov_b32 s46, 0x7c89e6b0 s_mov_b32 s49, 0x3f2a01a0 s_mov_b32 s48, 0x14761f6e s_mov_b32 s51, 0x3f56c16c s_mov_b32 s50, 0x1852b7b0 s_mov_b32 s53, 0x3f811111 s_mov_b32 s52, 0x11122322 s_mov_b32 s55, 0x3fa55555 s_mov_b32 s54, 0x555502a1 s_mov_b32 s57, 0x3fc55555 s_mov_b32 s56, 0x55555511 s_mov_b32 s59, 0x3fe00000 s_mov_b32 s58, 11 .LBB0_2: v_cmp_eq_u32_e32 vcc_lo, 0, v1 s_mov_b32 s28, s4 s_mov_b32 s36, s24 s_mov_b32 s38, s26 v_cndmask_b32_e32 v3, 0x400921f9, v0, vcc_lo v_cndmask_b32_e64 v2, 0xf01b866e, 0, vcc_lo s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_frexp_mant_f64_e32 v[4:5], v[2:3] v_cmp_gt_f64_e32 vcc_lo, s[4:5], v[4:5] v_cndmask_b32_e64 v6, 0, 1, vcc_lo s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ldexp_f64 v[4:5], v[4:5], v6 v_add_f64 v[6:7], v[4:5], 1.0 v_add_f64 v[12:13], v[4:5], -1.0 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_rcp_f64_e32 v[8:9], v[6:7] v_add_f64 v[14:15], v[6:7], -1.0 v_add_f64 v[4:5], v[4:5], -v[14:15] s_waitcnt_depctr 0xfff v_fma_f64 v[10:11], -v[6:7], v[8:9], 1.0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[8:9], v[10:11], v[8:9], v[8:9] v_fma_f64 v[10:11], -v[6:7], v[8:9], 1.0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[8:9], v[10:11], v[8:9], v[8:9] v_mul_f64 v[10:11], v[12:13], v[8:9] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_f64 v[16:17], v[6:7], v[10:11] v_fma_f64 v[6:7], v[10:11], v[6:7], -v[16:17] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[4:5], v[10:11], v[4:5], v[6:7] v_add_f64 v[6:7], v[16:17], v[4:5] s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_add_f64 v[14:15], v[12:13], -v[6:7] v_add_f64 v[16:17], v[6:7], -v[16:17] v_add_f64 v[12:13], v[12:13], -v[14:15] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[4:5], v[16:17], -v[4:5] v_add_f64 v[6:7], v[12:13], -v[6:7] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_f64 v[4:5], v[4:5], v[6:7] v_add_f64 v[4:5], v[14:15], v[4:5] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_f64 v[4:5], v[8:9], v[4:5] v_add_f64 v[6:7], v[10:11], v[4:5] s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_add_f64 v[8:9], v[6:7], -v[10:11] v_mul_f64 v[10:11], v[6:7], v[6:7] v_add_f64 v[4:5], v[4:5], -v[8:9] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f64 v[8:9], v[6:7], v[6:7], -v[10:11] v_add_f64 v[12:13], v[4:5], v[4:5] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[8:9], v[6:7], v[12:13], v[8:9] v_add_f64 v[12:13], v[10:11], v[8:9] s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_3) v_fma_f64 v[14:15], v[12:13], s[8:9], s[6:7] v_add_f64 v[10:11], v[12:13], -v[10:11] v_mul_f64 v[18:19], v[6:7], v[12:13] v_fma_f64 v[14:15], v[12:13], v[14:15], s[10:11] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_f64 v[8:9], v[8:9], -v[10:11] v_fma_f64 v[22:23], v[12:13], v[6:7], -v[18:19] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[14:15], v[12:13], v[14:15], s[12:13] v_fma_f64 v[14:15], v[12:13], v[14:15], s[14:15] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[14:15], v[12:13], v[14:15], s[16:17] v_fma_f64 v[14:15], v[12:13], v[14:15], s[18:19] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[14:15], v[12:13], v[14:15], s[20:21] v_fma_f64 v[14:15], v[12:13], v[14:15], s[22:23] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_f64 v[16:17], v[12:13], v[14:15] v_fma_f64 v[10:11], v[12:13], v[14:15], -v[16:17] v_fma_f64 v[12:13], v[12:13], v[4:5], v[22:23] v_ldexp_f64 v[4:5], v[4:5], 1 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_fma_f64 v[10:11], v[8:9], v[14:15], v[10:11] v_fma_f64 v[8:9], v[8:9], v[6:7], v[12:13] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[14:15], v[16:17], v[10:11] v_add_f64 v[12:13], v[18:19], v[8:9] s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_3) v_add_f64 v[20:21], v[14:15], s[4:5] v_add_f64 v[16:17], v[14:15], -v[16:17] v_add_f64 v[18:19], v[12:13], -v[18:19] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_f64 v[24:25], v[20:21], s[28:29] v_add_f64 v[10:11], v[10:11], -v[16:17] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_f64 v[8:9], v[8:9], -v[18:19] v_add_f64 v[14:15], v[14:15], -v[24:25] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_f64 v[10:11], v[10:11], s[30:31] v_add_f64 v[10:11], v[10:11], v[14:15] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_f64 v[14:15], v[20:21], v[10:11] v_add_f64 v[16:17], v[20:21], -v[14:15] v_mul_f64 v[20:21], v[12:13], v[14:15] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[10:11], v[10:11], v[16:17] v_fma_f64 v[16:17], v[12:13], v[14:15], -v[20:21] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[10:11], v[12:13], v[10:11], v[16:17] v_fma_f64 v[8:9], v[8:9], v[14:15], v[10:11] v_frexp_exp_i32_f64_e32 v10, v[2:3] v_ldexp_f64 v[2:3], v[6:7], 1 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_f64 v[6:7], v[20:21], v[8:9] v_subrev_co_ci_u32_e32 v10, vcc_lo, 0, v10, vcc_lo s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_3) v_cvt_f64_i32_e32 v[10:11], v10 v_add_f64 v[12:13], v[2:3], v[6:7] v_add_f64 v[14:15], v[6:7], -v[20:21] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_mul_f64 v[16:17], v[10:11], s[24:25] v_add_f64 v[2:3], v[12:13], -v[2:3] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_f64 v[8:9], v[8:9], -v[14:15] v_fma_f64 v[14:15], v[10:11], s[24:25], -v[16:17] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_f64 v[2:3], v[6:7], -v[2:3] v_add_f64 v[4:5], v[4:5], v[8:9] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f64 v[6:7], v[10:11], s[26:27], v[14:15] v_add_f64 v[2:3], v[4:5], v[2:3] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[4:5], v[16:17], v[6:7] v_add_f64 v[8:9], v[12:13], v[2:3] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[16:17], v[4:5], -v[16:17] v_add_f64 v[10:11], v[4:5], v[8:9] v_add_f64 v[12:13], v[8:9], -v[12:13] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_f64 v[6:7], v[6:7], -v[16:17] v_add_f64 v[14:15], v[10:11], -v[4:5] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[2:3], v[2:3], -v[12:13] v_add_f64 v[18:19], v[10:11], -v[14:15] v_add_f64 v[8:9], v[8:9], -v[14:15] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_f64 v[12:13], v[6:7], v[2:3] v_add_f64 v[4:5], v[4:5], -v[18:19] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_f64 v[4:5], v[8:9], v[4:5] v_add_f64 v[8:9], v[12:13], -v[6:7] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[4:5], v[12:13], v[4:5] v_add_f64 v[12:13], v[12:13], -v[8:9] v_add_f64 v[2:3], v[2:3], -v[8:9] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_f64 v[14:15], v[10:11], v[4:5] v_add_f64 v[6:7], v[6:7], -v[12:13] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[8:9], v[14:15], -v[10:11] v_add_f64 v[2:3], v[2:3], v[6:7] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_f64 v[4:5], v[4:5], -v[8:9] v_add_f64 v[2:3], v[2:3], v[4:5] v_cvt_f64_i32_e32 v[4:5], v1 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_f64 v[6:7], v[14:15], v[2:3] v_add_f64 v[8:9], v[6:7], -v[14:15] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_mul_f64 v[10:11], v[6:7], v[4:5] v_add_f64 v[2:3], v[2:3], -v[8:9] s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_fma_f64 v[6:7], v[4:5], v[6:7], -v[10:11] v_cmp_class_f64_e64 vcc_lo, v[10:11], 0x204 v_fma_f64 v[2:3], v[4:5], v[2:3], v[6:7] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_f64 v[4:5], v[10:11], v[2:3] v_dual_cndmask_b32 v7, v5, v11 :: v_dual_cndmask_b32 v6, v4, v10 v_add_f64 v[4:5], v[4:5], -v[10:11] s_delay_alu instid0(VALU_DEP_2) v_mul_f64 v[8:9], v[6:7], s[34:35] v_cmp_nlt_f64_e64 s0, 0x40900000, v[6:7] v_cmp_neq_f64_e64 vcc_lo, 0x7ff00000, |v[6:7]| v_cmp_ngt_f64_e64 s1, 0xc090cc00, v[6:7] v_add_f64 v[2:3], v[2:3], -v[4:5] v_rndne_f64_e32 v[8:9], v[8:9] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_4) v_dual_cndmask_b32 v3, 0, v3 :: v_dual_cndmask_b32 v2, 0, v2 s_and_b32 vcc_lo, s1, s0 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_fma_f64 v[12:13], v[8:9], s[36:37], v[6:7] v_cvt_i32_f64_e32 v16, v[8:9] v_fma_f64 v[12:13], v[8:9], s[38:39], v[12:13] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[14:15], v[12:13], s[42:43], s[40:41] v_fma_f64 v[14:15], v[12:13], v[14:15], s[44:45] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[14:15], v[12:13], v[14:15], s[46:47] v_fma_f64 v[14:15], v[12:13], v[14:15], s[48:49] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[14:15], v[12:13], v[14:15], s[50:51] v_fma_f64 v[14:15], v[12:13], v[14:15], s[52:53] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[14:15], v[12:13], v[14:15], s[54:55] v_fma_f64 v[14:15], v[12:13], v[14:15], s[56:57] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[14:15], v[12:13], v[14:15], s[58:59] v_fma_f64 v[14:15], v[12:13], v[14:15], 1.0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[8:9], v[12:13], v[14:15], 1.0 v_ldexp_f64 v[8:9], v[8:9], v16 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cndmask_b32_e64 v4, 0x7ff00000, v9, s0 v_cndmask_b32_e64 v5, 0, v4, s1 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_cndmask_b32_e32 v4, 0, v8, vcc_lo v_fma_f64 v[2:3], v[4:5], v[2:3], v[4:5] v_cmp_class_f64_e64 vcc_lo, v[4:5], 0x204 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_dual_cndmask_b32 v3, v3, v5 :: v_dual_cndmask_b32 v2, v2, v4 v_cmp_gt_f64_e64 s0, 0x10000000, |v[2:3]| s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cndmask_b32_e64 v4, 0, 1, s0 v_lshlrev_b32_e32 v4, 8, v4 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ldexp_f64 v[2:3], |v[2:3]|, v4 v_rsq_f64_e32 v[4:5], v[2:3] v_cmp_class_f64_e64 vcc_lo, v[2:3], 0x260 s_waitcnt_depctr 0xfff v_mul_f64 v[6:7], v[2:3], v[4:5] v_mul_f64 v[4:5], v[4:5], 0.5 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[8:9], -v[4:5], v[6:7], 0.5 v_fma_f64 v[6:7], v[6:7], v[8:9], v[6:7] v_fma_f64 v[4:5], v[4:5], v[8:9], v[4:5] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[8:9], -v[6:7], v[6:7], v[2:3] v_fma_f64 v[6:7], v[8:9], v[4:5], v[6:7] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[8:9], -v[6:7], v[6:7], v[2:3] v_fma_f64 v[4:5], v[8:9], v[4:5], v[6:7] v_cndmask_b32_e64 v6, 0, 0xffffff80, s0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ldexp_f64 v[4:5], v[4:5], v6 v_dual_cndmask_b32 v3, v5, v3 :: v_dual_cndmask_b32 v2, v4, v2 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_cvt_f32_f64_e32 v4, v[2:3] v_ashrrev_i32_e32 v2, 31, v1 v_lshlrev_b64 v[2:3], 2, v[1:2] v_add_nc_u32_e32 v1, s61, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_3) v_cmp_le_i32_e32 vcc_lo, s33, v1 v_add_co_u32 v2, s0, s2, v2 s_delay_alu instid0(VALU_DEP_1) v_add_co_ci_u32_e64 v3, s0, s3, v3, s0 s_or_b32 s60, vcc_lo, s60 global_store_b32 v[2:3], v4, off s_and_not1_b32 exec_lo, exec_lo, s60 s_cbranch_execnz .LBB0_2 .LBB0_3: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z6kernelPfi .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 272 .amdhsa_user_sgpr_count 15 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 0 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 0 .amdhsa_next_free_vgpr 26 .amdhsa_next_free_sgpr 62 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z6kernelPfi, .Lfunc_end0-_Z6kernelPfi .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .offset: 8 .size: 4 .value_kind: by_value - .offset: 16 .size: 4 .value_kind: hidden_block_count_x - .offset: 20 .size: 4 .value_kind: hidden_block_count_y - .offset: 24 .size: 4 .value_kind: hidden_block_count_z - .offset: 28 .size: 2 .value_kind: hidden_group_size_x - .offset: 30 .size: 2 .value_kind: hidden_group_size_y - .offset: 32 .size: 2 .value_kind: hidden_group_size_z - .offset: 34 .size: 2 .value_kind: hidden_remainder_x - .offset: 36 .size: 2 .value_kind: hidden_remainder_y - .offset: 38 .size: 2 .value_kind: hidden_remainder_z - .offset: 56 .size: 8 .value_kind: hidden_global_offset_x - .offset: 64 .size: 8 .value_kind: hidden_global_offset_y - .offset: 72 .size: 8 .value_kind: hidden_global_offset_z - .offset: 80 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 272 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z6kernelPfi .private_segment_fixed_size: 0 .sgpr_count: 64 .sgpr_spill_count: 0 .symbol: _Z6kernelPfi.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 26 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <hip/hip_runtime.h> #include <stdio.h> #include <pthread.h> const int N = 1 << 27; __global__ void kernel(float *x, int n) { int tid = threadIdx.x + blockIdx.x * blockDim.x; for (int i = tid; i < n; i += blockDim.x * gridDim.x) { x[i] = sqrt(pow(3.14159,i)); } } void *thread(void *args) { int * thread_data = (int*) args; const int num_streams = 8; hipStream_t streams[num_streams]; float *data[num_streams]; int deviceNum; hipError_t ret = hipGetDeviceCount(&deviceNum); ret = hipSetDevice(*thread_data); printf("device num: %d\n", *thread_data); for (int i = 0; i < num_streams; i++) { hipStreamCreate(&streams[i]); hipError_t ret0 = hipMalloc(&data[i], N * sizeof(float)); if (ret0 != hipSuccess) { printf("allocate failed\n"); return 0; } else { printf("%d MB\n", N*sizeof(float)/1024/1024); } // launch one worker kernel per stream kernel<<<1, 64, 0, streams[i]>>>(data[i], N); // launch a dummy kernel on the default stream kernel<<<1, 1>>>(0, 0); printf("finished stream syn %d\n", i); } printf("finished all stream %d\n"); hipDeviceReset(); printf("finished device reset %d\n"); return 0; } int main() { pthread_t threads[4]; int thread_data[4]; for(int t=0;t<4;t++){ printf("In main: creating thread %ld\n", t); thread_data[t] = t; int rc = pthread_create(&threads[t], NULL, thread, &thread_data[t]); if (rc){ printf("ERROR; return code from pthread_create() is %d\n", rc); exit(-1); } } for(int i = 0; i < 4; i++) pthread_join(threads[i], NULL); /* Last thing that main() should do */ pthread_exit(NULL); }
.text .file "defaultstream_devices.hip" .globl _Z21__device_stub__kernelPfi # -- Begin function _Z21__device_stub__kernelPfi .p2align 4, 0x90 .type _Z21__device_stub__kernelPfi,@function _Z21__device_stub__kernelPfi: # @_Z21__device_stub__kernelPfi .cfi_startproc # %bb.0: subq $88, %rsp .cfi_def_cfa_offset 96 movq %rdi, 56(%rsp) movl %esi, 4(%rsp) leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 4(%rsp), %rax movq %rax, 72(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 64(%rsp), %r9 movl $_Z6kernelPfi, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $104, %rsp .cfi_adjust_cfa_offset -104 retq .Lfunc_end0: .size _Z21__device_stub__kernelPfi, .Lfunc_end0-_Z21__device_stub__kernelPfi .cfi_endproc # -- End function .globl _Z6threadPv # -- Begin function _Z6threadPv .p2align 4, 0x90 .type _Z6threadPv,@function _Z6threadPv: # @_Z6threadPv .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 subq $232, %rsp .cfi_def_cfa_offset 288 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movq %rdi, %r14 movabsq $4294967297, %rbx # imm = 0x100000001 leaq 92(%rsp), %rdi callq hipGetDeviceCount movl (%r14), %edi callq hipSetDevice movl (%r14), %esi movl $.L.str, %edi xorl %eax, %eax callq printf leaq 63(%rbx), %rax movq %rax, 80(%rsp) # 8-byte Spill xorl %r14d, %r14d xorl %r12d, %r12d .p2align 4, 0x90 .LBB1_2: # =>This Inner Loop Header: Depth=1 leaq (%rsp,%r14), %r15 addq $160, %r15 movq %r15, %rdi callq hipStreamCreate leaq (%rsp,%r14), %rbp addq $96, %rbp movl $536870912, %esi # imm = 0x20000000 movq %rbp, %rdi callq hipMalloc movl %eax, %r13d testl %eax, %eax je .LBB1_4 # %bb.3: # in Loop: Header=BB1_2 Depth=1 movl $.Lstr, %edi callq puts@PLT jmp .LBB1_9 .p2align 4, 0x90 .LBB1_4: # in Loop: Header=BB1_2 Depth=1 movl $.L.str.2, %edi movl $512, %esi # imm = 0x200 xorl %eax, %eax callq printf movq (%r15), %r9 movq %rbx, %rdi movl $1, %esi movq 80(%rsp), %rdx # 8-byte Reload movl $1, %ecx xorl %r8d, %r8d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_6 # %bb.5: # in Loop: Header=BB1_2 Depth=1 movq (%rbp), %rax movq %rax, 56(%rsp) movl $134217728, 4(%rsp) # imm = 0x8000000 leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 4(%rsp), %rax movq %rax, 72(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d movl $_Z6kernelPfi, %edi leaq 64(%rsp), %r9 pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_6: # in Loop: Header=BB1_2 Depth=1 movq %rbx, %rdi movl $1, %esi movq %rbx, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_8 # %bb.7: # in Loop: Header=BB1_2 Depth=1 movq $0, 56(%rsp) movl $0, 4(%rsp) leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 4(%rsp), %rax movq %rax, 72(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d movl $_Z6kernelPfi, %edi leaq 64(%rsp), %r9 pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_8: # in Loop: Header=BB1_2 Depth=1 movl $.L.str.3, %edi movl %r12d, %esi xorl %eax, %eax callq printf .LBB1_9: # in Loop: Header=BB1_2 Depth=1 testl %r13d, %r13d jne .LBB1_11 # %bb.1: # in Loop: Header=BB1_2 Depth=1 incl %r12d addq $8, %r14 cmpq $64, %r14 jne .LBB1_2 # %bb.10: # %.critedge movl $.L.str.4, %edi xorl %eax, %eax callq printf callq hipDeviceReset movl $.L.str.5, %edi xorl %eax, %eax callq printf .LBB1_11: # %.loopexit xorl %eax, %eax addq $232, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end1: .size _Z6threadPv, .Lfunc_end1-_Z6threadPv .cfi_endproc # -- End function .globl main # -- Begin function main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %r15 .cfi_def_cfa_offset 16 pushq %r14 .cfi_def_cfa_offset 24 pushq %rbx .cfi_def_cfa_offset 32 subq $48, %rsp .cfi_def_cfa_offset 80 .cfi_offset %rbx, -32 .cfi_offset %r14, -24 .cfi_offset %r15, -16 leaq 16(%rsp), %rbx movq %rsp, %r14 xorl %r15d, %r15d .p2align 4, 0x90 .LBB2_1: # =>This Inner Loop Header: Depth=1 movl $.L.str.6, %edi movl %r15d, %esi xorl %eax, %eax callq printf movl %r15d, (%r14) movl $_Z6threadPv, %edx movq %rbx, %rdi xorl %esi, %esi movq %r14, %rcx callq pthread_create testl %eax, %eax jne .LBB2_6 # %bb.2: # in Loop: Header=BB2_1 Depth=1 incq %r15 addq $8, %rbx addq $4, %r14 cmpq $4, %r15 jne .LBB2_1 # %bb.3: # %.preheader.preheader xorl %ebx, %ebx .p2align 4, 0x90 .LBB2_4: # %.preheader # =>This Inner Loop Header: Depth=1 movq 16(%rsp,%rbx,8), %rdi xorl %esi, %esi callq pthread_join incq %rbx cmpq $4, %rbx jne .LBB2_4 # %bb.5: xorl %edi, %edi callq pthread_exit .LBB2_6: movl $.L.str.7, %edi movl %eax, %esi xorl %eax, %eax callq printf movl $-1, %edi callq exit .Lfunc_end2: .size main, .Lfunc_end2-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB3_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB3_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z6kernelPfi, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end3: .size __hip_module_ctor, .Lfunc_end3-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB4_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB4_2: retq .Lfunc_end4: .size __hip_module_dtor, .Lfunc_end4-__hip_module_dtor .cfi_endproc # -- End function .type _Z6kernelPfi,@object # @_Z6kernelPfi .section .rodata,"a",@progbits .globl _Z6kernelPfi .p2align 3, 0x0 _Z6kernelPfi: .quad _Z21__device_stub__kernelPfi .size _Z6kernelPfi, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "device num: %d\n" .size .L.str, 16 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "%d MB\n" .size .L.str.2, 7 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "finished stream syn %d\n" .size .L.str.3, 24 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "finished all stream %d\n" .size .L.str.4, 24 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "finished device reset %d\n" .size .L.str.5, 26 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz "In main: creating thread %ld\n" .size .L.str.6, 30 .type .L.str.7,@object # @.str.7 .L.str.7: .asciz "ERROR .size .L.str.7, 48 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z6kernelPfi" .size .L__unnamed_1, 13 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .type .Lstr,@object # @str .section .rodata.str1.1,"aMS",@progbits,1 .Lstr: .asciz "allocate failed" .size .Lstr, 16 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z21__device_stub__kernelPfi .addrsig_sym _Z6threadPv .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z6kernelPfi .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_00173385_00000000-6_defaultstream_devices.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2071: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2071: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "In main: creating thread %ld\n" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC1: .string "ERROR; return code from pthread_create() is %d\n" .text .globl main .type main, @function main: .LFB2068: .cfi_startproc endbr64 pushq %r14 .cfi_def_cfa_offset 16 .cfi_offset 14, -16 pushq %r13 .cfi_def_cfa_offset 24 .cfi_offset 13, -24 pushq %r12 .cfi_def_cfa_offset 32 .cfi_offset 12, -32 pushq %rbp .cfi_def_cfa_offset 40 .cfi_offset 6, -40 pushq %rbx .cfi_def_cfa_offset 48 .cfi_offset 3, -48 subq $64, %rsp .cfi_def_cfa_offset 112 movq %fs:40, %rax movq %rax, 56(%rsp) xorl %eax, %eax movq %rsp, %r12 movl $0, %ebp leaq .LC0(%rip), %r14 leaq _Z6threadPv(%rip), %r13 .L5: movl %ebp, %edx movq %r14, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl %ebp, (%r12) leaq 16(%rsp), %rbx leaq (%rbx,%rbp,8), %rdi movq %r12, %rcx movq %r13, %rdx movl $0, %esi call pthread_create@PLT testl %eax, %eax jne .L12 addq $1, %rbp addq $4, %r12 cmpq $4, %rbp jne .L5 leaq 32(%rbx), %rbp .L6: movq (%rbx), %rdi movl $0, %esi call pthread_join@PLT addq $8, %rbx cmpq %rbp, %rbx jne .L6 movq 56(%rsp), %rax subq %fs:40, %rax jne .L13 movl $0, %edi call pthread_exit@PLT .L12: movl %eax, %edx leaq .LC1(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $-1, %edi call exit@PLT .L13: call __stack_chk_fail@PLT .cfi_endproc .LFE2068: .size main, .-main .globl _Z26__device_stub__Z6kernelPfiPfi .type _Z26__device_stub__Z6kernelPfiPfi, @function _Z26__device_stub__Z6kernelPfiPfi: .LFB2093: .cfi_startproc endbr64 subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 8(%rsp) movl %esi, 4(%rsp) movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax leaq 8(%rsp), %rax movq %rax, 80(%rsp) leaq 4(%rsp), %rax movq %rax, 88(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) movl $1, 40(%rsp) movl $1, 44(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) leaq 24(%rsp), %rcx leaq 16(%rsp), %rdx leaq 44(%rsp), %rsi leaq 32(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L18 .L14: movq 104(%rsp), %rax subq %fs:40, %rax jne .L19 addq $120, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L18: .cfi_restore_state pushq 24(%rsp) .cfi_def_cfa_offset 136 pushq 24(%rsp) .cfi_def_cfa_offset 144 leaq 96(%rsp), %r9 movq 60(%rsp), %rcx movl 68(%rsp), %r8d movq 48(%rsp), %rsi movl 56(%rsp), %edx leaq _Z6kernelPfi(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 128 jmp .L14 .L19: call __stack_chk_fail@PLT .cfi_endproc .LFE2093: .size _Z26__device_stub__Z6kernelPfiPfi, .-_Z26__device_stub__Z6kernelPfiPfi .globl _Z6kernelPfi .type _Z6kernelPfi, @function _Z6kernelPfi: .LFB2094: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z26__device_stub__Z6kernelPfiPfi addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2094: .size _Z6kernelPfi, .-_Z6kernelPfi .section .rodata.str1.1 .LC2: .string "device num: %d\n" .LC3: .string "allocate failed\n" .LC4: .string "%d MB\n" .LC5: .string "finished stream syn %d\n" .LC6: .string "finished all stream %d\n" .LC7: .string "finished device reset %d\n" .text .globl _Z6threadPv .type _Z6threadPv, @function _Z6threadPv: .LFB2067: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $184, %rsp .cfi_def_cfa_offset 240 movq %rdi, %rbx movq %fs:40, %rax movq %rax, 168(%rsp) xorl %eax, %eax leaq 4(%rsp), %rdi call cudaGetDeviceCount@PLT movl (%rbx), %edi call cudaSetDevice@PLT movl (%rbx), %edx leaq .LC2(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT leaq 32(%rsp), %rbp movl $0, %ebx leaq 96(%rsp), %r13 leaq .LC4(%rip), %r15 leaq .LC5(%rip), %r14 jmp .L27 .L32: leaq .LC3(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT jmp .L24 .L33: movq 0(%r13,%rbx,8), %rdi movl $134217728, %esi call _Z26__device_stub__Z6kernelPfiPfi jmp .L25 .L26: movl %r12d, %edx movq %r14, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT addq $1, %rbx addq $8, %rbp cmpq $8, %rbx je .L31 .L27: movl %ebx, %r12d movq %rbp, %rdi call cudaStreamCreate@PLT leaq 0(%r13,%rbx,8), %rdi movl $536870912, %esi call cudaMalloc@PLT testl %eax, %eax jne .L32 movl $512, %edx movq %r15, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $64, 20(%rsp) movl $1, 24(%rsp) movl $1, 28(%rsp) movl $1, 8(%rsp) movl $1, 12(%rsp) movl $1, 16(%rsp) movq 0(%rbp), %r9 movl $0, %r8d movq 20(%rsp), %rdx movl $1, %ecx movq 8(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L33 .L25: movl $1, 20(%rsp) movl $1, 24(%rsp) movl $1, 28(%rsp) movl $1, 8(%rsp) movl $1, 12(%rsp) movl $1, 16(%rsp) movl $0, %r9d movl $0, %r8d movq 20(%rsp), %rdx movl $1, %ecx movq 8(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax jne .L26 movl $0, %esi movl $0, %edi call _Z26__device_stub__Z6kernelPfiPfi jmp .L26 .L31: leaq .LC6(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT call cudaDeviceReset@PLT leaq .LC7(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT .L24: movq 168(%rsp), %rax subq %fs:40, %rax jne .L34 movl $0, %eax addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L34: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2067: .size _Z6threadPv, .-_Z6threadPv .section .rodata.str1.1 .LC8: .string "_Z6kernelPfi" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2096: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC8(%rip), %rdx movq %rdx, %rcx leaq _Z6kernelPfi(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2096: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "defaultstream_devices.hip" .globl _Z21__device_stub__kernelPfi # -- Begin function _Z21__device_stub__kernelPfi .p2align 4, 0x90 .type _Z21__device_stub__kernelPfi,@function _Z21__device_stub__kernelPfi: # @_Z21__device_stub__kernelPfi .cfi_startproc # %bb.0: subq $88, %rsp .cfi_def_cfa_offset 96 movq %rdi, 56(%rsp) movl %esi, 4(%rsp) leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 4(%rsp), %rax movq %rax, 72(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d leaq 64(%rsp), %r9 movl $_Z6kernelPfi, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $104, %rsp .cfi_adjust_cfa_offset -104 retq .Lfunc_end0: .size _Z21__device_stub__kernelPfi, .Lfunc_end0-_Z21__device_stub__kernelPfi .cfi_endproc # -- End function .globl _Z6threadPv # -- Begin function _Z6threadPv .p2align 4, 0x90 .type _Z6threadPv,@function _Z6threadPv: # @_Z6threadPv .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 subq $232, %rsp .cfi_def_cfa_offset 288 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movq %rdi, %r14 movabsq $4294967297, %rbx # imm = 0x100000001 leaq 92(%rsp), %rdi callq hipGetDeviceCount movl (%r14), %edi callq hipSetDevice movl (%r14), %esi movl $.L.str, %edi xorl %eax, %eax callq printf leaq 63(%rbx), %rax movq %rax, 80(%rsp) # 8-byte Spill xorl %r14d, %r14d xorl %r12d, %r12d .p2align 4, 0x90 .LBB1_2: # =>This Inner Loop Header: Depth=1 leaq (%rsp,%r14), %r15 addq $160, %r15 movq %r15, %rdi callq hipStreamCreate leaq (%rsp,%r14), %rbp addq $96, %rbp movl $536870912, %esi # imm = 0x20000000 movq %rbp, %rdi callq hipMalloc movl %eax, %r13d testl %eax, %eax je .LBB1_4 # %bb.3: # in Loop: Header=BB1_2 Depth=1 movl $.Lstr, %edi callq puts@PLT jmp .LBB1_9 .p2align 4, 0x90 .LBB1_4: # in Loop: Header=BB1_2 Depth=1 movl $.L.str.2, %edi movl $512, %esi # imm = 0x200 xorl %eax, %eax callq printf movq (%r15), %r9 movq %rbx, %rdi movl $1, %esi movq 80(%rsp), %rdx # 8-byte Reload movl $1, %ecx xorl %r8d, %r8d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_6 # %bb.5: # in Loop: Header=BB1_2 Depth=1 movq (%rbp), %rax movq %rax, 56(%rsp) movl $134217728, 4(%rsp) # imm = 0x8000000 leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 4(%rsp), %rax movq %rax, 72(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d movl $_Z6kernelPfi, %edi leaq 64(%rsp), %r9 pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_6: # in Loop: Header=BB1_2 Depth=1 movq %rbx, %rdi movl $1, %esi movq %rbx, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_8 # %bb.7: # in Loop: Header=BB1_2 Depth=1 movq $0, 56(%rsp) movl $0, 4(%rsp) leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 4(%rsp), %rax movq %rax, 72(%rsp) leaq 40(%rsp), %rdi leaq 24(%rsp), %rsi leaq 16(%rsp), %rdx leaq 8(%rsp), %rcx callq __hipPopCallConfiguration movq 40(%rsp), %rsi movl 48(%rsp), %edx movq 24(%rsp), %rcx movl 32(%rsp), %r8d movl $_Z6kernelPfi, %edi leaq 64(%rsp), %r9 pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_8: # in Loop: Header=BB1_2 Depth=1 movl $.L.str.3, %edi movl %r12d, %esi xorl %eax, %eax callq printf .LBB1_9: # in Loop: Header=BB1_2 Depth=1 testl %r13d, %r13d jne .LBB1_11 # %bb.1: # in Loop: Header=BB1_2 Depth=1 incl %r12d addq $8, %r14 cmpq $64, %r14 jne .LBB1_2 # %bb.10: # %.critedge movl $.L.str.4, %edi xorl %eax, %eax callq printf callq hipDeviceReset movl $.L.str.5, %edi xorl %eax, %eax callq printf .LBB1_11: # %.loopexit xorl %eax, %eax addq $232, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end1: .size _Z6threadPv, .Lfunc_end1-_Z6threadPv .cfi_endproc # -- End function .globl main # -- Begin function main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %r15 .cfi_def_cfa_offset 16 pushq %r14 .cfi_def_cfa_offset 24 pushq %rbx .cfi_def_cfa_offset 32 subq $48, %rsp .cfi_def_cfa_offset 80 .cfi_offset %rbx, -32 .cfi_offset %r14, -24 .cfi_offset %r15, -16 leaq 16(%rsp), %rbx movq %rsp, %r14 xorl %r15d, %r15d .p2align 4, 0x90 .LBB2_1: # =>This Inner Loop Header: Depth=1 movl $.L.str.6, %edi movl %r15d, %esi xorl %eax, %eax callq printf movl %r15d, (%r14) movl $_Z6threadPv, %edx movq %rbx, %rdi xorl %esi, %esi movq %r14, %rcx callq pthread_create testl %eax, %eax jne .LBB2_6 # %bb.2: # in Loop: Header=BB2_1 Depth=1 incq %r15 addq $8, %rbx addq $4, %r14 cmpq $4, %r15 jne .LBB2_1 # %bb.3: # %.preheader.preheader xorl %ebx, %ebx .p2align 4, 0x90 .LBB2_4: # %.preheader # =>This Inner Loop Header: Depth=1 movq 16(%rsp,%rbx,8), %rdi xorl %esi, %esi callq pthread_join incq %rbx cmpq $4, %rbx jne .LBB2_4 # %bb.5: xorl %edi, %edi callq pthread_exit .LBB2_6: movl $.L.str.7, %edi movl %eax, %esi xorl %eax, %eax callq printf movl $-1, %edi callq exit .Lfunc_end2: .size main, .Lfunc_end2-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB3_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB3_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z6kernelPfi, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end3: .size __hip_module_ctor, .Lfunc_end3-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB4_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB4_2: retq .Lfunc_end4: .size __hip_module_dtor, .Lfunc_end4-__hip_module_dtor .cfi_endproc # -- End function .type _Z6kernelPfi,@object # @_Z6kernelPfi .section .rodata,"a",@progbits .globl _Z6kernelPfi .p2align 3, 0x0 _Z6kernelPfi: .quad _Z21__device_stub__kernelPfi .size _Z6kernelPfi, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "device num: %d\n" .size .L.str, 16 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "%d MB\n" .size .L.str.2, 7 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "finished stream syn %d\n" .size .L.str.3, 24 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "finished all stream %d\n" .size .L.str.4, 24 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "finished device reset %d\n" .size .L.str.5, 26 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz "In main: creating thread %ld\n" .size .L.str.6, 30 .type .L.str.7,@object # @.str.7 .L.str.7: .asciz "ERROR .size .L.str.7, 48 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z6kernelPfi" .size .L__unnamed_1, 13 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .type .Lstr,@object # @str .section .rodata.str1.1,"aMS",@progbits,1 .Lstr: .asciz "allocate failed" .size .Lstr, 16 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z21__device_stub__kernelPfi .addrsig_sym _Z6threadPv .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z6kernelPfi .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
#include <iostream> #include <stdio.h> #include <math.h> using namespace std; #define TILE_WIDTH 2 __global__ void MatrixMult(int m, int n, int k, float *a, float *b, float *c) { int row = threadIdx.y + blockIdx.y*blockDim.y; int col = threadIdx.x + blockIdx.x*blockDim.x; if((row < m) && (col < k)) { float temp = 0.0; for (int i = 0; i < n; ++i) { temp += a[row*n+i]*b[col+i*k]; } c[row*k+col] = temp; } } // main fn int main(void) { int m = 4; int n = 6; int k = 7; float* a = new float[m*n]; float* b = new float[n*k]; float* c = new float[m*k]; float *dev_a, *dev_b, *dev_c; dim3 dimGrid((k-1)/TILE_WIDTH+1,(m-1)/TILE_WIDTH+1,1); dim3 dimBlock(TILE_WIDTH,TILE_WIDTH,1); cudaMalloc((void**)&dev_a, m*n*sizeof(float)); cudaMalloc((void**)&dev_b, n*k*sizeof(float)); cudaMalloc((void**)&dev_c, m*k*sizeof(float)); for (int i=0; i<m*n; i++) { a[i] = sin((float) i); } for (int i=0; i<n*k; i++) { b[i] = cos((float) i); } cudaMemcpy(dev_a, a, m*n*sizeof(float), cudaMemcpyHostToDevice); cudaMemcpy(dev_b, b, n*k*sizeof(float), cudaMemcpyHostToDevice); MatrixMult<<<dimGrid,dimBlock>>>(m,n,k,dev_a,dev_b,dev_c); cudaMemcpy(c, dev_c, m*k*sizeof(float), cudaMemcpyDeviceToHost); cout<<"a matrix: \n"; for (int i=0; i<m; i++) { for (int j=0; j<n; j++) { cout<<a[n*i+j]<<" "; } cout<<"\n"; } cout<<"b matrix: \n"; for (int i=0; i<n; i++) { for (int j=0; j<k; j++) { cout<<b[k*i+j]<<" "; } cout<<"\n"; } cout<<"c matrix: \n"; for (int i=0; i<m; i++) { for (int j=0; j<k; j++) { cout<<c[k*i+j]<<" "; } cout<<"\n"; } cudaFree(dev_a); cudaFree(dev_b); cudaFree(dev_c); delete [] a; delete [] b; delete [] c; }
code for sm_80 Function : _Z10MatrixMultiiiPfS_S_ .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R0, SR_CTAID.X ; /* 0x0000000000007919 */ /* 0x000e280000002500 */ /*0020*/ S2R R5, SR_TID.X ; /* 0x0000000000057919 */ /* 0x000e280000002100 */ /*0030*/ S2R R3, SR_CTAID.Y ; /* 0x0000000000037919 */ /* 0x000e680000002600 */ /*0040*/ S2R R2, SR_TID.Y ; /* 0x0000000000027919 */ /* 0x000e620000002200 */ /*0050*/ IMAD R0, R0, c[0x0][0x0], R5 ; /* 0x0000000000007a24 */ /* 0x001fca00078e0205 */ /*0060*/ ISETP.GE.AND P0, PT, R0, c[0x0][0x168], PT ; /* 0x00005a0000007a0c */ /* 0x000fe20003f06270 */ /*0070*/ IMAD R3, R3, c[0x0][0x4], R2 ; /* 0x0000010003037a24 */ /* 0x002fca00078e0202 */ /*0080*/ ISETP.GE.OR P0, PT, R3, c[0x0][0x160], P0 ; /* 0x0000580003007a0c */ /* 0x000fda0000706670 */ /*0090*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*00a0*/ MOV R4, c[0x0][0x164] ; /* 0x0000590000047a02 */ /* 0x000fe20000000f00 */ /*00b0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*00c0*/ HFMA2.MMA R24, -RZ, RZ, 0, 0 ; /* 0x00000000ff187435 */ /* 0x000fe400000001ff */ /*00d0*/ ISETP.GE.AND P0, PT, R4, 0x1, PT ; /* 0x000000010400780c */ /* 0x000fda0003f06270 */ /*00e0*/ @!P0 BRA 0xc40 ; /* 0x00000b5000008947 */ /* 0x000fea0003800000 */ /*00f0*/ IADD3 R2, R4.reuse, -0x1, RZ ; /* 0xffffffff04027810 */ /* 0x040fe40007ffe0ff */ /*0100*/ LOP3.LUT R4, R4, 0x3, RZ, 0xc0, !PT ; /* 0x0000000304047812 */ /* 0x000fe400078ec0ff */ /*0110*/ ISETP.GE.U32.AND P0, PT, R2, 0x3, PT ; /* 0x000000030200780c */ /* 0x000fe40003f06070 */ /*0120*/ MOV R24, RZ ; /* 0x000000ff00187202 */ /* 0x000fe40000000f00 */ /*0130*/ MOV R2, RZ ; /* 0x000000ff00027202 */ /* 0x000fd20000000f00 */ /*0140*/ @!P0 BRA 0xb30 ; /* 0x000009e000008947 */ /* 0x000fea0003800000 */ /*0150*/ IADD3 R5, -R4, c[0x0][0x164], RZ ; /* 0x0000590004057a10 */ /* 0x000fe20007ffe1ff */ /*0160*/ HFMA2.MMA R9, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff097435 */ /* 0x000fe200000001ff */ /*0170*/ ULDC.64 UR6, c[0x0][0x170] ; /* 0x00005c0000067ab9 */ /* 0x000fe20000000a00 */ /*0180*/ HFMA2.MMA R2, -RZ, RZ, 0, 0 ; /* 0x00000000ff027435 */ /* 0x000fe200000001ff */ /*0190*/ ISETP.GT.AND P0, PT, R5, RZ, PT ; /* 0x000000ff0500720c */ /* 0x000fe20003f04270 */ /*01a0*/ IMAD R6, R3, c[0x0][0x164], RZ ; /* 0x0000590003067a24 */ /* 0x000fe200078e02ff */ /*01b0*/ MOV R24, RZ ; /* 0x000000ff00187202 */ /* 0x000fca0000000f00 */ /*01c0*/ IMAD.WIDE R8, R0, R9, c[0x0][0x178] ; /* 0x00005e0000087625 */ /* 0x000fcc00078e0209 */ /*01d0*/ @!P0 BRA 0x990 ; /* 0x000007b000008947 */ /* 0x000fea0003800000 */ /*01e0*/ ISETP.GT.AND P1, PT, R5, 0xc, PT ; /* 0x0000000c0500780c */ /* 0x000fe40003f24270 */ /*01f0*/ PLOP3.LUT P0, PT, PT, PT, PT, 0x80, 0x0 ; /* 0x000000000000781c */ /* 0x000fd60003f0f070 */ /*0200*/ @!P1 BRA 0x6c0 ; /* 0x000004b000009947 */ /* 0x000fea0003800000 */ /*0210*/ PLOP3.LUT P0, PT, PT, PT, PT, 0x8, 0x0 ; /* 0x000000000000781c */ /* 0x000fe40003f0e170 */ /*0220*/ MOV R12, UR6 ; /* 0x00000006000c7c02 */ /* 0x000fe20008000f00 */ /*0230*/ LDG.E R21, [R8.64] ; /* 0x0000000408157981 */ /* 0x0000a2000c1e1900 */ /*0240*/ MOV R13, UR7 ; /* 0x00000007000d7c02 */ /* 0x000fca0008000f00 */ /*0250*/ IMAD.WIDE R12, R6, 0x4, R12 ; /* 0x00000004060c7825 */ /* 0x000fca00078e020c */ /*0260*/ LDG.E R20, [R12.64] ; /* 0x000000040c147981 */ /* 0x000ea2000c1e1900 */ /*0270*/ MOV R7, c[0x0][0x168] ; /* 0x00005a0000077a02 */ /* 0x000fc60000000f00 */ /*0280*/ LDG.E R14, [R12.64+0x4] ; /* 0x000004040c0e7981 */ /* 0x000ee4000c1e1900 */ /*0290*/ IMAD.WIDE R10, R7.reuse, 0x4, R8 ; /* 0x00000004070a7825 */ /* 0x040fe400078e0208 */ /*02a0*/ LDG.E R27, [R12.64+0x8] ; /* 0x000008040c1b7981 */ /* 0x000f28000c1e1900 */ /*02b0*/ LDG.E R15, [R10.64] ; /* 0x000000040a0f7981 */ /* 0x0002e2000c1e1900 */ /*02c0*/ IMAD.WIDE R22, R7, 0x4, R10 ; /* 0x0000000407167825 */ /* 0x000fc600078e020a */ /*02d0*/ LDG.E R18, [R12.64+0xc] ; /* 0x00000c040c127981 */ /* 0x000f66000c1e1900 */ /*02e0*/ IMAD.WIDE R28, R7.reuse, 0x4, R22 ; /* 0x00000004071c7825 */ /* 0x040fe200078e0216 */ /*02f0*/ LDG.E R26, [R22.64] ; /* 0x00000004161a7981 */ /* 0x000328000c1e1900 */ /*0300*/ LDG.E R19, [R28.64] ; /* 0x000000041c137981 */ /* 0x000362000c1e1900 */ /*0310*/ IMAD.WIDE R16, R7, 0x4, R28 ; /* 0x0000000407107825 */ /* 0x000fc600078e021c */ /*0320*/ LDG.E R8, [R12.64+0x10] ; /* 0x000010040c087981 */ /* 0x001f68000c1e1900 */ /*0330*/ LDG.E R9, [R16.64] ; /* 0x0000000410097981 */ /* 0x000168000c1e1900 */ /*0340*/ LDG.E R10, [R12.64+0x14] ; /* 0x000014040c0a7981 */ /* 0x002f68000c1e1900 */ /*0350*/ LDG.E R28, [R12.64+0x1c] ; /* 0x00001c040c1c7981 */ /* 0x000f62000c1e1900 */ /*0360*/ IMAD.WIDE R16, R7, 0x4, R16 ; /* 0x0000000407107825 */ /* 0x001fca00078e0210 */ /*0370*/ LDG.E R11, [R16.64] ; /* 0x00000004100b7981 */ /* 0x000562000c1e1900 */ /*0380*/ IMAD.WIDE R22, R7, 0x4, R16 ; /* 0x0000000407167825 */ /* 0x000fc800078e0210 */ /*0390*/ FFMA R16, R21, R20, R24 ; /* 0x0000001415107223 */ /* 0x004fe40000000018 */ /*03a0*/ LDG.E R20, [R12.64+0x18] ; /* 0x000018040c147981 */ /* 0x000ea2000c1e1900 */ /*03b0*/ IMAD.WIDE R24, R7, 0x4, R22 ; /* 0x0000000407187825 */ /* 0x000fc600078e0216 */ /*03c0*/ LDG.E R21, [R22.64] ; /* 0x0000000416157981 */ /* 0x0000a8000c1e1900 */ /*03d0*/ LDG.E R29, [R24.64] ; /* 0x00000004181d7981 */ /* 0x0002a2000c1e1900 */ /*03e0*/ FFMA R16, R15, R14, R16 ; /* 0x0000000e0f107223 */ /* 0x008fe40000000010 */ /*03f0*/ IMAD.WIDE R14, R7.reuse, 0x4, R24 ; /* 0x00000004070e7825 */ /* 0x040fe200078e0218 */ /*0400*/ LDG.E R23, [R12.64+0x20] ; /* 0x000020040c177981 */ /* 0x001ee6000c1e1900 */ /*0410*/ FFMA R26, R26, R27, R16 ; /* 0x0000001b1a1a7223 */ /* 0x010fe20000000010 */ /*0420*/ LDG.E R25, [R12.64+0x24] ; /* 0x000024040c197981 */ /* 0x002f22000c1e1900 */ /*0430*/ IMAD.WIDE R16, R7, 0x4, R14 ; /* 0x0000000407107825 */ /* 0x000fc600078e020e */ /*0440*/ LDG.E R14, [R14.64] ; /* 0x000000040e0e7981 */ /* 0x0000e2000c1e1900 */ /*0450*/ FFMA R26, R19, R18, R26 ; /* 0x00000012131a7223 */ /* 0x020fe4000000001a */ /*0460*/ IMAD.WIDE R18, R7, 0x4, R16 ; /* 0x0000000407127825 */ /* 0x000fe200078e0210 */ /*0470*/ LDG.E R22, [R12.64+0x28] ; /* 0x000028040c167981 */ /* 0x000f66000c1e1900 */ /*0480*/ FFMA R26, R9, R8, R26 ; /* 0x00000008091a7223 */ /* 0x000fe2000000001a */ /*0490*/ LDG.E R16, [R16.64] ; /* 0x0000000410107981 */ /* 0x000322000c1e1900 */ /*04a0*/ IMAD.WIDE R8, R7, 0x4, R18 ; /* 0x0000000407087825 */ /* 0x000fc600078e0212 */ /*04b0*/ LDG.E R18, [R18.64] ; /* 0x0000000412127981 */ /* 0x000368000c1e1900 */ /*04c0*/ LDG.E R24, [R8.64] ; /* 0x0000000408187981 */ /* 0x000568000c1e1900 */ /*04d0*/ LDG.E R15, [R12.64+0x2c] ; /* 0x00002c040c0f7981 */ /* 0x001f62000c1e1900 */ /*04e0*/ FFMA R26, R11, R10, R26 ; /* 0x0000000a0b1a7223 */ /* 0x000fe4000000001a */ /*04f0*/ IMAD.WIDE R10, R7, 0x4, R8 ; /* 0x00000004070a7825 */ /* 0x000fe200078e0208 */ /*0500*/ LDG.E R17, [R12.64+0x30] ; /* 0x000030040c117981 */ /* 0x002f66000c1e1900 */ /*0510*/ FFMA R26, R21, R20, R26 ; /* 0x00000014151a7223 */ /* 0x004fc4000000001a */ /*0520*/ IMAD.WIDE R20, R7, 0x4, R10 ; /* 0x0000000407147825 */ /* 0x000fe400078e020a */ /*0530*/ LDG.E R10, [R10.64] ; /* 0x000000040a0a7981 */ /* 0x0000a4000c1e1900 */ /*0540*/ FFMA R28, R29, R28, R26 ; /* 0x0000001c1d1c7223 */ /* 0x000fe4000000001a */ /*0550*/ IMAD.WIDE R26, R7.reuse, 0x4, R20 ; /* 0x00000004071a7825 */ /* 0x040fe200078e0214 */ /*0560*/ LDG.E R29, [R12.64+0x34] ; /* 0x000034040c1d7981 */ /* 0x000ea8000c1e1900 */ /*0570*/ LDG.E R20, [R20.64] ; /* 0x0000000414147981 */ /* 0x0002a2000c1e1900 */ /*0580*/ IMAD.WIDE R8, R7, 0x4, R26 ; /* 0x0000000407087825 */ /* 0x000fc600078e021a */ /*0590*/ LDG.E R19, [R26.64] ; /* 0x000000041a137981 */ /* 0x0006a8000c1e1900 */ /*05a0*/ LDG.E R11, [R8.64] ; /* 0x00000004080b7981 */ /* 0x0010a8000c1e1900 */ /*05b0*/ LDG.E R21, [R12.64+0x38] ; /* 0x000038040c157981 */ /* 0x002ea8000c1e1900 */ /*05c0*/ LDG.E R26, [R12.64+0x3c] ; /* 0x00003c040c1a7981 */ /* 0x008ee2000c1e1900 */ /*05d0*/ FFMA R14, R14, R23, R28 ; /* 0x000000170e0e7223 */ /* 0x000fc8000000001c */ /*05e0*/ FFMA R25, R16, R25, R14 ; /* 0x0000001910197223 */ /* 0x010fe2000000000e */ /*05f0*/ IADD3 R5, R5, -0x10, RZ ; /* 0xfffffff005057810 */ /* 0x000fc60007ffe0ff */ /*0600*/ FFMA R18, R18, R22, R25 ; /* 0x0000001612127223 */ /* 0x020fe20000000019 */ /*0610*/ ISETP.GT.AND P1, PT, R5, 0xc, PT ; /* 0x0000000c0500780c */ /* 0x000fc60003f24270 */ /*0620*/ FFMA R15, R24, R15, R18 ; /* 0x0000000f180f7223 */ /* 0x000fe20000000012 */ /*0630*/ UIADD3 UR6, UP0, UR6, 0x40, URZ ; /* 0x0000004006067890 */ /* 0x000fe2000ff1e03f */ /*0640*/ IMAD.WIDE R8, R7, 0x4, R8 ; /* 0x0000000407087825 */ /* 0x001fc600078e0208 */ /*0650*/ UIADD3.X UR7, URZ, UR7, URZ, UP0, !UPT ; /* 0x000000073f077290 */ /* 0x000fe200087fe43f */ /*0660*/ IADD3 R2, R2, 0x10, RZ ; /* 0x0000001002027810 */ /* 0x000fe20007ffe0ff */ /*0670*/ FFMA R10, R10, R17, R15 ; /* 0x000000110a0a7223 */ /* 0x004fc8000000000f */ /*0680*/ FFMA R10, R20, R29, R10 ; /* 0x0000001d140a7223 */ /* 0x000fc8000000000a */ /*0690*/ FFMA R10, R19, R21, R10 ; /* 0x00000015130a7223 */ /* 0x000fc8000000000a */ /*06a0*/ FFMA R24, R11, R26, R10 ; /* 0x0000001a0b187223 */ /* 0x008fe2000000000a */ /*06b0*/ @P1 BRA 0x220 ; /* 0xfffffb6000001947 */ /* 0x000fea000383ffff */ /*06c0*/ ISETP.GT.AND P1, PT, R5, 0x4, PT ; /* 0x000000040500780c */ /* 0x000fda0003f24270 */ /*06d0*/ @!P1 BRA 0x970 ; /* 0x0000029000009947 */ /* 0x000fea0003800000 */ /*06e0*/ MOV R7, c[0x0][0x168] ; /* 0x00005a0000077a02 */ /* 0x000fe20000000f00 */ /*06f0*/ LDG.E R23, [R8.64] ; /* 0x0000000408177981 */ /* 0x0000a2000c1e1900 */ /*0700*/ MOV R10, UR6 ; /* 0x00000006000a7c02 */ /* 0x000fe40008000f00 */ /*0710*/ MOV R11, UR7 ; /* 0x00000007000b7c02 */ /* 0x000fe20008000f00 */ /*0720*/ IMAD.WIDE R16, R7, 0x4, R8 ; /* 0x0000000407107825 */ /* 0x000fc800078e0208 */ /*0730*/ IMAD.WIDE R10, R6, 0x4, R10 ; /* 0x00000004060a7825 */ /* 0x000fc800078e020a */ /*0740*/ IMAD.WIDE R12, R7.reuse, 0x4, R16 ; /* 0x00000004070c7825 */ /* 0x040fe200078e0210 */ /*0750*/ LDG.E R22, [R10.64] ; /* 0x000000040a167981 */ /* 0x000ea8000c1e1900 */ /*0760*/ LDG.E R16, [R16.64] ; /* 0x0000000410107981 */ /* 0x0002e2000c1e1900 */ /*0770*/ IMAD.WIDE R14, R7, 0x4, R12 ; /* 0x00000004070e7825 */ /* 0x000fc600078e020c */ /*0780*/ LDG.E R25, [R10.64+0x4] ; /* 0x000004040a197981 */ /* 0x000ee6000c1e1900 */ /*0790*/ IMAD.WIDE R18, R7.reuse, 0x4, R14 ; /* 0x0000000407127825 */ /* 0x040fe200078e020e */ /*07a0*/ LDG.E R26, [R12.64] ; /* 0x000000040c1a7981 */ /* 0x000968000c1e1900 */ /*07b0*/ LDG.E R27, [R10.64+0x8] ; /* 0x000008040a1b7981 */ /* 0x000f62000c1e1900 */ /*07c0*/ IMAD.WIDE R20, R7, 0x4, R18 ; /* 0x0000000407147825 */ /* 0x000fc600078e0212 */ /*07d0*/ LDG.E R14, [R14.64] ; /* 0x000000040e0e7981 */ /* 0x000368000c1e1900 */ /*07e0*/ LDG.E R29, [R10.64+0xc] ; /* 0x00000c040a1d7981 */ /* 0x000f62000c1e1900 */ /*07f0*/ IMAD.WIDE R8, R7, 0x4, R20 ; /* 0x0000000407087825 */ /* 0x001fc600078e0214 */ /*0800*/ LDG.E R18, [R18.64] ; /* 0x0000000412127981 */ /* 0x000168000c1e1900 */ /*0810*/ LDG.E R28, [R10.64+0x10] ; /* 0x000010040a1c7981 */ /* 0x000f62000c1e1900 */ /*0820*/ IMAD.WIDE R12, R7, 0x4, R8 ; /* 0x00000004070c7825 */ /* 0x010fc600078e0208 */ /*0830*/ LDG.E R20, [R20.64] ; /* 0x0000000414147981 */ /* 0x000968000c1e1900 */ /*0840*/ LDG.E R15, [R10.64+0x14] ; /* 0x000014040a0f7981 */ /* 0x002f68000c1e1900 */ /*0850*/ LDG.E R17, [R8.64] ; /* 0x0000000408117981 */ /* 0x000368000c1e1900 */ /*0860*/ LDG.E R21, [R10.64+0x1c] ; /* 0x00001c040a157981 */ /* 0x010f28000c1e1900 */ /*0870*/ LDG.E R19, [R12.64] ; /* 0x000000040c137981 */ /* 0x001f28000c1e1900 */ /*0880*/ LDG.E R8, [R10.64+0x18] ; /* 0x000018040a087981 */ /* 0x002f22000c1e1900 */ /*0890*/ UIADD3 UR6, UP0, UR6, 0x20, URZ ; /* 0x0000002006067890 */ /* 0x000fe2000ff1e03f */ /*08a0*/ PLOP3.LUT P0, PT, PT, PT, PT, 0x8, 0x0 ; /* 0x000000000000781c */ /* 0x000fc40003f0e170 */ /*08b0*/ IADD3 R2, R2, 0x8, RZ ; /* 0x0000000802027810 */ /* 0x000fe40007ffe0ff */ /*08c0*/ IADD3 R5, R5, -0x8, RZ ; /* 0xfffffff805057810 */ /* 0x000fe20007ffe0ff */ /*08d0*/ UIADD3.X UR7, URZ, UR7, URZ, UP0, !UPT ; /* 0x000000073f077290 */ /* 0x000fe200087fe43f */ /*08e0*/ FFMA R22, R23, R22, R24 ; /* 0x0000001617167223 */ /* 0x004fc80000000018 */ /*08f0*/ FFMA R16, R16, R25, R22 ; /* 0x0000001910107223 */ /* 0x008fc80000000016 */ /*0900*/ FFMA R16, R26, R27, R16 ; /* 0x0000001b1a107223 */ /* 0x020fc80000000010 */ /*0910*/ FFMA R29, R14, R29, R16 ; /* 0x0000001d0e1d7223 */ /* 0x000fc80000000010 */ /*0920*/ FFMA R18, R18, R28, R29 ; /* 0x0000001c12127223 */ /* 0x000fc8000000001d */ /*0930*/ FFMA R15, R20, R15, R18 ; /* 0x0000000f140f7223 */ /* 0x000fc80000000012 */ /*0940*/ FFMA R24, R17, R8, R15 ; /* 0x0000000811187223 */ /* 0x010fe4000000000f */ /*0950*/ IMAD.WIDE R8, R7, 0x4, R12 ; /* 0x0000000407087825 */ /* 0x000fc800078e020c */ /*0960*/ FFMA R24, R19, R21, R24 ; /* 0x0000001513187223 */ /* 0x000fe40000000018 */ /*0970*/ ISETP.NE.OR P0, PT, R5, RZ, P0 ; /* 0x000000ff0500720c */ /* 0x000fda0000705670 */ /*0980*/ @!P0 BRA 0xb30 ; /* 0x000001a000008947 */ /* 0x000fea0003800000 */ /*0990*/ MOV R10, UR6 ; /* 0x00000006000a7c02 */ /* 0x000fe40008000f00 */ /*09a0*/ MOV R11, UR7 ; /* 0x00000007000b7c02 */ /* 0x000fe40008000f00 */ /*09b0*/ MOV R7, c[0x0][0x168] ; /* 0x00005a0000077a02 */ /* 0x000fc60000000f00 */ /*09c0*/ IMAD.WIDE R10, R6, 0x4, R10 ; /* 0x00000004060a7825 */ /* 0x000fc800078e020a */ /*09d0*/ IMAD.WIDE R16, R7.reuse, 0x4, R8 ; /* 0x0000000407107825 */ /* 0x040fe200078e0208 */ /*09e0*/ LDG.E R18, [R10.64] ; /* 0x000000040a127981 */ /* 0x000ea8000c1e1900 */ /*09f0*/ LDG.E R9, [R8.64] ; /* 0x0000000408097981 */ /* 0x000ea2000c1e1900 */ /*0a00*/ IMAD.WIDE R12, R7, 0x4, R16 ; /* 0x00000004070c7825 */ /* 0x000fc600078e0210 */ /*0a10*/ LDG.E R17, [R16.64] ; /* 0x0000000410117981 */ /* 0x000ee8000c1e1900 */ /*0a20*/ LDG.E R19, [R10.64+0x4] ; /* 0x000004040a137981 */ /* 0x000ee2000c1e1900 */ /*0a30*/ IMAD.WIDE R14, R7, 0x4, R12 ; /* 0x00000004070e7825 */ /* 0x000fc600078e020c */ /*0a40*/ LDG.E R21, [R12.64] ; /* 0x000000040c157981 */ /* 0x000f28000c1e1900 */ /*0a50*/ LDG.E R20, [R10.64+0x8] ; /* 0x000008040a147981 */ /* 0x000f28000c1e1900 */ /*0a60*/ LDG.E R22, [R10.64+0xc] ; /* 0x00000c040a167981 */ /* 0x000f68000c1e1900 */ /*0a70*/ LDG.E R23, [R14.64] ; /* 0x000000040e177981 */ /* 0x000f62000c1e1900 */ /*0a80*/ IADD3 R5, R5, -0x4, RZ ; /* 0xfffffffc05057810 */ /* 0x000fc80007ffe0ff */ /*0a90*/ ISETP.NE.AND P0, PT, R5, RZ, PT ; /* 0x000000ff0500720c */ /* 0x000fe20003f05270 */ /*0aa0*/ UIADD3 UR6, UP0, UR6, 0x10, URZ ; /* 0x0000001006067890 */ /* 0x000fe2000ff1e03f */ /*0ab0*/ IADD3 R2, R2, 0x4, RZ ; /* 0x0000000402027810 */ /* 0x000fc60007ffe0ff */ /*0ac0*/ UIADD3.X UR7, URZ, UR7, URZ, UP0, !UPT ; /* 0x000000073f077290 */ /* 0x000fe200087fe43f */ /*0ad0*/ FFMA R18, R9, R18, R24 ; /* 0x0000001209127223 */ /* 0x004fc80000000018 */ /*0ae0*/ FFMA R18, R17, R19, R18 ; /* 0x0000001311127223 */ /* 0x008fe40000000012 */ /*0af0*/ IMAD.WIDE R8, R7, 0x4, R14 ; /* 0x0000000407087825 */ /* 0x000fc800078e020e */ /*0b00*/ FFMA R18, R21, R20, R18 ; /* 0x0000001415127223 */ /* 0x010fc80000000012 */ /*0b10*/ FFMA R24, R23, R22, R18 ; /* 0x0000001617187223 */ /* 0x020fe20000000012 */ /*0b20*/ @P0 BRA 0x990 ; /* 0xfffffe6000000947 */ /* 0x000fea000383ffff */ /*0b30*/ ISETP.NE.AND P0, PT, R4, RZ, PT ; /* 0x000000ff0400720c */ /* 0x000fda0003f05270 */ /*0b40*/ @!P0 BRA 0xc40 ; /* 0x000000f000008947 */ /* 0x000fea0003800000 */ /*0b50*/ HFMA2.MMA R9, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff097435 */ /* 0x000fe200000001ff */ /*0b60*/ IMAD R6, R3, c[0x0][0x164], R2 ; /* 0x0000590003067a24 */ /* 0x000fe400078e0202 */ /*0b70*/ IMAD R2, R2, c[0x0][0x168], R0 ; /* 0x00005a0002027a24 */ /* 0x000fce00078e0200 */ /*0b80*/ IMAD.WIDE R6, R6, R9, c[0x0][0x170] ; /* 0x00005c0006067625 */ /* 0x000fc800078e0209 */ /*0b90*/ IMAD.WIDE R8, R2, R9, c[0x0][0x178] ; /* 0x00005e0002087625 */ /* 0x000fca00078e0209 */ /*0ba0*/ LDG.E R5, [R8.64] ; /* 0x0000000408057981 */ /* 0x0000a8000c1e1900 */ /*0bb0*/ LDG.E R2, [R6.64] ; /* 0x0000000406027981 */ /* 0x0002a2000c1e1900 */ /*0bc0*/ IADD3 R4, R4, -0x1, RZ ; /* 0xffffffff04047810 */ /* 0x000fe40007ffe0ff */ /*0bd0*/ MOV R11, c[0x0][0x168] ; /* 0x00005a00000b7a02 */ /* 0x000fe40000000f00 */ /*0be0*/ ISETP.NE.AND P0, PT, R4, RZ, PT ; /* 0x000000ff0400720c */ /* 0x000fc60003f05270 */ /*0bf0*/ IMAD.WIDE R8, R11, 0x4, R8 ; /* 0x000000040b087825 */ /* 0x001fe200078e0208 */ /*0c00*/ IADD3 R6, P1, R6, 0x4, RZ ; /* 0x0000000406067810 */ /* 0x002fc80007f3e0ff */ /*0c10*/ IADD3.X R7, RZ, R7, RZ, P1, !PT ; /* 0x00000007ff077210 */ /* 0x000fe20000ffe4ff */ /*0c20*/ FFMA R24, R5, R2, R24 ; /* 0x0000000205187223 */ /* 0x004fc80000000018 */ /*0c30*/ @P0 BRA 0xba0 ; /* 0xffffff6000000947 */ /* 0x000fea000383ffff */ /*0c40*/ MOV R2, 0x4 ; /* 0x0000000400027802 */ /* 0x000fe20000000f00 */ /*0c50*/ IMAD R3, R3, c[0x0][0x168], R0 ; /* 0x00005a0003037a24 */ /* 0x000fc800078e0200 */ /*0c60*/ IMAD.WIDE R2, R3, R2, c[0x0][0x180] ; /* 0x0000600003027625 */ /* 0x000fca00078e0202 */ /*0c70*/ STG.E [R2.64], R24 ; /* 0x0000001802007986 */ /* 0x000fe2000c101904 */ /*0c80*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0c90*/ BRA 0xc90; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0ca0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0cb0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0cc0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0cd0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ce0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0cf0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d00*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d10*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d20*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d30*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d40*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d50*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d60*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d70*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include <iostream> #include <stdio.h> #include <math.h> using namespace std; #define TILE_WIDTH 2 __global__ void MatrixMult(int m, int n, int k, float *a, float *b, float *c) { int row = threadIdx.y + blockIdx.y*blockDim.y; int col = threadIdx.x + blockIdx.x*blockDim.x; if((row < m) && (col < k)) { float temp = 0.0; for (int i = 0; i < n; ++i) { temp += a[row*n+i]*b[col+i*k]; } c[row*k+col] = temp; } } // main fn int main(void) { int m = 4; int n = 6; int k = 7; float* a = new float[m*n]; float* b = new float[n*k]; float* c = new float[m*k]; float *dev_a, *dev_b, *dev_c; dim3 dimGrid((k-1)/TILE_WIDTH+1,(m-1)/TILE_WIDTH+1,1); dim3 dimBlock(TILE_WIDTH,TILE_WIDTH,1); cudaMalloc((void**)&dev_a, m*n*sizeof(float)); cudaMalloc((void**)&dev_b, n*k*sizeof(float)); cudaMalloc((void**)&dev_c, m*k*sizeof(float)); for (int i=0; i<m*n; i++) { a[i] = sin((float) i); } for (int i=0; i<n*k; i++) { b[i] = cos((float) i); } cudaMemcpy(dev_a, a, m*n*sizeof(float), cudaMemcpyHostToDevice); cudaMemcpy(dev_b, b, n*k*sizeof(float), cudaMemcpyHostToDevice); MatrixMult<<<dimGrid,dimBlock>>>(m,n,k,dev_a,dev_b,dev_c); cudaMemcpy(c, dev_c, m*k*sizeof(float), cudaMemcpyDeviceToHost); cout<<"a matrix: \n"; for (int i=0; i<m; i++) { for (int j=0; j<n; j++) { cout<<a[n*i+j]<<" "; } cout<<"\n"; } cout<<"b matrix: \n"; for (int i=0; i<n; i++) { for (int j=0; j<k; j++) { cout<<b[k*i+j]<<" "; } cout<<"\n"; } cout<<"c matrix: \n"; for (int i=0; i<m; i++) { for (int j=0; j<k; j++) { cout<<c[k*i+j]<<" "; } cout<<"\n"; } cudaFree(dev_a); cudaFree(dev_b); cudaFree(dev_c); delete [] a; delete [] b; delete [] c; }
.file "tmpxft_00101d16_00000000-6_matrix_mult.cudafe1.cpp" .text #APP .globl _ZSt21ios_base_library_initv #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB3672: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE3672: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_ .type _Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_, @function _Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_: .LFB3694: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movl %edi, 44(%rsp) movl %esi, 40(%rsp) movl %edx, 36(%rsp) movq %rcx, 24(%rsp) movq %r8, 16(%rsp) movq %r9, 8(%rsp) movq %fs:40, %rax movq %rax, 168(%rsp) xorl %eax, %eax leaq 44(%rsp), %rax movq %rax, 112(%rsp) leaq 40(%rsp), %rax movq %rax, 120(%rsp) leaq 36(%rsp), %rax movq %rax, 128(%rsp) leaq 24(%rsp), %rax movq %rax, 136(%rsp) leaq 16(%rsp), %rax movq %rax, 144(%rsp) leaq 8(%rsp), %rax movq %rax, 152(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) movl $1, 72(%rsp) movl $1, 76(%rsp) movl $1, 80(%rsp) movl $1, 84(%rsp) leaq 56(%rsp), %rcx leaq 48(%rsp), %rdx leaq 76(%rsp), %rsi leaq 64(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 168(%rsp), %rax subq %fs:40, %rax jne .L8 addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 56(%rsp) .cfi_def_cfa_offset 200 pushq 56(%rsp) .cfi_def_cfa_offset 208 leaq 128(%rsp), %r9 movq 92(%rsp), %rcx movl 100(%rsp), %r8d movq 80(%rsp), %rsi movl 88(%rsp), %edx leaq _Z10MatrixMultiiiPfS_S_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 192 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE3694: .size _Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_, .-_Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_ .globl _Z10MatrixMultiiiPfS_S_ .type _Z10MatrixMultiiiPfS_S_, @function _Z10MatrixMultiiiPfS_S_: .LFB3695: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE3695: .size _Z10MatrixMultiiiPfS_S_, .-_Z10MatrixMultiiiPfS_S_ .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "a matrix: \n" .LC1: .string " " .LC2: .string "\n" .LC3: .string "b matrix: \n" .LC4: .string "c matrix: \n" .text .globl main .type main, @function main: .LFB3669: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $88, %rsp .cfi_def_cfa_offset 144 movq %fs:40, %rax movq %rax, 72(%rsp) xorl %eax, %eax movl $96, %edi call _Znam@PLT movq %rax, %r13 movl $168, %edi call _Znam@PLT movq %rax, %rbp movl $112, %edi call _Znam@PLT movq %rax, 8(%rsp) movl $4, 48(%rsp) movl $2, 52(%rsp) movl $1, 56(%rsp) movl $2, 60(%rsp) movl $2, 64(%rsp) movl $1, 68(%rsp) leaq 24(%rsp), %rdi movl $96, %esi call cudaMalloc@PLT leaq 32(%rsp), %rdi movl $168, %esi call cudaMalloc@PLT leaq 40(%rsp), %rdi movl $112, %esi call cudaMalloc@PLT movl $0, %ebx .L12: pxor %xmm0, %xmm0 cvtsi2ssl %ebx, %xmm0 call sinf@PLT movss %xmm0, 0(%r13,%rbx,4) addq $1, %rbx cmpq $24, %rbx jne .L12 movl $0, %ebx .L13: pxor %xmm0, %xmm0 cvtsi2ssl %ebx, %xmm0 call cosf@PLT movss %xmm0, 0(%rbp,%rbx,4) addq $1, %rbx cmpq $42, %rbx jne .L13 movl $1, %ecx movl $96, %edx movq %r13, %rsi movq 24(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movl $168, %edx movq %rbp, %rsi movq 32(%rsp), %rdi call cudaMemcpy@PLT movl 68(%rsp), %ecx movl $0, %r9d movl $0, %r8d movq 60(%rsp), %rdx movq 48(%rsp), %rdi movl 56(%rsp), %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L31 .L14: movl $2, %ecx movl $112, %edx movq 40(%rsp), %rsi movq 8(%rsp), %rdi call cudaMemcpy@PLT leaq .LC0(%rip), %rsi leaq _ZSt4cout(%rip), %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %r13, %r12 movl $0, (%rsp) leaq _ZSt4cout(%rip), %r14 leaq .LC1(%rip), %r15 .L15: movl $0, %ebx .L16: pxor %xmm0, %xmm0 cvtss2sd (%r12,%rbx,4), %xmm0 movq %r14, %rdi call _ZNSo9_M_insertIdEERSoT_@PLT movq %rax, %rdi movl $1, %edx movq %r15, %rsi call _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l@PLT addq $1, %rbx cmpq $6, %rbx jne .L16 movl $1, %edx leaq .LC2(%rip), %rsi movq %r14, %rdi call _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l@PLT addl $6, (%rsp) movl (%rsp), %eax addq $24, %r12 cmpl $24, %eax jne .L15 leaq .LC3(%rip), %rsi leaq _ZSt4cout(%rip), %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rbp, %r12 leaq 168(%rbp), %rax movq %rax, (%rsp) leaq _ZSt4cout(%rip), %r14 leaq .LC1(%rip), %r15 .L18: movl $0, %ebx .L19: pxor %xmm0, %xmm0 cvtss2sd (%r12,%rbx,4), %xmm0 movq %r14, %rdi call _ZNSo9_M_insertIdEERSoT_@PLT movq %rax, %rdi movl $1, %edx movq %r15, %rsi call _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l@PLT addq $1, %rbx cmpq $7, %rbx jne .L19 movl $1, %edx leaq .LC2(%rip), %rsi movq %r14, %rdi call _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l@PLT addq $28, %r12 cmpq %r12, (%rsp) jne .L18 leaq .LC4(%rip), %rsi leaq _ZSt4cout(%rip), %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq 8(%rsp), %r12 movl $0, (%rsp) leaq _ZSt4cout(%rip), %r14 leaq .LC1(%rip), %r15 .L21: movl $0, %ebx .L22: pxor %xmm0, %xmm0 cvtss2sd (%r12,%rbx,4), %xmm0 movq %r14, %rdi call _ZNSo9_M_insertIdEERSoT_@PLT movq %rax, %rdi movl $1, %edx movq %r15, %rsi call _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l@PLT addq $1, %rbx cmpq $7, %rbx jne .L22 movl $1, %edx leaq .LC2(%rip), %rsi movq %r14, %rdi call _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l@PLT addl $7, (%rsp) movl (%rsp), %eax addq $28, %r12 cmpl $28, %eax jne .L21 movq 24(%rsp), %rdi call cudaFree@PLT movq 32(%rsp), %rdi call cudaFree@PLT movq 40(%rsp), %rdi call cudaFree@PLT movq %r13, %rdi call _ZdaPv@PLT movq %rbp, %rdi call _ZdaPv@PLT movq 8(%rsp), %rdi call _ZdaPv@PLT movq 72(%rsp), %rax subq %fs:40, %rax jne .L32 movl $0, %eax addq $88, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L31: .cfi_restore_state movq 40(%rsp), %r9 movq 32(%rsp), %r8 movq 24(%rsp), %rcx movl $7, %edx movl $6, %esi movl $4, %edi call _Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_ jmp .L14 .L32: call __stack_chk_fail@PLT .cfi_endproc .LFE3669: .size main, .-main .section .rodata.str1.1 .LC5: .string "_Z10MatrixMultiiiPfS_S_" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB3697: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC5(%rip), %rdx movq %rdx, %rcx leaq _Z10MatrixMultiiiPfS_S_(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE3697: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include <iostream> #include <stdio.h> #include <math.h> using namespace std; #define TILE_WIDTH 2 __global__ void MatrixMult(int m, int n, int k, float *a, float *b, float *c) { int row = threadIdx.y + blockIdx.y*blockDim.y; int col = threadIdx.x + blockIdx.x*blockDim.x; if((row < m) && (col < k)) { float temp = 0.0; for (int i = 0; i < n; ++i) { temp += a[row*n+i]*b[col+i*k]; } c[row*k+col] = temp; } } // main fn int main(void) { int m = 4; int n = 6; int k = 7; float* a = new float[m*n]; float* b = new float[n*k]; float* c = new float[m*k]; float *dev_a, *dev_b, *dev_c; dim3 dimGrid((k-1)/TILE_WIDTH+1,(m-1)/TILE_WIDTH+1,1); dim3 dimBlock(TILE_WIDTH,TILE_WIDTH,1); cudaMalloc((void**)&dev_a, m*n*sizeof(float)); cudaMalloc((void**)&dev_b, n*k*sizeof(float)); cudaMalloc((void**)&dev_c, m*k*sizeof(float)); for (int i=0; i<m*n; i++) { a[i] = sin((float) i); } for (int i=0; i<n*k; i++) { b[i] = cos((float) i); } cudaMemcpy(dev_a, a, m*n*sizeof(float), cudaMemcpyHostToDevice); cudaMemcpy(dev_b, b, n*k*sizeof(float), cudaMemcpyHostToDevice); MatrixMult<<<dimGrid,dimBlock>>>(m,n,k,dev_a,dev_b,dev_c); cudaMemcpy(c, dev_c, m*k*sizeof(float), cudaMemcpyDeviceToHost); cout<<"a matrix: \n"; for (int i=0; i<m; i++) { for (int j=0; j<n; j++) { cout<<a[n*i+j]<<" "; } cout<<"\n"; } cout<<"b matrix: \n"; for (int i=0; i<n; i++) { for (int j=0; j<k; j++) { cout<<b[k*i+j]<<" "; } cout<<"\n"; } cout<<"c matrix: \n"; for (int i=0; i<m; i++) { for (int j=0; j<k; j++) { cout<<c[k*i+j]<<" "; } cout<<"\n"; } cudaFree(dev_a); cudaFree(dev_b); cudaFree(dev_c); delete [] a; delete [] b; delete [] c; }
#include <hip/hip_runtime.h> #include <iostream> #include <stdio.h> #include <math.h> using namespace std; #define TILE_WIDTH 2 __global__ void MatrixMult(int m, int n, int k, float *a, float *b, float *c) { int row = threadIdx.y + blockIdx.y*blockDim.y; int col = threadIdx.x + blockIdx.x*blockDim.x; if((row < m) && (col < k)) { float temp = 0.0; for (int i = 0; i < n; ++i) { temp += a[row*n+i]*b[col+i*k]; } c[row*k+col] = temp; } } // main fn int main(void) { int m = 4; int n = 6; int k = 7; float* a = new float[m*n]; float* b = new float[n*k]; float* c = new float[m*k]; float *dev_a, *dev_b, *dev_c; dim3 dimGrid((k-1)/TILE_WIDTH+1,(m-1)/TILE_WIDTH+1,1); dim3 dimBlock(TILE_WIDTH,TILE_WIDTH,1); hipMalloc((void**)&dev_a, m*n*sizeof(float)); hipMalloc((void**)&dev_b, n*k*sizeof(float)); hipMalloc((void**)&dev_c, m*k*sizeof(float)); for (int i=0; i<m*n; i++) { a[i] = sin((float) i); } for (int i=0; i<n*k; i++) { b[i] = cos((float) i); } hipMemcpy(dev_a, a, m*n*sizeof(float), hipMemcpyHostToDevice); hipMemcpy(dev_b, b, n*k*sizeof(float), hipMemcpyHostToDevice); MatrixMult<<<dimGrid,dimBlock>>>(m,n,k,dev_a,dev_b,dev_c); hipMemcpy(c, dev_c, m*k*sizeof(float), hipMemcpyDeviceToHost); cout<<"a matrix: \n"; for (int i=0; i<m; i++) { for (int j=0; j<n; j++) { cout<<a[n*i+j]<<" "; } cout<<"\n"; } cout<<"b matrix: \n"; for (int i=0; i<n; i++) { for (int j=0; j<k; j++) { cout<<b[k*i+j]<<" "; } cout<<"\n"; } cout<<"c matrix: \n"; for (int i=0; i<m; i++) { for (int j=0; j<k; j++) { cout<<c[k*i+j]<<" "; } cout<<"\n"; } hipFree(dev_a); hipFree(dev_b); hipFree(dev_c); delete [] a; delete [] b; delete [] c; }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include <iostream> #include <stdio.h> #include <math.h> using namespace std; #define TILE_WIDTH 2 __global__ void MatrixMult(int m, int n, int k, float *a, float *b, float *c) { int row = threadIdx.y + blockIdx.y*blockDim.y; int col = threadIdx.x + blockIdx.x*blockDim.x; if((row < m) && (col < k)) { float temp = 0.0; for (int i = 0; i < n; ++i) { temp += a[row*n+i]*b[col+i*k]; } c[row*k+col] = temp; } } // main fn int main(void) { int m = 4; int n = 6; int k = 7; float* a = new float[m*n]; float* b = new float[n*k]; float* c = new float[m*k]; float *dev_a, *dev_b, *dev_c; dim3 dimGrid((k-1)/TILE_WIDTH+1,(m-1)/TILE_WIDTH+1,1); dim3 dimBlock(TILE_WIDTH,TILE_WIDTH,1); hipMalloc((void**)&dev_a, m*n*sizeof(float)); hipMalloc((void**)&dev_b, n*k*sizeof(float)); hipMalloc((void**)&dev_c, m*k*sizeof(float)); for (int i=0; i<m*n; i++) { a[i] = sin((float) i); } for (int i=0; i<n*k; i++) { b[i] = cos((float) i); } hipMemcpy(dev_a, a, m*n*sizeof(float), hipMemcpyHostToDevice); hipMemcpy(dev_b, b, n*k*sizeof(float), hipMemcpyHostToDevice); MatrixMult<<<dimGrid,dimBlock>>>(m,n,k,dev_a,dev_b,dev_c); hipMemcpy(c, dev_c, m*k*sizeof(float), hipMemcpyDeviceToHost); cout<<"a matrix: \n"; for (int i=0; i<m; i++) { for (int j=0; j<n; j++) { cout<<a[n*i+j]<<" "; } cout<<"\n"; } cout<<"b matrix: \n"; for (int i=0; i<n; i++) { for (int j=0; j<k; j++) { cout<<b[k*i+j]<<" "; } cout<<"\n"; } cout<<"c matrix: \n"; for (int i=0; i<m; i++) { for (int j=0; j<k; j++) { cout<<c[k*i+j]<<" "; } cout<<"\n"; } hipFree(dev_a); hipFree(dev_b); hipFree(dev_c); delete [] a; delete [] b; delete [] c; }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z10MatrixMultiiiPfS_S_ .globl _Z10MatrixMultiiiPfS_S_ .p2align 8 .type _Z10MatrixMultiiiPfS_S_,@function _Z10MatrixMultiiiPfS_S_: s_clause 0x2 s_load_b32 s2, s[0:1], 0x34 s_load_b32 s4, s[0:1], 0x0 s_load_b32 s3, s[0:1], 0x8 v_bfe_u32 v2, v0, 10, 10 v_and_b32_e32 v3, 0x3ff, v0 s_waitcnt lgkmcnt(0) s_lshr_b32 s5, s2, 16 s_and_b32 s2, s2, 0xffff s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_mad_u64_u32 v[0:1], null, s15, s5, v[2:3] v_mad_u64_u32 v[1:2], null, s14, s2, v[3:4] v_cmp_gt_i32_e32 vcc_lo, s4, v0 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_cmp_gt_i32_e64 s2, s3, v1 s_and_b32 s2, vcc_lo, s2 s_delay_alu instid0(SALU_CYCLE_1) s_and_saveexec_b32 s4, s2 s_cbranch_execz .LBB0_6 s_load_b32 s2, s[0:1], 0x4 s_waitcnt lgkmcnt(0) s_cmp_lt_i32 s2, 1 s_cbranch_scc1 .LBB0_4 s_load_b128 s[4:7], s[0:1], 0x10 v_mul_lo_u32 v2, v0, s2 v_mov_b32_e32 v6, 0 v_mov_b32_e32 v4, v1 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v3, 31, v2 v_lshlrev_b64 v[2:3], 2, v[2:3] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v2, vcc_lo, s4, v2 v_add_co_ci_u32_e32 v3, vcc_lo, s5, v3, vcc_lo .p2align 6 .LBB0_3: v_ashrrev_i32_e32 v5, 31, v4 s_add_i32 s2, s2, -1 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1) s_cmp_eq_u32 s2, 0 v_lshlrev_b64 v[7:8], 2, v[4:5] v_add_nc_u32_e32 v4, s3, v4 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_co_u32 v7, vcc_lo, s6, v7 v_add_co_ci_u32_e32 v8, vcc_lo, s7, v8, vcc_lo global_load_b32 v5, v[2:3], off global_load_b32 v7, v[7:8], off v_add_co_u32 v2, vcc_lo, v2, 4 v_add_co_ci_u32_e32 v3, vcc_lo, 0, v3, vcc_lo s_waitcnt vmcnt(0) v_fmac_f32_e32 v6, v5, v7 s_cbranch_scc0 .LBB0_3 s_branch .LBB0_5 .LBB0_4: v_mov_b32_e32 v6, 0 .LBB0_5: s_load_b64 s[0:1], s[0:1], 0x20 v_mad_u64_u32 v[2:3], null, v0, s3, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v3, 31, v2 v_lshlrev_b64 v[0:1], 2, v[2:3] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v0, vcc_lo, s0, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo global_store_b32 v[0:1], v6, off .LBB0_6: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z10MatrixMultiiiPfS_S_ .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 296 .amdhsa_user_sgpr_count 14 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 1 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 1 .amdhsa_next_free_vgpr 9 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z10MatrixMultiiiPfS_S_, .Lfunc_end0-_Z10MatrixMultiiiPfS_S_ .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .offset: 0 .size: 4 .value_kind: by_value - .offset: 4 .size: 4 .value_kind: by_value - .offset: 8 .size: 4 .value_kind: by_value - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 24 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 32 .size: 8 .value_kind: global_buffer - .offset: 40 .size: 4 .value_kind: hidden_block_count_x - .offset: 44 .size: 4 .value_kind: hidden_block_count_y - .offset: 48 .size: 4 .value_kind: hidden_block_count_z - .offset: 52 .size: 2 .value_kind: hidden_group_size_x - .offset: 54 .size: 2 .value_kind: hidden_group_size_y - .offset: 56 .size: 2 .value_kind: hidden_group_size_z - .offset: 58 .size: 2 .value_kind: hidden_remainder_x - .offset: 60 .size: 2 .value_kind: hidden_remainder_y - .offset: 62 .size: 2 .value_kind: hidden_remainder_z - .offset: 80 .size: 8 .value_kind: hidden_global_offset_x - .offset: 88 .size: 8 .value_kind: hidden_global_offset_y - .offset: 96 .size: 8 .value_kind: hidden_global_offset_z - .offset: 104 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 296 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z10MatrixMultiiiPfS_S_ .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z10MatrixMultiiiPfS_S_.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 9 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <hip/hip_runtime.h> #include <iostream> #include <stdio.h> #include <math.h> using namespace std; #define TILE_WIDTH 2 __global__ void MatrixMult(int m, int n, int k, float *a, float *b, float *c) { int row = threadIdx.y + blockIdx.y*blockDim.y; int col = threadIdx.x + blockIdx.x*blockDim.x; if((row < m) && (col < k)) { float temp = 0.0; for (int i = 0; i < n; ++i) { temp += a[row*n+i]*b[col+i*k]; } c[row*k+col] = temp; } } // main fn int main(void) { int m = 4; int n = 6; int k = 7; float* a = new float[m*n]; float* b = new float[n*k]; float* c = new float[m*k]; float *dev_a, *dev_b, *dev_c; dim3 dimGrid((k-1)/TILE_WIDTH+1,(m-1)/TILE_WIDTH+1,1); dim3 dimBlock(TILE_WIDTH,TILE_WIDTH,1); hipMalloc((void**)&dev_a, m*n*sizeof(float)); hipMalloc((void**)&dev_b, n*k*sizeof(float)); hipMalloc((void**)&dev_c, m*k*sizeof(float)); for (int i=0; i<m*n; i++) { a[i] = sin((float) i); } for (int i=0; i<n*k; i++) { b[i] = cos((float) i); } hipMemcpy(dev_a, a, m*n*sizeof(float), hipMemcpyHostToDevice); hipMemcpy(dev_b, b, n*k*sizeof(float), hipMemcpyHostToDevice); MatrixMult<<<dimGrid,dimBlock>>>(m,n,k,dev_a,dev_b,dev_c); hipMemcpy(c, dev_c, m*k*sizeof(float), hipMemcpyDeviceToHost); cout<<"a matrix: \n"; for (int i=0; i<m; i++) { for (int j=0; j<n; j++) { cout<<a[n*i+j]<<" "; } cout<<"\n"; } cout<<"b matrix: \n"; for (int i=0; i<n; i++) { for (int j=0; j<k; j++) { cout<<b[k*i+j]<<" "; } cout<<"\n"; } cout<<"c matrix: \n"; for (int i=0; i<m; i++) { for (int j=0; j<k; j++) { cout<<c[k*i+j]<<" "; } cout<<"\n"; } hipFree(dev_a); hipFree(dev_b); hipFree(dev_c); delete [] a; delete [] b; delete [] c; }
.text .file "matrix_mult.hip" # Start of file scope inline assembly .globl _ZSt21ios_base_library_initv # End of file scope inline assembly .globl _Z25__device_stub__MatrixMultiiiPfS_S_ # -- Begin function _Z25__device_stub__MatrixMultiiiPfS_S_ .p2align 4, 0x90 .type _Z25__device_stub__MatrixMultiiiPfS_S_,@function _Z25__device_stub__MatrixMultiiiPfS_S_: # @_Z25__device_stub__MatrixMultiiiPfS_S_ .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movl %edi, 20(%rsp) movl %esi, 16(%rsp) movl %edx, 12(%rsp) movq %rcx, 88(%rsp) movq %r8, 80(%rsp) movq %r9, 72(%rsp) leaq 20(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 12(%rsp), %rax movq %rax, 112(%rsp) leaq 88(%rsp), %rax movq %rax, 120(%rsp) leaq 80(%rsp), %rax movq %rax, 128(%rsp) leaq 72(%rsp), %rax movq %rax, 136(%rsp) leaq 56(%rsp), %rdi leaq 40(%rsp), %rsi leaq 32(%rsp), %rdx leaq 24(%rsp), %rcx callq __hipPopCallConfiguration movq 56(%rsp), %rsi movl 64(%rsp), %edx movq 40(%rsp), %rcx movl 48(%rsp), %r8d leaq 96(%rsp), %r9 movl $_Z10MatrixMultiiiPfS_S_, %edi pushq 24(%rsp) .cfi_adjust_cfa_offset 8 pushq 40(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $168, %rsp .cfi_adjust_cfa_offset -168 retq .Lfunc_end0: .size _Z25__device_stub__MatrixMultiiiPfS_S_, .Lfunc_end0-_Z25__device_stub__MatrixMultiiiPfS_S_ .cfi_endproc # -- End function .globl main # -- Begin function main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 subq $168, %rsp .cfi_def_cfa_offset 224 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl $96, %edi callq _Znam movq %rax, %rbx movl $168, %edi callq _Znam movq %rax, %r14 movl $112, %edi callq _Znam movq %rax, %r15 leaq 16(%rsp), %rdi movl $96, %esi callq hipMalloc leaq 8(%rsp), %rdi movl $168, %esi callq hipMalloc movq %rsp, %rdi movl $112, %esi callq hipMalloc xorl %r12d, %r12d .p2align 4, 0x90 .LBB1_1: # =>This Inner Loop Header: Depth=1 xorps %xmm0, %xmm0 cvtsi2ss %r12d, %xmm0 callq sinf movss %xmm0, (%rbx,%r12,4) incq %r12 cmpq $24, %r12 jne .LBB1_1 # %bb.2: # %.preheader90.preheader xorl %r12d, %r12d .p2align 4, 0x90 .LBB1_3: # %.preheader90 # =>This Inner Loop Header: Depth=1 xorps %xmm0, %xmm0 cvtsi2ss %r12d, %xmm0 callq cosf movss %xmm0, (%r14,%r12,4) incq %r12 cmpq $42, %r12 jne .LBB1_3 # %bb.4: movq 16(%rsp), %rdi movl $96, %edx movq %rbx, %rsi movl $1, %ecx callq hipMemcpy movq 8(%rsp), %rdi movl $168, %edx movq %r14, %rsi movl $1, %ecx callq hipMemcpy movabsq $8589934594, %rdx # imm = 0x200000002 leaq 2(%rdx), %rdi movl $1, %esi movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_6 # %bb.5: movq 16(%rsp), %rax movq 8(%rsp), %rcx movq (%rsp), %rdx movl $4, 36(%rsp) movl $6, 32(%rsp) movl $7, 28(%rsp) movq %rax, 104(%rsp) movq %rcx, 96(%rsp) movq %rdx, 88(%rsp) leaq 36(%rsp), %rax movq %rax, 112(%rsp) leaq 32(%rsp), %rax movq %rax, 120(%rsp) leaq 28(%rsp), %rax movq %rax, 128(%rsp) leaq 104(%rsp), %rax movq %rax, 136(%rsp) leaq 96(%rsp), %rax movq %rax, 144(%rsp) leaq 88(%rsp), %rax movq %rax, 152(%rsp) leaq 72(%rsp), %rdi leaq 56(%rsp), %rsi leaq 48(%rsp), %rdx leaq 40(%rsp), %rcx callq __hipPopCallConfiguration movq 72(%rsp), %rsi movl 80(%rsp), %edx movq 56(%rsp), %rcx movl 64(%rsp), %r8d leaq 112(%rsp), %r9 movl $_Z10MatrixMultiiiPfS_S_, %edi pushq 40(%rsp) .cfi_adjust_cfa_offset 8 pushq 56(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_6: movq (%rsp), %rsi movl $112, %edx movq %r15, %rdi movl $2, %ecx callq hipMemcpy movl $_ZSt4cout, %edi movl $.L.str, %esi movl $11, %edx callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l xorl %r12d, %r12d movq %rbx, %r13 .p2align 4, 0x90 .LBB1_7: # %.preheader89 # =>This Loop Header: Depth=1 # Child Loop BB1_8 Depth 2 xorl %ebp, %ebp .p2align 4, 0x90 .LBB1_8: # Parent Loop BB1_7 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r13,%rbp,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $_ZSt4cout, %edi callq _ZNSo9_M_insertIdEERSoT_ movl $.L.str.1, %esi movl $1, %edx movq %rax, %rdi callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l incq %rbp cmpq $6, %rbp jne .LBB1_8 # %bb.9: # in Loop: Header=BB1_7 Depth=1 movl $_ZSt4cout, %edi movl $.L.str.2, %esi movl $1, %edx callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l incq %r12 addq $24, %r13 cmpq $4, %r12 jne .LBB1_7 # %bb.10: movl $_ZSt4cout, %edi movl $.L.str.3, %esi movl $11, %edx callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l xorl %r12d, %r12d movq %r14, %r13 .p2align 4, 0x90 .LBB1_11: # %.preheader88 # =>This Loop Header: Depth=1 # Child Loop BB1_12 Depth 2 xorl %ebp, %ebp .p2align 4, 0x90 .LBB1_12: # Parent Loop BB1_11 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r13,%rbp,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $_ZSt4cout, %edi callq _ZNSo9_M_insertIdEERSoT_ movl $.L.str.1, %esi movl $1, %edx movq %rax, %rdi callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l incq %rbp cmpq $7, %rbp jne .LBB1_12 # %bb.13: # in Loop: Header=BB1_11 Depth=1 movl $_ZSt4cout, %edi movl $.L.str.2, %esi movl $1, %edx callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l incq %r12 addq $28, %r13 cmpq $6, %r12 jne .LBB1_11 # %bb.14: movl $_ZSt4cout, %edi movl $.L.str.4, %esi movl $11, %edx callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l xorl %r12d, %r12d movq %r15, %r13 .p2align 4, 0x90 .LBB1_15: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB1_16 Depth 2 xorl %ebp, %ebp .p2align 4, 0x90 .LBB1_16: # Parent Loop BB1_15 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r13,%rbp,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $_ZSt4cout, %edi callq _ZNSo9_M_insertIdEERSoT_ movl $.L.str.1, %esi movl $1, %edx movq %rax, %rdi callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l incq %rbp cmpq $7, %rbp jne .LBB1_16 # %bb.17: # in Loop: Header=BB1_15 Depth=1 movl $_ZSt4cout, %edi movl $.L.str.2, %esi movl $1, %edx callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l incq %r12 addq $28, %r13 cmpq $4, %r12 jne .LBB1_15 # %bb.18: movq 16(%rsp), %rdi callq hipFree movq 8(%rsp), %rdi callq hipFree movq (%rsp), %rdi callq hipFree movq %rbx, %rdi callq _ZdaPv movq %r14, %rdi callq _ZdaPv movq %r15, %rdi callq _ZdaPv xorl %eax, %eax addq $168, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end1: .size main, .Lfunc_end1-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB2_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB2_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z10MatrixMultiiiPfS_S_, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end2: .size __hip_module_ctor, .Lfunc_end2-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB3_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB3_2: retq .Lfunc_end3: .size __hip_module_dtor, .Lfunc_end3-__hip_module_dtor .cfi_endproc # -- End function .type _Z10MatrixMultiiiPfS_S_,@object # @_Z10MatrixMultiiiPfS_S_ .section .rodata,"a",@progbits .globl _Z10MatrixMultiiiPfS_S_ .p2align 3, 0x0 _Z10MatrixMultiiiPfS_S_: .quad _Z25__device_stub__MatrixMultiiiPfS_S_ .size _Z10MatrixMultiiiPfS_S_, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "a matrix: \n" .size .L.str, 12 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz " " .size .L.str.1, 2 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "\n" .size .L.str.2, 2 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "b matrix: \n" .size .L.str.3, 12 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "c matrix: \n" .size .L.str.4, 12 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z10MatrixMultiiiPfS_S_" .size .L__unnamed_1, 24 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z25__device_stub__MatrixMultiiiPfS_S_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z10MatrixMultiiiPfS_S_ .addrsig_sym _ZSt4cout .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA device assembly to AMD device assembly.
code for sm_80 Function : _Z10MatrixMultiiiPfS_S_ .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R0, SR_CTAID.X ; /* 0x0000000000007919 */ /* 0x000e280000002500 */ /*0020*/ S2R R5, SR_TID.X ; /* 0x0000000000057919 */ /* 0x000e280000002100 */ /*0030*/ S2R R3, SR_CTAID.Y ; /* 0x0000000000037919 */ /* 0x000e680000002600 */ /*0040*/ S2R R2, SR_TID.Y ; /* 0x0000000000027919 */ /* 0x000e620000002200 */ /*0050*/ IMAD R0, R0, c[0x0][0x0], R5 ; /* 0x0000000000007a24 */ /* 0x001fca00078e0205 */ /*0060*/ ISETP.GE.AND P0, PT, R0, c[0x0][0x168], PT ; /* 0x00005a0000007a0c */ /* 0x000fe20003f06270 */ /*0070*/ IMAD R3, R3, c[0x0][0x4], R2 ; /* 0x0000010003037a24 */ /* 0x002fca00078e0202 */ /*0080*/ ISETP.GE.OR P0, PT, R3, c[0x0][0x160], P0 ; /* 0x0000580003007a0c */ /* 0x000fda0000706670 */ /*0090*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*00a0*/ MOV R4, c[0x0][0x164] ; /* 0x0000590000047a02 */ /* 0x000fe20000000f00 */ /*00b0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*00c0*/ HFMA2.MMA R24, -RZ, RZ, 0, 0 ; /* 0x00000000ff187435 */ /* 0x000fe400000001ff */ /*00d0*/ ISETP.GE.AND P0, PT, R4, 0x1, PT ; /* 0x000000010400780c */ /* 0x000fda0003f06270 */ /*00e0*/ @!P0 BRA 0xc40 ; /* 0x00000b5000008947 */ /* 0x000fea0003800000 */ /*00f0*/ IADD3 R2, R4.reuse, -0x1, RZ ; /* 0xffffffff04027810 */ /* 0x040fe40007ffe0ff */ /*0100*/ LOP3.LUT R4, R4, 0x3, RZ, 0xc0, !PT ; /* 0x0000000304047812 */ /* 0x000fe400078ec0ff */ /*0110*/ ISETP.GE.U32.AND P0, PT, R2, 0x3, PT ; /* 0x000000030200780c */ /* 0x000fe40003f06070 */ /*0120*/ MOV R24, RZ ; /* 0x000000ff00187202 */ /* 0x000fe40000000f00 */ /*0130*/ MOV R2, RZ ; /* 0x000000ff00027202 */ /* 0x000fd20000000f00 */ /*0140*/ @!P0 BRA 0xb30 ; /* 0x000009e000008947 */ /* 0x000fea0003800000 */ /*0150*/ IADD3 R5, -R4, c[0x0][0x164], RZ ; /* 0x0000590004057a10 */ /* 0x000fe20007ffe1ff */ /*0160*/ HFMA2.MMA R9, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff097435 */ /* 0x000fe200000001ff */ /*0170*/ ULDC.64 UR6, c[0x0][0x170] ; /* 0x00005c0000067ab9 */ /* 0x000fe20000000a00 */ /*0180*/ HFMA2.MMA R2, -RZ, RZ, 0, 0 ; /* 0x00000000ff027435 */ /* 0x000fe200000001ff */ /*0190*/ ISETP.GT.AND P0, PT, R5, RZ, PT ; /* 0x000000ff0500720c */ /* 0x000fe20003f04270 */ /*01a0*/ IMAD R6, R3, c[0x0][0x164], RZ ; /* 0x0000590003067a24 */ /* 0x000fe200078e02ff */ /*01b0*/ MOV R24, RZ ; /* 0x000000ff00187202 */ /* 0x000fca0000000f00 */ /*01c0*/ IMAD.WIDE R8, R0, R9, c[0x0][0x178] ; /* 0x00005e0000087625 */ /* 0x000fcc00078e0209 */ /*01d0*/ @!P0 BRA 0x990 ; /* 0x000007b000008947 */ /* 0x000fea0003800000 */ /*01e0*/ ISETP.GT.AND P1, PT, R5, 0xc, PT ; /* 0x0000000c0500780c */ /* 0x000fe40003f24270 */ /*01f0*/ PLOP3.LUT P0, PT, PT, PT, PT, 0x80, 0x0 ; /* 0x000000000000781c */ /* 0x000fd60003f0f070 */ /*0200*/ @!P1 BRA 0x6c0 ; /* 0x000004b000009947 */ /* 0x000fea0003800000 */ /*0210*/ PLOP3.LUT P0, PT, PT, PT, PT, 0x8, 0x0 ; /* 0x000000000000781c */ /* 0x000fe40003f0e170 */ /*0220*/ MOV R12, UR6 ; /* 0x00000006000c7c02 */ /* 0x000fe20008000f00 */ /*0230*/ LDG.E R21, [R8.64] ; /* 0x0000000408157981 */ /* 0x0000a2000c1e1900 */ /*0240*/ MOV R13, UR7 ; /* 0x00000007000d7c02 */ /* 0x000fca0008000f00 */ /*0250*/ IMAD.WIDE R12, R6, 0x4, R12 ; /* 0x00000004060c7825 */ /* 0x000fca00078e020c */ /*0260*/ LDG.E R20, [R12.64] ; /* 0x000000040c147981 */ /* 0x000ea2000c1e1900 */ /*0270*/ MOV R7, c[0x0][0x168] ; /* 0x00005a0000077a02 */ /* 0x000fc60000000f00 */ /*0280*/ LDG.E R14, [R12.64+0x4] ; /* 0x000004040c0e7981 */ /* 0x000ee4000c1e1900 */ /*0290*/ IMAD.WIDE R10, R7.reuse, 0x4, R8 ; /* 0x00000004070a7825 */ /* 0x040fe400078e0208 */ /*02a0*/ LDG.E R27, [R12.64+0x8] ; /* 0x000008040c1b7981 */ /* 0x000f28000c1e1900 */ /*02b0*/ LDG.E R15, [R10.64] ; /* 0x000000040a0f7981 */ /* 0x0002e2000c1e1900 */ /*02c0*/ IMAD.WIDE R22, R7, 0x4, R10 ; /* 0x0000000407167825 */ /* 0x000fc600078e020a */ /*02d0*/ LDG.E R18, [R12.64+0xc] ; /* 0x00000c040c127981 */ /* 0x000f66000c1e1900 */ /*02e0*/ IMAD.WIDE R28, R7.reuse, 0x4, R22 ; /* 0x00000004071c7825 */ /* 0x040fe200078e0216 */ /*02f0*/ LDG.E R26, [R22.64] ; /* 0x00000004161a7981 */ /* 0x000328000c1e1900 */ /*0300*/ LDG.E R19, [R28.64] ; /* 0x000000041c137981 */ /* 0x000362000c1e1900 */ /*0310*/ IMAD.WIDE R16, R7, 0x4, R28 ; /* 0x0000000407107825 */ /* 0x000fc600078e021c */ /*0320*/ LDG.E R8, [R12.64+0x10] ; /* 0x000010040c087981 */ /* 0x001f68000c1e1900 */ /*0330*/ LDG.E R9, [R16.64] ; /* 0x0000000410097981 */ /* 0x000168000c1e1900 */ /*0340*/ LDG.E R10, [R12.64+0x14] ; /* 0x000014040c0a7981 */ /* 0x002f68000c1e1900 */ /*0350*/ LDG.E R28, [R12.64+0x1c] ; /* 0x00001c040c1c7981 */ /* 0x000f62000c1e1900 */ /*0360*/ IMAD.WIDE R16, R7, 0x4, R16 ; /* 0x0000000407107825 */ /* 0x001fca00078e0210 */ /*0370*/ LDG.E R11, [R16.64] ; /* 0x00000004100b7981 */ /* 0x000562000c1e1900 */ /*0380*/ IMAD.WIDE R22, R7, 0x4, R16 ; /* 0x0000000407167825 */ /* 0x000fc800078e0210 */ /*0390*/ FFMA R16, R21, R20, R24 ; /* 0x0000001415107223 */ /* 0x004fe40000000018 */ /*03a0*/ LDG.E R20, [R12.64+0x18] ; /* 0x000018040c147981 */ /* 0x000ea2000c1e1900 */ /*03b0*/ IMAD.WIDE R24, R7, 0x4, R22 ; /* 0x0000000407187825 */ /* 0x000fc600078e0216 */ /*03c0*/ LDG.E R21, [R22.64] ; /* 0x0000000416157981 */ /* 0x0000a8000c1e1900 */ /*03d0*/ LDG.E R29, [R24.64] ; /* 0x00000004181d7981 */ /* 0x0002a2000c1e1900 */ /*03e0*/ FFMA R16, R15, R14, R16 ; /* 0x0000000e0f107223 */ /* 0x008fe40000000010 */ /*03f0*/ IMAD.WIDE R14, R7.reuse, 0x4, R24 ; /* 0x00000004070e7825 */ /* 0x040fe200078e0218 */ /*0400*/ LDG.E R23, [R12.64+0x20] ; /* 0x000020040c177981 */ /* 0x001ee6000c1e1900 */ /*0410*/ FFMA R26, R26, R27, R16 ; /* 0x0000001b1a1a7223 */ /* 0x010fe20000000010 */ /*0420*/ LDG.E R25, [R12.64+0x24] ; /* 0x000024040c197981 */ /* 0x002f22000c1e1900 */ /*0430*/ IMAD.WIDE R16, R7, 0x4, R14 ; /* 0x0000000407107825 */ /* 0x000fc600078e020e */ /*0440*/ LDG.E R14, [R14.64] ; /* 0x000000040e0e7981 */ /* 0x0000e2000c1e1900 */ /*0450*/ FFMA R26, R19, R18, R26 ; /* 0x00000012131a7223 */ /* 0x020fe4000000001a */ /*0460*/ IMAD.WIDE R18, R7, 0x4, R16 ; /* 0x0000000407127825 */ /* 0x000fe200078e0210 */ /*0470*/ LDG.E R22, [R12.64+0x28] ; /* 0x000028040c167981 */ /* 0x000f66000c1e1900 */ /*0480*/ FFMA R26, R9, R8, R26 ; /* 0x00000008091a7223 */ /* 0x000fe2000000001a */ /*0490*/ LDG.E R16, [R16.64] ; /* 0x0000000410107981 */ /* 0x000322000c1e1900 */ /*04a0*/ IMAD.WIDE R8, R7, 0x4, R18 ; /* 0x0000000407087825 */ /* 0x000fc600078e0212 */ /*04b0*/ LDG.E R18, [R18.64] ; /* 0x0000000412127981 */ /* 0x000368000c1e1900 */ /*04c0*/ LDG.E R24, [R8.64] ; /* 0x0000000408187981 */ /* 0x000568000c1e1900 */ /*04d0*/ LDG.E R15, [R12.64+0x2c] ; /* 0x00002c040c0f7981 */ /* 0x001f62000c1e1900 */ /*04e0*/ FFMA R26, R11, R10, R26 ; /* 0x0000000a0b1a7223 */ /* 0x000fe4000000001a */ /*04f0*/ IMAD.WIDE R10, R7, 0x4, R8 ; /* 0x00000004070a7825 */ /* 0x000fe200078e0208 */ /*0500*/ LDG.E R17, [R12.64+0x30] ; /* 0x000030040c117981 */ /* 0x002f66000c1e1900 */ /*0510*/ FFMA R26, R21, R20, R26 ; /* 0x00000014151a7223 */ /* 0x004fc4000000001a */ /*0520*/ IMAD.WIDE R20, R7, 0x4, R10 ; /* 0x0000000407147825 */ /* 0x000fe400078e020a */ /*0530*/ LDG.E R10, [R10.64] ; /* 0x000000040a0a7981 */ /* 0x0000a4000c1e1900 */ /*0540*/ FFMA R28, R29, R28, R26 ; /* 0x0000001c1d1c7223 */ /* 0x000fe4000000001a */ /*0550*/ IMAD.WIDE R26, R7.reuse, 0x4, R20 ; /* 0x00000004071a7825 */ /* 0x040fe200078e0214 */ /*0560*/ LDG.E R29, [R12.64+0x34] ; /* 0x000034040c1d7981 */ /* 0x000ea8000c1e1900 */ /*0570*/ LDG.E R20, [R20.64] ; /* 0x0000000414147981 */ /* 0x0002a2000c1e1900 */ /*0580*/ IMAD.WIDE R8, R7, 0x4, R26 ; /* 0x0000000407087825 */ /* 0x000fc600078e021a */ /*0590*/ LDG.E R19, [R26.64] ; /* 0x000000041a137981 */ /* 0x0006a8000c1e1900 */ /*05a0*/ LDG.E R11, [R8.64] ; /* 0x00000004080b7981 */ /* 0x0010a8000c1e1900 */ /*05b0*/ LDG.E R21, [R12.64+0x38] ; /* 0x000038040c157981 */ /* 0x002ea8000c1e1900 */ /*05c0*/ LDG.E R26, [R12.64+0x3c] ; /* 0x00003c040c1a7981 */ /* 0x008ee2000c1e1900 */ /*05d0*/ FFMA R14, R14, R23, R28 ; /* 0x000000170e0e7223 */ /* 0x000fc8000000001c */ /*05e0*/ FFMA R25, R16, R25, R14 ; /* 0x0000001910197223 */ /* 0x010fe2000000000e */ /*05f0*/ IADD3 R5, R5, -0x10, RZ ; /* 0xfffffff005057810 */ /* 0x000fc60007ffe0ff */ /*0600*/ FFMA R18, R18, R22, R25 ; /* 0x0000001612127223 */ /* 0x020fe20000000019 */ /*0610*/ ISETP.GT.AND P1, PT, R5, 0xc, PT ; /* 0x0000000c0500780c */ /* 0x000fc60003f24270 */ /*0620*/ FFMA R15, R24, R15, R18 ; /* 0x0000000f180f7223 */ /* 0x000fe20000000012 */ /*0630*/ UIADD3 UR6, UP0, UR6, 0x40, URZ ; /* 0x0000004006067890 */ /* 0x000fe2000ff1e03f */ /*0640*/ IMAD.WIDE R8, R7, 0x4, R8 ; /* 0x0000000407087825 */ /* 0x001fc600078e0208 */ /*0650*/ UIADD3.X UR7, URZ, UR7, URZ, UP0, !UPT ; /* 0x000000073f077290 */ /* 0x000fe200087fe43f */ /*0660*/ IADD3 R2, R2, 0x10, RZ ; /* 0x0000001002027810 */ /* 0x000fe20007ffe0ff */ /*0670*/ FFMA R10, R10, R17, R15 ; /* 0x000000110a0a7223 */ /* 0x004fc8000000000f */ /*0680*/ FFMA R10, R20, R29, R10 ; /* 0x0000001d140a7223 */ /* 0x000fc8000000000a */ /*0690*/ FFMA R10, R19, R21, R10 ; /* 0x00000015130a7223 */ /* 0x000fc8000000000a */ /*06a0*/ FFMA R24, R11, R26, R10 ; /* 0x0000001a0b187223 */ /* 0x008fe2000000000a */ /*06b0*/ @P1 BRA 0x220 ; /* 0xfffffb6000001947 */ /* 0x000fea000383ffff */ /*06c0*/ ISETP.GT.AND P1, PT, R5, 0x4, PT ; /* 0x000000040500780c */ /* 0x000fda0003f24270 */ /*06d0*/ @!P1 BRA 0x970 ; /* 0x0000029000009947 */ /* 0x000fea0003800000 */ /*06e0*/ MOV R7, c[0x0][0x168] ; /* 0x00005a0000077a02 */ /* 0x000fe20000000f00 */ /*06f0*/ LDG.E R23, [R8.64] ; /* 0x0000000408177981 */ /* 0x0000a2000c1e1900 */ /*0700*/ MOV R10, UR6 ; /* 0x00000006000a7c02 */ /* 0x000fe40008000f00 */ /*0710*/ MOV R11, UR7 ; /* 0x00000007000b7c02 */ /* 0x000fe20008000f00 */ /*0720*/ IMAD.WIDE R16, R7, 0x4, R8 ; /* 0x0000000407107825 */ /* 0x000fc800078e0208 */ /*0730*/ IMAD.WIDE R10, R6, 0x4, R10 ; /* 0x00000004060a7825 */ /* 0x000fc800078e020a */ /*0740*/ IMAD.WIDE R12, R7.reuse, 0x4, R16 ; /* 0x00000004070c7825 */ /* 0x040fe200078e0210 */ /*0750*/ LDG.E R22, [R10.64] ; /* 0x000000040a167981 */ /* 0x000ea8000c1e1900 */ /*0760*/ LDG.E R16, [R16.64] ; /* 0x0000000410107981 */ /* 0x0002e2000c1e1900 */ /*0770*/ IMAD.WIDE R14, R7, 0x4, R12 ; /* 0x00000004070e7825 */ /* 0x000fc600078e020c */ /*0780*/ LDG.E R25, [R10.64+0x4] ; /* 0x000004040a197981 */ /* 0x000ee6000c1e1900 */ /*0790*/ IMAD.WIDE R18, R7.reuse, 0x4, R14 ; /* 0x0000000407127825 */ /* 0x040fe200078e020e */ /*07a0*/ LDG.E R26, [R12.64] ; /* 0x000000040c1a7981 */ /* 0x000968000c1e1900 */ /*07b0*/ LDG.E R27, [R10.64+0x8] ; /* 0x000008040a1b7981 */ /* 0x000f62000c1e1900 */ /*07c0*/ IMAD.WIDE R20, R7, 0x4, R18 ; /* 0x0000000407147825 */ /* 0x000fc600078e0212 */ /*07d0*/ LDG.E R14, [R14.64] ; /* 0x000000040e0e7981 */ /* 0x000368000c1e1900 */ /*07e0*/ LDG.E R29, [R10.64+0xc] ; /* 0x00000c040a1d7981 */ /* 0x000f62000c1e1900 */ /*07f0*/ IMAD.WIDE R8, R7, 0x4, R20 ; /* 0x0000000407087825 */ /* 0x001fc600078e0214 */ /*0800*/ LDG.E R18, [R18.64] ; /* 0x0000000412127981 */ /* 0x000168000c1e1900 */ /*0810*/ LDG.E R28, [R10.64+0x10] ; /* 0x000010040a1c7981 */ /* 0x000f62000c1e1900 */ /*0820*/ IMAD.WIDE R12, R7, 0x4, R8 ; /* 0x00000004070c7825 */ /* 0x010fc600078e0208 */ /*0830*/ LDG.E R20, [R20.64] ; /* 0x0000000414147981 */ /* 0x000968000c1e1900 */ /*0840*/ LDG.E R15, [R10.64+0x14] ; /* 0x000014040a0f7981 */ /* 0x002f68000c1e1900 */ /*0850*/ LDG.E R17, [R8.64] ; /* 0x0000000408117981 */ /* 0x000368000c1e1900 */ /*0860*/ LDG.E R21, [R10.64+0x1c] ; /* 0x00001c040a157981 */ /* 0x010f28000c1e1900 */ /*0870*/ LDG.E R19, [R12.64] ; /* 0x000000040c137981 */ /* 0x001f28000c1e1900 */ /*0880*/ LDG.E R8, [R10.64+0x18] ; /* 0x000018040a087981 */ /* 0x002f22000c1e1900 */ /*0890*/ UIADD3 UR6, UP0, UR6, 0x20, URZ ; /* 0x0000002006067890 */ /* 0x000fe2000ff1e03f */ /*08a0*/ PLOP3.LUT P0, PT, PT, PT, PT, 0x8, 0x0 ; /* 0x000000000000781c */ /* 0x000fc40003f0e170 */ /*08b0*/ IADD3 R2, R2, 0x8, RZ ; /* 0x0000000802027810 */ /* 0x000fe40007ffe0ff */ /*08c0*/ IADD3 R5, R5, -0x8, RZ ; /* 0xfffffff805057810 */ /* 0x000fe20007ffe0ff */ /*08d0*/ UIADD3.X UR7, URZ, UR7, URZ, UP0, !UPT ; /* 0x000000073f077290 */ /* 0x000fe200087fe43f */ /*08e0*/ FFMA R22, R23, R22, R24 ; /* 0x0000001617167223 */ /* 0x004fc80000000018 */ /*08f0*/ FFMA R16, R16, R25, R22 ; /* 0x0000001910107223 */ /* 0x008fc80000000016 */ /*0900*/ FFMA R16, R26, R27, R16 ; /* 0x0000001b1a107223 */ /* 0x020fc80000000010 */ /*0910*/ FFMA R29, R14, R29, R16 ; /* 0x0000001d0e1d7223 */ /* 0x000fc80000000010 */ /*0920*/ FFMA R18, R18, R28, R29 ; /* 0x0000001c12127223 */ /* 0x000fc8000000001d */ /*0930*/ FFMA R15, R20, R15, R18 ; /* 0x0000000f140f7223 */ /* 0x000fc80000000012 */ /*0940*/ FFMA R24, R17, R8, R15 ; /* 0x0000000811187223 */ /* 0x010fe4000000000f */ /*0950*/ IMAD.WIDE R8, R7, 0x4, R12 ; /* 0x0000000407087825 */ /* 0x000fc800078e020c */ /*0960*/ FFMA R24, R19, R21, R24 ; /* 0x0000001513187223 */ /* 0x000fe40000000018 */ /*0970*/ ISETP.NE.OR P0, PT, R5, RZ, P0 ; /* 0x000000ff0500720c */ /* 0x000fda0000705670 */ /*0980*/ @!P0 BRA 0xb30 ; /* 0x000001a000008947 */ /* 0x000fea0003800000 */ /*0990*/ MOV R10, UR6 ; /* 0x00000006000a7c02 */ /* 0x000fe40008000f00 */ /*09a0*/ MOV R11, UR7 ; /* 0x00000007000b7c02 */ /* 0x000fe40008000f00 */ /*09b0*/ MOV R7, c[0x0][0x168] ; /* 0x00005a0000077a02 */ /* 0x000fc60000000f00 */ /*09c0*/ IMAD.WIDE R10, R6, 0x4, R10 ; /* 0x00000004060a7825 */ /* 0x000fc800078e020a */ /*09d0*/ IMAD.WIDE R16, R7.reuse, 0x4, R8 ; /* 0x0000000407107825 */ /* 0x040fe200078e0208 */ /*09e0*/ LDG.E R18, [R10.64] ; /* 0x000000040a127981 */ /* 0x000ea8000c1e1900 */ /*09f0*/ LDG.E R9, [R8.64] ; /* 0x0000000408097981 */ /* 0x000ea2000c1e1900 */ /*0a00*/ IMAD.WIDE R12, R7, 0x4, R16 ; /* 0x00000004070c7825 */ /* 0x000fc600078e0210 */ /*0a10*/ LDG.E R17, [R16.64] ; /* 0x0000000410117981 */ /* 0x000ee8000c1e1900 */ /*0a20*/ LDG.E R19, [R10.64+0x4] ; /* 0x000004040a137981 */ /* 0x000ee2000c1e1900 */ /*0a30*/ IMAD.WIDE R14, R7, 0x4, R12 ; /* 0x00000004070e7825 */ /* 0x000fc600078e020c */ /*0a40*/ LDG.E R21, [R12.64] ; /* 0x000000040c157981 */ /* 0x000f28000c1e1900 */ /*0a50*/ LDG.E R20, [R10.64+0x8] ; /* 0x000008040a147981 */ /* 0x000f28000c1e1900 */ /*0a60*/ LDG.E R22, [R10.64+0xc] ; /* 0x00000c040a167981 */ /* 0x000f68000c1e1900 */ /*0a70*/ LDG.E R23, [R14.64] ; /* 0x000000040e177981 */ /* 0x000f62000c1e1900 */ /*0a80*/ IADD3 R5, R5, -0x4, RZ ; /* 0xfffffffc05057810 */ /* 0x000fc80007ffe0ff */ /*0a90*/ ISETP.NE.AND P0, PT, R5, RZ, PT ; /* 0x000000ff0500720c */ /* 0x000fe20003f05270 */ /*0aa0*/ UIADD3 UR6, UP0, UR6, 0x10, URZ ; /* 0x0000001006067890 */ /* 0x000fe2000ff1e03f */ /*0ab0*/ IADD3 R2, R2, 0x4, RZ ; /* 0x0000000402027810 */ /* 0x000fc60007ffe0ff */ /*0ac0*/ UIADD3.X UR7, URZ, UR7, URZ, UP0, !UPT ; /* 0x000000073f077290 */ /* 0x000fe200087fe43f */ /*0ad0*/ FFMA R18, R9, R18, R24 ; /* 0x0000001209127223 */ /* 0x004fc80000000018 */ /*0ae0*/ FFMA R18, R17, R19, R18 ; /* 0x0000001311127223 */ /* 0x008fe40000000012 */ /*0af0*/ IMAD.WIDE R8, R7, 0x4, R14 ; /* 0x0000000407087825 */ /* 0x000fc800078e020e */ /*0b00*/ FFMA R18, R21, R20, R18 ; /* 0x0000001415127223 */ /* 0x010fc80000000012 */ /*0b10*/ FFMA R24, R23, R22, R18 ; /* 0x0000001617187223 */ /* 0x020fe20000000012 */ /*0b20*/ @P0 BRA 0x990 ; /* 0xfffffe6000000947 */ /* 0x000fea000383ffff */ /*0b30*/ ISETP.NE.AND P0, PT, R4, RZ, PT ; /* 0x000000ff0400720c */ /* 0x000fda0003f05270 */ /*0b40*/ @!P0 BRA 0xc40 ; /* 0x000000f000008947 */ /* 0x000fea0003800000 */ /*0b50*/ HFMA2.MMA R9, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff097435 */ /* 0x000fe200000001ff */ /*0b60*/ IMAD R6, R3, c[0x0][0x164], R2 ; /* 0x0000590003067a24 */ /* 0x000fe400078e0202 */ /*0b70*/ IMAD R2, R2, c[0x0][0x168], R0 ; /* 0x00005a0002027a24 */ /* 0x000fce00078e0200 */ /*0b80*/ IMAD.WIDE R6, R6, R9, c[0x0][0x170] ; /* 0x00005c0006067625 */ /* 0x000fc800078e0209 */ /*0b90*/ IMAD.WIDE R8, R2, R9, c[0x0][0x178] ; /* 0x00005e0002087625 */ /* 0x000fca00078e0209 */ /*0ba0*/ LDG.E R5, [R8.64] ; /* 0x0000000408057981 */ /* 0x0000a8000c1e1900 */ /*0bb0*/ LDG.E R2, [R6.64] ; /* 0x0000000406027981 */ /* 0x0002a2000c1e1900 */ /*0bc0*/ IADD3 R4, R4, -0x1, RZ ; /* 0xffffffff04047810 */ /* 0x000fe40007ffe0ff */ /*0bd0*/ MOV R11, c[0x0][0x168] ; /* 0x00005a00000b7a02 */ /* 0x000fe40000000f00 */ /*0be0*/ ISETP.NE.AND P0, PT, R4, RZ, PT ; /* 0x000000ff0400720c */ /* 0x000fc60003f05270 */ /*0bf0*/ IMAD.WIDE R8, R11, 0x4, R8 ; /* 0x000000040b087825 */ /* 0x001fe200078e0208 */ /*0c00*/ IADD3 R6, P1, R6, 0x4, RZ ; /* 0x0000000406067810 */ /* 0x002fc80007f3e0ff */ /*0c10*/ IADD3.X R7, RZ, R7, RZ, P1, !PT ; /* 0x00000007ff077210 */ /* 0x000fe20000ffe4ff */ /*0c20*/ FFMA R24, R5, R2, R24 ; /* 0x0000000205187223 */ /* 0x004fc80000000018 */ /*0c30*/ @P0 BRA 0xba0 ; /* 0xffffff6000000947 */ /* 0x000fea000383ffff */ /*0c40*/ MOV R2, 0x4 ; /* 0x0000000400027802 */ /* 0x000fe20000000f00 */ /*0c50*/ IMAD R3, R3, c[0x0][0x168], R0 ; /* 0x00005a0003037a24 */ /* 0x000fc800078e0200 */ /*0c60*/ IMAD.WIDE R2, R3, R2, c[0x0][0x180] ; /* 0x0000600003027625 */ /* 0x000fca00078e0202 */ /*0c70*/ STG.E [R2.64], R24 ; /* 0x0000001802007986 */ /* 0x000fe2000c101904 */ /*0c80*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0c90*/ BRA 0xc90; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0ca0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0cb0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0cc0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0cd0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ce0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0cf0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d00*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d10*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d20*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d30*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d40*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d50*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d60*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0d70*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z10MatrixMultiiiPfS_S_ .globl _Z10MatrixMultiiiPfS_S_ .p2align 8 .type _Z10MatrixMultiiiPfS_S_,@function _Z10MatrixMultiiiPfS_S_: s_clause 0x2 s_load_b32 s2, s[0:1], 0x34 s_load_b32 s4, s[0:1], 0x0 s_load_b32 s3, s[0:1], 0x8 v_bfe_u32 v2, v0, 10, 10 v_and_b32_e32 v3, 0x3ff, v0 s_waitcnt lgkmcnt(0) s_lshr_b32 s5, s2, 16 s_and_b32 s2, s2, 0xffff s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_mad_u64_u32 v[0:1], null, s15, s5, v[2:3] v_mad_u64_u32 v[1:2], null, s14, s2, v[3:4] v_cmp_gt_i32_e32 vcc_lo, s4, v0 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_cmp_gt_i32_e64 s2, s3, v1 s_and_b32 s2, vcc_lo, s2 s_delay_alu instid0(SALU_CYCLE_1) s_and_saveexec_b32 s4, s2 s_cbranch_execz .LBB0_6 s_load_b32 s2, s[0:1], 0x4 s_waitcnt lgkmcnt(0) s_cmp_lt_i32 s2, 1 s_cbranch_scc1 .LBB0_4 s_load_b128 s[4:7], s[0:1], 0x10 v_mul_lo_u32 v2, v0, s2 v_mov_b32_e32 v6, 0 v_mov_b32_e32 v4, v1 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v3, 31, v2 v_lshlrev_b64 v[2:3], 2, v[2:3] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v2, vcc_lo, s4, v2 v_add_co_ci_u32_e32 v3, vcc_lo, s5, v3, vcc_lo .p2align 6 .LBB0_3: v_ashrrev_i32_e32 v5, 31, v4 s_add_i32 s2, s2, -1 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1) s_cmp_eq_u32 s2, 0 v_lshlrev_b64 v[7:8], 2, v[4:5] v_add_nc_u32_e32 v4, s3, v4 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_co_u32 v7, vcc_lo, s6, v7 v_add_co_ci_u32_e32 v8, vcc_lo, s7, v8, vcc_lo global_load_b32 v5, v[2:3], off global_load_b32 v7, v[7:8], off v_add_co_u32 v2, vcc_lo, v2, 4 v_add_co_ci_u32_e32 v3, vcc_lo, 0, v3, vcc_lo s_waitcnt vmcnt(0) v_fmac_f32_e32 v6, v5, v7 s_cbranch_scc0 .LBB0_3 s_branch .LBB0_5 .LBB0_4: v_mov_b32_e32 v6, 0 .LBB0_5: s_load_b64 s[0:1], s[0:1], 0x20 v_mad_u64_u32 v[2:3], null, v0, s3, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v3, 31, v2 v_lshlrev_b64 v[0:1], 2, v[2:3] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v0, vcc_lo, s0, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo global_store_b32 v[0:1], v6, off .LBB0_6: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z10MatrixMultiiiPfS_S_ .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 296 .amdhsa_user_sgpr_count 14 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 1 .amdhsa_system_sgpr_workgroup_id_z 0 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 1 .amdhsa_next_free_vgpr 9 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z10MatrixMultiiiPfS_S_, .Lfunc_end0-_Z10MatrixMultiiiPfS_S_ .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .offset: 0 .size: 4 .value_kind: by_value - .offset: 4 .size: 4 .value_kind: by_value - .offset: 8 .size: 4 .value_kind: by_value - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 24 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 32 .size: 8 .value_kind: global_buffer - .offset: 40 .size: 4 .value_kind: hidden_block_count_x - .offset: 44 .size: 4 .value_kind: hidden_block_count_y - .offset: 48 .size: 4 .value_kind: hidden_block_count_z - .offset: 52 .size: 2 .value_kind: hidden_group_size_x - .offset: 54 .size: 2 .value_kind: hidden_group_size_y - .offset: 56 .size: 2 .value_kind: hidden_group_size_z - .offset: 58 .size: 2 .value_kind: hidden_remainder_x - .offset: 60 .size: 2 .value_kind: hidden_remainder_y - .offset: 62 .size: 2 .value_kind: hidden_remainder_z - .offset: 80 .size: 8 .value_kind: hidden_global_offset_x - .offset: 88 .size: 8 .value_kind: hidden_global_offset_y - .offset: 96 .size: 8 .value_kind: hidden_global_offset_z - .offset: 104 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 296 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z10MatrixMultiiiPfS_S_ .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z10MatrixMultiiiPfS_S_.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 9 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_00101d16_00000000-6_matrix_mult.cudafe1.cpp" .text #APP .globl _ZSt21ios_base_library_initv #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB3672: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE3672: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_ .type _Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_, @function _Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_: .LFB3694: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movl %edi, 44(%rsp) movl %esi, 40(%rsp) movl %edx, 36(%rsp) movq %rcx, 24(%rsp) movq %r8, 16(%rsp) movq %r9, 8(%rsp) movq %fs:40, %rax movq %rax, 168(%rsp) xorl %eax, %eax leaq 44(%rsp), %rax movq %rax, 112(%rsp) leaq 40(%rsp), %rax movq %rax, 120(%rsp) leaq 36(%rsp), %rax movq %rax, 128(%rsp) leaq 24(%rsp), %rax movq %rax, 136(%rsp) leaq 16(%rsp), %rax movq %rax, 144(%rsp) leaq 8(%rsp), %rax movq %rax, 152(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) movl $1, 72(%rsp) movl $1, 76(%rsp) movl $1, 80(%rsp) movl $1, 84(%rsp) leaq 56(%rsp), %rcx leaq 48(%rsp), %rdx leaq 76(%rsp), %rsi leaq 64(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 168(%rsp), %rax subq %fs:40, %rax jne .L8 addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 56(%rsp) .cfi_def_cfa_offset 200 pushq 56(%rsp) .cfi_def_cfa_offset 208 leaq 128(%rsp), %r9 movq 92(%rsp), %rcx movl 100(%rsp), %r8d movq 80(%rsp), %rsi movl 88(%rsp), %edx leaq _Z10MatrixMultiiiPfS_S_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 192 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE3694: .size _Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_, .-_Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_ .globl _Z10MatrixMultiiiPfS_S_ .type _Z10MatrixMultiiiPfS_S_, @function _Z10MatrixMultiiiPfS_S_: .LFB3695: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE3695: .size _Z10MatrixMultiiiPfS_S_, .-_Z10MatrixMultiiiPfS_S_ .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "a matrix: \n" .LC1: .string " " .LC2: .string "\n" .LC3: .string "b matrix: \n" .LC4: .string "c matrix: \n" .text .globl main .type main, @function main: .LFB3669: .cfi_startproc endbr64 pushq %r15 .cfi_def_cfa_offset 16 .cfi_offset 15, -16 pushq %r14 .cfi_def_cfa_offset 24 .cfi_offset 14, -24 pushq %r13 .cfi_def_cfa_offset 32 .cfi_offset 13, -32 pushq %r12 .cfi_def_cfa_offset 40 .cfi_offset 12, -40 pushq %rbp .cfi_def_cfa_offset 48 .cfi_offset 6, -48 pushq %rbx .cfi_def_cfa_offset 56 .cfi_offset 3, -56 subq $88, %rsp .cfi_def_cfa_offset 144 movq %fs:40, %rax movq %rax, 72(%rsp) xorl %eax, %eax movl $96, %edi call _Znam@PLT movq %rax, %r13 movl $168, %edi call _Znam@PLT movq %rax, %rbp movl $112, %edi call _Znam@PLT movq %rax, 8(%rsp) movl $4, 48(%rsp) movl $2, 52(%rsp) movl $1, 56(%rsp) movl $2, 60(%rsp) movl $2, 64(%rsp) movl $1, 68(%rsp) leaq 24(%rsp), %rdi movl $96, %esi call cudaMalloc@PLT leaq 32(%rsp), %rdi movl $168, %esi call cudaMalloc@PLT leaq 40(%rsp), %rdi movl $112, %esi call cudaMalloc@PLT movl $0, %ebx .L12: pxor %xmm0, %xmm0 cvtsi2ssl %ebx, %xmm0 call sinf@PLT movss %xmm0, 0(%r13,%rbx,4) addq $1, %rbx cmpq $24, %rbx jne .L12 movl $0, %ebx .L13: pxor %xmm0, %xmm0 cvtsi2ssl %ebx, %xmm0 call cosf@PLT movss %xmm0, 0(%rbp,%rbx,4) addq $1, %rbx cmpq $42, %rbx jne .L13 movl $1, %ecx movl $96, %edx movq %r13, %rsi movq 24(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movl $168, %edx movq %rbp, %rsi movq 32(%rsp), %rdi call cudaMemcpy@PLT movl 68(%rsp), %ecx movl $0, %r9d movl $0, %r8d movq 60(%rsp), %rdx movq 48(%rsp), %rdi movl 56(%rsp), %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L31 .L14: movl $2, %ecx movl $112, %edx movq 40(%rsp), %rsi movq 8(%rsp), %rdi call cudaMemcpy@PLT leaq .LC0(%rip), %rsi leaq _ZSt4cout(%rip), %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %r13, %r12 movl $0, (%rsp) leaq _ZSt4cout(%rip), %r14 leaq .LC1(%rip), %r15 .L15: movl $0, %ebx .L16: pxor %xmm0, %xmm0 cvtss2sd (%r12,%rbx,4), %xmm0 movq %r14, %rdi call _ZNSo9_M_insertIdEERSoT_@PLT movq %rax, %rdi movl $1, %edx movq %r15, %rsi call _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l@PLT addq $1, %rbx cmpq $6, %rbx jne .L16 movl $1, %edx leaq .LC2(%rip), %rsi movq %r14, %rdi call _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l@PLT addl $6, (%rsp) movl (%rsp), %eax addq $24, %r12 cmpl $24, %eax jne .L15 leaq .LC3(%rip), %rsi leaq _ZSt4cout(%rip), %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rbp, %r12 leaq 168(%rbp), %rax movq %rax, (%rsp) leaq _ZSt4cout(%rip), %r14 leaq .LC1(%rip), %r15 .L18: movl $0, %ebx .L19: pxor %xmm0, %xmm0 cvtss2sd (%r12,%rbx,4), %xmm0 movq %r14, %rdi call _ZNSo9_M_insertIdEERSoT_@PLT movq %rax, %rdi movl $1, %edx movq %r15, %rsi call _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l@PLT addq $1, %rbx cmpq $7, %rbx jne .L19 movl $1, %edx leaq .LC2(%rip), %rsi movq %r14, %rdi call _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l@PLT addq $28, %r12 cmpq %r12, (%rsp) jne .L18 leaq .LC4(%rip), %rsi leaq _ZSt4cout(%rip), %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq 8(%rsp), %r12 movl $0, (%rsp) leaq _ZSt4cout(%rip), %r14 leaq .LC1(%rip), %r15 .L21: movl $0, %ebx .L22: pxor %xmm0, %xmm0 cvtss2sd (%r12,%rbx,4), %xmm0 movq %r14, %rdi call _ZNSo9_M_insertIdEERSoT_@PLT movq %rax, %rdi movl $1, %edx movq %r15, %rsi call _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l@PLT addq $1, %rbx cmpq $7, %rbx jne .L22 movl $1, %edx leaq .LC2(%rip), %rsi movq %r14, %rdi call _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l@PLT addl $7, (%rsp) movl (%rsp), %eax addq $28, %r12 cmpl $28, %eax jne .L21 movq 24(%rsp), %rdi call cudaFree@PLT movq 32(%rsp), %rdi call cudaFree@PLT movq 40(%rsp), %rdi call cudaFree@PLT movq %r13, %rdi call _ZdaPv@PLT movq %rbp, %rdi call _ZdaPv@PLT movq 8(%rsp), %rdi call _ZdaPv@PLT movq 72(%rsp), %rax subq %fs:40, %rax jne .L32 movl $0, %eax addq $88, %rsp .cfi_remember_state .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %rbp .cfi_def_cfa_offset 40 popq %r12 .cfi_def_cfa_offset 32 popq %r13 .cfi_def_cfa_offset 24 popq %r14 .cfi_def_cfa_offset 16 popq %r15 .cfi_def_cfa_offset 8 ret .L31: .cfi_restore_state movq 40(%rsp), %r9 movq 32(%rsp), %r8 movq 24(%rsp), %rcx movl $7, %edx movl $6, %esi movl $4, %edi call _Z37__device_stub__Z10MatrixMultiiiPfS_S_iiiPfS_S_ jmp .L14 .L32: call __stack_chk_fail@PLT .cfi_endproc .LFE3669: .size main, .-main .section .rodata.str1.1 .LC5: .string "_Z10MatrixMultiiiPfS_S_" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB3697: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC5(%rip), %rdx movq %rdx, %rcx leaq _Z10MatrixMultiiiPfS_S_(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE3697: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "matrix_mult.hip" # Start of file scope inline assembly .globl _ZSt21ios_base_library_initv # End of file scope inline assembly .globl _Z25__device_stub__MatrixMultiiiPfS_S_ # -- Begin function _Z25__device_stub__MatrixMultiiiPfS_S_ .p2align 4, 0x90 .type _Z25__device_stub__MatrixMultiiiPfS_S_,@function _Z25__device_stub__MatrixMultiiiPfS_S_: # @_Z25__device_stub__MatrixMultiiiPfS_S_ .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movl %edi, 20(%rsp) movl %esi, 16(%rsp) movl %edx, 12(%rsp) movq %rcx, 88(%rsp) movq %r8, 80(%rsp) movq %r9, 72(%rsp) leaq 20(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 12(%rsp), %rax movq %rax, 112(%rsp) leaq 88(%rsp), %rax movq %rax, 120(%rsp) leaq 80(%rsp), %rax movq %rax, 128(%rsp) leaq 72(%rsp), %rax movq %rax, 136(%rsp) leaq 56(%rsp), %rdi leaq 40(%rsp), %rsi leaq 32(%rsp), %rdx leaq 24(%rsp), %rcx callq __hipPopCallConfiguration movq 56(%rsp), %rsi movl 64(%rsp), %edx movq 40(%rsp), %rcx movl 48(%rsp), %r8d leaq 96(%rsp), %r9 movl $_Z10MatrixMultiiiPfS_S_, %edi pushq 24(%rsp) .cfi_adjust_cfa_offset 8 pushq 40(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $168, %rsp .cfi_adjust_cfa_offset -168 retq .Lfunc_end0: .size _Z25__device_stub__MatrixMultiiiPfS_S_, .Lfunc_end0-_Z25__device_stub__MatrixMultiiiPfS_S_ .cfi_endproc # -- End function .globl main # -- Begin function main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rbp .cfi_def_cfa_offset 16 pushq %r15 .cfi_def_cfa_offset 24 pushq %r14 .cfi_def_cfa_offset 32 pushq %r13 .cfi_def_cfa_offset 40 pushq %r12 .cfi_def_cfa_offset 48 pushq %rbx .cfi_def_cfa_offset 56 subq $168, %rsp .cfi_def_cfa_offset 224 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 movl $96, %edi callq _Znam movq %rax, %rbx movl $168, %edi callq _Znam movq %rax, %r14 movl $112, %edi callq _Znam movq %rax, %r15 leaq 16(%rsp), %rdi movl $96, %esi callq hipMalloc leaq 8(%rsp), %rdi movl $168, %esi callq hipMalloc movq %rsp, %rdi movl $112, %esi callq hipMalloc xorl %r12d, %r12d .p2align 4, 0x90 .LBB1_1: # =>This Inner Loop Header: Depth=1 xorps %xmm0, %xmm0 cvtsi2ss %r12d, %xmm0 callq sinf movss %xmm0, (%rbx,%r12,4) incq %r12 cmpq $24, %r12 jne .LBB1_1 # %bb.2: # %.preheader90.preheader xorl %r12d, %r12d .p2align 4, 0x90 .LBB1_3: # %.preheader90 # =>This Inner Loop Header: Depth=1 xorps %xmm0, %xmm0 cvtsi2ss %r12d, %xmm0 callq cosf movss %xmm0, (%r14,%r12,4) incq %r12 cmpq $42, %r12 jne .LBB1_3 # %bb.4: movq 16(%rsp), %rdi movl $96, %edx movq %rbx, %rsi movl $1, %ecx callq hipMemcpy movq 8(%rsp), %rdi movl $168, %edx movq %r14, %rsi movl $1, %ecx callq hipMemcpy movabsq $8589934594, %rdx # imm = 0x200000002 leaq 2(%rdx), %rdi movl $1, %esi movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_6 # %bb.5: movq 16(%rsp), %rax movq 8(%rsp), %rcx movq (%rsp), %rdx movl $4, 36(%rsp) movl $6, 32(%rsp) movl $7, 28(%rsp) movq %rax, 104(%rsp) movq %rcx, 96(%rsp) movq %rdx, 88(%rsp) leaq 36(%rsp), %rax movq %rax, 112(%rsp) leaq 32(%rsp), %rax movq %rax, 120(%rsp) leaq 28(%rsp), %rax movq %rax, 128(%rsp) leaq 104(%rsp), %rax movq %rax, 136(%rsp) leaq 96(%rsp), %rax movq %rax, 144(%rsp) leaq 88(%rsp), %rax movq %rax, 152(%rsp) leaq 72(%rsp), %rdi leaq 56(%rsp), %rsi leaq 48(%rsp), %rdx leaq 40(%rsp), %rcx callq __hipPopCallConfiguration movq 72(%rsp), %rsi movl 80(%rsp), %edx movq 56(%rsp), %rcx movl 64(%rsp), %r8d leaq 112(%rsp), %r9 movl $_Z10MatrixMultiiiPfS_S_, %edi pushq 40(%rsp) .cfi_adjust_cfa_offset 8 pushq 56(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_6: movq (%rsp), %rsi movl $112, %edx movq %r15, %rdi movl $2, %ecx callq hipMemcpy movl $_ZSt4cout, %edi movl $.L.str, %esi movl $11, %edx callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l xorl %r12d, %r12d movq %rbx, %r13 .p2align 4, 0x90 .LBB1_7: # %.preheader89 # =>This Loop Header: Depth=1 # Child Loop BB1_8 Depth 2 xorl %ebp, %ebp .p2align 4, 0x90 .LBB1_8: # Parent Loop BB1_7 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r13,%rbp,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $_ZSt4cout, %edi callq _ZNSo9_M_insertIdEERSoT_ movl $.L.str.1, %esi movl $1, %edx movq %rax, %rdi callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l incq %rbp cmpq $6, %rbp jne .LBB1_8 # %bb.9: # in Loop: Header=BB1_7 Depth=1 movl $_ZSt4cout, %edi movl $.L.str.2, %esi movl $1, %edx callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l incq %r12 addq $24, %r13 cmpq $4, %r12 jne .LBB1_7 # %bb.10: movl $_ZSt4cout, %edi movl $.L.str.3, %esi movl $11, %edx callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l xorl %r12d, %r12d movq %r14, %r13 .p2align 4, 0x90 .LBB1_11: # %.preheader88 # =>This Loop Header: Depth=1 # Child Loop BB1_12 Depth 2 xorl %ebp, %ebp .p2align 4, 0x90 .LBB1_12: # Parent Loop BB1_11 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r13,%rbp,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $_ZSt4cout, %edi callq _ZNSo9_M_insertIdEERSoT_ movl $.L.str.1, %esi movl $1, %edx movq %rax, %rdi callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l incq %rbp cmpq $7, %rbp jne .LBB1_12 # %bb.13: # in Loop: Header=BB1_11 Depth=1 movl $_ZSt4cout, %edi movl $.L.str.2, %esi movl $1, %edx callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l incq %r12 addq $28, %r13 cmpq $6, %r12 jne .LBB1_11 # %bb.14: movl $_ZSt4cout, %edi movl $.L.str.4, %esi movl $11, %edx callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l xorl %r12d, %r12d movq %r15, %r13 .p2align 4, 0x90 .LBB1_15: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB1_16 Depth 2 xorl %ebp, %ebp .p2align 4, 0x90 .LBB1_16: # Parent Loop BB1_15 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r13,%rbp,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $_ZSt4cout, %edi callq _ZNSo9_M_insertIdEERSoT_ movl $.L.str.1, %esi movl $1, %edx movq %rax, %rdi callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l incq %rbp cmpq $7, %rbp jne .LBB1_16 # %bb.17: # in Loop: Header=BB1_15 Depth=1 movl $_ZSt4cout, %edi movl $.L.str.2, %esi movl $1, %edx callq _ZSt16__ostream_insertIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_PKS3_l incq %r12 addq $28, %r13 cmpq $4, %r12 jne .LBB1_15 # %bb.18: movq 16(%rsp), %rdi callq hipFree movq 8(%rsp), %rdi callq hipFree movq (%rsp), %rdi callq hipFree movq %rbx, %rdi callq _ZdaPv movq %r14, %rdi callq _ZdaPv movq %r15, %rdi callq _ZdaPv xorl %eax, %eax addq $168, %rsp .cfi_def_cfa_offset 56 popq %rbx .cfi_def_cfa_offset 48 popq %r12 .cfi_def_cfa_offset 40 popq %r13 .cfi_def_cfa_offset 32 popq %r14 .cfi_def_cfa_offset 24 popq %r15 .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 retq .Lfunc_end1: .size main, .Lfunc_end1-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB2_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB2_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z10MatrixMultiiiPfS_S_, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end2: .size __hip_module_ctor, .Lfunc_end2-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB3_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB3_2: retq .Lfunc_end3: .size __hip_module_dtor, .Lfunc_end3-__hip_module_dtor .cfi_endproc # -- End function .type _Z10MatrixMultiiiPfS_S_,@object # @_Z10MatrixMultiiiPfS_S_ .section .rodata,"a",@progbits .globl _Z10MatrixMultiiiPfS_S_ .p2align 3, 0x0 _Z10MatrixMultiiiPfS_S_: .quad _Z25__device_stub__MatrixMultiiiPfS_S_ .size _Z10MatrixMultiiiPfS_S_, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "a matrix: \n" .size .L.str, 12 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz " " .size .L.str.1, 2 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "\n" .size .L.str.2, 2 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "b matrix: \n" .size .L.str.3, 12 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "c matrix: \n" .size .L.str.4, 12 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z10MatrixMultiiiPfS_S_" .size .L__unnamed_1, 24 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z25__device_stub__MatrixMultiiiPfS_S_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z10MatrixMultiiiPfS_S_ .addrsig_sym _ZSt4cout .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include "includes.h" __global__ void erosionLayers3DKernel( unsigned short *d_dst, unsigned short *d_src, int w, int h, int d, int kernel_radius ) { __shared__ unsigned short smem[ER_LAYERS_BLOCKDIM_X][ER_LAYERS_BLOCKDIM_Y][(ER_LAYERS_RESULT_STEPS + 2 * ER_LAYERS_HALO_STEPS) * ER_LAYERS_BLOCKDIM_Z + 1]; unsigned short *smem_thread = smem[threadIdx.x][threadIdx.y]; //Offset to the upper halo edge const int baseX = blockIdx.x * ER_LAYERS_BLOCKDIM_X + threadIdx.x; const int baseY = blockIdx.y * ER_LAYERS_BLOCKDIM_Y + threadIdx.y; const int baseZ = (blockIdx.z * ER_LAYERS_RESULT_STEPS - ER_LAYERS_HALO_STEPS) * ER_LAYERS_BLOCKDIM_Z + threadIdx.z; d_src += (baseZ * h + baseY) * w + baseX; d_dst += (baseZ * h + baseY) * w + baseX; const int pitch = w*h; //Main data #pragma unroll for (int i = ER_LAYERS_HALO_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z] = d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch]; } //Upper halo #pragma unroll for (int i = 0; i < ER_LAYERS_HALO_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z] = (baseZ + i * ER_LAYERS_BLOCKDIM_Z >= 0) ? d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch] : 0; } //Lower halo #pragma unroll for (int i = ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS + ER_LAYERS_HALO_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z]= (baseZ + i * ER_LAYERS_BLOCKDIM_Z < d) ? d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch] : 0; } //Compute and store results __syncthreads(); #pragma unroll for (int i = ER_LAYERS_HALO_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i++) { unsigned short *smem_kern = &smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z - kernel_radius]; unsigned short val = smem_kern[0]; //#pragma unroll for (int j = 1; j <= 2*kernel_radius; j++) { val = min(val, smem_kern[j]); } d_dst[i * ER_LAYERS_BLOCKDIM_Z * pitch] = val; } }
.file "tmpxft_00012e33_00000000-6_erosionLayers3DKernel.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2029: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2029: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z47__device_stub__Z21erosionLayers3DKernelPtS_iiiiPtS_iiii .type _Z47__device_stub__Z21erosionLayers3DKernelPtS_iiiiPtS_iiii, @function _Z47__device_stub__Z21erosionLayers3DKernelPtS_iiiiPtS_iiii: .LFB2051: .cfi_startproc endbr64 subq $168, %rsp .cfi_def_cfa_offset 176 movq %rdi, 24(%rsp) movq %rsi, 16(%rsp) movl %edx, 12(%rsp) movl %ecx, 8(%rsp) movl %r8d, 4(%rsp) movl %r9d, (%rsp) movq %fs:40, %rax movq %rax, 152(%rsp) xorl %eax, %eax leaq 24(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 12(%rsp), %rax movq %rax, 112(%rsp) leaq 8(%rsp), %rax movq %rax, 120(%rsp) leaq 4(%rsp), %rax movq %rax, 128(%rsp) movq %rsp, %rax movq %rax, 136(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 56(%rsp) movl $1, 60(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) leaq 40(%rsp), %rcx leaq 32(%rsp), %rdx leaq 60(%rsp), %rsi leaq 48(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 152(%rsp), %rax subq %fs:40, %rax jne .L8 addq $168, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 40(%rsp) .cfi_def_cfa_offset 184 pushq 40(%rsp) .cfi_def_cfa_offset 192 leaq 112(%rsp), %r9 movq 76(%rsp), %rcx movl 84(%rsp), %r8d movq 64(%rsp), %rsi movl 72(%rsp), %edx leaq _Z21erosionLayers3DKernelPtS_iiii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 176 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z47__device_stub__Z21erosionLayers3DKernelPtS_iiiiPtS_iiii, .-_Z47__device_stub__Z21erosionLayers3DKernelPtS_iiiiPtS_iiii .globl _Z21erosionLayers3DKernelPtS_iiii .type _Z21erosionLayers3DKernelPtS_iiii, @function _Z21erosionLayers3DKernelPtS_iiii: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z47__device_stub__Z21erosionLayers3DKernelPtS_iiiiPtS_iiii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z21erosionLayers3DKernelPtS_iiii, .-_Z21erosionLayers3DKernelPtS_iiii .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "_Z21erosionLayers3DKernelPtS_iiii" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2054: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC0(%rip), %rdx movq %rdx, %rcx leaq _Z21erosionLayers3DKernelPtS_iiii(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2054: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include "includes.h" __global__ void erosionLayers3DKernel( unsigned short *d_dst, unsigned short *d_src, int w, int h, int d, int kernel_radius ) { __shared__ unsigned short smem[ER_LAYERS_BLOCKDIM_X][ER_LAYERS_BLOCKDIM_Y][(ER_LAYERS_RESULT_STEPS + 2 * ER_LAYERS_HALO_STEPS) * ER_LAYERS_BLOCKDIM_Z + 1]; unsigned short *smem_thread = smem[threadIdx.x][threadIdx.y]; //Offset to the upper halo edge const int baseX = blockIdx.x * ER_LAYERS_BLOCKDIM_X + threadIdx.x; const int baseY = blockIdx.y * ER_LAYERS_BLOCKDIM_Y + threadIdx.y; const int baseZ = (blockIdx.z * ER_LAYERS_RESULT_STEPS - ER_LAYERS_HALO_STEPS) * ER_LAYERS_BLOCKDIM_Z + threadIdx.z; d_src += (baseZ * h + baseY) * w + baseX; d_dst += (baseZ * h + baseY) * w + baseX; const int pitch = w*h; //Main data #pragma unroll for (int i = ER_LAYERS_HALO_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z] = d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch]; } //Upper halo #pragma unroll for (int i = 0; i < ER_LAYERS_HALO_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z] = (baseZ + i * ER_LAYERS_BLOCKDIM_Z >= 0) ? d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch] : 0; } //Lower halo #pragma unroll for (int i = ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS + ER_LAYERS_HALO_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z]= (baseZ + i * ER_LAYERS_BLOCKDIM_Z < d) ? d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch] : 0; } //Compute and store results __syncthreads(); #pragma unroll for (int i = ER_LAYERS_HALO_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i++) { unsigned short *smem_kern = &smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z - kernel_radius]; unsigned short val = smem_kern[0]; //#pragma unroll for (int j = 1; j <= 2*kernel_radius; j++) { val = min(val, smem_kern[j]); } d_dst[i * ER_LAYERS_BLOCKDIM_Z * pitch] = val; } }
#include <hip/hip_runtime.h> #include "includes.h" __global__ void erosionLayers3DKernel( unsigned short *d_dst, unsigned short *d_src, int w, int h, int d, int kernel_radius ) { __shared__ unsigned short smem[ER_LAYERS_BLOCKDIM_X][ER_LAYERS_BLOCKDIM_Y][(ER_LAYERS_RESULT_STEPS + 2 * ER_LAYERS_HALO_STEPS) * ER_LAYERS_BLOCKDIM_Z + 1]; unsigned short *smem_thread = smem[threadIdx.x][threadIdx.y]; //Offset to the upper halo edge const int baseX = blockIdx.x * ER_LAYERS_BLOCKDIM_X + threadIdx.x; const int baseY = blockIdx.y * ER_LAYERS_BLOCKDIM_Y + threadIdx.y; const int baseZ = (blockIdx.z * ER_LAYERS_RESULT_STEPS - ER_LAYERS_HALO_STEPS) * ER_LAYERS_BLOCKDIM_Z + threadIdx.z; d_src += (baseZ * h + baseY) * w + baseX; d_dst += (baseZ * h + baseY) * w + baseX; const int pitch = w*h; //Main data #pragma unroll for (int i = ER_LAYERS_HALO_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z] = d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch]; } //Upper halo #pragma unroll for (int i = 0; i < ER_LAYERS_HALO_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z] = (baseZ + i * ER_LAYERS_BLOCKDIM_Z >= 0) ? d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch] : 0; } //Lower halo #pragma unroll for (int i = ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS + ER_LAYERS_HALO_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z]= (baseZ + i * ER_LAYERS_BLOCKDIM_Z < d) ? d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch] : 0; } //Compute and store results __syncthreads(); #pragma unroll for (int i = ER_LAYERS_HALO_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i++) { unsigned short *smem_kern = &smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z - kernel_radius]; unsigned short val = smem_kern[0]; //#pragma unroll for (int j = 1; j <= 2*kernel_radius; j++) { val = min(val, smem_kern[j]); } d_dst[i * ER_LAYERS_BLOCKDIM_Z * pitch] = val; } }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include "includes.h" __global__ void erosionLayers3DKernel( unsigned short *d_dst, unsigned short *d_src, int w, int h, int d, int kernel_radius ) { __shared__ unsigned short smem[ER_LAYERS_BLOCKDIM_X][ER_LAYERS_BLOCKDIM_Y][(ER_LAYERS_RESULT_STEPS + 2 * ER_LAYERS_HALO_STEPS) * ER_LAYERS_BLOCKDIM_Z + 1]; unsigned short *smem_thread = smem[threadIdx.x][threadIdx.y]; //Offset to the upper halo edge const int baseX = blockIdx.x * ER_LAYERS_BLOCKDIM_X + threadIdx.x; const int baseY = blockIdx.y * ER_LAYERS_BLOCKDIM_Y + threadIdx.y; const int baseZ = (blockIdx.z * ER_LAYERS_RESULT_STEPS - ER_LAYERS_HALO_STEPS) * ER_LAYERS_BLOCKDIM_Z + threadIdx.z; d_src += (baseZ * h + baseY) * w + baseX; d_dst += (baseZ * h + baseY) * w + baseX; const int pitch = w*h; //Main data #pragma unroll for (int i = ER_LAYERS_HALO_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z] = d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch]; } //Upper halo #pragma unroll for (int i = 0; i < ER_LAYERS_HALO_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z] = (baseZ + i * ER_LAYERS_BLOCKDIM_Z >= 0) ? d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch] : 0; } //Lower halo #pragma unroll for (int i = ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS + ER_LAYERS_HALO_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z]= (baseZ + i * ER_LAYERS_BLOCKDIM_Z < d) ? d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch] : 0; } //Compute and store results __syncthreads(); #pragma unroll for (int i = ER_LAYERS_HALO_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i++) { unsigned short *smem_kern = &smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z - kernel_radius]; unsigned short val = smem_kern[0]; //#pragma unroll for (int j = 1; j <= 2*kernel_radius; j++) { val = min(val, smem_kern[j]); } d_dst[i * ER_LAYERS_BLOCKDIM_Z * pitch] = val; } }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z21erosionLayers3DKernelPtS_iiii .globl _Z21erosionLayers3DKernelPtS_iiii .p2align 8 .type _Z21erosionLayers3DKernelPtS_iiii,@function _Z21erosionLayers3DKernelPtS_iiii: s_load_b128 s[4:7], s[0:1], 0x8 v_bfe_u32 v4, v0, 20, 10 s_lshl_b32 s2, s15, 3 v_bfe_u32 v5, v0, 10, 10 v_and_b32_e32 v6, 0x3ff, v0 s_mov_b32 s3, exec_lo v_add3_u32 v8, v4, s2, -8 s_lshl_b32 s2, s14, 3 v_mul_u32_u24_e32 v7, 50, v5 s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_lo_u32 v1, v8, s7 v_add3_u32 v1, s2, v5, v1 s_lshl_b32 s2, s13, 3 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_lo_u32 v1, v1, s6 v_add3_u32 v0, s2, v6, v1 s_mul_i32 s2, s7, s6 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) s_lshl_b32 s6, s2, 3 s_mul_i32 s8, s2, 12 v_ashrrev_i32_e32 v1, 31, v0 s_ashr_i32 s7, s6, 31 s_ashr_i32 s9, s8, 31 s_lshl_b64 s[6:7], s[6:7], 1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[2:3], 1, v[0:1] v_add_co_u32 v2, vcc_lo, s4, v2 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_add_co_ci_u32_e32 v3, vcc_lo, s5, v3, vcc_lo s_lshl_b64 s[4:5], s[8:9], 1 v_add_co_u32 v9, vcc_lo, v2, s6 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v10, vcc_lo, s7, v3, vcc_lo v_add_co_u32 v11, vcc_lo, v2, s4 v_add_co_ci_u32_e32 v12, vcc_lo, s5, v3, vcc_lo s_clause 0x1 global_load_u16 v13, v[9:10], off global_load_u16 v12, v[11:12], off v_mov_b32_e32 v11, 0 v_mad_u32_u24 v7, v6, 0x190, v7 v_mov_b32_e32 v9, 0 s_delay_alu instid0(VALU_DEP_2) v_lshl_add_u32 v10, v4, 1, v7 s_waitcnt vmcnt(1) ds_store_b16 v10, v13 offset:16 s_waitcnt vmcnt(0) ds_store_b16 v10, v12 offset:24 v_cmpx_lt_i32_e32 -1, v8 s_cbranch_execz .LBB0_2 global_load_u16 v9, v[2:3], off .LBB0_2: s_or_b32 exec_lo, exec_lo, s3 s_delay_alu instid0(SALU_CYCLE_1) s_mov_b32 s3, exec_lo s_waitcnt vmcnt(0) ds_store_b16 v10, v9 v_cmpx_lt_i32_e32 -5, v8 s_cbranch_execz .LBB0_4 s_lshl_b32 s4, s2, 2 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_ashr_i32 s5, s4, 31 s_lshl_b64 s[4:5], s[4:5], 1 s_delay_alu instid0(SALU_CYCLE_1) v_add_co_u32 v11, vcc_lo, v2, s4 v_add_co_ci_u32_e32 v12, vcc_lo, s5, v3, vcc_lo global_load_u16 v11, v[11:12], off .LBB0_4: s_or_b32 exec_lo, exec_lo, s3 s_load_b32 s3, s[0:1], 0x18 v_dual_mov_b32 v9, 0 :: v_dual_add_nc_u32 v12, 16, v8 s_waitcnt vmcnt(0) ds_store_b16 v10, v11 offset:8 s_waitcnt lgkmcnt(0) v_cmp_gt_i32_e32 vcc_lo, s3, v12 v_mov_b32_e32 v12, 0 s_and_saveexec_b32 s4, vcc_lo s_cbranch_execz .LBB0_6 s_lshl_b32 s6, s2, 4 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_ashr_i32 s7, s6, 31 s_lshl_b64 s[6:7], s[6:7], 1 s_delay_alu instid0(SALU_CYCLE_1) v_add_co_u32 v10, vcc_lo, v2, s6 v_add_co_ci_u32_e32 v11, vcc_lo, s7, v3, vcc_lo global_load_u16 v12, v[10:11], off .LBB0_6: s_or_b32 exec_lo, exec_lo, s4 v_add_nc_u32_e32 v10, 20, v8 v_lshl_add_u32 v8, v4, 1, v7 s_delay_alu instid0(VALU_DEP_2) v_cmp_gt_i32_e32 vcc_lo, s3, v10 s_waitcnt vmcnt(0) ds_store_b16 v8, v12 offset:32 s_and_saveexec_b32 s3, vcc_lo s_cbranch_execz .LBB0_8 s_mul_i32 s4, s2, 20 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_ashr_i32 s5, s4, 31 s_lshl_b64 s[4:5], s[4:5], 1 s_delay_alu instid0(SALU_CYCLE_1) v_add_co_u32 v2, vcc_lo, v2, s4 v_add_co_ci_u32_e32 v3, vcc_lo, s5, v3, vcc_lo global_load_u16 v9, v[2:3], off .LBB0_8: s_or_b32 exec_lo, exec_lo, s3 s_load_b32 s6, s[0:1], 0x1c s_waitcnt vmcnt(0) ds_store_b16 v8, v9 offset:40 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv v_subrev_nc_u32_e32 v3, s6, v4 s_lshl_b32 s4, s6, 1 s_cmp_gt_i32 s6, 0 s_cselect_b32 s5, -1, 0 s_delay_alu instid0(VALU_DEP_1) v_lshl_add_u32 v2, v3, 1, v7 s_max_i32 s3, s4, 1 s_cmp_lt_i32 s6, 1 ds_load_u16 v2, v2 offset:16 s_cbranch_scc1 .LBB0_11 v_mul_u32_u24_e32 v8, 0x190, v6 v_mul_u32_u24_e32 v9, 50, v5 v_lshlrev_b32_e32 v10, 1, v4 s_mov_b32 s6, s3 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add3_u32 v8, v8, v9, v10 v_subrev_nc_u32_e32 v8, s4, v8 s_delay_alu instid0(VALU_DEP_1) v_add_nc_u32_e32 v8, 18, v8 .LBB0_10: ds_load_u16 v9, v8 v_add_nc_u32_e32 v8, 2, v8 s_add_i32 s6, s6, -1 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_eq_u32 s6, 0 s_waitcnt lgkmcnt(0) v_min_u16 v2, v2, v9 s_cbranch_scc0 .LBB0_10 .LBB0_11: s_load_b64 s[0:1], s[0:1], 0x0 v_lshl_add_u32 v3, v3, 1, v7 v_lshlrev_b64 v[7:8], 1, v[0:1] s_lshl_b32 s6, s2, 3 s_delay_alu instid0(SALU_CYCLE_1) s_ashr_i32 s7, s6, 31 ds_load_u16 v0, v3 offset:24 s_lshl_b64 s[6:7], s[6:7], 1 s_waitcnt lgkmcnt(0) v_add_co_u32 v1, vcc_lo, s0, v7 v_add_co_ci_u32_e32 v3, vcc_lo, s1, v8, vcc_lo s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v7, vcc_lo, v1, s6 v_add_co_ci_u32_e32 v8, vcc_lo, s7, v3, vcc_lo s_and_not1_b32 vcc_lo, exec_lo, s5 global_store_b16 v[7:8], v2, off s_cbranch_vccnz .LBB0_14 v_mul_u32_u24_e32 v2, 0x190, v6 v_mul_u32_u24_e32 v5, 50, v5 v_lshlrev_b32_e32 v4, 1, v4 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add3_u32 v2, v2, v5, v4 v_subrev_nc_u32_e32 v2, s4, v2 s_delay_alu instid0(VALU_DEP_1) v_add_nc_u32_e32 v2, 26, v2 .LBB0_13: ds_load_u16 v4, v2 v_add_nc_u32_e32 v2, 2, v2 s_add_i32 s3, s3, -1 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_lg_u32 s3, 0 s_waitcnt lgkmcnt(0) v_min_u16 v0, v0, v4 s_cbranch_scc1 .LBB0_13 .LBB0_14: s_mul_i32 s0, s2, 12 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_ashr_i32 s1, s0, 31 s_lshl_b64 s[0:1], s[0:1], 1 s_delay_alu instid0(SALU_CYCLE_1) v_add_co_u32 v1, vcc_lo, v1, s0 v_add_co_ci_u32_e32 v2, vcc_lo, s1, v3, vcc_lo global_store_b16 v[1:2], v0, off s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z21erosionLayers3DKernelPtS_iiii .amdhsa_group_segment_fixed_size 3200 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 32 .amdhsa_user_sgpr_count 13 .amdhsa_user_sgpr_dispatch_ptr 0 .amdhsa_user_sgpr_queue_ptr 0 .amdhsa_user_sgpr_kernarg_segment_ptr 1 .amdhsa_user_sgpr_dispatch_id 0 .amdhsa_user_sgpr_private_segment_size 0 .amdhsa_wavefront_size32 1 .amdhsa_uses_dynamic_stack 0 .amdhsa_enable_private_segment 0 .amdhsa_system_sgpr_workgroup_id_x 1 .amdhsa_system_sgpr_workgroup_id_y 1 .amdhsa_system_sgpr_workgroup_id_z 1 .amdhsa_system_sgpr_workgroup_info 0 .amdhsa_system_vgpr_workitem_id 2 .amdhsa_next_free_vgpr 14 .amdhsa_next_free_sgpr 16 .amdhsa_float_round_mode_32 0 .amdhsa_float_round_mode_16_64 0 .amdhsa_float_denorm_mode_32 3 .amdhsa_float_denorm_mode_16_64 3 .amdhsa_dx10_clamp 1 .amdhsa_ieee_mode 1 .amdhsa_fp16_overflow 0 .amdhsa_workgroup_processor_mode 1 .amdhsa_memory_ordered 1 .amdhsa_forward_progress 0 .amdhsa_shared_vgpr_count 0 .amdhsa_exception_fp_ieee_invalid_op 0 .amdhsa_exception_fp_denorm_src 0 .amdhsa_exception_fp_ieee_div_zero 0 .amdhsa_exception_fp_ieee_overflow 0 .amdhsa_exception_fp_ieee_underflow 0 .amdhsa_exception_fp_ieee_inexact 0 .amdhsa_exception_int_div_zero 0 .end_amdhsa_kernel .text .Lfunc_end0: .size _Z21erosionLayers3DKernelPtS_iiii, .Lfunc_end0-_Z21erosionLayers3DKernelPtS_iiii .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .type __hip_cuid_,@object .section .bss,"aw",@nobits .globl __hip_cuid_ __hip_cuid_: .byte 0 .size __hip_cuid_, 1 .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym __hip_cuid_ .amdgpu_metadata --- amdhsa.kernels: - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 8 .size: 8 .value_kind: global_buffer - .offset: 16 .size: 4 .value_kind: by_value - .offset: 20 .size: 4 .value_kind: by_value - .offset: 24 .size: 4 .value_kind: by_value - .offset: 28 .size: 4 .value_kind: by_value .group_segment_fixed_size: 3200 .kernarg_segment_align: 8 .kernarg_segment_size: 32 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z21erosionLayers3DKernelPtS_iiii .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z21erosionLayers3DKernelPtS_iiii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 14 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 amdhsa.target: amdgcn-amd-amdhsa--gfx1100 amdhsa.version: - 1 - 2 ... .end_amdgpu_metadata
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <hip/hip_runtime.h> #include "includes.h" __global__ void erosionLayers3DKernel( unsigned short *d_dst, unsigned short *d_src, int w, int h, int d, int kernel_radius ) { __shared__ unsigned short smem[ER_LAYERS_BLOCKDIM_X][ER_LAYERS_BLOCKDIM_Y][(ER_LAYERS_RESULT_STEPS + 2 * ER_LAYERS_HALO_STEPS) * ER_LAYERS_BLOCKDIM_Z + 1]; unsigned short *smem_thread = smem[threadIdx.x][threadIdx.y]; //Offset to the upper halo edge const int baseX = blockIdx.x * ER_LAYERS_BLOCKDIM_X + threadIdx.x; const int baseY = blockIdx.y * ER_LAYERS_BLOCKDIM_Y + threadIdx.y; const int baseZ = (blockIdx.z * ER_LAYERS_RESULT_STEPS - ER_LAYERS_HALO_STEPS) * ER_LAYERS_BLOCKDIM_Z + threadIdx.z; d_src += (baseZ * h + baseY) * w + baseX; d_dst += (baseZ * h + baseY) * w + baseX; const int pitch = w*h; //Main data #pragma unroll for (int i = ER_LAYERS_HALO_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z] = d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch]; } //Upper halo #pragma unroll for (int i = 0; i < ER_LAYERS_HALO_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z] = (baseZ + i * ER_LAYERS_BLOCKDIM_Z >= 0) ? d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch] : 0; } //Lower halo #pragma unroll for (int i = ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS + ER_LAYERS_HALO_STEPS; i++) { smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z]= (baseZ + i * ER_LAYERS_BLOCKDIM_Z < d) ? d_src[i * ER_LAYERS_BLOCKDIM_Z * pitch] : 0; } //Compute and store results __syncthreads(); #pragma unroll for (int i = ER_LAYERS_HALO_STEPS; i < ER_LAYERS_HALO_STEPS + ER_LAYERS_RESULT_STEPS; i++) { unsigned short *smem_kern = &smem_thread[threadIdx.z + i * ER_LAYERS_BLOCKDIM_Z - kernel_radius]; unsigned short val = smem_kern[0]; //#pragma unroll for (int j = 1; j <= 2*kernel_radius; j++) { val = min(val, smem_kern[j]); } d_dst[i * ER_LAYERS_BLOCKDIM_Z * pitch] = val; } }
.text .file "erosionLayers3DKernel.hip" .globl _Z36__device_stub__erosionLayers3DKernelPtS_iiii # -- Begin function _Z36__device_stub__erosionLayers3DKernelPtS_iiii .p2align 4, 0x90 .type _Z36__device_stub__erosionLayers3DKernelPtS_iiii,@function _Z36__device_stub__erosionLayers3DKernelPtS_iiii: # @_Z36__device_stub__erosionLayers3DKernelPtS_iiii .cfi_startproc # %bb.0: subq $136, %rsp .cfi_def_cfa_offset 144 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movl %edx, 12(%rsp) movl %ecx, 8(%rsp) movl %r8d, 4(%rsp) movl %r9d, (%rsp) leaq 72(%rsp), %rax movq %rax, 80(%rsp) leaq 64(%rsp), %rax movq %rax, 88(%rsp) leaq 12(%rsp), %rax movq %rax, 96(%rsp) leaq 8(%rsp), %rax movq %rax, 104(%rsp) leaq 4(%rsp), %rax movq %rax, 112(%rsp) movq %rsp, %rax movq %rax, 120(%rsp) leaq 48(%rsp), %rdi leaq 32(%rsp), %rsi leaq 24(%rsp), %rdx leaq 16(%rsp), %rcx callq __hipPopCallConfiguration movq 48(%rsp), %rsi movl 56(%rsp), %edx movq 32(%rsp), %rcx movl 40(%rsp), %r8d leaq 80(%rsp), %r9 movl $_Z21erosionLayers3DKernelPtS_iiii, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $152, %rsp .cfi_adjust_cfa_offset -152 retq .Lfunc_end0: .size _Z36__device_stub__erosionLayers3DKernelPtS_iiii, .Lfunc_end0-_Z36__device_stub__erosionLayers3DKernelPtS_iiii .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB1_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB1_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z21erosionLayers3DKernelPtS_iiii, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end1: .size __hip_module_ctor, .Lfunc_end1-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB2_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB2_2: retq .Lfunc_end2: .size __hip_module_dtor, .Lfunc_end2-__hip_module_dtor .cfi_endproc # -- End function .type _Z21erosionLayers3DKernelPtS_iiii,@object # @_Z21erosionLayers3DKernelPtS_iiii .section .rodata,"a",@progbits .globl _Z21erosionLayers3DKernelPtS_iiii .p2align 3, 0x0 _Z21erosionLayers3DKernelPtS_iiii: .quad _Z36__device_stub__erosionLayers3DKernelPtS_iiii .size _Z21erosionLayers3DKernelPtS_iiii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z21erosionLayers3DKernelPtS_iiii" .size .L__unnamed_1, 34 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z36__device_stub__erosionLayers3DKernelPtS_iiii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z21erosionLayers3DKernelPtS_iiii .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_00012e33_00000000-6_erosionLayers3DKernel.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2029: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2029: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z47__device_stub__Z21erosionLayers3DKernelPtS_iiiiPtS_iiii .type _Z47__device_stub__Z21erosionLayers3DKernelPtS_iiiiPtS_iiii, @function _Z47__device_stub__Z21erosionLayers3DKernelPtS_iiiiPtS_iiii: .LFB2051: .cfi_startproc endbr64 subq $168, %rsp .cfi_def_cfa_offset 176 movq %rdi, 24(%rsp) movq %rsi, 16(%rsp) movl %edx, 12(%rsp) movl %ecx, 8(%rsp) movl %r8d, 4(%rsp) movl %r9d, (%rsp) movq %fs:40, %rax movq %rax, 152(%rsp) xorl %eax, %eax leaq 24(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 12(%rsp), %rax movq %rax, 112(%rsp) leaq 8(%rsp), %rax movq %rax, 120(%rsp) leaq 4(%rsp), %rax movq %rax, 128(%rsp) movq %rsp, %rax movq %rax, 136(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 56(%rsp) movl $1, 60(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) leaq 40(%rsp), %rcx leaq 32(%rsp), %rdx leaq 60(%rsp), %rsi leaq 48(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 152(%rsp), %rax subq %fs:40, %rax jne .L8 addq $168, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 40(%rsp) .cfi_def_cfa_offset 184 pushq 40(%rsp) .cfi_def_cfa_offset 192 leaq 112(%rsp), %r9 movq 76(%rsp), %rcx movl 84(%rsp), %r8d movq 64(%rsp), %rsi movl 72(%rsp), %edx leaq _Z21erosionLayers3DKernelPtS_iiii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 176 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z47__device_stub__Z21erosionLayers3DKernelPtS_iiiiPtS_iiii, .-_Z47__device_stub__Z21erosionLayers3DKernelPtS_iiiiPtS_iiii .globl _Z21erosionLayers3DKernelPtS_iiii .type _Z21erosionLayers3DKernelPtS_iiii, @function _Z21erosionLayers3DKernelPtS_iiii: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z47__device_stub__Z21erosionLayers3DKernelPtS_iiiiPtS_iiii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z21erosionLayers3DKernelPtS_iiii, .-_Z21erosionLayers3DKernelPtS_iiii .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "_Z21erosionLayers3DKernelPtS_iiii" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2054: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC0(%rip), %rdx movq %rdx, %rcx leaq _Z21erosionLayers3DKernelPtS_iiii(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2054: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "erosionLayers3DKernel.hip" .globl _Z36__device_stub__erosionLayers3DKernelPtS_iiii # -- Begin function _Z36__device_stub__erosionLayers3DKernelPtS_iiii .p2align 4, 0x90 .type _Z36__device_stub__erosionLayers3DKernelPtS_iiii,@function _Z36__device_stub__erosionLayers3DKernelPtS_iiii: # @_Z36__device_stub__erosionLayers3DKernelPtS_iiii .cfi_startproc # %bb.0: subq $136, %rsp .cfi_def_cfa_offset 144 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movl %edx, 12(%rsp) movl %ecx, 8(%rsp) movl %r8d, 4(%rsp) movl %r9d, (%rsp) leaq 72(%rsp), %rax movq %rax, 80(%rsp) leaq 64(%rsp), %rax movq %rax, 88(%rsp) leaq 12(%rsp), %rax movq %rax, 96(%rsp) leaq 8(%rsp), %rax movq %rax, 104(%rsp) leaq 4(%rsp), %rax movq %rax, 112(%rsp) movq %rsp, %rax movq %rax, 120(%rsp) leaq 48(%rsp), %rdi leaq 32(%rsp), %rsi leaq 24(%rsp), %rdx leaq 16(%rsp), %rcx callq __hipPopCallConfiguration movq 48(%rsp), %rsi movl 56(%rsp), %edx movq 32(%rsp), %rcx movl 40(%rsp), %r8d leaq 80(%rsp), %r9 movl $_Z21erosionLayers3DKernelPtS_iiii, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $152, %rsp .cfi_adjust_cfa_offset -152 retq .Lfunc_end0: .size _Z36__device_stub__erosionLayers3DKernelPtS_iiii, .Lfunc_end0-_Z36__device_stub__erosionLayers3DKernelPtS_iiii .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB1_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB1_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z21erosionLayers3DKernelPtS_iiii, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end1: .size __hip_module_ctor, .Lfunc_end1-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB2_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB2_2: retq .Lfunc_end2: .size __hip_module_dtor, .Lfunc_end2-__hip_module_dtor .cfi_endproc # -- End function .type _Z21erosionLayers3DKernelPtS_iiii,@object # @_Z21erosionLayers3DKernelPtS_iiii .section .rodata,"a",@progbits .globl _Z21erosionLayers3DKernelPtS_iiii .p2align 3, 0x0 _Z21erosionLayers3DKernelPtS_iiii: .quad _Z36__device_stub__erosionLayers3DKernelPtS_iiii .size _Z21erosionLayers3DKernelPtS_iiii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z21erosionLayers3DKernelPtS_iiii" .size .L__unnamed_1, 34 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z36__device_stub__erosionLayers3DKernelPtS_iiii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z21erosionLayers3DKernelPtS_iiii .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
#include "stdio.h" __global__ void helloFromGPU(void) { printf("Hello World from thread %d\n", threadIdx.x); } int main() { printf("Hello World from CPU\n"); helloFromGPU<<<10, 10>>>(); cudaDeviceReset(); // cudaDeviceSynchronize(); return 0; }
code for sm_80 Function : _Z12helloFromGPUv .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ IMAD.MOV.U32 R1, RZ, RZ, c[0x0][0x28] ; /* 0x00000a00ff017624 */ /* 0x000fc800078e00ff */ /*0010*/ S2R R8, SR_TID.X ; /* 0x0000000000087919 */ /* 0x000e220000002100 */ /*0020*/ IADD3 R1, R1, -0x8, RZ ; /* 0xfffffff801017810 */ /* 0x000fe20007ffe0ff */ /*0030*/ IMAD.MOV.U32 R4, RZ, RZ, c[0x4][0x8] ; /* 0x01000200ff047624 */ /* 0x000fe200078e00ff */ /*0040*/ MOV R0, 0x0 ; /* 0x0000000000007802 */ /* 0x000fe20000000f00 */ /*0050*/ IMAD.MOV.U32 R5, RZ, RZ, c[0x4][0xc] ; /* 0x01000300ff057624 */ /* 0x000fe200078e00ff */ /*0060*/ IADD3 R6, P0, R1, c[0x0][0x20], RZ ; /* 0x0000080001067a10 */ /* 0x000fe40007f1e0ff */ /*0070*/ LDC.64 R2, c[0x4][R0] ; /* 0x0100000000027b82 */ /* 0x0002a60000000a00 */ /*0080*/ IMAD.X R7, RZ, RZ, c[0x0][0x24], P0 ; /* 0x00000900ff077624 */ /* 0x000fe200000e06ff */ /*0090*/ STL [R1], R8 ; /* 0x0000000801007387 */ /* 0x0013e80000100800 */ /*00a0*/ LEPC R8 ; /* 0x000000000008734e */ /* 0x002fc60000000000 */ /*00b0*/ MOV R11, 0x120 ; /* 0x00000120000b7802 */ /* 0x000fe40000000f00 */ /*00c0*/ MOV R20, 0xa0 ; /* 0x000000a000147802 */ /* 0x000fc40000000f00 */ /*00d0*/ MOV R21, 0x0 ; /* 0x0000000000157802 */ /* 0x000fe40000000f00 */ /*00e0*/ MOV R0, 0x0 ; /* 0x0000000000007802 */ /* 0x000fe40000000f00 */ /*00f0*/ IADD3 R20, P0, P1, -R20, R11, R8 ; /* 0x0000000b14147210 */ /* 0x000fc8000791e108 */ /*0100*/ IADD3.X R21, ~R0, R21, R9, P0, P1 ; /* 0x0000001500157210 */ /* 0x000fc800007e2509 */ /*0110*/ CALL.ABS.NOINC R2 ; /* 0x0000000002007343 */ /* 0x004fea0003c00000 */ /*0120*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0130*/ BRA 0x130; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0140*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0150*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0160*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0170*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0180*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0190*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include "stdio.h" __global__ void helloFromGPU(void) { printf("Hello World from thread %d\n", threadIdx.x); } int main() { printf("Hello World from CPU\n"); helloFromGPU<<<10, 10>>>(); cudaDeviceReset(); // cudaDeviceSynchronize(); return 0; }
.file "tmpxft_000586f4_00000000-6_hello.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2060: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2060: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z31__device_stub__Z12helloFromGPUvv .type _Z31__device_stub__Z12helloFromGPUvv, @function _Z31__device_stub__Z12helloFromGPUvv: .LFB2082: .cfi_startproc endbr64 subq $88, %rsp .cfi_def_cfa_offset 96 movq %fs:40, %rax movq %rax, 72(%rsp) xorl %eax, %eax movl $1, 16(%rsp) movl $1, 20(%rsp) movl $1, 24(%rsp) movl $1, 28(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) leaq 8(%rsp), %rcx movq %rsp, %rdx leaq 28(%rsp), %rsi leaq 16(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 72(%rsp), %rax subq %fs:40, %rax jne .L8 addq $88, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 8(%rsp) .cfi_def_cfa_offset 104 pushq 8(%rsp) .cfi_def_cfa_offset 112 leaq 80(%rsp), %r9 movq 44(%rsp), %rcx movl 52(%rsp), %r8d movq 32(%rsp), %rsi movl 40(%rsp), %edx leaq _Z12helloFromGPUv(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 96 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _Z31__device_stub__Z12helloFromGPUvv, .-_Z31__device_stub__Z12helloFromGPUvv .globl _Z12helloFromGPUv .type _Z12helloFromGPUv, @function _Z12helloFromGPUv: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z31__device_stub__Z12helloFromGPUvv addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _Z12helloFromGPUv, .-_Z12helloFromGPUv .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "Hello World from CPU\n" .text .globl main .type main, @function main: .LFB2057: .cfi_startproc endbr64 subq $40, %rsp .cfi_def_cfa_offset 48 leaq .LC0(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $10, 20(%rsp) movl $1, 24(%rsp) movl $10, 8(%rsp) movl $1, 12(%rsp) movl $0, %r9d movl $0, %r8d movq 20(%rsp), %rdx movl $1, %ecx movq 8(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L14 .L12: call cudaDeviceReset@PLT movl $0, %eax addq $40, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L14: .cfi_restore_state call _Z31__device_stub__Z12helloFromGPUvv jmp .L12 .cfi_endproc .LFE2057: .size main, .-main .section .rodata.str1.1 .LC1: .string "_Z12helloFromGPUv" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2085: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC1(%rip), %rdx movq %rdx, %rcx leaq _Z12helloFromGPUv(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2085: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
You are a helpful coding assistant specialized in converting CUDA source code to HIP/AMD source code.
#include "stdio.h" __global__ void helloFromGPU(void) { printf("Hello World from thread %d\n", threadIdx.x); } int main() { printf("Hello World from CPU\n"); helloFromGPU<<<10, 10>>>(); cudaDeviceReset(); // cudaDeviceSynchronize(); return 0; }
#include <hip/hip_runtime.h> #include "stdio.h" __global__ void helloFromGPU(void) { printf("Hello World from thread %d\n", threadIdx.x); } int main() { printf("Hello World from CPU\n"); helloFromGPU<<<10, 10>>>(); hipDeviceReset(); // cudaDeviceSynchronize(); return 0; }
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <hip/hip_runtime.h> #include "stdio.h" __global__ void helloFromGPU(void) { printf("Hello World from thread %d\n", threadIdx.x); } int main() { printf("Hello World from CPU\n"); helloFromGPU<<<10, 10>>>(); hipDeviceReset(); // cudaDeviceSynchronize(); return 0; }
.text .file "hello.hip" .globl _Z27__device_stub__helloFromGPUv # -- Begin function _Z27__device_stub__helloFromGPUv .p2align 4, 0x90 .type _Z27__device_stub__helloFromGPUv,@function _Z27__device_stub__helloFromGPUv: # @_Z27__device_stub__helloFromGPUv .cfi_startproc # %bb.0: subq $56, %rsp .cfi_def_cfa_offset 64 leaq 32(%rsp), %rdi leaq 16(%rsp), %rsi leaq 8(%rsp), %rdx movq %rsp, %rcx callq __hipPopCallConfiguration movq 32(%rsp), %rsi movl 40(%rsp), %edx movq 16(%rsp), %rcx movl 24(%rsp), %r8d leaq 48(%rsp), %r9 movl $_Z12helloFromGPUv, %edi pushq (%rsp) .cfi_adjust_cfa_offset 8 pushq 16(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $72, %rsp .cfi_adjust_cfa_offset -72 retq .Lfunc_end0: .size _Z27__device_stub__helloFromGPUv, .Lfunc_end0-_Z27__device_stub__helloFromGPUv .cfi_endproc # -- End function .globl main # -- Begin function main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: subq $56, %rsp .cfi_def_cfa_offset 64 movl $.Lstr, %edi callq puts@PLT movabsq $4294967306, %rdi # imm = 0x10000000A movl $1, %esi movq %rdi, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_2 # %bb.1: leaq 32(%rsp), %rdi leaq 16(%rsp), %rsi leaq 8(%rsp), %rdx movq %rsp, %rcx callq __hipPopCallConfiguration movq 32(%rsp), %rsi movl 40(%rsp), %edx movq 16(%rsp), %rcx movl 24(%rsp), %r8d leaq 48(%rsp), %r9 movl $_Z12helloFromGPUv, %edi pushq (%rsp) .cfi_adjust_cfa_offset 8 pushq 16(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_2: callq hipDeviceReset xorl %eax, %eax addq $56, %rsp .cfi_def_cfa_offset 8 retq .Lfunc_end1: .size main, .Lfunc_end1-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB2_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB2_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z12helloFromGPUv, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end2: .size __hip_module_ctor, .Lfunc_end2-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB3_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB3_2: retq .Lfunc_end3: .size __hip_module_dtor, .Lfunc_end3-__hip_module_dtor .cfi_endproc # -- End function .type _Z12helloFromGPUv,@object # @_Z12helloFromGPUv .section .rodata,"a",@progbits .globl _Z12helloFromGPUv .p2align 3, 0x0 _Z12helloFromGPUv: .quad _Z27__device_stub__helloFromGPUv .size _Z12helloFromGPUv, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z12helloFromGPUv" .size .L__unnamed_1, 18 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .type .Lstr,@object # @str .section .rodata.str1.1,"aMS",@progbits,1 .Lstr: .asciz "Hello World from CPU" .size .Lstr, 21 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z27__device_stub__helloFromGPUv .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z12helloFromGPUv .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting from CUDA host assembly to AMD host assembly.
.file "tmpxft_000586f4_00000000-6_hello.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2060: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2060: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z31__device_stub__Z12helloFromGPUvv .type _Z31__device_stub__Z12helloFromGPUvv, @function _Z31__device_stub__Z12helloFromGPUvv: .LFB2082: .cfi_startproc endbr64 subq $88, %rsp .cfi_def_cfa_offset 96 movq %fs:40, %rax movq %rax, 72(%rsp) xorl %eax, %eax movl $1, 16(%rsp) movl $1, 20(%rsp) movl $1, 24(%rsp) movl $1, 28(%rsp) movl $1, 32(%rsp) movl $1, 36(%rsp) leaq 8(%rsp), %rcx movq %rsp, %rdx leaq 28(%rsp), %rsi leaq 16(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 72(%rsp), %rax subq %fs:40, %rax jne .L8 addq $88, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 8(%rsp) .cfi_def_cfa_offset 104 pushq 8(%rsp) .cfi_def_cfa_offset 112 leaq 80(%rsp), %r9 movq 44(%rsp), %rcx movl 52(%rsp), %r8d movq 32(%rsp), %rsi movl 40(%rsp), %edx leaq _Z12helloFromGPUv(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 96 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _Z31__device_stub__Z12helloFromGPUvv, .-_Z31__device_stub__Z12helloFromGPUvv .globl _Z12helloFromGPUv .type _Z12helloFromGPUv, @function _Z12helloFromGPUv: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z31__device_stub__Z12helloFromGPUvv addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _Z12helloFromGPUv, .-_Z12helloFromGPUv .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "Hello World from CPU\n" .text .globl main .type main, @function main: .LFB2057: .cfi_startproc endbr64 subq $40, %rsp .cfi_def_cfa_offset 48 leaq .LC0(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $10, 20(%rsp) movl $1, 24(%rsp) movl $10, 8(%rsp) movl $1, 12(%rsp) movl $0, %r9d movl $0, %r8d movq 20(%rsp), %rdx movl $1, %ecx movq 8(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L14 .L12: call cudaDeviceReset@PLT movl $0, %eax addq $40, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L14: .cfi_restore_state call _Z31__device_stub__Z12helloFromGPUvv jmp .L12 .cfi_endproc .LFE2057: .size main, .-main .section .rodata.str1.1 .LC1: .string "_Z12helloFromGPUv" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2085: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC1(%rip), %rdx movq %rdx, %rcx leaq _Z12helloFromGPUv(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2085: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4:
.text .file "hello.hip" .globl _Z27__device_stub__helloFromGPUv # -- Begin function _Z27__device_stub__helloFromGPUv .p2align 4, 0x90 .type _Z27__device_stub__helloFromGPUv,@function _Z27__device_stub__helloFromGPUv: # @_Z27__device_stub__helloFromGPUv .cfi_startproc # %bb.0: subq $56, %rsp .cfi_def_cfa_offset 64 leaq 32(%rsp), %rdi leaq 16(%rsp), %rsi leaq 8(%rsp), %rdx movq %rsp, %rcx callq __hipPopCallConfiguration movq 32(%rsp), %rsi movl 40(%rsp), %edx movq 16(%rsp), %rcx movl 24(%rsp), %r8d leaq 48(%rsp), %r9 movl $_Z12helloFromGPUv, %edi pushq (%rsp) .cfi_adjust_cfa_offset 8 pushq 16(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $72, %rsp .cfi_adjust_cfa_offset -72 retq .Lfunc_end0: .size _Z27__device_stub__helloFromGPUv, .Lfunc_end0-_Z27__device_stub__helloFromGPUv .cfi_endproc # -- End function .globl main # -- Begin function main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: subq $56, %rsp .cfi_def_cfa_offset 64 movl $.Lstr, %edi callq puts@PLT movabsq $4294967306, %rdi # imm = 0x10000000A movl $1, %esi movq %rdi, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_2 # %bb.1: leaq 32(%rsp), %rdi leaq 16(%rsp), %rsi leaq 8(%rsp), %rdx movq %rsp, %rcx callq __hipPopCallConfiguration movq 32(%rsp), %rsi movl 40(%rsp), %edx movq 16(%rsp), %rcx movl 24(%rsp), %r8d leaq 48(%rsp), %r9 movl $_Z12helloFromGPUv, %edi pushq (%rsp) .cfi_adjust_cfa_offset 8 pushq 16(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_2: callq hipDeviceReset xorl %eax, %eax addq $56, %rsp .cfi_def_cfa_offset 8 retq .Lfunc_end1: .size main, .Lfunc_end1-main .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_ctor .type __hip_module_ctor,@function __hip_module_ctor: # @__hip_module_ctor .cfi_startproc # %bb.0: subq $40, %rsp .cfi_def_cfa_offset 48 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB2_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB2_2: movq __hip_gpubin_handle(%rip), %rdi xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z12helloFromGPUv, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $40, %rsp .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end2: .size __hip_module_ctor, .Lfunc_end2-__hip_module_ctor .cfi_endproc # -- End function .p2align 4, 0x90 # -- Begin function __hip_module_dtor .type __hip_module_dtor,@function __hip_module_dtor: # @__hip_module_dtor .cfi_startproc # %bb.0: movq __hip_gpubin_handle(%rip), %rdi testq %rdi, %rdi je .LBB3_2 # %bb.1: pushq %rax .cfi_def_cfa_offset 16 callq __hipUnregisterFatBinary movq $0, __hip_gpubin_handle(%rip) addq $8, %rsp .cfi_def_cfa_offset 8 .LBB3_2: retq .Lfunc_end3: .size __hip_module_dtor, .Lfunc_end3-__hip_module_dtor .cfi_endproc # -- End function .type _Z12helloFromGPUv,@object # @_Z12helloFromGPUv .section .rodata,"a",@progbits .globl _Z12helloFromGPUv .p2align 3, 0x0 _Z12helloFromGPUv: .quad _Z27__device_stub__helloFromGPUv .size _Z12helloFromGPUv, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z12helloFromGPUv" .size .L__unnamed_1, 18 .type __hip_fatbin_wrapper,@object # @__hip_fatbin_wrapper .section .hipFatBinSegment,"a",@progbits .p2align 3, 0x0 __hip_fatbin_wrapper: .long 1212764230 # 0x48495046 .long 1 # 0x1 .quad __hip_fatbin .quad 0 .size __hip_fatbin_wrapper, 24 .type __hip_gpubin_handle,@object # @__hip_gpubin_handle .local __hip_gpubin_handle .comm __hip_gpubin_handle,8,8 .section .init_array,"aw",@init_array .p2align 3, 0x0 .quad __hip_module_ctor .type __hip_cuid_,@object # @__hip_cuid_ .bss .globl __hip_cuid_ __hip_cuid_: .byte 0 # 0x0 .size __hip_cuid_, 1 .type .Lstr,@object # @str .section .rodata.str1.1,"aMS",@progbits,1 .Lstr: .asciz "Hello World from CPU" .size .Lstr, 21 .section ".linker-options","e",@llvm_linker_options .ident "AMD clang version 18.0.0git (https://github.com/RadeonOpenCompute/llvm-project roc-6.3.2 25012 e5bf7e55c91490b07c49d8960fa7983d864936c4)" .section ".note.GNU-stack","",@progbits .addrsig .addrsig_sym _Z27__device_stub__helloFromGPUv .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z12helloFromGPUv .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
#include "includes.h" // VectorAdd.cu #define N 10 // size of vectors #define B 1 // blocks in the grid #define T 10 // threads in a block __global__ void add (int *a,int *b, int *c) { int tid = blockIdx.x * blockDim.x + threadIdx.x; if(tid < N) { c[tid] = a[tid]+b[tid]; } }
code for sm_80 Function : _Z3addPiS_S_ .headerflags @"EF_CUDA_TEXMODE_UNIFIED EF_CUDA_64BIT_ADDRESS EF_CUDA_SM80 EF_CUDA_VIRTUAL_SM(EF_CUDA_SM80)" /*0000*/ MOV R1, c[0x0][0x28] ; /* 0x00000a0000017a02 */ /* 0x000fe40000000f00 */ /*0010*/ S2R R6, SR_CTAID.X ; /* 0x0000000000067919 */ /* 0x000e280000002500 */ /*0020*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e240000002100 */ /*0030*/ IMAD R6, R6, c[0x0][0x0], R3 ; /* 0x0000000006067a24 */ /* 0x001fca00078e0203 */ /*0040*/ ISETP.GT.AND P0, PT, R6, 0x9, PT ; /* 0x000000090600780c */ /* 0x000fda0003f04270 */ /*0050*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0060*/ HFMA2.MMA R7, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff077435 */ /* 0x000fe200000001ff */ /*0070*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fd20000000a00 */ /*0080*/ IMAD.WIDE R4, R6, R7, c[0x0][0x168] ; /* 0x00005a0006047625 */ /* 0x000fc800078e0207 */ /*0090*/ IMAD.WIDE R2, R6.reuse, R7.reuse, c[0x0][0x160] ; /* 0x0000580006027625 */ /* 0x0c0fe400078e0207 */ /*00a0*/ LDG.E R4, [R4.64] ; /* 0x0000000404047981 */ /* 0x000ea8000c1e1900 */ /*00b0*/ LDG.E R3, [R2.64] ; /* 0x0000000402037981 */ /* 0x000ea2000c1e1900 */ /*00c0*/ IMAD.WIDE R6, R6, R7, c[0x0][0x170] ; /* 0x00005c0006067625 */ /* 0x000fe200078e0207 */ /*00d0*/ IADD3 R9, R4, R3, RZ ; /* 0x0000000304097210 */ /* 0x004fca0007ffe0ff */ /*00e0*/ STG.E [R6.64], R9 ; /* 0x0000000906007986 */ /* 0x000fe2000c101904 */ /*00f0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0100*/ BRA 0x100; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0110*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0120*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0130*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0140*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0150*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0160*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0170*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0180*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0190*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*01f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include "includes.h" // VectorAdd.cu #define N 10 // size of vectors #define B 1 // blocks in the grid #define T 10 // threads in a block __global__ void add (int *a,int *b, int *c) { int tid = blockIdx.x * blockDim.x + threadIdx.x; if(tid < N) { c[tid] = a[tid]+b[tid]; } }
.file "tmpxft_001a9806_00000000-6_add.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2029: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaUnregisterFatBinary@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2029: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z26__device_stub__Z3addPiS_S_PiS_S_ .type _Z26__device_stub__Z3addPiS_S_PiS_S_, @function _Z26__device_stub__Z3addPiS_S_PiS_S_: .LFB2051: .cfi_startproc endbr64 subq $136, %rsp .cfi_def_cfa_offset 144 movq %rdi, 24(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movq %fs:40, %rax movq %rax, 120(%rsp) xorl %eax, %eax leaq 24(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) movl $1, 48(%rsp) movl $1, 52(%rsp) movl $1, 56(%rsp) movl $1, 60(%rsp) movl $1, 64(%rsp) movl $1, 68(%rsp) leaq 40(%rsp), %rcx leaq 32(%rsp), %rdx leaq 60(%rsp), %rsi leaq 48(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 120(%rsp), %rax subq %fs:40, %rax jne .L8 addq $136, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 40(%rsp) .cfi_def_cfa_offset 152 pushq 40(%rsp) .cfi_def_cfa_offset 160 leaq 112(%rsp), %r9 movq 76(%rsp), %rcx movl 84(%rsp), %r8d movq 64(%rsp), %rsi movl 72(%rsp), %edx leaq _Z3addPiS_S_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 144 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z26__device_stub__Z3addPiS_S_PiS_S_, .-_Z26__device_stub__Z3addPiS_S_PiS_S_ .globl _Z3addPiS_S_ .type _Z3addPiS_S_, @function _Z3addPiS_S_: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z26__device_stub__Z3addPiS_S_PiS_S_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z3addPiS_S_, .-_Z3addPiS_S_ .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "_Z3addPiS_S_" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2054: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rdi movq %rax, _ZL20__cudaFatCubinHandle(%rip) pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 pushq $0 .cfi_def_cfa_offset 40 pushq $0 .cfi_def_cfa_offset 48 movl $0, %r9d movl $-1, %r8d leaq .LC0(%rip), %rdx movq %rdx, %rcx leaq _Z3addPiS_S_(%rip), %rsi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2054: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .nvFatBinSegment,"aw" .align 8 .type _ZL15__fatDeviceText, @object .size _ZL15__fatDeviceText, 24 _ZL15__fatDeviceText: .long 1180844977 .long 1 .quad fatbinData .quad 0 .local _ZL20__cudaFatCubinHandle .comm _ZL20__cudaFatCubinHandle,8,8 .ident "GCC: (Ubuntu 13.3.0-6ubuntu2~24.04) 13.3.0" .section .note.GNU-stack,"",@progbits .section .note.gnu.property,"a" .align 8 .long 1f - 0f .long 4f - 1f .long 5 0: .string "GNU" 1: .align 8 .long 0xc0000002 .long 3f - 2f 2: .long 0x3 3: .align 8 4: