system
stringclasses
7 values
input
stringlengths
10
94.5k
output
stringlengths
14
55.5k
You are a helpful coding assistant specialized in converting CUDA source code to CUDA device assembly.
#include <cuda.h> #include <cuda_runtime_api.h> #include <stdio.h> __global__ void kernel( int *a, int dimx, int dimy ) { int ix = blockIdx.x*blockDim.x + threadIdx.x; int iy = blockIdx.y*blockDim.y + threadIdx.y; int idx = iy*dimx + ix; a[idx] = a[idx]+1; } int main() { int dimx = 16, dimy = 16; int num_bytes = dimx*dimy*sizeof(int); int *d_a=0, *h_a=0; // device and host pointers h_a = (int*)malloc(num_bytes); cudaMalloc( (void**)&d_a, num_bytes ); if( 0==h_a || 0==d_a ) { printf("couldn't allocate memory\n"); return 1; } cudaMemset( d_a, 0, num_bytes ); dim3 grid, block; block.x = 4; block.y = 4; grid.x = dimx / block.x; grid.y = dimy / block.y; kernel<<<grid, block>>>( d_a, dimx, dimy ); cudaMemcpy(h_a, d_a, num_bytes, cudaMemcpyDeviceToHost ); free( h_a ); cudaFree( d_a ); return 0; }
code for sm_80 Function : _Z6kernelPiii .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*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fc60000000a00 */ /*0030*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e280000002100 */ /*0040*/ S2R R2, SR_CTAID.Y ; /* 0x0000000000027919 */ /* 0x000e680000002600 */ /*0050*/ S2R R5, SR_TID.Y ; /* 0x0000000000057919 */ /* 0x000e620000002200 */ /*0060*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x001fe400078e0203 */ /*0070*/ IMAD R3, R2, c[0x0][0x4], R5 ; /* 0x0000010002037a24 */ /* 0x002fe200078e0205 */ /*0080*/ HFMA2.MMA R2, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff027435 */ /* 0x000fc600000001ff */ /*0090*/ IMAD R3, R3, c[0x0][0x168], R0 ; /* 0x00005a0003037a24 */ /* 0x000fce00078e0200 */ /*00a0*/ IMAD.WIDE R2, R3, R2, c[0x0][0x160] ; /* 0x0000580003027625 */ /* 0x000fca00078e0202 */ /*00b0*/ LDG.E R0, [R2.64] ; /* 0x0000000402007981 */ /* 0x000ea4000c1e1900 */ /*00c0*/ IADD3 R5, R0, 0x1, RZ ; /* 0x0000000100057810 */ /* 0x004fca0007ffe0ff */ /*00d0*/ STG.E [R2.64], R5 ; /* 0x0000000502007986 */ /* 0x000fe2000c101904 */ /*00e0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*00f0*/ BRA 0xf0; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0100*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*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 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include <cuda.h> #include <cuda_runtime_api.h> #include <stdio.h> __global__ void kernel( int *a, int dimx, int dimy ) { int ix = blockIdx.x*blockDim.x + threadIdx.x; int iy = blockIdx.y*blockDim.y + threadIdx.y; int idx = iy*dimx + ix; a[idx] = a[idx]+1; } int main() { int dimx = 16, dimy = 16; int num_bytes = dimx*dimy*sizeof(int); int *d_a=0, *h_a=0; // device and host pointers h_a = (int*)malloc(num_bytes); cudaMalloc( (void**)&d_a, num_bytes ); if( 0==h_a || 0==d_a ) { printf("couldn't allocate memory\n"); return 1; } cudaMemset( d_a, 0, num_bytes ); dim3 grid, block; block.x = 4; block.y = 4; grid.x = dimx / block.x; grid.y = dimy / block.y; kernel<<<grid, block>>>( d_a, dimx, dimy ); cudaMemcpy(h_a, d_a, num_bytes, cudaMemcpyDeviceToHost ); free( h_a ); cudaFree( d_a ); return 0; }
.file "tmpxft_001b5f56_00000000-6_example.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.str1.1,"aMS",@progbits,1 .LC0: .string "couldn't allocate memory\n" .text .globl main .type main, @function main: .LFB2057: .cfi_startproc endbr64 subq $24, %rsp .cfi_def_cfa_offset 32 movq %fs:40, %rax movq %rax, 8(%rsp) xorl %eax, %eax movq $0, (%rsp) movq %rsp, %rdi movl $1024, %esi call cudaMalloc@PLT leaq .LC0(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 8(%rsp), %rax subq %fs:40, %rax jne .L6 movl $1, %eax addq $24, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L6: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size main, .-main .globl _Z27__device_stub__Z6kernelPiiiPiii .type _Z27__device_stub__Z6kernelPiiiPiii, @function _Z27__device_stub__Z6kernelPiiiPiii: .LFB2082: .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 .L11 .L7: movq 104(%rsp), %rax subq %fs:40, %rax jne .L12 addq $120, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L11: .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 _Z6kernelPiii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 128 jmp .L7 .L12: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _Z27__device_stub__Z6kernelPiiiPiii, .-_Z27__device_stub__Z6kernelPiiiPiii .globl _Z6kernelPiii .type _Z6kernelPiii, @function _Z6kernelPiii: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z27__device_stub__Z6kernelPiiiPiii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _Z6kernelPiii, .-_Z6kernelPiii .section .rodata.str1.1 .LC1: .string "_Z6kernelPiii" .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 _Z6kernelPiii(%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 <cuda.h> #include <cuda_runtime_api.h> #include <stdio.h> __global__ void kernel( int *a, int dimx, int dimy ) { int ix = blockIdx.x*blockDim.x + threadIdx.x; int iy = blockIdx.y*blockDim.y + threadIdx.y; int idx = iy*dimx + ix; a[idx] = a[idx]+1; } int main() { int dimx = 16, dimy = 16; int num_bytes = dimx*dimy*sizeof(int); int *d_a=0, *h_a=0; // device and host pointers h_a = (int*)malloc(num_bytes); cudaMalloc( (void**)&d_a, num_bytes ); if( 0==h_a || 0==d_a ) { printf("couldn't allocate memory\n"); return 1; } cudaMemset( d_a, 0, num_bytes ); dim3 grid, block; block.x = 4; block.y = 4; grid.x = dimx / block.x; grid.y = dimy / block.y; kernel<<<grid, block>>>( d_a, dimx, dimy ); cudaMemcpy(h_a, d_a, num_bytes, cudaMemcpyDeviceToHost ); free( h_a ); cudaFree( d_a ); return 0; }
#include <hip/hip_runtime.h> #include <hip/hip_runtime_api.h> #include <stdio.h> __global__ void kernel( int *a, int dimx, int dimy ) { int ix = blockIdx.x*blockDim.x + threadIdx.x; int iy = blockIdx.y*blockDim.y + threadIdx.y; int idx = iy*dimx + ix; a[idx] = a[idx]+1; } int main() { int dimx = 16, dimy = 16; int num_bytes = dimx*dimy*sizeof(int); int *d_a=0, *h_a=0; // device and host pointers h_a = (int*)malloc(num_bytes); hipMalloc( (void**)&d_a, num_bytes ); if( 0==h_a || 0==d_a ) { printf("couldn't allocate memory\n"); return 1; } hipMemset( d_a, 0, num_bytes ); dim3 grid, block; block.x = 4; block.y = 4; grid.x = dimx / block.x; grid.y = dimy / block.y; kernel<<<grid, block>>>( d_a, dimx, dimy ); hipMemcpy(h_a, d_a, num_bytes, hipMemcpyDeviceToHost ); free( h_a ); hipFree( d_a ); return 0; }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include <hip/hip_runtime_api.h> #include <stdio.h> __global__ void kernel( int *a, int dimx, int dimy ) { int ix = blockIdx.x*blockDim.x + threadIdx.x; int iy = blockIdx.y*blockDim.y + threadIdx.y; int idx = iy*dimx + ix; a[idx] = a[idx]+1; } int main() { int dimx = 16, dimy = 16; int num_bytes = dimx*dimy*sizeof(int); int *d_a=0, *h_a=0; // device and host pointers h_a = (int*)malloc(num_bytes); hipMalloc( (void**)&d_a, num_bytes ); if( 0==h_a || 0==d_a ) { printf("couldn't allocate memory\n"); return 1; } hipMemset( d_a, 0, num_bytes ); dim3 grid, block; block.x = 4; block.y = 4; grid.x = dimx / block.x; grid.y = dimy / block.y; kernel<<<grid, block>>>( d_a, dimx, dimy ); hipMemcpy(h_a, d_a, num_bytes, hipMemcpyDeviceToHost ); free( h_a ); hipFree( d_a ); return 0; }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z6kernelPiii .globl _Z6kernelPiii .p2align 8 .type _Z6kernelPiii,@function _Z6kernelPiii: s_clause 0x1 s_load_b32 s2, s[0:1], 0x1c s_load_b32 s3, s[0:1], 0x8 v_bfe_u32 v1, v0, 10, 10 v_and_b32_e32 v0, 0x3ff, v0 s_load_b64 s[0:1], s[0:1], 0x0 s_waitcnt lgkmcnt(0) s_lshr_b32 s4, s2, 16 s_and_b32 s2, s2, 0xffff v_mad_u64_u32 v[2:3], null, s15, s4, v[1:2] s_mul_i32 s14, s14, s2 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_lo_u32 v1, v2, s3 v_add3_u32 v0, s14, v0, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v1, 31, v0 v_lshlrev_b64 v[0:1], 2, v[0:1] 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_load_b32 v2, v[0:1], off s_waitcnt vmcnt(0) v_add_nc_u32_e32 v2, 1, v2 global_store_b32 v[0:1], v2, off s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z6kernelPiii .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 272 .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 4 .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 _Z6kernelPiii, .Lfunc_end0-_Z6kernelPiii .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: _Z6kernelPiii .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z6kernelPiii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 4 .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 <hip/hip_runtime_api.h> #include <stdio.h> __global__ void kernel( int *a, int dimx, int dimy ) { int ix = blockIdx.x*blockDim.x + threadIdx.x; int iy = blockIdx.y*blockDim.y + threadIdx.y; int idx = iy*dimx + ix; a[idx] = a[idx]+1; } int main() { int dimx = 16, dimy = 16; int num_bytes = dimx*dimy*sizeof(int); int *d_a=0, *h_a=0; // device and host pointers h_a = (int*)malloc(num_bytes); hipMalloc( (void**)&d_a, num_bytes ); if( 0==h_a || 0==d_a ) { printf("couldn't allocate memory\n"); return 1; } hipMemset( d_a, 0, num_bytes ); dim3 grid, block; block.x = 4; block.y = 4; grid.x = dimx / block.x; grid.y = dimy / block.y; kernel<<<grid, block>>>( d_a, dimx, dimy ); hipMemcpy(h_a, d_a, num_bytes, hipMemcpyDeviceToHost ); free( h_a ); hipFree( d_a ); return 0; }
.text .file "example.hip" .globl _Z21__device_stub__kernelPiii # -- Begin function _Z21__device_stub__kernelPiii .p2align 4, 0x90 .type _Z21__device_stub__kernelPiii,@function _Z21__device_stub__kernelPiii: # @_Z21__device_stub__kernelPiii .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 $_Z6kernelPiii, %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__kernelPiii, .Lfunc_end0-_Z21__device_stub__kernelPiii .cfi_endproc # -- End function .globl main # -- Begin function main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rax .cfi_def_cfa_offset 16 movq $0, (%rsp) movq %rsp, %rdi movl $1024, %esi # imm = 0x400 callq hipMalloc movl $.Lstr, %edi callq puts@PLT movl $1, %eax popq %rcx .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 $_Z6kernelPiii, %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 _Z6kernelPiii,@object # @_Z6kernelPiii .section .rodata,"a",@progbits .globl _Z6kernelPiii .p2align 3, 0x0 _Z6kernelPiii: .quad _Z21__device_stub__kernelPiii .size _Z6kernelPiii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z6kernelPiii" .size .L__unnamed_1, 14 .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 "couldn't allocate memory" .size .Lstr, 25 .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__kernelPiii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z6kernelPiii .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 : _Z6kernelPiii .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*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fc60000000a00 */ /*0030*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e280000002100 */ /*0040*/ S2R R2, SR_CTAID.Y ; /* 0x0000000000027919 */ /* 0x000e680000002600 */ /*0050*/ S2R R5, SR_TID.Y ; /* 0x0000000000057919 */ /* 0x000e620000002200 */ /*0060*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x001fe400078e0203 */ /*0070*/ IMAD R3, R2, c[0x0][0x4], R5 ; /* 0x0000010002037a24 */ /* 0x002fe200078e0205 */ /*0080*/ HFMA2.MMA R2, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff027435 */ /* 0x000fc600000001ff */ /*0090*/ IMAD R3, R3, c[0x0][0x168], R0 ; /* 0x00005a0003037a24 */ /* 0x000fce00078e0200 */ /*00a0*/ IMAD.WIDE R2, R3, R2, c[0x0][0x160] ; /* 0x0000580003027625 */ /* 0x000fca00078e0202 */ /*00b0*/ LDG.E R0, [R2.64] ; /* 0x0000000402007981 */ /* 0x000ea4000c1e1900 */ /*00c0*/ IADD3 R5, R0, 0x1, RZ ; /* 0x0000000100057810 */ /* 0x004fca0007ffe0ff */ /*00d0*/ STG.E [R2.64], R5 ; /* 0x0000000502007986 */ /* 0x000fe2000c101904 */ /*00e0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*00f0*/ BRA 0xf0; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0100*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*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 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z6kernelPiii .globl _Z6kernelPiii .p2align 8 .type _Z6kernelPiii,@function _Z6kernelPiii: s_clause 0x1 s_load_b32 s2, s[0:1], 0x1c s_load_b32 s3, s[0:1], 0x8 v_bfe_u32 v1, v0, 10, 10 v_and_b32_e32 v0, 0x3ff, v0 s_load_b64 s[0:1], s[0:1], 0x0 s_waitcnt lgkmcnt(0) s_lshr_b32 s4, s2, 16 s_and_b32 s2, s2, 0xffff v_mad_u64_u32 v[2:3], null, s15, s4, v[1:2] s_mul_i32 s14, s14, s2 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_lo_u32 v1, v2, s3 v_add3_u32 v0, s14, v0, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v1, 31, v0 v_lshlrev_b64 v[0:1], 2, v[0:1] 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_load_b32 v2, v[0:1], off s_waitcnt vmcnt(0) v_add_nc_u32_e32 v2, 1, v2 global_store_b32 v[0:1], v2, off s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z6kernelPiii .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 272 .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 4 .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 _Z6kernelPiii, .Lfunc_end0-_Z6kernelPiii .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: _Z6kernelPiii .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z6kernelPiii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 4 .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_001b5f56_00000000-6_example.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.str1.1,"aMS",@progbits,1 .LC0: .string "couldn't allocate memory\n" .text .globl main .type main, @function main: .LFB2057: .cfi_startproc endbr64 subq $24, %rsp .cfi_def_cfa_offset 32 movq %fs:40, %rax movq %rax, 8(%rsp) xorl %eax, %eax movq $0, (%rsp) movq %rsp, %rdi movl $1024, %esi call cudaMalloc@PLT leaq .LC0(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 8(%rsp), %rax subq %fs:40, %rax jne .L6 movl $1, %eax addq $24, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L6: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size main, .-main .globl _Z27__device_stub__Z6kernelPiiiPiii .type _Z27__device_stub__Z6kernelPiiiPiii, @function _Z27__device_stub__Z6kernelPiiiPiii: .LFB2082: .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 .L11 .L7: movq 104(%rsp), %rax subq %fs:40, %rax jne .L12 addq $120, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L11: .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 _Z6kernelPiii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 128 jmp .L7 .L12: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _Z27__device_stub__Z6kernelPiiiPiii, .-_Z27__device_stub__Z6kernelPiiiPiii .globl _Z6kernelPiii .type _Z6kernelPiii, @function _Z6kernelPiii: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z27__device_stub__Z6kernelPiiiPiii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _Z6kernelPiii, .-_Z6kernelPiii .section .rodata.str1.1 .LC1: .string "_Z6kernelPiii" .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 _Z6kernelPiii(%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 "example.hip" .globl _Z21__device_stub__kernelPiii # -- Begin function _Z21__device_stub__kernelPiii .p2align 4, 0x90 .type _Z21__device_stub__kernelPiii,@function _Z21__device_stub__kernelPiii: # @_Z21__device_stub__kernelPiii .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 $_Z6kernelPiii, %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__kernelPiii, .Lfunc_end0-_Z21__device_stub__kernelPiii .cfi_endproc # -- End function .globl main # -- Begin function main .p2align 4, 0x90 .type main,@function main: # @main .cfi_startproc # %bb.0: pushq %rax .cfi_def_cfa_offset 16 movq $0, (%rsp) movq %rsp, %rdi movl $1024, %esi # imm = 0x400 callq hipMalloc movl $.Lstr, %edi callq puts@PLT movl $1, %eax popq %rcx .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 $_Z6kernelPiii, %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 _Z6kernelPiii,@object # @_Z6kernelPiii .section .rodata,"a",@progbits .globl _Z6kernelPiii .p2align 3, 0x0 _Z6kernelPiii: .quad _Z21__device_stub__kernelPiii .size _Z6kernelPiii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z6kernelPiii" .size .L__unnamed_1, 14 .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 "couldn't allocate memory" .size .Lstr, 25 .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__kernelPiii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z6kernelPiii .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.
extern "C" __global__ void setDomainPoolKernel( int nBatch, int rbs, int nDegree, int nD, int dScale, int rScale, int expansion, int isCenAlign, float regularize, float *data,float *dataRev, // array of data and reverse data // arrays pointer float *DA, float *AA, float *IA, // pointer of array of pointer to pointer of array in arrays, nevermind i just stun you. // p(i) = data(i + size(data)) float **DP, float **AP, float **IP ) { int taskIdx = blockIdx.x * blockDim.x + threadIdx.x; if (taskIdx < nBatch){ // initialize domain arrays // array structure // DA = rbs-rows , 1 + D_1^1 + D_2^1 + D_3^1 + ... + D_ds^2 // nCoeff is power of Domains. start from power 0 // nDegree determine the number of degree (maximum degree + 1) // nD determine the number of domain blocks // dScale determing the scale of domain size compare to rbs // dpOffset determine the number of bytes padded for each Array int nCoeff = ((nDegree - 1) * nD + 1); // pointing array of pointers to array const int daOffset = taskIdx * rbs * rScale * nCoeff; const int aaOffset = taskIdx * nCoeff * nCoeff; const int iaOffset = taskIdx * nCoeff * nCoeff; DP[taskIdx] = &DA[daOffset]; AP[taskIdx] = &AA[aaOffset]; IP[taskIdx] = &IA[iaOffset]; // initialize covariance matrix with regularization for(int i = 0; i < nCoeff * nCoeff; i++){ AA[aaOffset + i] = 0.0f; } for(int i = 0; i < nCoeff * nCoeff; i+= nCoeff+1){ // set diagonal to regularization parameter AA[aaOffset + i] = regularize * regularize; } // initialize first column covariance matrix for(int i = 0; i < rbs * rScale; i++){ DA[daOffset + i] = 1.0f; // power 0 } int dIdx = taskIdx % (nBatch/2); // for each block number dn for(int dn = 1; dn <= nD; dn++){ // set reference domain block // compute sumScale int sumScale = 0; for(int k = 1; k <= nD && k < dn; k++){ sumScale += (int) powf( (float) dScale, (float) (1 + expansion * (k - 1))) ; } int dnScale = (int) powf( (float) dScale, (float) (1 + expansion * (dn - 1))); //int dnIdx = dIdx + rbs * sumScale; // * domian location factor int dnIdx = dIdx; if( isCenAlign == 0 ){ // if left aligned dnIdx = dIdx + rbs * sumScale + 1; } else { // if center aligned dnIdx = dIdx + rbs/2 * (1 - dnScale) + 1; } int padDA = rbs * rScale * dn; // number of row of DA // initialize column dn-th index for(int i = 0; i < rbs * rScale; i++){ DA[daOffset + padDA + i] = 0.0f; // power 1 } // construct DA from domain blocks at power 1 // copy elements for(int i = 0; i < rbs * rScale; i++){ int datIdx = dnIdx + i*dnScale/rScale; if(datIdx >=0 && datIdx < (nBatch/2)){ if(taskIdx < (nBatch/2)){ DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + data[datIdx]; }else{ // gen reverse domain DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + dataRev[datIdx]; } } } // // handling if domain blocks are larger than rbs (by downsample) for(int ds = 1; ds < dnScale / rScale; ds++){ // vec sumation for(int i = 0; i < rbs * rScale; i++){ int datIdx = dnIdx + ds + i*dnScale/rScale; if(datIdx >=0 && datIdx < (nBatch/2)){ if(taskIdx < (nBatch/2)){ DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + data[datIdx]; }else{ // gen reverse domain DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + dataRev[datIdx]; } } } } // vec scalig after resample for(int i = 0; i < rbs * rScale; i++){ DA[daOffset + padDA + i] = DA[daOffset + padDA + i] / (dnScale/rScale); } // calculate next degree for(int deg = 2; deg <= nDegree - 1; deg++){ int degPad = rbs * nD * (deg - 2) + rbs * dn; int nextDegPad = rbs * nD * (deg - 1) + rbs * dn; for(int i = 0; i < rbs * rScale; i++){ // power n>=2 // D^n = D^1 * D^(n-1) DA[daOffset + nextDegPad + i] = DA[daOffset + rbs*dn + i] * DA[daOffset + degPad + i] ; } } } } }
.file "tmpxft_000c1457_00000000-6_setDomainPoolKernel.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 _Z65__device_stub__Z19setDomainPoolKerneliiiiiiiifPfS_S_S_S_PS_S0_S0_iiiiiiiifPfS_S_S_S_PS_S0_S0_ .type _Z65__device_stub__Z19setDomainPoolKerneliiiiiiiifPfS_S_S_S_PS_S0_S0_iiiiiiiifPfS_S_S_S_PS_S0_S0_, @function _Z65__device_stub__Z19setDomainPoolKerneliiiiiiiifPfS_S_S_S_PS_S0_S0_iiiiiiiifPfS_S_S_S_PS_S0_S0_: .LFB2051: .cfi_startproc endbr64 subq $312, %rsp .cfi_def_cfa_offset 320 movl %edi, 92(%rsp) movl %esi, 88(%rsp) movl %edx, 84(%rsp) movl %ecx, 80(%rsp) movl %r8d, 76(%rsp) movl %r9d, 72(%rsp) movss %xmm0, 68(%rsp) movq 336(%rsp), %rax movq %rax, 56(%rsp) movq 344(%rsp), %rax movq %rax, 48(%rsp) movq 352(%rsp), %rax movq %rax, 40(%rsp) movq 360(%rsp), %rax movq %rax, 32(%rsp) movq 368(%rsp), %rax movq %rax, 24(%rsp) movq 376(%rsp), %rax movq %rax, 16(%rsp) movq 384(%rsp), %rax movq %rax, 8(%rsp) movq 392(%rsp), %rax movq %rax, (%rsp) movq %fs:40, %rax movq %rax, 296(%rsp) xorl %eax, %eax leaq 92(%rsp), %rax movq %rax, 160(%rsp) leaq 88(%rsp), %rax movq %rax, 168(%rsp) leaq 84(%rsp), %rax movq %rax, 176(%rsp) leaq 80(%rsp), %rax movq %rax, 184(%rsp) leaq 76(%rsp), %rax movq %rax, 192(%rsp) leaq 72(%rsp), %rax movq %rax, 200(%rsp) leaq 320(%rsp), %rax movq %rax, 208(%rsp) leaq 328(%rsp), %rax movq %rax, 216(%rsp) leaq 68(%rsp), %rax movq %rax, 224(%rsp) leaq 56(%rsp), %rax movq %rax, 232(%rsp) leaq 48(%rsp), %rax movq %rax, 240(%rsp) leaq 40(%rsp), %rax movq %rax, 248(%rsp) leaq 32(%rsp), %rax movq %rax, 256(%rsp) leaq 24(%rsp), %rax movq %rax, 264(%rsp) leaq 16(%rsp), %rax movq %rax, 272(%rsp) leaq 8(%rsp), %rax movq %rax, 280(%rsp) movq %rsp, %rax movq %rax, 288(%rsp) movl $1, 112(%rsp) movl $1, 116(%rsp) movl $1, 120(%rsp) movl $1, 124(%rsp) movl $1, 128(%rsp) movl $1, 132(%rsp) leaq 104(%rsp), %rcx leaq 96(%rsp), %rdx leaq 124(%rsp), %rsi leaq 112(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 296(%rsp), %rax subq %fs:40, %rax jne .L8 addq $312, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 104(%rsp) .cfi_def_cfa_offset 328 pushq 104(%rsp) .cfi_def_cfa_offset 336 leaq 176(%rsp), %r9 movq 140(%rsp), %rcx movl 148(%rsp), %r8d movq 128(%rsp), %rsi movl 136(%rsp), %edx leaq setDomainPoolKernel(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 320 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z65__device_stub__Z19setDomainPoolKerneliiiiiiiifPfS_S_S_S_PS_S0_S0_iiiiiiiifPfS_S_S_S_PS_S0_S0_, .-_Z65__device_stub__Z19setDomainPoolKerneliiiiiiiifPfS_S_S_S_PS_S0_S0_iiiiiiiifPfS_S_S_S_PS_S0_S0_ .globl setDomainPoolKernel .type setDomainPoolKernel, @function setDomainPoolKernel: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 pushq 88(%rsp) .cfi_def_cfa_offset 24 pushq 88(%rsp) .cfi_def_cfa_offset 32 pushq 88(%rsp) .cfi_def_cfa_offset 40 pushq 88(%rsp) .cfi_def_cfa_offset 48 pushq 88(%rsp) .cfi_def_cfa_offset 56 pushq 88(%rsp) .cfi_def_cfa_offset 64 pushq 88(%rsp) .cfi_def_cfa_offset 72 pushq 88(%rsp) .cfi_def_cfa_offset 80 movl 88(%rsp), %eax pushq %rax .cfi_def_cfa_offset 88 movl 88(%rsp), %eax pushq %rax .cfi_def_cfa_offset 96 call _Z65__device_stub__Z19setDomainPoolKerneliiiiiiiifPfS_S_S_S_PS_S0_S0_iiiiiiiifPfS_S_S_S_PS_S0_S0_ addq $88, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size setDomainPoolKernel, .-setDomainPoolKernel .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "setDomainPoolKernel" .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 setDomainPoolKernel(%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.
extern "C" __global__ void setDomainPoolKernel( int nBatch, int rbs, int nDegree, int nD, int dScale, int rScale, int expansion, int isCenAlign, float regularize, float *data,float *dataRev, // array of data and reverse data // arrays pointer float *DA, float *AA, float *IA, // pointer of array of pointer to pointer of array in arrays, nevermind i just stun you. // p(i) = data(i + size(data)) float **DP, float **AP, float **IP ) { int taskIdx = blockIdx.x * blockDim.x + threadIdx.x; if (taskIdx < nBatch){ // initialize domain arrays // array structure // DA = rbs-rows , 1 + D_1^1 + D_2^1 + D_3^1 + ... + D_ds^2 // nCoeff is power of Domains. start from power 0 // nDegree determine the number of degree (maximum degree + 1) // nD determine the number of domain blocks // dScale determing the scale of domain size compare to rbs // dpOffset determine the number of bytes padded for each Array int nCoeff = ((nDegree - 1) * nD + 1); // pointing array of pointers to array const int daOffset = taskIdx * rbs * rScale * nCoeff; const int aaOffset = taskIdx * nCoeff * nCoeff; const int iaOffset = taskIdx * nCoeff * nCoeff; DP[taskIdx] = &DA[daOffset]; AP[taskIdx] = &AA[aaOffset]; IP[taskIdx] = &IA[iaOffset]; // initialize covariance matrix with regularization for(int i = 0; i < nCoeff * nCoeff; i++){ AA[aaOffset + i] = 0.0f; } for(int i = 0; i < nCoeff * nCoeff; i+= nCoeff+1){ // set diagonal to regularization parameter AA[aaOffset + i] = regularize * regularize; } // initialize first column covariance matrix for(int i = 0; i < rbs * rScale; i++){ DA[daOffset + i] = 1.0f; // power 0 } int dIdx = taskIdx % (nBatch/2); // for each block number dn for(int dn = 1; dn <= nD; dn++){ // set reference domain block // compute sumScale int sumScale = 0; for(int k = 1; k <= nD && k < dn; k++){ sumScale += (int) powf( (float) dScale, (float) (1 + expansion * (k - 1))) ; } int dnScale = (int) powf( (float) dScale, (float) (1 + expansion * (dn - 1))); //int dnIdx = dIdx + rbs * sumScale; // * domian location factor int dnIdx = dIdx; if( isCenAlign == 0 ){ // if left aligned dnIdx = dIdx + rbs * sumScale + 1; } else { // if center aligned dnIdx = dIdx + rbs/2 * (1 - dnScale) + 1; } int padDA = rbs * rScale * dn; // number of row of DA // initialize column dn-th index for(int i = 0; i < rbs * rScale; i++){ DA[daOffset + padDA + i] = 0.0f; // power 1 } // construct DA from domain blocks at power 1 // copy elements for(int i = 0; i < rbs * rScale; i++){ int datIdx = dnIdx + i*dnScale/rScale; if(datIdx >=0 && datIdx < (nBatch/2)){ if(taskIdx < (nBatch/2)){ DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + data[datIdx]; }else{ // gen reverse domain DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + dataRev[datIdx]; } } } // // handling if domain blocks are larger than rbs (by downsample) for(int ds = 1; ds < dnScale / rScale; ds++){ // vec sumation for(int i = 0; i < rbs * rScale; i++){ int datIdx = dnIdx + ds + i*dnScale/rScale; if(datIdx >=0 && datIdx < (nBatch/2)){ if(taskIdx < (nBatch/2)){ DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + data[datIdx]; }else{ // gen reverse domain DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + dataRev[datIdx]; } } } } // vec scalig after resample for(int i = 0; i < rbs * rScale; i++){ DA[daOffset + padDA + i] = DA[daOffset + padDA + i] / (dnScale/rScale); } // calculate next degree for(int deg = 2; deg <= nDegree - 1; deg++){ int degPad = rbs * nD * (deg - 2) + rbs * dn; int nextDegPad = rbs * nD * (deg - 1) + rbs * dn; for(int i = 0; i < rbs * rScale; i++){ // power n>=2 // D^n = D^1 * D^(n-1) DA[daOffset + nextDegPad + i] = DA[daOffset + rbs*dn + i] * DA[daOffset + degPad + i] ; } } } } }
#include <hip/hip_runtime.h> extern "C" __global__ void setDomainPoolKernel( int nBatch, int rbs, int nDegree, int nD, int dScale, int rScale, int expansion, int isCenAlign, float regularize, float *data,float *dataRev, // array of data and reverse data // arrays pointer float *DA, float *AA, float *IA, // pointer of array of pointer to pointer of array in arrays, nevermind i just stun you. // p(i) = data(i + size(data)) float **DP, float **AP, float **IP ) { int taskIdx = blockIdx.x * blockDim.x + threadIdx.x; if (taskIdx < nBatch){ // initialize domain arrays // array structure // DA = rbs-rows , 1 + D_1^1 + D_2^1 + D_3^1 + ... + D_ds^2 // nCoeff is power of Domains. start from power 0 // nDegree determine the number of degree (maximum degree + 1) // nD determine the number of domain blocks // dScale determing the scale of domain size compare to rbs // dpOffset determine the number of bytes padded for each Array int nCoeff = ((nDegree - 1) * nD + 1); // pointing array of pointers to array const int daOffset = taskIdx * rbs * rScale * nCoeff; const int aaOffset = taskIdx * nCoeff * nCoeff; const int iaOffset = taskIdx * nCoeff * nCoeff; DP[taskIdx] = &DA[daOffset]; AP[taskIdx] = &AA[aaOffset]; IP[taskIdx] = &IA[iaOffset]; // initialize covariance matrix with regularization for(int i = 0; i < nCoeff * nCoeff; i++){ AA[aaOffset + i] = 0.0f; } for(int i = 0; i < nCoeff * nCoeff; i+= nCoeff+1){ // set diagonal to regularization parameter AA[aaOffset + i] = regularize * regularize; } // initialize first column covariance matrix for(int i = 0; i < rbs * rScale; i++){ DA[daOffset + i] = 1.0f; // power 0 } int dIdx = taskIdx % (nBatch/2); // for each block number dn for(int dn = 1; dn <= nD; dn++){ // set reference domain block // compute sumScale int sumScale = 0; for(int k = 1; k <= nD && k < dn; k++){ sumScale += (int) powf( (float) dScale, (float) (1 + expansion * (k - 1))) ; } int dnScale = (int) powf( (float) dScale, (float) (1 + expansion * (dn - 1))); //int dnIdx = dIdx + rbs * sumScale; // * domian location factor int dnIdx = dIdx; if( isCenAlign == 0 ){ // if left aligned dnIdx = dIdx + rbs * sumScale + 1; } else { // if center aligned dnIdx = dIdx + rbs/2 * (1 - dnScale) + 1; } int padDA = rbs * rScale * dn; // number of row of DA // initialize column dn-th index for(int i = 0; i < rbs * rScale; i++){ DA[daOffset + padDA + i] = 0.0f; // power 1 } // construct DA from domain blocks at power 1 // copy elements for(int i = 0; i < rbs * rScale; i++){ int datIdx = dnIdx + i*dnScale/rScale; if(datIdx >=0 && datIdx < (nBatch/2)){ if(taskIdx < (nBatch/2)){ DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + data[datIdx]; }else{ // gen reverse domain DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + dataRev[datIdx]; } } } // // handling if domain blocks are larger than rbs (by downsample) for(int ds = 1; ds < dnScale / rScale; ds++){ // vec sumation for(int i = 0; i < rbs * rScale; i++){ int datIdx = dnIdx + ds + i*dnScale/rScale; if(datIdx >=0 && datIdx < (nBatch/2)){ if(taskIdx < (nBatch/2)){ DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + data[datIdx]; }else{ // gen reverse domain DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + dataRev[datIdx]; } } } } // vec scalig after resample for(int i = 0; i < rbs * rScale; i++){ DA[daOffset + padDA + i] = DA[daOffset + padDA + i] / (dnScale/rScale); } // calculate next degree for(int deg = 2; deg <= nDegree - 1; deg++){ int degPad = rbs * nD * (deg - 2) + rbs * dn; int nextDegPad = rbs * nD * (deg - 1) + rbs * dn; for(int i = 0; i < rbs * rScale; i++){ // power n>=2 // D^n = D^1 * D^(n-1) DA[daOffset + nextDegPad + i] = DA[daOffset + rbs*dn + i] * DA[daOffset + degPad + i] ; } } } } }
You are a helpful coding assistant specialized in converting HIP source code to AMD host assembly.
#include <hip/hip_runtime.h> extern "C" __global__ void setDomainPoolKernel( int nBatch, int rbs, int nDegree, int nD, int dScale, int rScale, int expansion, int isCenAlign, float regularize, float *data,float *dataRev, // array of data and reverse data // arrays pointer float *DA, float *AA, float *IA, // pointer of array of pointer to pointer of array in arrays, nevermind i just stun you. // p(i) = data(i + size(data)) float **DP, float **AP, float **IP ) { int taskIdx = blockIdx.x * blockDim.x + threadIdx.x; if (taskIdx < nBatch){ // initialize domain arrays // array structure // DA = rbs-rows , 1 + D_1^1 + D_2^1 + D_3^1 + ... + D_ds^2 // nCoeff is power of Domains. start from power 0 // nDegree determine the number of degree (maximum degree + 1) // nD determine the number of domain blocks // dScale determing the scale of domain size compare to rbs // dpOffset determine the number of bytes padded for each Array int nCoeff = ((nDegree - 1) * nD + 1); // pointing array of pointers to array const int daOffset = taskIdx * rbs * rScale * nCoeff; const int aaOffset = taskIdx * nCoeff * nCoeff; const int iaOffset = taskIdx * nCoeff * nCoeff; DP[taskIdx] = &DA[daOffset]; AP[taskIdx] = &AA[aaOffset]; IP[taskIdx] = &IA[iaOffset]; // initialize covariance matrix with regularization for(int i = 0; i < nCoeff * nCoeff; i++){ AA[aaOffset + i] = 0.0f; } for(int i = 0; i < nCoeff * nCoeff; i+= nCoeff+1){ // set diagonal to regularization parameter AA[aaOffset + i] = regularize * regularize; } // initialize first column covariance matrix for(int i = 0; i < rbs * rScale; i++){ DA[daOffset + i] = 1.0f; // power 0 } int dIdx = taskIdx % (nBatch/2); // for each block number dn for(int dn = 1; dn <= nD; dn++){ // set reference domain block // compute sumScale int sumScale = 0; for(int k = 1; k <= nD && k < dn; k++){ sumScale += (int) powf( (float) dScale, (float) (1 + expansion * (k - 1))) ; } int dnScale = (int) powf( (float) dScale, (float) (1 + expansion * (dn - 1))); //int dnIdx = dIdx + rbs * sumScale; // * domian location factor int dnIdx = dIdx; if( isCenAlign == 0 ){ // if left aligned dnIdx = dIdx + rbs * sumScale + 1; } else { // if center aligned dnIdx = dIdx + rbs/2 * (1 - dnScale) + 1; } int padDA = rbs * rScale * dn; // number of row of DA // initialize column dn-th index for(int i = 0; i < rbs * rScale; i++){ DA[daOffset + padDA + i] = 0.0f; // power 1 } // construct DA from domain blocks at power 1 // copy elements for(int i = 0; i < rbs * rScale; i++){ int datIdx = dnIdx + i*dnScale/rScale; if(datIdx >=0 && datIdx < (nBatch/2)){ if(taskIdx < (nBatch/2)){ DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + data[datIdx]; }else{ // gen reverse domain DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + dataRev[datIdx]; } } } // // handling if domain blocks are larger than rbs (by downsample) for(int ds = 1; ds < dnScale / rScale; ds++){ // vec sumation for(int i = 0; i < rbs * rScale; i++){ int datIdx = dnIdx + ds + i*dnScale/rScale; if(datIdx >=0 && datIdx < (nBatch/2)){ if(taskIdx < (nBatch/2)){ DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + data[datIdx]; }else{ // gen reverse domain DA[daOffset + padDA + i] = DA[daOffset + padDA + i] + dataRev[datIdx]; } } } } // vec scalig after resample for(int i = 0; i < rbs * rScale; i++){ DA[daOffset + padDA + i] = DA[daOffset + padDA + i] / (dnScale/rScale); } // calculate next degree for(int deg = 2; deg <= nDegree - 1; deg++){ int degPad = rbs * nD * (deg - 2) + rbs * dn; int nextDegPad = rbs * nD * (deg - 1) + rbs * dn; for(int i = 0; i < rbs * rScale; i++){ // power n>=2 // D^n = D^1 * D^(n-1) DA[daOffset + nextDegPad + i] = DA[daOffset + rbs*dn + i] * DA[daOffset + degPad + i] ; } } } } }
.text .file "setDomainPoolKernel.hip" .globl __device_stub__setDomainPoolKernel # -- Begin function __device_stub__setDomainPoolKernel .p2align 4, 0x90 .type __device_stub__setDomainPoolKernel,@function __device_stub__setDomainPoolKernel: # @__device_stub__setDomainPoolKernel .cfi_startproc # %bb.0: subq $216, %rsp .cfi_def_cfa_offset 224 movl %edi, 28(%rsp) movl %esi, 24(%rsp) movl %edx, 20(%rsp) movl %ecx, 16(%rsp) movl %r8d, 12(%rsp) movl %r9d, 8(%rsp) movss %xmm0, 4(%rsp) leaq 28(%rsp), %rax movq %rax, 80(%rsp) leaq 24(%rsp), %rax movq %rax, 88(%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 8(%rsp), %rax movq %rax, 120(%rsp) leaq 224(%rsp), %rax movq %rax, 128(%rsp) leaq 232(%rsp), %rax movq %rax, 136(%rsp) leaq 4(%rsp), %rax movq %rax, 144(%rsp) leaq 240(%rsp), %rax movq %rax, 152(%rsp) leaq 248(%rsp), %rax movq %rax, 160(%rsp) leaq 256(%rsp), %rax movq %rax, 168(%rsp) leaq 264(%rsp), %rax movq %rax, 176(%rsp) leaq 272(%rsp), %rax movq %rax, 184(%rsp) leaq 280(%rsp), %rax movq %rax, 192(%rsp) leaq 288(%rsp), %rax movq %rax, 200(%rsp) leaq 296(%rsp), %rax movq %rax, 208(%rsp) leaq 64(%rsp), %rdi leaq 48(%rsp), %rsi leaq 40(%rsp), %rdx leaq 32(%rsp), %rcx callq __hipPopCallConfiguration movq 64(%rsp), %rsi movl 72(%rsp), %edx movq 48(%rsp), %rcx movl 56(%rsp), %r8d leaq 80(%rsp), %r9 movl $setDomainPoolKernel, %edi pushq 32(%rsp) .cfi_adjust_cfa_offset 8 pushq 48(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $232, %rsp .cfi_adjust_cfa_offset -232 retq .Lfunc_end0: .size __device_stub__setDomainPoolKernel, .Lfunc_end0-__device_stub__setDomainPoolKernel .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 $setDomainPoolKernel, %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 setDomainPoolKernel,@object # @setDomainPoolKernel .section .rodata,"a",@progbits .globl setDomainPoolKernel .p2align 3, 0x0 setDomainPoolKernel: .quad __device_stub__setDomainPoolKernel .size setDomainPoolKernel, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "setDomainPoolKernel" .size .L__unnamed_1, 20 .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 __device_stub__setDomainPoolKernel .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym setDomainPoolKernel .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_000c1457_00000000-6_setDomainPoolKernel.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 _Z65__device_stub__Z19setDomainPoolKerneliiiiiiiifPfS_S_S_S_PS_S0_S0_iiiiiiiifPfS_S_S_S_PS_S0_S0_ .type _Z65__device_stub__Z19setDomainPoolKerneliiiiiiiifPfS_S_S_S_PS_S0_S0_iiiiiiiifPfS_S_S_S_PS_S0_S0_, @function _Z65__device_stub__Z19setDomainPoolKerneliiiiiiiifPfS_S_S_S_PS_S0_S0_iiiiiiiifPfS_S_S_S_PS_S0_S0_: .LFB2051: .cfi_startproc endbr64 subq $312, %rsp .cfi_def_cfa_offset 320 movl %edi, 92(%rsp) movl %esi, 88(%rsp) movl %edx, 84(%rsp) movl %ecx, 80(%rsp) movl %r8d, 76(%rsp) movl %r9d, 72(%rsp) movss %xmm0, 68(%rsp) movq 336(%rsp), %rax movq %rax, 56(%rsp) movq 344(%rsp), %rax movq %rax, 48(%rsp) movq 352(%rsp), %rax movq %rax, 40(%rsp) movq 360(%rsp), %rax movq %rax, 32(%rsp) movq 368(%rsp), %rax movq %rax, 24(%rsp) movq 376(%rsp), %rax movq %rax, 16(%rsp) movq 384(%rsp), %rax movq %rax, 8(%rsp) movq 392(%rsp), %rax movq %rax, (%rsp) movq %fs:40, %rax movq %rax, 296(%rsp) xorl %eax, %eax leaq 92(%rsp), %rax movq %rax, 160(%rsp) leaq 88(%rsp), %rax movq %rax, 168(%rsp) leaq 84(%rsp), %rax movq %rax, 176(%rsp) leaq 80(%rsp), %rax movq %rax, 184(%rsp) leaq 76(%rsp), %rax movq %rax, 192(%rsp) leaq 72(%rsp), %rax movq %rax, 200(%rsp) leaq 320(%rsp), %rax movq %rax, 208(%rsp) leaq 328(%rsp), %rax movq %rax, 216(%rsp) leaq 68(%rsp), %rax movq %rax, 224(%rsp) leaq 56(%rsp), %rax movq %rax, 232(%rsp) leaq 48(%rsp), %rax movq %rax, 240(%rsp) leaq 40(%rsp), %rax movq %rax, 248(%rsp) leaq 32(%rsp), %rax movq %rax, 256(%rsp) leaq 24(%rsp), %rax movq %rax, 264(%rsp) leaq 16(%rsp), %rax movq %rax, 272(%rsp) leaq 8(%rsp), %rax movq %rax, 280(%rsp) movq %rsp, %rax movq %rax, 288(%rsp) movl $1, 112(%rsp) movl $1, 116(%rsp) movl $1, 120(%rsp) movl $1, 124(%rsp) movl $1, 128(%rsp) movl $1, 132(%rsp) leaq 104(%rsp), %rcx leaq 96(%rsp), %rdx leaq 124(%rsp), %rsi leaq 112(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 296(%rsp), %rax subq %fs:40, %rax jne .L8 addq $312, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 104(%rsp) .cfi_def_cfa_offset 328 pushq 104(%rsp) .cfi_def_cfa_offset 336 leaq 176(%rsp), %r9 movq 140(%rsp), %rcx movl 148(%rsp), %r8d movq 128(%rsp), %rsi movl 136(%rsp), %edx leaq setDomainPoolKernel(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 320 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2051: .size _Z65__device_stub__Z19setDomainPoolKerneliiiiiiiifPfS_S_S_S_PS_S0_S0_iiiiiiiifPfS_S_S_S_PS_S0_S0_, .-_Z65__device_stub__Z19setDomainPoolKerneliiiiiiiifPfS_S_S_S_PS_S0_S0_iiiiiiiifPfS_S_S_S_PS_S0_S0_ .globl setDomainPoolKernel .type setDomainPoolKernel, @function setDomainPoolKernel: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 pushq 88(%rsp) .cfi_def_cfa_offset 24 pushq 88(%rsp) .cfi_def_cfa_offset 32 pushq 88(%rsp) .cfi_def_cfa_offset 40 pushq 88(%rsp) .cfi_def_cfa_offset 48 pushq 88(%rsp) .cfi_def_cfa_offset 56 pushq 88(%rsp) .cfi_def_cfa_offset 64 pushq 88(%rsp) .cfi_def_cfa_offset 72 pushq 88(%rsp) .cfi_def_cfa_offset 80 movl 88(%rsp), %eax pushq %rax .cfi_def_cfa_offset 88 movl 88(%rsp), %eax pushq %rax .cfi_def_cfa_offset 96 call _Z65__device_stub__Z19setDomainPoolKerneliiiiiiiifPfS_S_S_S_PS_S0_S0_iiiiiiiifPfS_S_S_S_PS_S0_S0_ addq $88, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size setDomainPoolKernel, .-setDomainPoolKernel .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "setDomainPoolKernel" .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 setDomainPoolKernel(%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 "setDomainPoolKernel.hip" .globl __device_stub__setDomainPoolKernel # -- Begin function __device_stub__setDomainPoolKernel .p2align 4, 0x90 .type __device_stub__setDomainPoolKernel,@function __device_stub__setDomainPoolKernel: # @__device_stub__setDomainPoolKernel .cfi_startproc # %bb.0: subq $216, %rsp .cfi_def_cfa_offset 224 movl %edi, 28(%rsp) movl %esi, 24(%rsp) movl %edx, 20(%rsp) movl %ecx, 16(%rsp) movl %r8d, 12(%rsp) movl %r9d, 8(%rsp) movss %xmm0, 4(%rsp) leaq 28(%rsp), %rax movq %rax, 80(%rsp) leaq 24(%rsp), %rax movq %rax, 88(%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 8(%rsp), %rax movq %rax, 120(%rsp) leaq 224(%rsp), %rax movq %rax, 128(%rsp) leaq 232(%rsp), %rax movq %rax, 136(%rsp) leaq 4(%rsp), %rax movq %rax, 144(%rsp) leaq 240(%rsp), %rax movq %rax, 152(%rsp) leaq 248(%rsp), %rax movq %rax, 160(%rsp) leaq 256(%rsp), %rax movq %rax, 168(%rsp) leaq 264(%rsp), %rax movq %rax, 176(%rsp) leaq 272(%rsp), %rax movq %rax, 184(%rsp) leaq 280(%rsp), %rax movq %rax, 192(%rsp) leaq 288(%rsp), %rax movq %rax, 200(%rsp) leaq 296(%rsp), %rax movq %rax, 208(%rsp) leaq 64(%rsp), %rdi leaq 48(%rsp), %rsi leaq 40(%rsp), %rdx leaq 32(%rsp), %rcx callq __hipPopCallConfiguration movq 64(%rsp), %rsi movl 72(%rsp), %edx movq 48(%rsp), %rcx movl 56(%rsp), %r8d leaq 80(%rsp), %r9 movl $setDomainPoolKernel, %edi pushq 32(%rsp) .cfi_adjust_cfa_offset 8 pushq 48(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $232, %rsp .cfi_adjust_cfa_offset -232 retq .Lfunc_end0: .size __device_stub__setDomainPoolKernel, .Lfunc_end0-__device_stub__setDomainPoolKernel .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 $setDomainPoolKernel, %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 setDomainPoolKernel,@object # @setDomainPoolKernel .section .rodata,"a",@progbits .globl setDomainPoolKernel .p2align 3, 0x0 setDomainPoolKernel: .quad __device_stub__setDomainPoolKernel .size setDomainPoolKernel, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "setDomainPoolKernel" .size .L__unnamed_1, 20 .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 __device_stub__setDomainPoolKernel .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym setDomainPoolKernel .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.
#define REORDER 1 #define GOOD_WEATHER 0 #define BAD_WEATHER 1 #define TAG_Car 0 #define TAG_Pedestrian 1 #include <stdio.h> #include <stdlib.h> #include <time.h> //#include <random> //#include <array> #include <algorithm> #define NUM_CARS 4096 #define NUM_PEDS 16384 #define NUM_STREETS 500 #define MAX_CONNECTIONS 10 #define MAX_LEN 25 using namespace std; // Define this to turn on error checking #define CUDA_ERROR_CHECK #define CudaSafeCall( err ) __cudaSafeCall( err, __FILE__, __LINE__ ) #define CudaCheckError() __cudaCheckError( __FILE__, __LINE__ ) inline void __cudaSafeCall( cudaError err, const char *file, const int line ) { #ifdef CUDA_ERROR_CHECK if ( cudaSuccess != err ) { fprintf( stderr, "cudaSafeCall() failed at %s:%i : %s\n", file, line, cudaGetErrorString( err ) ); exit( -1 ); } #endif return; } inline void __cudaCheckError( const char *file, const int line ) { #ifdef CUDA_ERROR_CHECK cudaError err = cudaGetLastError(); if ( cudaSuccess != err ) { fprintf( stderr, "cudaCheckError() failed at %s:%i : %s\n", file, line, cudaGetErrorString( err ) ); exit( -1 ); } // More careful checking. However, this will affect performance. // Comment away if needed. err = cudaDeviceSynchronize(); if( cudaSuccess != err ) { fprintf( stderr, "cudaCheckError() with sync failed at %s:%i : %s\n", file, line, cudaGetErrorString( err ) ); exit( -1 ); } #endif return; } __device__ float *d_Actor_progress; __device__ int *d_Actor_street; __device__ float *d_Car_max_velocity; __device__ float *d_Street_length; __device__ float *d_Street_max_velocity; __device__ int *d_Street_neighbors; __device__ int *d_Array_Street_size; __device__ int *d_Array_Street_offset; __device__ int *d_Array_Street_arrays; __device__ int *d_input_actor_tag; __device__ int *d_input_actor_id; __device__ int *d_jobs; __device__ int *d_randomn; __device__ void method_Car_move(int actor_id, int weather) { float weather_multiplier; if (weather == GOOD_WEATHER) { weather_multiplier = 1.0; } else if (weather == BAD_WEATHER) { weather_multiplier = 0.75; } float speed = min(d_Car_max_velocity[actor_id], d_Street_max_velocity[d_Actor_street[actor_id]]) * weather_multiplier; d_Actor_progress[actor_id] = d_Actor_progress[actor_id] + (speed / 60.0); /* 1 tick = 1 minute */ if (d_Actor_progress[actor_id] >= d_Street_length[d_Actor_street[actor_id]]) { // move to different street int array_id = d_Street_neighbors[d_Actor_street[actor_id]]; int neighbor_index = d_randomn[d_Actor_street[actor_id]] % d_Array_Street_size[array_id]; d_Actor_street[actor_id] = d_Array_Street_arrays[d_Array_Street_offset[array_id] + neighbor_index]; d_Actor_progress[actor_id] = 0.0f; } } __device__ void method_Pedestrian_move(int actor_id, int weather) { float speed = d_randomn[((int) (d_Actor_progress[actor_id]*d_Actor_progress[actor_id])) % NUM_STREETS] % 7 - 2; d_Actor_progress[actor_id] = d_Actor_progress[actor_id] + (speed / 60.0); if (d_Actor_progress[actor_id] >= d_Street_length[d_Actor_street[actor_id]]) { // move to different street int array_id = d_Street_neighbors[d_Actor_street[actor_id]]; int neighbor_index = d_randomn[d_Actor_street[actor_id]] % d_Array_Street_size[array_id]; d_Actor_street[actor_id] = d_Array_Street_arrays[d_Array_Street_offset[array_id] + neighbor_index]; d_Actor_progress[actor_id] = 0.0f; } } __device__ void block(int actor_tag, int actor_id, int weather, int ticks) { for (int i = 0; i < ticks; i++) { switch (actor_tag) { case TAG_Car: method_Car_move(actor_id, weather); break; case TAG_Pedestrian: method_Pedestrian_move(actor_id, weather); break; } } } __global__ void kernel(int weather, int ticks, float *v_d_Actor_progress, int *v_d_Actor_street, float *v_d_Car_max_velocity, float *v_d_Street_length, float *v_d_Street_max_velocity, int *v_d_Street_neighbors, int *v_d_Array_Street_size, int *v_d_Array_Street_offset, int *v_d_Array_Street_arrays, int *v_d_input_actor_tag, int *v_d_input_actor_id, int *v_d_jobs, int *v_d_randomn) { d_Actor_progress = v_d_Actor_progress; d_Actor_street = v_d_Actor_street; d_Car_max_velocity = v_d_Car_max_velocity; d_Street_length = v_d_Street_length; d_Street_max_velocity = v_d_Street_max_velocity; d_Street_neighbors = v_d_Street_neighbors; d_Array_Street_size = v_d_Array_Street_size; d_Array_Street_offset = v_d_Array_Street_offset; d_Array_Street_arrays = v_d_Array_Street_arrays; d_input_actor_tag = v_d_input_actor_tag; d_input_actor_id = v_d_input_actor_id; d_jobs = v_d_jobs; d_randomn = v_d_randomn; int tid = blockIdx.x * blockDim.x + threadIdx.x; __syncthreads(); #if (REORDER) block(d_input_actor_tag[d_jobs[tid]], d_input_actor_id[d_jobs[tid]], weather, ticks); #else block(d_input_actor_tag[tid], d_input_actor_id[tid], weather, ticks); #endif } int main() { printf("Setting up scenario...\n"); srand(42); // streets float *Street_length = new float[NUM_STREETS]; float *Street_max_velocity = new float[NUM_STREETS]; int *Street_neighbors = new int[NUM_STREETS]; for (int i = 0; i < NUM_STREETS; i++) { Street_length[i] = rand() % MAX_LEN + 1; Street_max_velocity[i] = rand() % 40 + 45; /* speed between 45 and 105 */ Street_neighbors[i] = i; } // neighbors int *Array_Street_offset = new int[NUM_STREETS]; int *Array_Street_size = new int[NUM_STREETS]; int num_connections = 0; for (int i = 0; i < NUM_STREETS; i++) { Array_Street_offset[i] = num_connections; int connections = rand() % MAX_CONNECTIONS + 1; Array_Street_size[i] = connections; num_connections += connections; } int *Array_Street_arrays = new int[num_connections]; for (int i = 0; i < num_connections; i++) { Array_Street_arrays[i] = rand() % NUM_STREETS; } // actors int *Actor_street = new int[NUM_PEDS + NUM_CARS]; float *Actor_progress = new float[NUM_PEDS + NUM_CARS]; float *Car_max_velocity = new float[NUM_CARS + NUM_PEDS]; int *Actor_tag = new int[NUM_PEDS + NUM_CARS]; int *Actor_id = new int[NUM_PEDS + NUM_CARS]; for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { Actor_street[i] = rand() % NUM_STREETS; Actor_progress[i] = rand() % 10; Car_max_velocity[i] = rand() % 20 + 65; } for (int i = 0; i < NUM_PEDS; i++) { Actor_tag[i] = TAG_Pedestrian; Actor_id[i] = i; } for (int i = NUM_PEDS; i < NUM_PEDS + NUM_CARS; i++) { Actor_tag[i] = TAG_Car; Actor_id[i] = i; } std::srand(42); #if !(REORDER) random_shuffle(Actor_tag, Actor_tag + NUM_CARS + NUM_PEDS); #endif // jobs (dummy) int *jobs = new int[NUM_PEDS + NUM_CARS]; for (int i = 0; i < NUM_CARS + NUM_PEDS; i++) { jobs[i] = i; } // random numbers int *randomn = new int[NUM_STREETS]; for (int i = 0; i < NUM_STREETS; i++) { randomn[i] = rand() % NUM_STREETS; } printf("Scenario set up.\n"); printf("Copying data to GPU...\n"); float *v_d_Actor_progress; int *v_d_Actor_street; float *v_d_Car_max_velocity; float *v_d_Street_length; float *v_d_Street_max_velocity; int *v_d_Street_neighbors; int *v_d_Array_Street_size; int *v_d_Array_Street_offset; int *v_d_Array_Street_arrays; int *v_d_input_actor_tag; int *v_d_input_actor_id; int *v_d_jobs; int *v_d_randomn; CudaSafeCall(cudaMalloc((void**) &v_d_Actor_progress, sizeof(float) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(cudaMalloc((void**) &v_d_Actor_street, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(cudaMalloc((void**) &v_d_Car_max_velocity, sizeof(float) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(cudaMalloc((void**) &v_d_Street_length, sizeof(float) * NUM_STREETS)); CudaSafeCall(cudaMalloc((void**) &v_d_Street_max_velocity, sizeof(float) * NUM_STREETS)); CudaSafeCall(cudaMalloc((void**) &v_d_Street_neighbors, sizeof(int) * NUM_STREETS)); CudaSafeCall(cudaMalloc((void**) &v_d_Array_Street_size, sizeof(int) * NUM_STREETS)); CudaSafeCall(cudaMalloc((void**) &v_d_Array_Street_offset, sizeof(int) * NUM_STREETS)); CudaSafeCall(cudaMalloc((void**) &v_d_Array_Street_arrays, sizeof(int) * num_connections)); CudaSafeCall(cudaMalloc((void**) &v_d_input_actor_tag, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(cudaMalloc((void**) &v_d_input_actor_id, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(cudaMalloc((void**) &v_d_jobs, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(cudaMalloc((void**) &v_d_randomn, sizeof(int) * NUM_STREETS)); CudaSafeCall(cudaMemcpy(v_d_Actor_progress, &Actor_progress[0], sizeof(float) * (NUM_PEDS + NUM_CARS), cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Actor_street, &Actor_street[0], sizeof(int) * (NUM_PEDS + NUM_CARS), cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Car_max_velocity, &Car_max_velocity[0], sizeof(float) * (NUM_PEDS + NUM_CARS), cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Street_length, &Street_length[0], sizeof(float) * NUM_STREETS, cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Street_max_velocity, &Street_max_velocity[0], sizeof(float) * NUM_STREETS, cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Street_neighbors, &Street_neighbors[0], sizeof(int) * NUM_STREETS, cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Array_Street_size, &Array_Street_size[0], sizeof(int) * NUM_STREETS, cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Array_Street_offset, &Array_Street_offset[0], sizeof(int) * NUM_STREETS, cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Array_Street_arrays, &Array_Street_arrays[0], sizeof(int) * num_connections, cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_input_actor_tag, &Actor_tag[0], sizeof(int) * (NUM_PEDS + NUM_CARS), cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_input_actor_id, &Actor_id[0], sizeof(int) * (NUM_PEDS + NUM_CARS), cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_jobs, &jobs[0], sizeof(int) * (NUM_PEDS + NUM_CARS), cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_randomn, &randomn[0], sizeof(int) * NUM_STREETS, cudaMemcpyHostToDevice)); printf("Finished copying data.\n"); cudaEvent_t start, stop; cudaEventCreate(&start); cudaEventCreate(&stop); printf("Launching kernel...\n"); cudaEventRecord(start); kernel<<<dim3(32), dim3((NUM_PEDS + NUM_CARS) / 32)>>>(GOOD_WEATHER, 1000000, v_d_Actor_progress, v_d_Actor_street, v_d_Car_max_velocity, v_d_Street_length, v_d_Street_max_velocity, v_d_Street_neighbors, v_d_Array_Street_size, v_d_Array_Street_offset, v_d_Array_Street_arrays, v_d_input_actor_tag, v_d_input_actor_id, v_d_jobs, v_d_randomn); CudaCheckError(); cudaEventRecord(stop); cudaEventSynchronize(stop); float milliseconds = 0; cudaEventElapsedTime(&milliseconds, start, stop); CudaCheckError(); printf("Kernel finished.\n"); cudaMemcpy(Actor_progress, v_d_Actor_progress, sizeof(float) * (NUM_PEDS + NUM_CARS), cudaMemcpyDeviceToHost); for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { // printf(" %f ", Actor_progress[i]); } cudaMemcpy(Actor_street, v_d_Actor_street, sizeof(int) * (NUM_PEDS + NUM_CARS), cudaMemcpyDeviceToHost); for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { // printf(" %i ", Actor_street[i]); } printf("\n\n\nElapsed time millis: %f\n", milliseconds); }
.file "tmpxft_0009e46a_00000000-6_benchmark_column_reorder.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2345: .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 .LFE2345: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .section .rodata._Z14__cudaSafeCall9cudaErrorPKci.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "cudaSafeCall() failed at %s:%i : %s\n" .section .text._Z14__cudaSafeCall9cudaErrorPKci,"axG",@progbits,_Z14__cudaSafeCall9cudaErrorPKci,comdat .weak _Z14__cudaSafeCall9cudaErrorPKci .type _Z14__cudaSafeCall9cudaErrorPKci, @function _Z14__cudaSafeCall9cudaErrorPKci: .LFB2337: .cfi_startproc endbr64 testl %edi, %edi jne .L8 ret .L8: 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 movq %rsi, %rbx movl %edx, %ebp call cudaGetErrorString@PLT movq %rax, %r9 movl %ebp, %r8d movq %rbx, %rcx leaq .LC0(%rip), %rdx movl $2, %esi movq stderr(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movl $-1, %edi call exit@PLT .cfi_endproc .LFE2337: .size _Z14__cudaSafeCall9cudaErrorPKci, .-_Z14__cudaSafeCall9cudaErrorPKci .section .rodata._Z16__cudaCheckErrorPKci.str1.8,"aMS",@progbits,1 .align 8 .LC1: .string "cudaCheckError() failed at %s:%i : %s\n" .align 8 .LC2: .string "cudaCheckError() with sync failed at %s:%i : %s\n" .section .text._Z16__cudaCheckErrorPKci,"axG",@progbits,_Z16__cudaCheckErrorPKci,comdat .weak _Z16__cudaCheckErrorPKci .type _Z16__cudaCheckErrorPKci, @function _Z16__cudaCheckErrorPKci: .LFB2338: .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 movq %rdi, %rbx movl %esi, %ebp call cudaGetLastError@PLT testl %eax, %eax jne .L13 call cudaDeviceSynchronize@PLT testl %eax, %eax jne .L14 addq $8, %rsp .cfi_remember_state .cfi_def_cfa_offset 24 popq %rbx .cfi_def_cfa_offset 16 popq %rbp .cfi_def_cfa_offset 8 ret .L13: .cfi_restore_state movl %eax, %edi call cudaGetErrorString@PLT movq %rax, %r9 movl %ebp, %r8d movq %rbx, %rcx leaq .LC1(%rip), %rdx movl $2, %esi movq stderr(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movl $-1, %edi call exit@PLT .L14: movl %eax, %edi call cudaGetErrorString@PLT movq %rax, %r9 movl %ebp, %r8d movq %rbx, %rcx leaq .LC2(%rip), %rdx movl $2, %esi movq stderr(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movl $-1, %edi call exit@PLT .cfi_endproc .LFE2338: .size _Z16__cudaCheckErrorPKci, .-_Z16__cudaCheckErrorPKci .text .globl _Z15method_Car_moveii .type _Z15method_Car_moveii, @function _Z15method_Car_moveii: .LFB2339: .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 .LFE2339: .size _Z15method_Car_moveii, .-_Z15method_Car_moveii .globl _Z22method_Pedestrian_moveii .type _Z22method_Pedestrian_moveii, @function _Z22method_Pedestrian_moveii: .LFB2340: .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 .LFE2340: .size _Z22method_Pedestrian_moveii, .-_Z22method_Pedestrian_moveii .globl _Z5blockiiii .type _Z5blockiiii, @function _Z5blockiiii: .LFB2341: .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 .LFE2341: .size _Z5blockiiii, .-_Z5blockiiii .globl _Z59__device_stub__Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_iiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .type _Z59__device_stub__Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_iiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_, @function _Z59__device_stub__Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_iiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_: .LFB2367: .cfi_startproc endbr64 subq $312, %rsp .cfi_def_cfa_offset 320 movl %edi, 108(%rsp) movl %esi, 104(%rsp) movq %rdx, 96(%rsp) movq %rcx, 88(%rsp) movq %r8, 80(%rsp) movq %r9, 72(%rsp) movq 320(%rsp), %rax movq %rax, 64(%rsp) movq 328(%rsp), %rax movq %rax, 56(%rsp) movq 336(%rsp), %rax movq %rax, 48(%rsp) movq 344(%rsp), %rax movq %rax, 40(%rsp) movq 352(%rsp), %rax movq %rax, 32(%rsp) movq 360(%rsp), %rax movq %rax, 24(%rsp) movq 368(%rsp), %rax movq %rax, 16(%rsp) movq 376(%rsp), %rax movq %rax, 8(%rsp) movq 384(%rsp), %rax movq %rax, (%rsp) movq %fs:40, %rax movq %rax, 296(%rsp) xorl %eax, %eax leaq 108(%rsp), %rax movq %rax, 176(%rsp) leaq 104(%rsp), %rax movq %rax, 184(%rsp) leaq 96(%rsp), %rax movq %rax, 192(%rsp) leaq 88(%rsp), %rax movq %rax, 200(%rsp) leaq 80(%rsp), %rax movq %rax, 208(%rsp) leaq 72(%rsp), %rax movq %rax, 216(%rsp) leaq 64(%rsp), %rax movq %rax, 224(%rsp) leaq 56(%rsp), %rax movq %rax, 232(%rsp) leaq 48(%rsp), %rax movq %rax, 240(%rsp) leaq 40(%rsp), %rax movq %rax, 248(%rsp) leaq 32(%rsp), %rax movq %rax, 256(%rsp) leaq 24(%rsp), %rax movq %rax, 264(%rsp) leaq 16(%rsp), %rax movq %rax, 272(%rsp) leaq 8(%rsp), %rax movq %rax, 280(%rsp) movq %rsp, %rax movq %rax, 288(%rsp) movl $1, 128(%rsp) movl $1, 132(%rsp) movl $1, 136(%rsp) movl $1, 140(%rsp) movl $1, 144(%rsp) movl $1, 148(%rsp) leaq 120(%rsp), %rcx leaq 112(%rsp), %rdx leaq 140(%rsp), %rsi leaq 128(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L25 .L21: movq 296(%rsp), %rax subq %fs:40, %rax jne .L26 addq $312, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L25: .cfi_restore_state pushq 120(%rsp) .cfi_def_cfa_offset 328 pushq 120(%rsp) .cfi_def_cfa_offset 336 leaq 192(%rsp), %r9 movq 156(%rsp), %rcx movl 164(%rsp), %r8d movq 144(%rsp), %rsi movl 152(%rsp), %edx leaq _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 320 jmp .L21 .L26: call __stack_chk_fail@PLT .cfi_endproc .LFE2367: .size _Z59__device_stub__Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_iiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_, .-_Z59__device_stub__Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_iiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .globl _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .type _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_, @function _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_: .LFB2368: .cfi_startproc endbr64 subq $16, %rsp .cfi_def_cfa_offset 24 pushq 88(%rsp) .cfi_def_cfa_offset 32 pushq 88(%rsp) .cfi_def_cfa_offset 40 pushq 88(%rsp) .cfi_def_cfa_offset 48 pushq 88(%rsp) .cfi_def_cfa_offset 56 pushq 88(%rsp) .cfi_def_cfa_offset 64 pushq 88(%rsp) .cfi_def_cfa_offset 72 pushq 88(%rsp) .cfi_def_cfa_offset 80 pushq 88(%rsp) .cfi_def_cfa_offset 88 pushq 88(%rsp) .cfi_def_cfa_offset 96 call _Z59__device_stub__Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_iiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ addq $88, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2368: .size _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_, .-_Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .section .rodata.str1.1,"aMS",@progbits,1 .LC3: .string "Setting up scenario...\n" .LC4: .string "Scenario set up.\n" .LC5: .string "Copying data to GPU...\n" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC6: .string "/home/ubuntu/Datasets/stackv2/train-structured/prg-titech/array2016-paper/master/benchmarks/benchmark_column_reorder.cu" .section .rodata.str1.1 .LC7: .string "Finished copying data.\n" .LC8: .string "Launching kernel...\n" .LC10: .string "Kernel finished.\n" .LC11: .string "\n\n\nElapsed time millis: %f\n" .text .globl main .type main, @function main: .LFB2342: .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 $248, %rsp .cfi_def_cfa_offset 304 movq %fs:40, %rax movq %rax, 232(%rsp) xorl %eax, %eax leaq .LC3(%rip), %rsi movl $2, %edi call __printf_chk@PLT movl $42, %edi call srand@PLT movl $2000, %edi call _Znam@PLT movq %rax, %rbp movq %rax, 8(%rsp) movl $2000, %edi call _Znam@PLT movq %rax, %r12 movq %rax, 16(%rsp) movl $2000, %edi call _Znam@PLT movq %rax, %r13 movq %rax, 24(%rsp) movl $0, %ebx .L30: call rand@PLT movslq %eax, %rdx imulq $1374389535, %rdx, %rdx sarq $35, %rdx movl %eax, %ecx sarl $31, %ecx subl %ecx, %edx leal (%rdx,%rdx,4), %edx leal (%rdx,%rdx,4), %edx subl %edx, %eax addl $1, %eax pxor %xmm0, %xmm0 cvtsi2ssl %eax, %xmm0 movss %xmm0, 0(%rbp,%rbx,4) call rand@PLT movslq %eax, %rdx imulq $1717986919, %rdx, %rdx sarq $36, %rdx movl %eax, %ecx sarl $31, %ecx subl %ecx, %edx leal (%rdx,%rdx,4), %edx sall $3, %edx subl %edx, %eax addl $45, %eax pxor %xmm0, %xmm0 cvtsi2ssl %eax, %xmm0 movss %xmm0, (%r12,%rbx,4) movl %ebx, 0(%r13,%rbx,4) addq $1, %rbx cmpq $500, %rbx jne .L30 movl $2000, %edi call _Znam@PLT movq %rax, %r12 movq %rax, 32(%rsp) movl $2000, %edi call _Znam@PLT movq %rax, %r13 movq %rax, 40(%rsp) movl $0, %ebx movl $0, %ebp .L31: movl %ebp, (%r12,%rbx) call rand@PLT movslq %eax, %rdx imulq $1717986919, %rdx, %rdx sarq $34, %rdx movl %eax, %ecx sarl $31, %ecx subl %ecx, %edx leal (%rdx,%rdx,4), %edx addl %edx, %edx subl %edx, %eax addl $1, %eax movl %eax, 0(%r13,%rbx) addl %eax, %ebp addq $4, %rbx cmpq $2000, %rbx jne .L31 movslq %ebp, %rax movabsq $2305843009213693950, %rdx cmpq %rax, %rdx jb .L32 leaq 0(,%rax,4), %r14 movq %r14, (%rsp) movq %r14, %rdi call _Znam@PLT movq %rax, 48(%rsp) movq %rax, %rbx leaq (%r14,%rax), %r12 testl %ebp, %ebp jle .L34 .L36: call rand@PLT movslq %eax, %rdx imulq $274877907, %rdx, %rdx sarq $37, %rdx movl %eax, %ecx sarl $31, %ecx subl %ecx, %edx imull $500, %edx, %edx subl %edx, %eax movl %eax, (%rbx) addq $4, %rbx cmpq %r12, %rbx jne .L36 .L34: movl $81920, %edi call _Znam@PLT movq %rax, %r15 movl $81920, %edi call _Znam@PLT movq %rax, %r13 movl $81920, %edi call _Znam@PLT movq %rax, %r14 movl $81920, %edi call _Znam@PLT movq %rax, %r12 movl $81920, %edi call _Znam@PLT movq %rax, %rbp movl $0, %ebx .L37: call rand@PLT movslq %eax, %rdx imulq $274877907, %rdx, %rdx sarq $37, %rdx movl %eax, %ecx sarl $31, %ecx subl %ecx, %edx imull $500, %edx, %edx subl %edx, %eax movl %eax, (%r15,%rbx) call rand@PLT movslq %eax, %rdx imulq $1717986919, %rdx, %rdx sarq $34, %rdx movl %eax, %ecx sarl $31, %ecx subl %ecx, %edx leal (%rdx,%rdx,4), %edx addl %edx, %edx subl %edx, %eax pxor %xmm0, %xmm0 cvtsi2ssl %eax, %xmm0 movss %xmm0, 0(%r13,%rbx) call rand@PLT movslq %eax, %rdx imulq $1717986919, %rdx, %rdx sarq $35, %rdx movl %eax, %ecx sarl $31, %ecx subl %ecx, %edx leal (%rdx,%rdx,4), %edx sall $2, %edx subl %edx, %eax addl $65, %eax pxor %xmm0, %xmm0 cvtsi2ssl %eax, %xmm0 movss %xmm0, (%r14,%rbx) addq $4, %rbx cmpq $81920, %rbx jne .L37 movl $0, %eax .L38: movl $1, (%r12,%rax,4) movl %eax, 0(%rbp,%rax,4) addq $1, %rax cmpq $16384, %rax jne .L38 .L39: movl $0, (%r12,%rax,4) movl %eax, 0(%rbp,%rax,4) addq $1, %rax cmpq $20480, %rax jne .L39 movl $42, %edi call srand@PLT movl $81920, %edi call _Znam@PLT movq %rax, %rdx movq %rax, 56(%rsp) movl $0, %eax .L40: movl %eax, (%rdx,%rax,4) addq $1, %rax cmpq $20480, %rax jne .L40 movl $2000, %edi call _Znam@PLT movq %rax, 64(%rsp) movq %rax, %rbx addq $2000, %rax movq %r15, 72(%rsp) movq %rax, %r15 .L41: call rand@PLT movslq %eax, %rdx imulq $274877907, %rdx, %rdx sarq $37, %rdx movl %eax, %ecx sarl $31, %ecx subl %ecx, %edx imull $500, %edx, %edx subl %edx, %eax movl %eax, (%rbx) addq $4, %rbx cmpq %rbx, %r15 jne .L41 movq 72(%rsp), %r15 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 leaq 88(%rsp), %rdi movl $81920, %esi call cudaMalloc@PLT movl %eax, %edi movl $269, %edx leaq .LC6(%rip), %rbx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq 96(%rsp), %rdi movl $81920, %esi call cudaMalloc@PLT movl %eax, %edi movl $270, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq 104(%rsp), %rdi movl $81920, %esi call cudaMalloc@PLT movl %eax, %edi movl $271, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq 112(%rsp), %rdi movl $2000, %esi call cudaMalloc@PLT movl %eax, %edi movl $272, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq 120(%rsp), %rdi movl $2000, %esi call cudaMalloc@PLT movl %eax, %edi movl $273, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq 128(%rsp), %rdi movl $2000, %esi call cudaMalloc@PLT movl %eax, %edi movl $274, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq 136(%rsp), %rdi movl $2000, %esi call cudaMalloc@PLT movl %eax, %edi movl $275, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq 144(%rsp), %rdi movl $2000, %esi call cudaMalloc@PLT movl %eax, %edi movl $276, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq 152(%rsp), %rdi movq (%rsp), %rsi call cudaMalloc@PLT movl %eax, %edi movl $277, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq 160(%rsp), %rdi movl $81920, %esi call cudaMalloc@PLT movl %eax, %edi movl $278, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq 168(%rsp), %rdi movl $81920, %esi call cudaMalloc@PLT movl %eax, %edi movl $279, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq 176(%rsp), %rdi movl $81920, %esi call cudaMalloc@PLT movl %eax, %edi movl $280, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq 184(%rsp), %rdi movl $2000, %esi call cudaMalloc@PLT movl %eax, %edi movl $281, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movl $81920, %edx movq %r13, %rsi movq 88(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $283, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movl $81920, %edx movq %r15, %rsi movq 96(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $284, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movl $81920, %edx movq %r14, %rsi movq 104(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $285, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movl $2000, %edx movq 8(%rsp), %rsi movq 112(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $286, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movl $2000, %edx movq 16(%rsp), %rsi movq 120(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $287, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movl $2000, %edx movq 24(%rsp), %rsi movq 128(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $288, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movl $2000, %edx movq 40(%rsp), %rsi movq 136(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $289, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movl $2000, %edx movq 32(%rsp), %rsi movq 144(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $290, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movq (%rsp), %rdx movq 48(%rsp), %rsi movq 152(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $291, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movl $81920, %edx movq %r12, %rsi movq 160(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $292, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movl $81920, %edx movq %rbp, %rsi movq 168(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $293, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movl $81920, %edx movq 56(%rsp), %rsi movq 176(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $294, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci movl $1, %ecx movl $2000, %edx movq 64(%rsp), %rsi movq 184(%rsp), %rdi call cudaMemcpy@PLT movl %eax, %edi movl $295, %edx movq %rbx, %rsi call _Z14__cudaSafeCall9cudaErrorPKci leaq .LC7(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT leaq 192(%rsp), %rdi call cudaEventCreate@PLT leaq 200(%rsp), %rdi call cudaEventCreate@PLT leaq .LC8(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $0, %esi movq 192(%rsp), %rdi call cudaEventRecord@PLT movl $640, 220(%rsp) movl $1, 224(%rsp) movl $1, 228(%rsp) movl $32, 208(%rsp) movl $1, 212(%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 je .L59 .L42: movl $309, %esi leaq .LC6(%rip), %rbx movq %rbx, %rdi call _Z16__cudaCheckErrorPKci movl $0, %esi movq 200(%rsp), %rdi call cudaEventRecord@PLT movq 200(%rsp), %rdi call cudaEventSynchronize@PLT movl $0x00000000, 220(%rsp) leaq 220(%rsp), %rdi movq 200(%rsp), %rdx movq 192(%rsp), %rsi call cudaEventElapsedTime@PLT movl $316, %esi movq %rbx, %rdi call _Z16__cudaCheckErrorPKci leaq .LC10(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $2, %ecx movl $81920, %edx movq 88(%rsp), %rsi movq %r13, %rdi call cudaMemcpy@PLT movl $20480, %eax .L43: subl $1, %eax jne .L43 movl $2, %ecx movl $81920, %edx movq 96(%rsp), %rsi movq %r15, %rdi call cudaMemcpy@PLT movl $20480, %eax .L44: subl $1, %eax jne .L44 pxor %xmm0, %xmm0 cvtss2sd 220(%rsp), %xmm0 leaq .LC11(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movq 232(%rsp), %rax subq %fs:40, %rax jne .L60 movl $0, %eax addq $248, %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 .L32: .cfi_restore_state movq 232(%rsp), %rax subq %fs:40, %rax je .L35 call __stack_chk_fail@PLT .L35: call __cxa_throw_bad_array_new_length@PLT .L59: subq $8, %rsp .cfi_def_cfa_offset 312 pushq 192(%rsp) .cfi_def_cfa_offset 320 pushq 192(%rsp) .cfi_def_cfa_offset 328 pushq 192(%rsp) .cfi_def_cfa_offset 336 pushq 192(%rsp) .cfi_def_cfa_offset 344 pushq 192(%rsp) .cfi_def_cfa_offset 352 pushq 192(%rsp) .cfi_def_cfa_offset 360 pushq 192(%rsp) .cfi_def_cfa_offset 368 pushq 192(%rsp) .cfi_def_cfa_offset 376 pushq 192(%rsp) .cfi_def_cfa_offset 384 movq 192(%rsp), %r9 movq 184(%rsp), %r8 movq 176(%rsp), %rcx movq 168(%rsp), %rdx movl $1000000, %esi movl $0, %edi call _Z59__device_stub__Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_iiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ addq $80, %rsp .cfi_def_cfa_offset 304 jmp .L42 .L60: call __stack_chk_fail@PLT .cfi_endproc .LFE2342: .size main, .-main .section .rodata.str1.8 .align 8 .LC12: .string "_Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_" .section .rodata.str1.1 .LC13: .string "d_Actor_progress" .LC14: .string "d_Actor_street" .LC15: .string "d_Car_max_velocity" .LC16: .string "d_Street_length" .LC17: .string "d_Street_max_velocity" .LC18: .string "d_Street_neighbors" .LC19: .string "d_Array_Street_size" .LC20: .string "d_Array_Street_offset" .LC21: .string "d_Array_Street_arrays" .LC22: .string "d_input_actor_tag" .LC23: .string "d_input_actor_id" .LC24: .string "d_jobs" .LC25: .string "d_randomn" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2370: .cfi_startproc endbr64 pushq %rbx .cfi_def_cfa_offset 16 .cfi_offset 3, -16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rbx 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 _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_(%rip), %rsi movq %rax, %rdi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC13(%rip), %rdx movq %rdx, %rcx leaq _ZL16d_Actor_progress(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC14(%rip), %rdx movq %rdx, %rcx leaq _ZL14d_Actor_street(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC15(%rip), %rdx movq %rdx, %rcx leaq _ZL18d_Car_max_velocity(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC16(%rip), %rdx movq %rdx, %rcx leaq _ZL15d_Street_length(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC17(%rip), %rdx movq %rdx, %rcx leaq _ZL21d_Street_max_velocity(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC18(%rip), %rdx movq %rdx, %rcx leaq _ZL18d_Street_neighbors(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC19(%rip), %rdx movq %rdx, %rcx leaq _ZL19d_Array_Street_size(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC20(%rip), %rdx movq %rdx, %rcx leaq _ZL21d_Array_Street_offset(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC21(%rip), %rdx movq %rdx, %rcx leaq _ZL21d_Array_Street_arrays(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC22(%rip), %rdx movq %rdx, %rcx leaq _ZL17d_input_actor_tag(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC23(%rip), %rdx movq %rdx, %rcx leaq _ZL16d_input_actor_id(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC24(%rip), %rdx movq %rdx, %rcx leaq _ZL6d_jobs(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 pushq $0 .cfi_def_cfa_offset 24 pushq $0 .cfi_def_cfa_offset 32 movl $8, %r9d movl $0, %r8d leaq .LC25(%rip), %rdx movq %rdx, %rcx leaq _ZL9d_randomn(%rip), %rsi movq %rbx, %rdi call __cudaRegisterVar@PLT addq $16, %rsp .cfi_def_cfa_offset 16 movq _ZL20__cudaFatCubinHandle(%rip), %rdi call __cudaRegisterFatBinaryEnd@PLT leaq _ZL26__cudaUnregisterBinaryUtilv(%rip), %rdi call atexit@PLT popq %rbx .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2370: .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 .local _ZL9d_randomn .comm _ZL9d_randomn,8,8 .local _ZL6d_jobs .comm _ZL6d_jobs,8,8 .local _ZL16d_input_actor_id .comm _ZL16d_input_actor_id,8,8 .local _ZL17d_input_actor_tag .comm _ZL17d_input_actor_tag,8,8 .local _ZL21d_Array_Street_arrays .comm _ZL21d_Array_Street_arrays,8,8 .local _ZL21d_Array_Street_offset .comm _ZL21d_Array_Street_offset,8,8 .local _ZL19d_Array_Street_size .comm _ZL19d_Array_Street_size,8,8 .local _ZL18d_Street_neighbors .comm _ZL18d_Street_neighbors,8,8 .local _ZL21d_Street_max_velocity .comm _ZL21d_Street_max_velocity,8,8 .local _ZL15d_Street_length .comm _ZL15d_Street_length,8,8 .local _ZL18d_Car_max_velocity .comm _ZL18d_Car_max_velocity,8,8 .local _ZL14d_Actor_street .comm _ZL14d_Actor_street,8,8 .local _ZL16d_Actor_progress .comm _ZL16d_Actor_progress,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.
#define REORDER 1 #define GOOD_WEATHER 0 #define BAD_WEATHER 1 #define TAG_Car 0 #define TAG_Pedestrian 1 #include <stdio.h> #include <stdlib.h> #include <time.h> //#include <random> //#include <array> #include <algorithm> #define NUM_CARS 4096 #define NUM_PEDS 16384 #define NUM_STREETS 500 #define MAX_CONNECTIONS 10 #define MAX_LEN 25 using namespace std; // Define this to turn on error checking #define CUDA_ERROR_CHECK #define CudaSafeCall( err ) __cudaSafeCall( err, __FILE__, __LINE__ ) #define CudaCheckError() __cudaCheckError( __FILE__, __LINE__ ) inline void __cudaSafeCall( cudaError err, const char *file, const int line ) { #ifdef CUDA_ERROR_CHECK if ( cudaSuccess != err ) { fprintf( stderr, "cudaSafeCall() failed at %s:%i : %s\n", file, line, cudaGetErrorString( err ) ); exit( -1 ); } #endif return; } inline void __cudaCheckError( const char *file, const int line ) { #ifdef CUDA_ERROR_CHECK cudaError err = cudaGetLastError(); if ( cudaSuccess != err ) { fprintf( stderr, "cudaCheckError() failed at %s:%i : %s\n", file, line, cudaGetErrorString( err ) ); exit( -1 ); } // More careful checking. However, this will affect performance. // Comment away if needed. err = cudaDeviceSynchronize(); if( cudaSuccess != err ) { fprintf( stderr, "cudaCheckError() with sync failed at %s:%i : %s\n", file, line, cudaGetErrorString( err ) ); exit( -1 ); } #endif return; } __device__ float *d_Actor_progress; __device__ int *d_Actor_street; __device__ float *d_Car_max_velocity; __device__ float *d_Street_length; __device__ float *d_Street_max_velocity; __device__ int *d_Street_neighbors; __device__ int *d_Array_Street_size; __device__ int *d_Array_Street_offset; __device__ int *d_Array_Street_arrays; __device__ int *d_input_actor_tag; __device__ int *d_input_actor_id; __device__ int *d_jobs; __device__ int *d_randomn; __device__ void method_Car_move(int actor_id, int weather) { float weather_multiplier; if (weather == GOOD_WEATHER) { weather_multiplier = 1.0; } else if (weather == BAD_WEATHER) { weather_multiplier = 0.75; } float speed = min(d_Car_max_velocity[actor_id], d_Street_max_velocity[d_Actor_street[actor_id]]) * weather_multiplier; d_Actor_progress[actor_id] = d_Actor_progress[actor_id] + (speed / 60.0); /* 1 tick = 1 minute */ if (d_Actor_progress[actor_id] >= d_Street_length[d_Actor_street[actor_id]]) { // move to different street int array_id = d_Street_neighbors[d_Actor_street[actor_id]]; int neighbor_index = d_randomn[d_Actor_street[actor_id]] % d_Array_Street_size[array_id]; d_Actor_street[actor_id] = d_Array_Street_arrays[d_Array_Street_offset[array_id] + neighbor_index]; d_Actor_progress[actor_id] = 0.0f; } } __device__ void method_Pedestrian_move(int actor_id, int weather) { float speed = d_randomn[((int) (d_Actor_progress[actor_id]*d_Actor_progress[actor_id])) % NUM_STREETS] % 7 - 2; d_Actor_progress[actor_id] = d_Actor_progress[actor_id] + (speed / 60.0); if (d_Actor_progress[actor_id] >= d_Street_length[d_Actor_street[actor_id]]) { // move to different street int array_id = d_Street_neighbors[d_Actor_street[actor_id]]; int neighbor_index = d_randomn[d_Actor_street[actor_id]] % d_Array_Street_size[array_id]; d_Actor_street[actor_id] = d_Array_Street_arrays[d_Array_Street_offset[array_id] + neighbor_index]; d_Actor_progress[actor_id] = 0.0f; } } __device__ void block(int actor_tag, int actor_id, int weather, int ticks) { for (int i = 0; i < ticks; i++) { switch (actor_tag) { case TAG_Car: method_Car_move(actor_id, weather); break; case TAG_Pedestrian: method_Pedestrian_move(actor_id, weather); break; } } } __global__ void kernel(int weather, int ticks, float *v_d_Actor_progress, int *v_d_Actor_street, float *v_d_Car_max_velocity, float *v_d_Street_length, float *v_d_Street_max_velocity, int *v_d_Street_neighbors, int *v_d_Array_Street_size, int *v_d_Array_Street_offset, int *v_d_Array_Street_arrays, int *v_d_input_actor_tag, int *v_d_input_actor_id, int *v_d_jobs, int *v_d_randomn) { d_Actor_progress = v_d_Actor_progress; d_Actor_street = v_d_Actor_street; d_Car_max_velocity = v_d_Car_max_velocity; d_Street_length = v_d_Street_length; d_Street_max_velocity = v_d_Street_max_velocity; d_Street_neighbors = v_d_Street_neighbors; d_Array_Street_size = v_d_Array_Street_size; d_Array_Street_offset = v_d_Array_Street_offset; d_Array_Street_arrays = v_d_Array_Street_arrays; d_input_actor_tag = v_d_input_actor_tag; d_input_actor_id = v_d_input_actor_id; d_jobs = v_d_jobs; d_randomn = v_d_randomn; int tid = blockIdx.x * blockDim.x + threadIdx.x; __syncthreads(); #if (REORDER) block(d_input_actor_tag[d_jobs[tid]], d_input_actor_id[d_jobs[tid]], weather, ticks); #else block(d_input_actor_tag[tid], d_input_actor_id[tid], weather, ticks); #endif } int main() { printf("Setting up scenario...\n"); srand(42); // streets float *Street_length = new float[NUM_STREETS]; float *Street_max_velocity = new float[NUM_STREETS]; int *Street_neighbors = new int[NUM_STREETS]; for (int i = 0; i < NUM_STREETS; i++) { Street_length[i] = rand() % MAX_LEN + 1; Street_max_velocity[i] = rand() % 40 + 45; /* speed between 45 and 105 */ Street_neighbors[i] = i; } // neighbors int *Array_Street_offset = new int[NUM_STREETS]; int *Array_Street_size = new int[NUM_STREETS]; int num_connections = 0; for (int i = 0; i < NUM_STREETS; i++) { Array_Street_offset[i] = num_connections; int connections = rand() % MAX_CONNECTIONS + 1; Array_Street_size[i] = connections; num_connections += connections; } int *Array_Street_arrays = new int[num_connections]; for (int i = 0; i < num_connections; i++) { Array_Street_arrays[i] = rand() % NUM_STREETS; } // actors int *Actor_street = new int[NUM_PEDS + NUM_CARS]; float *Actor_progress = new float[NUM_PEDS + NUM_CARS]; float *Car_max_velocity = new float[NUM_CARS + NUM_PEDS]; int *Actor_tag = new int[NUM_PEDS + NUM_CARS]; int *Actor_id = new int[NUM_PEDS + NUM_CARS]; for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { Actor_street[i] = rand() % NUM_STREETS; Actor_progress[i] = rand() % 10; Car_max_velocity[i] = rand() % 20 + 65; } for (int i = 0; i < NUM_PEDS; i++) { Actor_tag[i] = TAG_Pedestrian; Actor_id[i] = i; } for (int i = NUM_PEDS; i < NUM_PEDS + NUM_CARS; i++) { Actor_tag[i] = TAG_Car; Actor_id[i] = i; } std::srand(42); #if !(REORDER) random_shuffle(Actor_tag, Actor_tag + NUM_CARS + NUM_PEDS); #endif // jobs (dummy) int *jobs = new int[NUM_PEDS + NUM_CARS]; for (int i = 0; i < NUM_CARS + NUM_PEDS; i++) { jobs[i] = i; } // random numbers int *randomn = new int[NUM_STREETS]; for (int i = 0; i < NUM_STREETS; i++) { randomn[i] = rand() % NUM_STREETS; } printf("Scenario set up.\n"); printf("Copying data to GPU...\n"); float *v_d_Actor_progress; int *v_d_Actor_street; float *v_d_Car_max_velocity; float *v_d_Street_length; float *v_d_Street_max_velocity; int *v_d_Street_neighbors; int *v_d_Array_Street_size; int *v_d_Array_Street_offset; int *v_d_Array_Street_arrays; int *v_d_input_actor_tag; int *v_d_input_actor_id; int *v_d_jobs; int *v_d_randomn; CudaSafeCall(cudaMalloc((void**) &v_d_Actor_progress, sizeof(float) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(cudaMalloc((void**) &v_d_Actor_street, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(cudaMalloc((void**) &v_d_Car_max_velocity, sizeof(float) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(cudaMalloc((void**) &v_d_Street_length, sizeof(float) * NUM_STREETS)); CudaSafeCall(cudaMalloc((void**) &v_d_Street_max_velocity, sizeof(float) * NUM_STREETS)); CudaSafeCall(cudaMalloc((void**) &v_d_Street_neighbors, sizeof(int) * NUM_STREETS)); CudaSafeCall(cudaMalloc((void**) &v_d_Array_Street_size, sizeof(int) * NUM_STREETS)); CudaSafeCall(cudaMalloc((void**) &v_d_Array_Street_offset, sizeof(int) * NUM_STREETS)); CudaSafeCall(cudaMalloc((void**) &v_d_Array_Street_arrays, sizeof(int) * num_connections)); CudaSafeCall(cudaMalloc((void**) &v_d_input_actor_tag, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(cudaMalloc((void**) &v_d_input_actor_id, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(cudaMalloc((void**) &v_d_jobs, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(cudaMalloc((void**) &v_d_randomn, sizeof(int) * NUM_STREETS)); CudaSafeCall(cudaMemcpy(v_d_Actor_progress, &Actor_progress[0], sizeof(float) * (NUM_PEDS + NUM_CARS), cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Actor_street, &Actor_street[0], sizeof(int) * (NUM_PEDS + NUM_CARS), cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Car_max_velocity, &Car_max_velocity[0], sizeof(float) * (NUM_PEDS + NUM_CARS), cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Street_length, &Street_length[0], sizeof(float) * NUM_STREETS, cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Street_max_velocity, &Street_max_velocity[0], sizeof(float) * NUM_STREETS, cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Street_neighbors, &Street_neighbors[0], sizeof(int) * NUM_STREETS, cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Array_Street_size, &Array_Street_size[0], sizeof(int) * NUM_STREETS, cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Array_Street_offset, &Array_Street_offset[0], sizeof(int) * NUM_STREETS, cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_Array_Street_arrays, &Array_Street_arrays[0], sizeof(int) * num_connections, cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_input_actor_tag, &Actor_tag[0], sizeof(int) * (NUM_PEDS + NUM_CARS), cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_input_actor_id, &Actor_id[0], sizeof(int) * (NUM_PEDS + NUM_CARS), cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_jobs, &jobs[0], sizeof(int) * (NUM_PEDS + NUM_CARS), cudaMemcpyHostToDevice)); CudaSafeCall(cudaMemcpy(v_d_randomn, &randomn[0], sizeof(int) * NUM_STREETS, cudaMemcpyHostToDevice)); printf("Finished copying data.\n"); cudaEvent_t start, stop; cudaEventCreate(&start); cudaEventCreate(&stop); printf("Launching kernel...\n"); cudaEventRecord(start); kernel<<<dim3(32), dim3((NUM_PEDS + NUM_CARS) / 32)>>>(GOOD_WEATHER, 1000000, v_d_Actor_progress, v_d_Actor_street, v_d_Car_max_velocity, v_d_Street_length, v_d_Street_max_velocity, v_d_Street_neighbors, v_d_Array_Street_size, v_d_Array_Street_offset, v_d_Array_Street_arrays, v_d_input_actor_tag, v_d_input_actor_id, v_d_jobs, v_d_randomn); CudaCheckError(); cudaEventRecord(stop); cudaEventSynchronize(stop); float milliseconds = 0; cudaEventElapsedTime(&milliseconds, start, stop); CudaCheckError(); printf("Kernel finished.\n"); cudaMemcpy(Actor_progress, v_d_Actor_progress, sizeof(float) * (NUM_PEDS + NUM_CARS), cudaMemcpyDeviceToHost); for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { // printf(" %f ", Actor_progress[i]); } cudaMemcpy(Actor_street, v_d_Actor_street, sizeof(int) * (NUM_PEDS + NUM_CARS), cudaMemcpyDeviceToHost); for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { // printf(" %i ", Actor_street[i]); } printf("\n\n\nElapsed time millis: %f\n", milliseconds); }
#define REORDER 1 #define GOOD_WEATHER 0 #define BAD_WEATHER 1 #define TAG_Car 0 #define TAG_Pedestrian 1 #include <hip/hip_runtime.h> #include <stdio.h> #include <stdlib.h> #include <time.h> //#include <random> //#include <array> #include <algorithm> #define NUM_CARS 4096 #define NUM_PEDS 16384 #define NUM_STREETS 500 #define MAX_CONNECTIONS 10 #define MAX_LEN 25 using namespace std; // Define this to turn on error checking #define CUDA_ERROR_CHECK #define CudaSafeCall( err ) __cudaSafeCall( err, __FILE__, __LINE__ ) #define CudaCheckError() __cudaCheckError( __FILE__, __LINE__ ) inline void __cudaSafeCall( hipError_t err, const char *file, const int line ) { #ifdef CUDA_ERROR_CHECK if ( hipSuccess != err ) { fprintf( stderr, "cudaSafeCall() failed at %s:%i : %s\n", file, line, hipGetErrorString( err ) ); exit( -1 ); } #endif return; } inline void __cudaCheckError( const char *file, const int line ) { #ifdef CUDA_ERROR_CHECK hipError_t err = hipGetLastError(); if ( hipSuccess != err ) { fprintf( stderr, "cudaCheckError() failed at %s:%i : %s\n", file, line, hipGetErrorString( err ) ); exit( -1 ); } // More careful checking. However, this will affect performance. // Comment away if needed. err = hipDeviceSynchronize(); if( hipSuccess != err ) { fprintf( stderr, "cudaCheckError() with sync failed at %s:%i : %s\n", file, line, hipGetErrorString( err ) ); exit( -1 ); } #endif return; } __device__ float *d_Actor_progress; __device__ int *d_Actor_street; __device__ float *d_Car_max_velocity; __device__ float *d_Street_length; __device__ float *d_Street_max_velocity; __device__ int *d_Street_neighbors; __device__ int *d_Array_Street_size; __device__ int *d_Array_Street_offset; __device__ int *d_Array_Street_arrays; __device__ int *d_input_actor_tag; __device__ int *d_input_actor_id; __device__ int *d_jobs; __device__ int *d_randomn; __device__ void method_Car_move(int actor_id, int weather) { float weather_multiplier; if (weather == GOOD_WEATHER) { weather_multiplier = 1.0; } else if (weather == BAD_WEATHER) { weather_multiplier = 0.75; } float speed = min(d_Car_max_velocity[actor_id], d_Street_max_velocity[d_Actor_street[actor_id]]) * weather_multiplier; d_Actor_progress[actor_id] = d_Actor_progress[actor_id] + (speed / 60.0); /* 1 tick = 1 minute */ if (d_Actor_progress[actor_id] >= d_Street_length[d_Actor_street[actor_id]]) { // move to different street int array_id = d_Street_neighbors[d_Actor_street[actor_id]]; int neighbor_index = d_randomn[d_Actor_street[actor_id]] % d_Array_Street_size[array_id]; d_Actor_street[actor_id] = d_Array_Street_arrays[d_Array_Street_offset[array_id] + neighbor_index]; d_Actor_progress[actor_id] = 0.0f; } } __device__ void method_Pedestrian_move(int actor_id, int weather) { float speed = d_randomn[((int) (d_Actor_progress[actor_id]*d_Actor_progress[actor_id])) % NUM_STREETS] % 7 - 2; d_Actor_progress[actor_id] = d_Actor_progress[actor_id] + (speed / 60.0); if (d_Actor_progress[actor_id] >= d_Street_length[d_Actor_street[actor_id]]) { // move to different street int array_id = d_Street_neighbors[d_Actor_street[actor_id]]; int neighbor_index = d_randomn[d_Actor_street[actor_id]] % d_Array_Street_size[array_id]; d_Actor_street[actor_id] = d_Array_Street_arrays[d_Array_Street_offset[array_id] + neighbor_index]; d_Actor_progress[actor_id] = 0.0f; } } __device__ void block(int actor_tag, int actor_id, int weather, int ticks) { for (int i = 0; i < ticks; i++) { switch (actor_tag) { case TAG_Car: method_Car_move(actor_id, weather); break; case TAG_Pedestrian: method_Pedestrian_move(actor_id, weather); break; } } } __global__ void kernel(int weather, int ticks, float *v_d_Actor_progress, int *v_d_Actor_street, float *v_d_Car_max_velocity, float *v_d_Street_length, float *v_d_Street_max_velocity, int *v_d_Street_neighbors, int *v_d_Array_Street_size, int *v_d_Array_Street_offset, int *v_d_Array_Street_arrays, int *v_d_input_actor_tag, int *v_d_input_actor_id, int *v_d_jobs, int *v_d_randomn) { d_Actor_progress = v_d_Actor_progress; d_Actor_street = v_d_Actor_street; d_Car_max_velocity = v_d_Car_max_velocity; d_Street_length = v_d_Street_length; d_Street_max_velocity = v_d_Street_max_velocity; d_Street_neighbors = v_d_Street_neighbors; d_Array_Street_size = v_d_Array_Street_size; d_Array_Street_offset = v_d_Array_Street_offset; d_Array_Street_arrays = v_d_Array_Street_arrays; d_input_actor_tag = v_d_input_actor_tag; d_input_actor_id = v_d_input_actor_id; d_jobs = v_d_jobs; d_randomn = v_d_randomn; int tid = blockIdx.x * blockDim.x + threadIdx.x; __syncthreads(); #if (REORDER) block(d_input_actor_tag[d_jobs[tid]], d_input_actor_id[d_jobs[tid]], weather, ticks); #else block(d_input_actor_tag[tid], d_input_actor_id[tid], weather, ticks); #endif } int main() { printf("Setting up scenario...\n"); srand(42); // streets float *Street_length = new float[NUM_STREETS]; float *Street_max_velocity = new float[NUM_STREETS]; int *Street_neighbors = new int[NUM_STREETS]; for (int i = 0; i < NUM_STREETS; i++) { Street_length[i] = rand() % MAX_LEN + 1; Street_max_velocity[i] = rand() % 40 + 45; /* speed between 45 and 105 */ Street_neighbors[i] = i; } // neighbors int *Array_Street_offset = new int[NUM_STREETS]; int *Array_Street_size = new int[NUM_STREETS]; int num_connections = 0; for (int i = 0; i < NUM_STREETS; i++) { Array_Street_offset[i] = num_connections; int connections = rand() % MAX_CONNECTIONS + 1; Array_Street_size[i] = connections; num_connections += connections; } int *Array_Street_arrays = new int[num_connections]; for (int i = 0; i < num_connections; i++) { Array_Street_arrays[i] = rand() % NUM_STREETS; } // actors int *Actor_street = new int[NUM_PEDS + NUM_CARS]; float *Actor_progress = new float[NUM_PEDS + NUM_CARS]; float *Car_max_velocity = new float[NUM_CARS + NUM_PEDS]; int *Actor_tag = new int[NUM_PEDS + NUM_CARS]; int *Actor_id = new int[NUM_PEDS + NUM_CARS]; for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { Actor_street[i] = rand() % NUM_STREETS; Actor_progress[i] = rand() % 10; Car_max_velocity[i] = rand() % 20 + 65; } for (int i = 0; i < NUM_PEDS; i++) { Actor_tag[i] = TAG_Pedestrian; Actor_id[i] = i; } for (int i = NUM_PEDS; i < NUM_PEDS + NUM_CARS; i++) { Actor_tag[i] = TAG_Car; Actor_id[i] = i; } std::srand(42); #if !(REORDER) random_shuffle(Actor_tag, Actor_tag + NUM_CARS + NUM_PEDS); #endif // jobs (dummy) int *jobs = new int[NUM_PEDS + NUM_CARS]; for (int i = 0; i < NUM_CARS + NUM_PEDS; i++) { jobs[i] = i; } // random numbers int *randomn = new int[NUM_STREETS]; for (int i = 0; i < NUM_STREETS; i++) { randomn[i] = rand() % NUM_STREETS; } printf("Scenario set up.\n"); printf("Copying data to GPU...\n"); float *v_d_Actor_progress; int *v_d_Actor_street; float *v_d_Car_max_velocity; float *v_d_Street_length; float *v_d_Street_max_velocity; int *v_d_Street_neighbors; int *v_d_Array_Street_size; int *v_d_Array_Street_offset; int *v_d_Array_Street_arrays; int *v_d_input_actor_tag; int *v_d_input_actor_id; int *v_d_jobs; int *v_d_randomn; CudaSafeCall(hipMalloc((void**) &v_d_Actor_progress, sizeof(float) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_Actor_street, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_Car_max_velocity, sizeof(float) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_Street_length, sizeof(float) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Street_max_velocity, sizeof(float) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Street_neighbors, sizeof(int) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Array_Street_size, sizeof(int) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Array_Street_offset, sizeof(int) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Array_Street_arrays, sizeof(int) * num_connections)); CudaSafeCall(hipMalloc((void**) &v_d_input_actor_tag, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_input_actor_id, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_jobs, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_randomn, sizeof(int) * NUM_STREETS)); CudaSafeCall(hipMemcpy(v_d_Actor_progress, &Actor_progress[0], sizeof(float) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Actor_street, &Actor_street[0], sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Car_max_velocity, &Car_max_velocity[0], sizeof(float) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Street_length, &Street_length[0], sizeof(float) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Street_max_velocity, &Street_max_velocity[0], sizeof(float) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Street_neighbors, &Street_neighbors[0], sizeof(int) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Array_Street_size, &Array_Street_size[0], sizeof(int) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Array_Street_offset, &Array_Street_offset[0], sizeof(int) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Array_Street_arrays, &Array_Street_arrays[0], sizeof(int) * num_connections, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_input_actor_tag, &Actor_tag[0], sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_input_actor_id, &Actor_id[0], sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_jobs, &jobs[0], sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_randomn, &randomn[0], sizeof(int) * NUM_STREETS, hipMemcpyHostToDevice)); printf("Finished copying data.\n"); hipEvent_t start, stop; hipEventCreate(&start); hipEventCreate(&stop); printf("Launching kernel...\n"); hipEventRecord(start); kernel<<<dim3(32), dim3((NUM_PEDS + NUM_CARS) / 32)>>>(GOOD_WEATHER, 1000000, v_d_Actor_progress, v_d_Actor_street, v_d_Car_max_velocity, v_d_Street_length, v_d_Street_max_velocity, v_d_Street_neighbors, v_d_Array_Street_size, v_d_Array_Street_offset, v_d_Array_Street_arrays, v_d_input_actor_tag, v_d_input_actor_id, v_d_jobs, v_d_randomn); CudaCheckError(); hipEventRecord(stop); hipEventSynchronize(stop); float milliseconds = 0; hipEventElapsedTime(&milliseconds, start, stop); CudaCheckError(); printf("Kernel finished.\n"); hipMemcpy(Actor_progress, v_d_Actor_progress, sizeof(float) * (NUM_PEDS + NUM_CARS), hipMemcpyDeviceToHost); for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { // printf(" %f ", Actor_progress[i]); } hipMemcpy(Actor_street, v_d_Actor_street, sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyDeviceToHost); for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { // printf(" %i ", Actor_street[i]); } printf("\n\n\nElapsed time millis: %f\n", milliseconds); }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#define REORDER 1 #define GOOD_WEATHER 0 #define BAD_WEATHER 1 #define TAG_Car 0 #define TAG_Pedestrian 1 #include <hip/hip_runtime.h> #include <stdio.h> #include <stdlib.h> #include <time.h> //#include <random> //#include <array> #include <algorithm> #define NUM_CARS 4096 #define NUM_PEDS 16384 #define NUM_STREETS 500 #define MAX_CONNECTIONS 10 #define MAX_LEN 25 using namespace std; // Define this to turn on error checking #define CUDA_ERROR_CHECK #define CudaSafeCall( err ) __cudaSafeCall( err, __FILE__, __LINE__ ) #define CudaCheckError() __cudaCheckError( __FILE__, __LINE__ ) inline void __cudaSafeCall( hipError_t err, const char *file, const int line ) { #ifdef CUDA_ERROR_CHECK if ( hipSuccess != err ) { fprintf( stderr, "cudaSafeCall() failed at %s:%i : %s\n", file, line, hipGetErrorString( err ) ); exit( -1 ); } #endif return; } inline void __cudaCheckError( const char *file, const int line ) { #ifdef CUDA_ERROR_CHECK hipError_t err = hipGetLastError(); if ( hipSuccess != err ) { fprintf( stderr, "cudaCheckError() failed at %s:%i : %s\n", file, line, hipGetErrorString( err ) ); exit( -1 ); } // More careful checking. However, this will affect performance. // Comment away if needed. err = hipDeviceSynchronize(); if( hipSuccess != err ) { fprintf( stderr, "cudaCheckError() with sync failed at %s:%i : %s\n", file, line, hipGetErrorString( err ) ); exit( -1 ); } #endif return; } __device__ float *d_Actor_progress; __device__ int *d_Actor_street; __device__ float *d_Car_max_velocity; __device__ float *d_Street_length; __device__ float *d_Street_max_velocity; __device__ int *d_Street_neighbors; __device__ int *d_Array_Street_size; __device__ int *d_Array_Street_offset; __device__ int *d_Array_Street_arrays; __device__ int *d_input_actor_tag; __device__ int *d_input_actor_id; __device__ int *d_jobs; __device__ int *d_randomn; __device__ void method_Car_move(int actor_id, int weather) { float weather_multiplier; if (weather == GOOD_WEATHER) { weather_multiplier = 1.0; } else if (weather == BAD_WEATHER) { weather_multiplier = 0.75; } float speed = min(d_Car_max_velocity[actor_id], d_Street_max_velocity[d_Actor_street[actor_id]]) * weather_multiplier; d_Actor_progress[actor_id] = d_Actor_progress[actor_id] + (speed / 60.0); /* 1 tick = 1 minute */ if (d_Actor_progress[actor_id] >= d_Street_length[d_Actor_street[actor_id]]) { // move to different street int array_id = d_Street_neighbors[d_Actor_street[actor_id]]; int neighbor_index = d_randomn[d_Actor_street[actor_id]] % d_Array_Street_size[array_id]; d_Actor_street[actor_id] = d_Array_Street_arrays[d_Array_Street_offset[array_id] + neighbor_index]; d_Actor_progress[actor_id] = 0.0f; } } __device__ void method_Pedestrian_move(int actor_id, int weather) { float speed = d_randomn[((int) (d_Actor_progress[actor_id]*d_Actor_progress[actor_id])) % NUM_STREETS] % 7 - 2; d_Actor_progress[actor_id] = d_Actor_progress[actor_id] + (speed / 60.0); if (d_Actor_progress[actor_id] >= d_Street_length[d_Actor_street[actor_id]]) { // move to different street int array_id = d_Street_neighbors[d_Actor_street[actor_id]]; int neighbor_index = d_randomn[d_Actor_street[actor_id]] % d_Array_Street_size[array_id]; d_Actor_street[actor_id] = d_Array_Street_arrays[d_Array_Street_offset[array_id] + neighbor_index]; d_Actor_progress[actor_id] = 0.0f; } } __device__ void block(int actor_tag, int actor_id, int weather, int ticks) { for (int i = 0; i < ticks; i++) { switch (actor_tag) { case TAG_Car: method_Car_move(actor_id, weather); break; case TAG_Pedestrian: method_Pedestrian_move(actor_id, weather); break; } } } __global__ void kernel(int weather, int ticks, float *v_d_Actor_progress, int *v_d_Actor_street, float *v_d_Car_max_velocity, float *v_d_Street_length, float *v_d_Street_max_velocity, int *v_d_Street_neighbors, int *v_d_Array_Street_size, int *v_d_Array_Street_offset, int *v_d_Array_Street_arrays, int *v_d_input_actor_tag, int *v_d_input_actor_id, int *v_d_jobs, int *v_d_randomn) { d_Actor_progress = v_d_Actor_progress; d_Actor_street = v_d_Actor_street; d_Car_max_velocity = v_d_Car_max_velocity; d_Street_length = v_d_Street_length; d_Street_max_velocity = v_d_Street_max_velocity; d_Street_neighbors = v_d_Street_neighbors; d_Array_Street_size = v_d_Array_Street_size; d_Array_Street_offset = v_d_Array_Street_offset; d_Array_Street_arrays = v_d_Array_Street_arrays; d_input_actor_tag = v_d_input_actor_tag; d_input_actor_id = v_d_input_actor_id; d_jobs = v_d_jobs; d_randomn = v_d_randomn; int tid = blockIdx.x * blockDim.x + threadIdx.x; __syncthreads(); #if (REORDER) block(d_input_actor_tag[d_jobs[tid]], d_input_actor_id[d_jobs[tid]], weather, ticks); #else block(d_input_actor_tag[tid], d_input_actor_id[tid], weather, ticks); #endif } int main() { printf("Setting up scenario...\n"); srand(42); // streets float *Street_length = new float[NUM_STREETS]; float *Street_max_velocity = new float[NUM_STREETS]; int *Street_neighbors = new int[NUM_STREETS]; for (int i = 0; i < NUM_STREETS; i++) { Street_length[i] = rand() % MAX_LEN + 1; Street_max_velocity[i] = rand() % 40 + 45; /* speed between 45 and 105 */ Street_neighbors[i] = i; } // neighbors int *Array_Street_offset = new int[NUM_STREETS]; int *Array_Street_size = new int[NUM_STREETS]; int num_connections = 0; for (int i = 0; i < NUM_STREETS; i++) { Array_Street_offset[i] = num_connections; int connections = rand() % MAX_CONNECTIONS + 1; Array_Street_size[i] = connections; num_connections += connections; } int *Array_Street_arrays = new int[num_connections]; for (int i = 0; i < num_connections; i++) { Array_Street_arrays[i] = rand() % NUM_STREETS; } // actors int *Actor_street = new int[NUM_PEDS + NUM_CARS]; float *Actor_progress = new float[NUM_PEDS + NUM_CARS]; float *Car_max_velocity = new float[NUM_CARS + NUM_PEDS]; int *Actor_tag = new int[NUM_PEDS + NUM_CARS]; int *Actor_id = new int[NUM_PEDS + NUM_CARS]; for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { Actor_street[i] = rand() % NUM_STREETS; Actor_progress[i] = rand() % 10; Car_max_velocity[i] = rand() % 20 + 65; } for (int i = 0; i < NUM_PEDS; i++) { Actor_tag[i] = TAG_Pedestrian; Actor_id[i] = i; } for (int i = NUM_PEDS; i < NUM_PEDS + NUM_CARS; i++) { Actor_tag[i] = TAG_Car; Actor_id[i] = i; } std::srand(42); #if !(REORDER) random_shuffle(Actor_tag, Actor_tag + NUM_CARS + NUM_PEDS); #endif // jobs (dummy) int *jobs = new int[NUM_PEDS + NUM_CARS]; for (int i = 0; i < NUM_CARS + NUM_PEDS; i++) { jobs[i] = i; } // random numbers int *randomn = new int[NUM_STREETS]; for (int i = 0; i < NUM_STREETS; i++) { randomn[i] = rand() % NUM_STREETS; } printf("Scenario set up.\n"); printf("Copying data to GPU...\n"); float *v_d_Actor_progress; int *v_d_Actor_street; float *v_d_Car_max_velocity; float *v_d_Street_length; float *v_d_Street_max_velocity; int *v_d_Street_neighbors; int *v_d_Array_Street_size; int *v_d_Array_Street_offset; int *v_d_Array_Street_arrays; int *v_d_input_actor_tag; int *v_d_input_actor_id; int *v_d_jobs; int *v_d_randomn; CudaSafeCall(hipMalloc((void**) &v_d_Actor_progress, sizeof(float) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_Actor_street, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_Car_max_velocity, sizeof(float) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_Street_length, sizeof(float) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Street_max_velocity, sizeof(float) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Street_neighbors, sizeof(int) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Array_Street_size, sizeof(int) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Array_Street_offset, sizeof(int) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Array_Street_arrays, sizeof(int) * num_connections)); CudaSafeCall(hipMalloc((void**) &v_d_input_actor_tag, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_input_actor_id, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_jobs, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_randomn, sizeof(int) * NUM_STREETS)); CudaSafeCall(hipMemcpy(v_d_Actor_progress, &Actor_progress[0], sizeof(float) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Actor_street, &Actor_street[0], sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Car_max_velocity, &Car_max_velocity[0], sizeof(float) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Street_length, &Street_length[0], sizeof(float) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Street_max_velocity, &Street_max_velocity[0], sizeof(float) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Street_neighbors, &Street_neighbors[0], sizeof(int) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Array_Street_size, &Array_Street_size[0], sizeof(int) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Array_Street_offset, &Array_Street_offset[0], sizeof(int) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Array_Street_arrays, &Array_Street_arrays[0], sizeof(int) * num_connections, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_input_actor_tag, &Actor_tag[0], sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_input_actor_id, &Actor_id[0], sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_jobs, &jobs[0], sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_randomn, &randomn[0], sizeof(int) * NUM_STREETS, hipMemcpyHostToDevice)); printf("Finished copying data.\n"); hipEvent_t start, stop; hipEventCreate(&start); hipEventCreate(&stop); printf("Launching kernel...\n"); hipEventRecord(start); kernel<<<dim3(32), dim3((NUM_PEDS + NUM_CARS) / 32)>>>(GOOD_WEATHER, 1000000, v_d_Actor_progress, v_d_Actor_street, v_d_Car_max_velocity, v_d_Street_length, v_d_Street_max_velocity, v_d_Street_neighbors, v_d_Array_Street_size, v_d_Array_Street_offset, v_d_Array_Street_arrays, v_d_input_actor_tag, v_d_input_actor_id, v_d_jobs, v_d_randomn); CudaCheckError(); hipEventRecord(stop); hipEventSynchronize(stop); float milliseconds = 0; hipEventElapsedTime(&milliseconds, start, stop); CudaCheckError(); printf("Kernel finished.\n"); hipMemcpy(Actor_progress, v_d_Actor_progress, sizeof(float) * (NUM_PEDS + NUM_CARS), hipMemcpyDeviceToHost); for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { // printf(" %f ", Actor_progress[i]); } hipMemcpy(Actor_street, v_d_Actor_street, sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyDeviceToHost); for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { // printf(" %i ", Actor_street[i]); } printf("\n\n\nElapsed time millis: %f\n", milliseconds); }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .globl _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .p2align 8 .type _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_,@function _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_: s_clause 0x4 s_load_b256 s[4:11], s[0:1], 0x8 s_load_b256 s[16:23], s[0:1], 0x28 s_load_b256 s[40:47], s[0:1], 0x48 s_load_b32 s14, s[0:1], 0x4 s_load_b64 s[38:39], s[0:1], 0x68 s_getpc_b64 s[36:37] s_add_u32 s36, s36, d_Actor_progress@rel32@lo+4 s_addc_u32 s37, s37, d_Actor_progress@rel32@hi+12 s_getpc_b64 s[30:31] s_add_u32 s30, s30, d_Actor_street@rel32@lo+4 s_addc_u32 s31, s31, d_Actor_street@rel32@hi+12 s_getpc_b64 s[12:13] s_add_u32 s12, s12, d_Car_max_velocity@rel32@lo+4 s_addc_u32 s13, s13, d_Car_max_velocity@rel32@hi+12 v_mov_b32_e32 v22, 0 s_getpc_b64 s[26:27] s_add_u32 s26, s26, d_Street_length@rel32@lo+4 s_addc_u32 s27, s27, d_Street_length@rel32@hi+12 s_getpc_b64 s[2:3] s_add_u32 s2, s2, d_Street_max_velocity@rel32@lo+4 s_addc_u32 s3, s3, d_Street_max_velocity@rel32@hi+12 s_getpc_b64 s[24:25] s_add_u32 s24, s24, d_Street_neighbors@rel32@lo+4 s_addc_u32 s25, s25, d_Street_neighbors@rel32@hi+12 s_getpc_b64 s[28:29] s_add_u32 s28, s28, d_Array_Street_size@rel32@lo+4 s_addc_u32 s29, s29, d_Array_Street_size@rel32@hi+12 s_getpc_b64 s[34:35] s_add_u32 s34, s34, d_Array_Street_offset@rel32@lo+4 s_addc_u32 s35, s35, d_Array_Street_offset@rel32@hi+12 s_waitcnt lgkmcnt(0) v_dual_mov_b32 v1, s4 :: v_dual_mov_b32 v2, s5 v_dual_mov_b32 v3, s6 :: v_dual_mov_b32 v4, s7 v_dual_mov_b32 v5, s8 :: v_dual_mov_b32 v6, s9 v_dual_mov_b32 v7, s10 :: v_dual_mov_b32 v8, s11 v_dual_mov_b32 v9, s16 :: v_dual_mov_b32 v10, s17 v_mov_b32_e32 v11, s18 s_getpc_b64 s[4:5] s_add_u32 s4, s4, d_Array_Street_arrays@rel32@lo+4 s_addc_u32 s5, s5, d_Array_Street_arrays@rel32@hi+12 v_dual_mov_b32 v12, s19 :: v_dual_mov_b32 v13, s20 s_getpc_b64 s[8:9] s_add_u32 s8, s8, d_input_actor_tag@rel32@lo+4 s_addc_u32 s9, s9, d_input_actor_tag@rel32@hi+12 v_dual_mov_b32 v14, s21 :: v_dual_mov_b32 v15, s22 s_getpc_b64 s[10:11] s_add_u32 s10, s10, d_input_actor_id@rel32@lo+4 s_addc_u32 s11, s11, d_input_actor_id@rel32@hi+12 v_mov_b32_e32 v16, s23 s_clause 0x7 global_store_b64 v22, v[1:2], s[36:37] global_store_b64 v22, v[3:4], s[30:31] global_store_b64 v22, v[5:6], s[12:13] global_store_b64 v22, v[7:8], s[26:27] global_store_b64 v22, v[9:10], s[2:3] global_store_b64 v22, v[11:12], s[24:25] global_store_b64 v22, v[13:14], s[28:29] global_store_b64 v22, v[15:16], s[34:35] v_mov_b32_e32 v1, s40 s_getpc_b64 s[16:17] s_add_u32 s16, s16, d_jobs@rel32@lo+4 s_addc_u32 s17, s17, d_jobs@rel32@hi+12 v_dual_mov_b32 v2, s41 :: v_dual_mov_b32 v3, s42 s_getpc_b64 s[6:7] s_add_u32 s6, s6, d_randomn@rel32@lo+4 s_addc_u32 s7, s7, d_randomn@rel32@hi+12 v_dual_mov_b32 v4, s43 :: v_dual_mov_b32 v5, s44 s_add_u32 s18, s0, 0x70 v_dual_mov_b32 v6, s45 :: v_dual_mov_b32 v7, s46 s_addc_u32 s19, s1, 0 v_dual_mov_b32 v8, s47 :: v_dual_mov_b32 v9, s38 s_cmp_lt_i32 s14, 1 v_mov_b32_e32 v10, s39 s_clause 0x4 global_store_b64 v22, v[1:2], s[4:5] global_store_b64 v22, v[3:4], s[8:9] global_store_b64 v22, v[5:6], s[10:11] global_store_b64 v22, v[7:8], s[16:17] global_store_b64 v22, v[9:10], s[6:7] s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv s_cbranch_scc1 .LBB0_13 s_clause 0x1 global_load_b64 v[1:2], v22, s[16:17] global_load_b64 v[5:6], v22, s[10:11] s_load_b32 s16, s[18:19], 0xc s_load_b32 s0, s[0:1], 0x0 s_waitcnt lgkmcnt(0) s_and_b32 s16, s16, 0xffff s_cmp_eq_u32 s0, 1 s_cselect_b32 s0, -1, 0 s_waitcnt vmcnt(1) v_mad_u64_u32 v[3:4], null, s15, s16, v[0:1] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v4, 31, v3 v_lshlrev_b64 v[3:4], 2, v[3:4] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v0, vcc_lo, v1, v3 v_add_co_ci_u32_e32 v1, vcc_lo, v2, v4, vcc_lo flat_load_b32 v0, v[0:1] global_load_b64 v[7:8], v22, s[8:9] s_waitcnt vmcnt(1) lgkmcnt(0) v_ashrrev_i32_e32 v1, 31, v0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[9:10], 2, v[0:1] v_add_co_u32 v0, vcc_lo, v5, v9 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v1, vcc_lo, v6, v10, vcc_lo s_waitcnt vmcnt(0) v_add_co_u32 v23, vcc_lo, v7, v9 v_add_co_ci_u32_e32 v24, vcc_lo, v8, v10, vcc_lo flat_load_b32 v14, v[0:1] s_clause 0x9 global_load_b64 v[16:17], v22, s[36:37] global_load_b64 v[0:1], v22, s[6:7] global_load_b64 v[18:19], v22, s[30:31] global_load_b64 v[2:3], v22, s[4:5] global_load_b64 v[4:5], v22, s[26:27] global_load_b64 v[20:21], v22, s[12:13] global_load_b64 v[6:7], v22, s[28:29] global_load_b64 v[8:9], v22, s[34:35] global_load_b64 v[10:11], v22, s[24:25] global_load_b64 v[12:13], v22, s[2:3] flat_load_b32 v23, v[23:24] v_cndmask_b32_e64 v24, 1.0, 0x3f400000, s0 s_waitcnt vmcnt(11) lgkmcnt(1) v_ashrrev_i32_e32 v15, 31, v14 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[25:26], 2, v[14:15] s_waitcnt vmcnt(10) v_add_co_u32 v14, vcc_lo, v16, v25 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v15, vcc_lo, v17, v26, vcc_lo s_waitcnt vmcnt(8) v_add_co_u32 v16, vcc_lo, v18, v25 v_add_co_ci_u32_e32 v17, vcc_lo, v19, v26, vcc_lo s_waitcnt vmcnt(5) v_add_co_u32 v18, vcc_lo, v20, v25 v_add_co_ci_u32_e32 v19, vcc_lo, v21, v26, vcc_lo s_branch .LBB0_3 .LBB0_2: s_or_b32 exec_lo, exec_lo, s0 s_add_i32 s14, s14, -1 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_lg_u32 s14, 0 s_cbranch_scc0 .LBB0_13 .LBB0_3: s_mov_b32 s1, 0 s_mov_b32 s0, exec_lo s_waitcnt vmcnt(0) lgkmcnt(0) v_cmpx_lt_i32_e32 0, v23 s_xor_b32 s0, exec_lo, s0 s_cbranch_execz .LBB0_7 s_mov_b32 s2, 0 s_mov_b32 s1, exec_lo v_cmpx_eq_u32_e32 1, v23 s_cbranch_execz .LBB0_6 flat_load_b32 v35, v[14:15] s_waitcnt vmcnt(0) lgkmcnt(0) v_mul_f32_e32 v20, v35, v35 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cvt_i32_f32_e32 v20, v20 v_mul_hi_i32 v21, v20, 0x10624dd3 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshrrev_b32_e32 v25, 31, v21 v_ashrrev_i32_e32 v21, 5, v21 v_add_nc_u32_e32 v21, v21, v25 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_lo_u32 v21, v21, 0x1f4 v_sub_nc_u32_e32 v20, v20, v21 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v21, 31, v20 v_lshlrev_b64 v[20:21], 2, v[20:21] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v20, vcc_lo, v0, v20 v_add_co_ci_u32_e32 v21, vcc_lo, v1, v21, vcc_lo flat_load_b32 v20, v[20:21] s_waitcnt vmcnt(0) lgkmcnt(0) v_mul_hi_i32 v21, v20, 0x92492493 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_nc_u32_e32 v21, v21, v20 v_lshrrev_b32_e32 v25, 31, v21 v_ashrrev_i32_e32 v21, 2, v21 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_nc_u32_e32 v21, v21, v25 v_mul_lo_u32 v21, v21, 7 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_sub_nc_u32_e32 v20, v20, v21 v_add_nc_u32_e32 v20, -2, v20 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cvt_f32_i32_e32 v20, v20 v_cvt_f64_f32_e32 v[25:26], v20 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_div_scale_f64 v[27:28], null, 0x404e0000, 0x404e0000, v[25:26] v_rcp_f64_e32 v[20:21], v[27:28] s_waitcnt_depctr 0xfff v_fma_f64 v[29:30], -v[27:28], v[20:21], 1.0 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_1) v_fma_f64 v[29:30], v[20:21], v[29:30], v[20:21] flat_load_b32 v20, v[16:17] v_fma_f64 v[31:32], -v[27:28], v[29:30], 1.0 v_fma_f64 v[29:30], v[29:30], v[31:32], v[29:30] v_div_scale_f64 v[31:32], vcc_lo, v[25:26], 0x404e0000, v[25:26] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_f64 v[33:34], v[31:32], v[29:30] v_fma_f64 v[27:28], -v[27:28], v[33:34], v[31:32] s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_div_fmas_f64 v[27:28], v[27:28], v[29:30], v[33:34] v_cvt_f64_f32_e32 v[29:30], v35 v_div_fixup_f64 v[25:26], v[27:28], 0x404e0000, v[25:26] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_f64 v[25:26], v[25:26], v[29:30] v_cvt_f32_f64_e32 v27, v[25:26] s_waitcnt vmcnt(0) lgkmcnt(0) v_ashrrev_i32_e32 v21, 31, v20 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[25:26], 2, v[20:21] v_add_co_u32 v25, vcc_lo, v4, v25 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v26, vcc_lo, v5, v26, vcc_lo flat_store_b32 v[14:15], v27 flat_load_b32 v25, v[25:26] s_waitcnt vmcnt(0) lgkmcnt(0) v_cmp_le_f32_e32 vcc_lo, v25, v27 s_and_b32 s2, vcc_lo, exec_lo .LBB0_6: s_or_b32 exec_lo, exec_lo, s1 s_delay_alu instid0(SALU_CYCLE_1) s_and_b32 s1, s2, exec_lo .LBB0_7: s_and_not1_saveexec_b32 s0, s0 s_cbranch_execz .LBB0_11 s_mov_b32 s3, s1 s_mov_b32 s2, exec_lo v_cmpx_eq_u32_e32 0, v23 s_cbranch_execz .LBB0_10 flat_load_b32 v20, v[16:17] s_and_not1_b32 s3, s1, exec_lo s_waitcnt vmcnt(0) lgkmcnt(0) v_ashrrev_i32_e32 v21, 31, v20 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[25:26], 2, v[20:21] v_add_co_u32 v27, vcc_lo, v12, v25 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v28, vcc_lo, v13, v26, vcc_lo flat_load_b32 v29, v[18:19] flat_load_b32 v27, v[27:28] flat_load_b32 v37, v[14:15] s_waitcnt vmcnt(1) lgkmcnt(1) v_dual_max_f32 v28, v29, v29 :: v_dual_max_f32 v27, v27, v27 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_min_f32_e32 v27, v28, v27 v_mul_f32_e32 v27, v24, v27 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cvt_f64_f32_e32 v[27:28], v27 v_div_scale_f64 v[29:30], null, 0x404e0000, 0x404e0000, v[27:28] s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_1) v_rcp_f64_e32 v[31:32], v[29:30] s_waitcnt_depctr 0xfff v_fma_f64 v[33:34], -v[29:30], v[31:32], 1.0 v_fma_f64 v[31:32], v[31:32], v[33:34], v[31:32] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[33:34], -v[29:30], v[31:32], 1.0 v_fma_f64 v[31:32], v[31:32], v[33:34], v[31:32] v_div_scale_f64 v[33:34], vcc_lo, v[27:28], 0x404e0000, v[27:28] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_f64 v[35:36], v[33:34], v[31:32] v_fma_f64 v[29:30], -v[29:30], v[35:36], v[33:34] s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_4) | instid1(VALU_DEP_4) v_div_fmas_f64 v[29:30], v[29:30], v[31:32], v[35:36] s_waitcnt vmcnt(0) lgkmcnt(0) v_cvt_f64_f32_e32 v[31:32], v37 v_add_co_u32 v25, vcc_lo, v4, v25 v_add_co_ci_u32_e32 v26, vcc_lo, v5, v26, vcc_lo v_div_fixup_f64 v[27:28], v[29:30], 0x404e0000, v[27:28] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_f64 v[27:28], v[27:28], v[31:32] v_cvt_f32_f64_e32 v27, v[27:28] flat_store_b32 v[14:15], v27 flat_load_b32 v25, v[25:26] s_waitcnt vmcnt(0) lgkmcnt(0) v_cmp_le_f32_e32 vcc_lo, v25, v27 s_and_b32 s4, vcc_lo, exec_lo s_delay_alu instid0(SALU_CYCLE_1) s_or_b32 s3, s3, s4 .LBB0_10: s_or_b32 exec_lo, exec_lo, s2 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_1) | instid1(SALU_CYCLE_1) s_and_not1_b32 s1, s1, exec_lo s_and_b32 s2, s3, exec_lo s_or_b32 s1, s1, s2 .LBB0_11: s_or_b32 exec_lo, exec_lo, s0 s_and_saveexec_b32 s0, s1 s_cbranch_execz .LBB0_2 v_lshlrev_b64 v[20:21], 2, v[20:21] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v25, vcc_lo, v10, v20 v_add_co_ci_u32_e32 v26, vcc_lo, v11, v21, vcc_lo flat_load_b32 v25, v[25:26] s_waitcnt vmcnt(0) lgkmcnt(0) v_ashrrev_i32_e32 v26, 31, v25 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[25:26], 2, v[25:26] v_add_co_u32 v27, vcc_lo, v6, v25 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v28, vcc_lo, v7, v26, vcc_lo v_add_co_u32 v20, vcc_lo, v0, v20 v_add_co_ci_u32_e32 v21, vcc_lo, v1, v21, vcc_lo flat_load_b32 v27, v[27:28] flat_load_b32 v28, v[20:21] v_add_co_u32 v20, vcc_lo, v8, v25 v_add_co_ci_u32_e32 v21, vcc_lo, v9, v26, vcc_lo flat_load_b32 v20, v[20:21] s_waitcnt vmcnt(2) lgkmcnt(2) v_ashrrev_i32_e32 v21, 31, v27 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_2) v_add_nc_u32_e32 v25, v27, v21 s_waitcnt vmcnt(1) lgkmcnt(1) v_ashrrev_i32_e32 v27, 31, v28 v_xor_b32_e32 v21, v25, v21 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_nc_u32_e32 v28, v28, v27 v_cvt_f32_u32_e32 v25, v21 v_sub_nc_u32_e32 v26, 0, v21 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_xor_b32_e32 v28, v28, v27 v_rcp_iflag_f32_e32 v25, v25 s_waitcnt_depctr 0xfff v_mul_f32_e32 v25, 0x4f7ffffe, v25 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cvt_u32_f32_e32 v25, v25 v_mul_lo_u32 v26, v26, v25 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_hi_u32 v26, v25, v26 v_add_nc_u32_e32 v25, v25, v26 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_hi_u32 v25, v28, v25 v_mul_lo_u32 v25, v25, v21 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_sub_nc_u32_e32 v25, v28, v25 v_sub_nc_u32_e32 v26, v25, v21 v_cmp_ge_u32_e32 vcc_lo, v25, v21 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_cndmask_b32_e32 v25, v25, v26, vcc_lo v_sub_nc_u32_e32 v26, v25, v21 v_cmp_ge_u32_e32 vcc_lo, v25, v21 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_cndmask_b32_e32 v21, v25, v26, vcc_lo v_xor_b32_e32 v21, v21, v27 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_sub_nc_u32_e32 v21, v21, v27 s_waitcnt vmcnt(0) lgkmcnt(0) v_add_nc_u32_e32 v20, v20, v21 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v21, 31, v20 v_lshlrev_b64 v[20:21], 2, v[20:21] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v20, vcc_lo, v2, v20 v_add_co_ci_u32_e32 v21, vcc_lo, v3, v21, vcc_lo flat_load_b32 v20, v[20:21] s_waitcnt vmcnt(0) lgkmcnt(0) flat_store_b32 v[16:17], v20 flat_store_b32 v[14:15], v22 s_branch .LBB0_2 .LBB0_13: s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 368 .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 38 .amdhsa_next_free_sgpr 48 .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 _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_, .Lfunc_end0-_Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .section .AMDGPU.csdata,"",@progbits .text .p2alignl 7, 3214868480 .fill 96, 4, 3214868480 .protected d_Actor_progress .type d_Actor_progress,@object .section .bss,"aw",@nobits .globl d_Actor_progress .p2align 3, 0x0 d_Actor_progress: .quad 0 .size d_Actor_progress, 8 .protected d_Actor_street .type d_Actor_street,@object .globl d_Actor_street .p2align 3, 0x0 d_Actor_street: .quad 0 .size d_Actor_street, 8 .protected d_Car_max_velocity .type d_Car_max_velocity,@object .globl d_Car_max_velocity .p2align 3, 0x0 d_Car_max_velocity: .quad 0 .size d_Car_max_velocity, 8 .protected d_Street_length .type d_Street_length,@object .globl d_Street_length .p2align 3, 0x0 d_Street_length: .quad 0 .size d_Street_length, 8 .protected d_Street_max_velocity .type d_Street_max_velocity,@object .globl d_Street_max_velocity .p2align 3, 0x0 d_Street_max_velocity: .quad 0 .size d_Street_max_velocity, 8 .protected d_Street_neighbors .type d_Street_neighbors,@object .globl d_Street_neighbors .p2align 3, 0x0 d_Street_neighbors: .quad 0 .size d_Street_neighbors, 8 .protected d_Array_Street_size .type d_Array_Street_size,@object .globl d_Array_Street_size .p2align 3, 0x0 d_Array_Street_size: .quad 0 .size d_Array_Street_size, 8 .protected d_Array_Street_offset .type d_Array_Street_offset,@object .globl d_Array_Street_offset .p2align 3, 0x0 d_Array_Street_offset: .quad 0 .size d_Array_Street_offset, 8 .protected d_Array_Street_arrays .type d_Array_Street_arrays,@object .globl d_Array_Street_arrays .p2align 3, 0x0 d_Array_Street_arrays: .quad 0 .size d_Array_Street_arrays, 8 .protected d_input_actor_tag .type d_input_actor_tag,@object .globl d_input_actor_tag .p2align 3, 0x0 d_input_actor_tag: .quad 0 .size d_input_actor_tag, 8 .protected d_input_actor_id .type d_input_actor_id,@object .globl d_input_actor_id .p2align 3, 0x0 d_input_actor_id: .quad 0 .size d_input_actor_id, 8 .protected d_jobs .type d_jobs,@object .globl d_jobs .p2align 3, 0x0 d_jobs: .quad 0 .size d_jobs, 8 .protected d_randomn .type d_randomn,@object .globl d_randomn .p2align 3, 0x0 d_randomn: .quad 0 .size d_randomn, 8 .type __hip_cuid_,@object .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 d_Actor_progress .addrsig_sym d_Actor_street .addrsig_sym d_Car_max_velocity .addrsig_sym d_Street_length .addrsig_sym d_Street_max_velocity .addrsig_sym d_Street_neighbors .addrsig_sym d_Array_Street_size .addrsig_sym d_Array_Street_offset .addrsig_sym d_Array_Street_arrays .addrsig_sym d_input_actor_tag .addrsig_sym d_input_actor_id .addrsig_sym d_jobs .addrsig_sym d_randomn .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 - .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 - .address_space: global .offset: 40 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 48 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 56 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 64 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 72 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 80 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 88 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 96 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 104 .size: 8 .value_kind: global_buffer - .offset: 112 .size: 4 .value_kind: hidden_block_count_x - .offset: 116 .size: 4 .value_kind: hidden_block_count_y - .offset: 120 .size: 4 .value_kind: hidden_block_count_z - .offset: 124 .size: 2 .value_kind: hidden_group_size_x - .offset: 126 .size: 2 .value_kind: hidden_group_size_y - .offset: 128 .size: 2 .value_kind: hidden_group_size_z - .offset: 130 .size: 2 .value_kind: hidden_remainder_x - .offset: 132 .size: 2 .value_kind: hidden_remainder_y - .offset: 134 .size: 2 .value_kind: hidden_remainder_z - .offset: 152 .size: 8 .value_kind: hidden_global_offset_x - .offset: 160 .size: 8 .value_kind: hidden_global_offset_y - .offset: 168 .size: 8 .value_kind: hidden_global_offset_z - .offset: 176 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 368 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .private_segment_fixed_size: 0 .sgpr_count: 50 .sgpr_spill_count: 0 .symbol: _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 38 .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.
#define REORDER 1 #define GOOD_WEATHER 0 #define BAD_WEATHER 1 #define TAG_Car 0 #define TAG_Pedestrian 1 #include <hip/hip_runtime.h> #include <stdio.h> #include <stdlib.h> #include <time.h> //#include <random> //#include <array> #include <algorithm> #define NUM_CARS 4096 #define NUM_PEDS 16384 #define NUM_STREETS 500 #define MAX_CONNECTIONS 10 #define MAX_LEN 25 using namespace std; // Define this to turn on error checking #define CUDA_ERROR_CHECK #define CudaSafeCall( err ) __cudaSafeCall( err, __FILE__, __LINE__ ) #define CudaCheckError() __cudaCheckError( __FILE__, __LINE__ ) inline void __cudaSafeCall( hipError_t err, const char *file, const int line ) { #ifdef CUDA_ERROR_CHECK if ( hipSuccess != err ) { fprintf( stderr, "cudaSafeCall() failed at %s:%i : %s\n", file, line, hipGetErrorString( err ) ); exit( -1 ); } #endif return; } inline void __cudaCheckError( const char *file, const int line ) { #ifdef CUDA_ERROR_CHECK hipError_t err = hipGetLastError(); if ( hipSuccess != err ) { fprintf( stderr, "cudaCheckError() failed at %s:%i : %s\n", file, line, hipGetErrorString( err ) ); exit( -1 ); } // More careful checking. However, this will affect performance. // Comment away if needed. err = hipDeviceSynchronize(); if( hipSuccess != err ) { fprintf( stderr, "cudaCheckError() with sync failed at %s:%i : %s\n", file, line, hipGetErrorString( err ) ); exit( -1 ); } #endif return; } __device__ float *d_Actor_progress; __device__ int *d_Actor_street; __device__ float *d_Car_max_velocity; __device__ float *d_Street_length; __device__ float *d_Street_max_velocity; __device__ int *d_Street_neighbors; __device__ int *d_Array_Street_size; __device__ int *d_Array_Street_offset; __device__ int *d_Array_Street_arrays; __device__ int *d_input_actor_tag; __device__ int *d_input_actor_id; __device__ int *d_jobs; __device__ int *d_randomn; __device__ void method_Car_move(int actor_id, int weather) { float weather_multiplier; if (weather == GOOD_WEATHER) { weather_multiplier = 1.0; } else if (weather == BAD_WEATHER) { weather_multiplier = 0.75; } float speed = min(d_Car_max_velocity[actor_id], d_Street_max_velocity[d_Actor_street[actor_id]]) * weather_multiplier; d_Actor_progress[actor_id] = d_Actor_progress[actor_id] + (speed / 60.0); /* 1 tick = 1 minute */ if (d_Actor_progress[actor_id] >= d_Street_length[d_Actor_street[actor_id]]) { // move to different street int array_id = d_Street_neighbors[d_Actor_street[actor_id]]; int neighbor_index = d_randomn[d_Actor_street[actor_id]] % d_Array_Street_size[array_id]; d_Actor_street[actor_id] = d_Array_Street_arrays[d_Array_Street_offset[array_id] + neighbor_index]; d_Actor_progress[actor_id] = 0.0f; } } __device__ void method_Pedestrian_move(int actor_id, int weather) { float speed = d_randomn[((int) (d_Actor_progress[actor_id]*d_Actor_progress[actor_id])) % NUM_STREETS] % 7 - 2; d_Actor_progress[actor_id] = d_Actor_progress[actor_id] + (speed / 60.0); if (d_Actor_progress[actor_id] >= d_Street_length[d_Actor_street[actor_id]]) { // move to different street int array_id = d_Street_neighbors[d_Actor_street[actor_id]]; int neighbor_index = d_randomn[d_Actor_street[actor_id]] % d_Array_Street_size[array_id]; d_Actor_street[actor_id] = d_Array_Street_arrays[d_Array_Street_offset[array_id] + neighbor_index]; d_Actor_progress[actor_id] = 0.0f; } } __device__ void block(int actor_tag, int actor_id, int weather, int ticks) { for (int i = 0; i < ticks; i++) { switch (actor_tag) { case TAG_Car: method_Car_move(actor_id, weather); break; case TAG_Pedestrian: method_Pedestrian_move(actor_id, weather); break; } } } __global__ void kernel(int weather, int ticks, float *v_d_Actor_progress, int *v_d_Actor_street, float *v_d_Car_max_velocity, float *v_d_Street_length, float *v_d_Street_max_velocity, int *v_d_Street_neighbors, int *v_d_Array_Street_size, int *v_d_Array_Street_offset, int *v_d_Array_Street_arrays, int *v_d_input_actor_tag, int *v_d_input_actor_id, int *v_d_jobs, int *v_d_randomn) { d_Actor_progress = v_d_Actor_progress; d_Actor_street = v_d_Actor_street; d_Car_max_velocity = v_d_Car_max_velocity; d_Street_length = v_d_Street_length; d_Street_max_velocity = v_d_Street_max_velocity; d_Street_neighbors = v_d_Street_neighbors; d_Array_Street_size = v_d_Array_Street_size; d_Array_Street_offset = v_d_Array_Street_offset; d_Array_Street_arrays = v_d_Array_Street_arrays; d_input_actor_tag = v_d_input_actor_tag; d_input_actor_id = v_d_input_actor_id; d_jobs = v_d_jobs; d_randomn = v_d_randomn; int tid = blockIdx.x * blockDim.x + threadIdx.x; __syncthreads(); #if (REORDER) block(d_input_actor_tag[d_jobs[tid]], d_input_actor_id[d_jobs[tid]], weather, ticks); #else block(d_input_actor_tag[tid], d_input_actor_id[tid], weather, ticks); #endif } int main() { printf("Setting up scenario...\n"); srand(42); // streets float *Street_length = new float[NUM_STREETS]; float *Street_max_velocity = new float[NUM_STREETS]; int *Street_neighbors = new int[NUM_STREETS]; for (int i = 0; i < NUM_STREETS; i++) { Street_length[i] = rand() % MAX_LEN + 1; Street_max_velocity[i] = rand() % 40 + 45; /* speed between 45 and 105 */ Street_neighbors[i] = i; } // neighbors int *Array_Street_offset = new int[NUM_STREETS]; int *Array_Street_size = new int[NUM_STREETS]; int num_connections = 0; for (int i = 0; i < NUM_STREETS; i++) { Array_Street_offset[i] = num_connections; int connections = rand() % MAX_CONNECTIONS + 1; Array_Street_size[i] = connections; num_connections += connections; } int *Array_Street_arrays = new int[num_connections]; for (int i = 0; i < num_connections; i++) { Array_Street_arrays[i] = rand() % NUM_STREETS; } // actors int *Actor_street = new int[NUM_PEDS + NUM_CARS]; float *Actor_progress = new float[NUM_PEDS + NUM_CARS]; float *Car_max_velocity = new float[NUM_CARS + NUM_PEDS]; int *Actor_tag = new int[NUM_PEDS + NUM_CARS]; int *Actor_id = new int[NUM_PEDS + NUM_CARS]; for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { Actor_street[i] = rand() % NUM_STREETS; Actor_progress[i] = rand() % 10; Car_max_velocity[i] = rand() % 20 + 65; } for (int i = 0; i < NUM_PEDS; i++) { Actor_tag[i] = TAG_Pedestrian; Actor_id[i] = i; } for (int i = NUM_PEDS; i < NUM_PEDS + NUM_CARS; i++) { Actor_tag[i] = TAG_Car; Actor_id[i] = i; } std::srand(42); #if !(REORDER) random_shuffle(Actor_tag, Actor_tag + NUM_CARS + NUM_PEDS); #endif // jobs (dummy) int *jobs = new int[NUM_PEDS + NUM_CARS]; for (int i = 0; i < NUM_CARS + NUM_PEDS; i++) { jobs[i] = i; } // random numbers int *randomn = new int[NUM_STREETS]; for (int i = 0; i < NUM_STREETS; i++) { randomn[i] = rand() % NUM_STREETS; } printf("Scenario set up.\n"); printf("Copying data to GPU...\n"); float *v_d_Actor_progress; int *v_d_Actor_street; float *v_d_Car_max_velocity; float *v_d_Street_length; float *v_d_Street_max_velocity; int *v_d_Street_neighbors; int *v_d_Array_Street_size; int *v_d_Array_Street_offset; int *v_d_Array_Street_arrays; int *v_d_input_actor_tag; int *v_d_input_actor_id; int *v_d_jobs; int *v_d_randomn; CudaSafeCall(hipMalloc((void**) &v_d_Actor_progress, sizeof(float) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_Actor_street, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_Car_max_velocity, sizeof(float) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_Street_length, sizeof(float) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Street_max_velocity, sizeof(float) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Street_neighbors, sizeof(int) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Array_Street_size, sizeof(int) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Array_Street_offset, sizeof(int) * NUM_STREETS)); CudaSafeCall(hipMalloc((void**) &v_d_Array_Street_arrays, sizeof(int) * num_connections)); CudaSafeCall(hipMalloc((void**) &v_d_input_actor_tag, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_input_actor_id, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_jobs, sizeof(int) * (NUM_PEDS + NUM_CARS))); CudaSafeCall(hipMalloc((void**) &v_d_randomn, sizeof(int) * NUM_STREETS)); CudaSafeCall(hipMemcpy(v_d_Actor_progress, &Actor_progress[0], sizeof(float) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Actor_street, &Actor_street[0], sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Car_max_velocity, &Car_max_velocity[0], sizeof(float) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Street_length, &Street_length[0], sizeof(float) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Street_max_velocity, &Street_max_velocity[0], sizeof(float) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Street_neighbors, &Street_neighbors[0], sizeof(int) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Array_Street_size, &Array_Street_size[0], sizeof(int) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Array_Street_offset, &Array_Street_offset[0], sizeof(int) * NUM_STREETS, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_Array_Street_arrays, &Array_Street_arrays[0], sizeof(int) * num_connections, hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_input_actor_tag, &Actor_tag[0], sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_input_actor_id, &Actor_id[0], sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_jobs, &jobs[0], sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyHostToDevice)); CudaSafeCall(hipMemcpy(v_d_randomn, &randomn[0], sizeof(int) * NUM_STREETS, hipMemcpyHostToDevice)); printf("Finished copying data.\n"); hipEvent_t start, stop; hipEventCreate(&start); hipEventCreate(&stop); printf("Launching kernel...\n"); hipEventRecord(start); kernel<<<dim3(32), dim3((NUM_PEDS + NUM_CARS) / 32)>>>(GOOD_WEATHER, 1000000, v_d_Actor_progress, v_d_Actor_street, v_d_Car_max_velocity, v_d_Street_length, v_d_Street_max_velocity, v_d_Street_neighbors, v_d_Array_Street_size, v_d_Array_Street_offset, v_d_Array_Street_arrays, v_d_input_actor_tag, v_d_input_actor_id, v_d_jobs, v_d_randomn); CudaCheckError(); hipEventRecord(stop); hipEventSynchronize(stop); float milliseconds = 0; hipEventElapsedTime(&milliseconds, start, stop); CudaCheckError(); printf("Kernel finished.\n"); hipMemcpy(Actor_progress, v_d_Actor_progress, sizeof(float) * (NUM_PEDS + NUM_CARS), hipMemcpyDeviceToHost); for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { // printf(" %f ", Actor_progress[i]); } hipMemcpy(Actor_street, v_d_Actor_street, sizeof(int) * (NUM_PEDS + NUM_CARS), hipMemcpyDeviceToHost); for (int i = 0; i < NUM_PEDS + NUM_CARS; i++) { // printf(" %i ", Actor_street[i]); } printf("\n\n\nElapsed time millis: %f\n", milliseconds); }
.text .file "benchmark_column_reorder.hip" .globl _Z21__device_stub__kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ # -- Begin function _Z21__device_stub__kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .p2align 4, 0x90 .type _Z21__device_stub__kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_,@function _Z21__device_stub__kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_: # @_Z21__device_stub__kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .cfi_startproc # %bb.0: subq $216, %rsp .cfi_def_cfa_offset 224 movl %edi, 12(%rsp) movl %esi, 8(%rsp) movq %rdx, 88(%rsp) movq %rcx, 80(%rsp) movq %r8, 72(%rsp) movq %r9, 64(%rsp) leaq 12(%rsp), %rax movq %rax, 96(%rsp) leaq 8(%rsp), %rax movq %rax, 104(%rsp) leaq 88(%rsp), %rax movq %rax, 112(%rsp) leaq 80(%rsp), %rax movq %rax, 120(%rsp) leaq 72(%rsp), %rax movq %rax, 128(%rsp) leaq 64(%rsp), %rax movq %rax, 136(%rsp) leaq 224(%rsp), %rax movq %rax, 144(%rsp) leaq 232(%rsp), %rax movq %rax, 152(%rsp) leaq 240(%rsp), %rax movq %rax, 160(%rsp) leaq 248(%rsp), %rax movq %rax, 168(%rsp) leaq 256(%rsp), %rax movq %rax, 176(%rsp) leaq 264(%rsp), %rax movq %rax, 184(%rsp) leaq 272(%rsp), %rax movq %rax, 192(%rsp) leaq 280(%rsp), %rax movq %rax, 200(%rsp) leaq 288(%rsp), %rax movq %rax, 208(%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 96(%rsp), %r9 movl $_Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $232, %rsp .cfi_adjust_cfa_offset -232 retq .Lfunc_end0: .size _Z21__device_stub__kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_, .Lfunc_end0-_Z21__device_stub__kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .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 $472, %rsp # imm = 0x1D8 .cfi_def_cfa_offset 528 .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 $.Lstr, %edi callq puts@PLT movl $42, %edi callq srand movl $2000, %edi # imm = 0x7D0 callq _Znam movq %rax, %r12 movl $2000, %edi # imm = 0x7D0 callq _Znam movq %rax, %r15 movl $2000, %edi # imm = 0x7D0 callq _Znam movq %rax, %r14 xorl %ebx, %ebx .p2align 4, 0x90 .LBB1_1: # =>This Inner Loop Header: Depth=1 callq rand cltq imulq $1374389535, %rax, %rcx # imm = 0x51EB851F movq %rcx, %rdx shrq $63, %rdx sarq $35, %rcx addl %edx, %ecx leal (%rcx,%rcx,4), %ecx leal (%rcx,%rcx,4), %ecx negl %ecx addl %ecx, %eax incl %eax xorps %xmm0, %xmm0 cvtsi2ss %eax, %xmm0 movss %xmm0, (%r12,%rbx,4) callq rand cltq imulq $1717986919, %rax, %rcx # imm = 0x66666667 movq %rcx, %rdx shrq $63, %rdx sarq $36, %rcx addl %edx, %ecx shll $3, %ecx leal (%rcx,%rcx,4), %ecx negl %ecx addl %ecx, %eax addl $45, %eax xorps %xmm0, %xmm0 cvtsi2ss %eax, %xmm0 movss %xmm0, (%r15,%rbx,4) movl %ebx, (%r14,%rbx,4) incq %rbx cmpq $500, %rbx # imm = 0x1F4 jne .LBB1_1 # %bb.2: movq %r15, 184(%rsp) # 8-byte Spill movq %r14, 176(%rsp) # 8-byte Spill movl $2000, %edi # imm = 0x7D0 callq _Znam movq %rax, %rbx movl $2000, %edi # imm = 0x7D0 callq _Znam movq %rax, %r13 xorl %r15d, %r15d xorl %r14d, %r14d .p2align 4, 0x90 .LBB1_3: # =>This Inner Loop Header: Depth=1 movl %r14d, (%rbx,%r15,4) callq rand cltq imulq $1717986919, %rax, %rcx # imm = 0x66666667 movq %rcx, %rdx shrq $63, %rdx sarq $34, %rcx addl %edx, %ecx addl %ecx, %ecx leal (%rcx,%rcx,4), %ecx movl %eax, %edx subl %ecx, %edx negl %ecx addl %ecx, %eax incl %eax movl %eax, (%r13,%r15,4) addl %edx, %r14d incl %r14d incq %r15 cmpq $500, %r15 # imm = 0x1F4 jne .LBB1_3 # %bb.4: movq %r13, 168(%rsp) # 8-byte Spill movq %rbx, 160(%rsp) # 8-byte Spill movq %r12, 192(%rsp) # 8-byte Spill movslq %r14d, %rbx leaq (,%rbx,4), %rax testl %ebx, %ebx movq $-1, %rdi movq %rax, 48(%rsp) # 8-byte Spill cmovnsq %rax, %rdi callq _Znam movq %rax, %r15 testl %ebx, %ebx jle .LBB1_7 # %bb.5: # %.lr.ph.preheader movl %r14d, %r14d xorl %ebx, %ebx .p2align 4, 0x90 .LBB1_6: # %.lr.ph # =>This Inner Loop Header: Depth=1 callq rand cltq imulq $274877907, %rax, %rcx # imm = 0x10624DD3 movq %rcx, %rdx shrq $63, %rdx sarq $37, %rcx addl %edx, %ecx imull $500, %ecx, %ecx # imm = 0x1F4 subl %ecx, %eax movl %eax, (%r15,%rbx,4) incq %rbx cmpq %rbx, %r14 jne .LBB1_6 .LBB1_7: # %._crit_edge movq %r15, 152(%rsp) # 8-byte Spill movl $81920, %edi # imm = 0x14000 callq _Znam movq %rax, %rbx movl $81920, %edi # imm = 0x14000 callq _Znam movq %rax, %rbp movl $81920, %edi # imm = 0x14000 callq _Znam movq %rax, %r12 movl $81920, %edi # imm = 0x14000 callq _Znam movq %rax, %r15 movl $81920, %edi # imm = 0x14000 callq _Znam movq %rax, %r14 xorl %r13d, %r13d .p2align 4, 0x90 .LBB1_8: # =>This Inner Loop Header: Depth=1 callq rand cltq imulq $274877907, %rax, %rcx # imm = 0x10624DD3 movq %rcx, %rdx shrq $63, %rdx sarq $37, %rcx addl %edx, %ecx imull $500, %ecx, %ecx # imm = 0x1F4 subl %ecx, %eax movl %eax, (%rbx,%r13,4) callq rand cltq imulq $1717986919, %rax, %rcx # imm = 0x66666667 movq %rcx, %rdx shrq $63, %rdx sarq $34, %rcx addl %edx, %ecx addl %ecx, %ecx leal (%rcx,%rcx,4), %ecx subl %ecx, %eax xorps %xmm0, %xmm0 cvtsi2ss %eax, %xmm0 movss %xmm0, (%rbp,%r13,4) callq rand cltq imulq $1717986919, %rax, %rcx # imm = 0x66666667 movq %rcx, %rdx shrq $63, %rdx sarq $35, %rcx addl %edx, %ecx shll $2, %ecx leal (%rcx,%rcx,4), %ecx negl %ecx addl %ecx, %eax addl $65, %eax xorps %xmm0, %xmm0 cvtsi2ss %eax, %xmm0 movss %xmm0, (%r12,%r13,4) incq %r13 cmpq $20480, %r13 # imm = 0x5000 jne .LBB1_8 # %bb.9: # %.preheader151.preheader movq %rbx, 32(%rsp) # 8-byte Spill xorl %eax, %eax .p2align 4, 0x90 .LBB1_10: # %.preheader151 # =>This Inner Loop Header: Depth=1 movl $1, (%r15,%rax,4) movl %eax, (%r14,%rax,4) incq %rax cmpq $16384, %rax # imm = 0x4000 jne .LBB1_10 # %bb.11: # %.preheader.preheader movq %r15, %rdi addq $65536, %rdi # imm = 0x10000 movl $16384, %ebx # imm = 0x4000 movl $16384, %edx # imm = 0x4000 xorl %esi, %esi callq memset@PLT .p2align 4, 0x90 .LBB1_12: # %.preheader # =>This Inner Loop Header: Depth=1 movl %ebx, (%r14,%rbx,4) incq %rbx cmpq $20480, %rbx # imm = 0x5000 jne .LBB1_12 # %bb.13: movq %rbp, 40(%rsp) # 8-byte Spill movl $42, %edi callq srand movl $81920, %edi # imm = 0x14000 callq _Znam movq %rax, %r13 xorl %eax, %eax .p2align 4, 0x90 .LBB1_14: # =>This Inner Loop Header: Depth=1 movl %eax, (%r13,%rax,4) incq %rax cmpq $20480, %rax # imm = 0x5000 jne .LBB1_14 # %bb.15: movl $2000, %edi # imm = 0x7D0 callq _Znam movq %rax, %rbp xorl %ebx, %ebx .p2align 4, 0x90 .LBB1_16: # =>This Inner Loop Header: Depth=1 callq rand cltq imulq $274877907, %rax, %rcx # imm = 0x10624DD3 movq %rcx, %rdx shrq $63, %rdx sarq $37, %rcx addl %edx, %ecx imull $500, %ecx, %ecx # imm = 0x1F4 subl %ecx, %eax movl %eax, (%rbp,%rbx,4) incq %rbx cmpq $500, %rbx # imm = 0x1F4 jne .LBB1_16 # %bb.17: movl $.Lstr.1, %edi callq puts@PLT movl $.Lstr.2, %edi callq puts@PLT leaq 16(%rsp), %rdi movl $81920, %esi # imm = 0x14000 callq hipMalloc testl %eax, %eax jne .LBB1_18 # %bb.19: # %_Z14__cudaSafeCall10hipError_tPKci.exit leaq 8(%rsp), %rdi movl $81920, %esi # imm = 0x14000 callq hipMalloc testl %eax, %eax movq 32(%rsp), %rbx # 8-byte Reload jne .LBB1_20 # %bb.21: # %_Z14__cudaSafeCall10hipError_tPKci.exit95 leaq 144(%rsp), %rdi movl $81920, %esi # imm = 0x14000 callq hipMalloc testl %eax, %eax jne .LBB1_22 # %bb.23: # %_Z14__cudaSafeCall10hipError_tPKci.exit97 leaq 136(%rsp), %rdi movl $2000, %esi # imm = 0x7D0 callq hipMalloc testl %eax, %eax jne .LBB1_24 # %bb.25: # %_Z14__cudaSafeCall10hipError_tPKci.exit99 leaq 128(%rsp), %rdi movl $2000, %esi # imm = 0x7D0 callq hipMalloc testl %eax, %eax jne .LBB1_26 # %bb.27: # %_Z14__cudaSafeCall10hipError_tPKci.exit101 leaq 120(%rsp), %rdi movl $2000, %esi # imm = 0x7D0 callq hipMalloc testl %eax, %eax jne .LBB1_28 # %bb.29: # %_Z14__cudaSafeCall10hipError_tPKci.exit103 leaq 112(%rsp), %rdi movl $2000, %esi # imm = 0x7D0 callq hipMalloc testl %eax, %eax jne .LBB1_30 # %bb.31: # %_Z14__cudaSafeCall10hipError_tPKci.exit105 leaq 104(%rsp), %rdi movl $2000, %esi # imm = 0x7D0 callq hipMalloc testl %eax, %eax jne .LBB1_32 # %bb.33: # %_Z14__cudaSafeCall10hipError_tPKci.exit107 leaq 96(%rsp), %rdi movq 48(%rsp), %rsi # 8-byte Reload callq hipMalloc testl %eax, %eax jne .LBB1_34 # %bb.35: # %_Z14__cudaSafeCall10hipError_tPKci.exit109 leaq 88(%rsp), %rdi movl $81920, %esi # imm = 0x14000 callq hipMalloc testl %eax, %eax jne .LBB1_36 # %bb.37: # %_Z14__cudaSafeCall10hipError_tPKci.exit111 leaq 80(%rsp), %rdi movl $81920, %esi # imm = 0x14000 callq hipMalloc testl %eax, %eax jne .LBB1_38 # %bb.39: # %_Z14__cudaSafeCall10hipError_tPKci.exit113 leaq 72(%rsp), %rdi movl $81920, %esi # imm = 0x14000 callq hipMalloc testl %eax, %eax jne .LBB1_40 # %bb.41: # %_Z14__cudaSafeCall10hipError_tPKci.exit115 leaq 64(%rsp), %rdi movl $2000, %esi # imm = 0x7D0 callq hipMalloc testl %eax, %eax jne .LBB1_42 # %bb.43: # %_Z14__cudaSafeCall10hipError_tPKci.exit117 movq 16(%rsp), %rdi movl $81920, %edx # imm = 0x14000 movq 40(%rsp), %rsi # 8-byte Reload movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_44 # %bb.45: # %_Z14__cudaSafeCall10hipError_tPKci.exit119 movq 8(%rsp), %rdi movl $81920, %edx # imm = 0x14000 movq %rbx, %rsi movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_46 # %bb.47: # %_Z14__cudaSafeCall10hipError_tPKci.exit121 movq 144(%rsp), %rdi movl $81920, %edx # imm = 0x14000 movq %r12, %rsi movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_48 # %bb.49: # %_Z14__cudaSafeCall10hipError_tPKci.exit123 movq 136(%rsp), %rdi movl $2000, %edx # imm = 0x7D0 movq 192(%rsp), %rsi # 8-byte Reload movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_50 # %bb.51: # %_Z14__cudaSafeCall10hipError_tPKci.exit125 movq 128(%rsp), %rdi movl $2000, %edx # imm = 0x7D0 movq 184(%rsp), %rsi # 8-byte Reload movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_52 # %bb.53: # %_Z14__cudaSafeCall10hipError_tPKci.exit127 movq 120(%rsp), %rdi movl $2000, %edx # imm = 0x7D0 movq 176(%rsp), %rsi # 8-byte Reload movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_54 # %bb.55: # %_Z14__cudaSafeCall10hipError_tPKci.exit129 movq 112(%rsp), %rdi movl $2000, %edx # imm = 0x7D0 movq 168(%rsp), %rsi # 8-byte Reload movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_56 # %bb.57: # %_Z14__cudaSafeCall10hipError_tPKci.exit131 movq 104(%rsp), %rdi movl $2000, %edx # imm = 0x7D0 movq 160(%rsp), %rsi # 8-byte Reload movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_58 # %bb.59: # %_Z14__cudaSafeCall10hipError_tPKci.exit133 movq 96(%rsp), %rdi movq 152(%rsp), %rsi # 8-byte Reload movq 48(%rsp), %rdx # 8-byte Reload movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_60 # %bb.61: # %_Z14__cudaSafeCall10hipError_tPKci.exit135 movq 88(%rsp), %rdi movl $81920, %edx # imm = 0x14000 movq %r15, %rsi movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_62 # %bb.63: # %_Z14__cudaSafeCall10hipError_tPKci.exit137 movq 80(%rsp), %rdi movl $81920, %edx # imm = 0x14000 movq %r14, %rsi movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_64 # %bb.65: # %_Z14__cudaSafeCall10hipError_tPKci.exit139 movq 72(%rsp), %rdi movl $81920, %edx # imm = 0x14000 movq %r13, %rsi movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_66 # %bb.67: # %_Z14__cudaSafeCall10hipError_tPKci.exit141 movq 64(%rsp), %rdi movl $2000, %edx # imm = 0x7D0 movq %rbp, %rsi movl $1, %ecx callq hipMemcpy testl %eax, %eax jne .LBB1_68 # %bb.69: # %_Z14__cudaSafeCall10hipError_tPKci.exit143 movl $.Lstr.3, %edi callq puts@PLT leaq 56(%rsp), %rdi callq hipEventCreate movq %rsp, %rdi callq hipEventCreate movl $.Lstr.4, %edi callq puts@PLT movq 56(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movabsq $4294967328, %rdi # imm = 0x100000020 leaq 608(%rdi), %rdx movl $1, %esi movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax movq 40(%rsp), %r13 # 8-byte Reload jne .LBB1_71 # %bb.70: movq 16(%rsp), %rax movq 8(%rsp), %rcx movq 144(%rsp), %rdx movq 136(%rsp), %rsi movq 128(%rsp), %rdi movq 120(%rsp), %r8 movq 112(%rsp), %r9 movq 104(%rsp), %r10 movq 96(%rsp), %r11 movq 88(%rsp), %rbx movq 80(%rsp), %r14 movq 72(%rsp), %r15 movq 64(%rsp), %r12 movl $0, 28(%rsp) movl $1000000, 24(%rsp) # imm = 0xF4240 movq %rax, 344(%rsp) movq %rcx, 336(%rsp) movq %rdx, 328(%rsp) movq %rsi, 320(%rsp) movq %rdi, 312(%rsp) movq %r8, 304(%rsp) movq %r9, 296(%rsp) movq %r10, 288(%rsp) movq %r11, 280(%rsp) movq %rbx, 272(%rsp) movq 32(%rsp), %rbx # 8-byte Reload movq %r14, 264(%rsp) movq %r15, 256(%rsp) movq %r12, 248(%rsp) leaq 28(%rsp), %rax movq %rax, 352(%rsp) leaq 24(%rsp), %rax movq %rax, 360(%rsp) leaq 344(%rsp), %rax movq %rax, 368(%rsp) leaq 336(%rsp), %rax movq %rax, 376(%rsp) leaq 328(%rsp), %rax movq %rax, 384(%rsp) leaq 320(%rsp), %rax movq %rax, 392(%rsp) leaq 312(%rsp), %rax movq %rax, 400(%rsp) leaq 304(%rsp), %rax movq %rax, 408(%rsp) leaq 296(%rsp), %rax movq %rax, 416(%rsp) leaq 288(%rsp), %rax movq %rax, 424(%rsp) leaq 280(%rsp), %rax movq %rax, 432(%rsp) leaq 272(%rsp), %rax movq %rax, 440(%rsp) leaq 264(%rsp), %rax movq %rax, 448(%rsp) leaq 256(%rsp), %rax movq %rax, 456(%rsp) leaq 248(%rsp), %rax movq %rax, 464(%rsp) leaq 232(%rsp), %rdi leaq 216(%rsp), %rsi leaq 208(%rsp), %rdx leaq 200(%rsp), %rcx callq __hipPopCallConfiguration movq 232(%rsp), %rsi movl 240(%rsp), %edx movq 216(%rsp), %rcx movl 224(%rsp), %r8d leaq 352(%rsp), %r9 movl $_Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_, %edi pushq 200(%rsp) .cfi_adjust_cfa_offset 8 pushq 216(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_71: callq hipGetLastError testl %eax, %eax jne .LBB1_72 # %bb.75: callq hipDeviceSynchronize testl %eax, %eax jne .LBB1_76 # %bb.77: # %_Z16__cudaCheckErrorPKci.exit movq (%rsp), %rdi xorl %esi, %esi callq hipEventRecord movq (%rsp), %rdi callq hipEventSynchronize movl $0, 352(%rsp) movq 56(%rsp), %rsi movq (%rsp), %rdx leaq 352(%rsp), %rdi callq hipEventElapsedTime callq hipGetLastError testl %eax, %eax jne .LBB1_78 # %bb.80: callq hipDeviceSynchronize testl %eax, %eax jne .LBB1_81 # %bb.82: # %_Z16__cudaCheckErrorPKci.exit147 movl $.Lstr.5, %edi callq puts@PLT movq 16(%rsp), %rsi movl $81920, %edx # imm = 0x14000 movq %r13, %rdi movl $2, %ecx callq hipMemcpy movq 8(%rsp), %rsi movl $81920, %edx # imm = 0x14000 movq %rbx, %rdi movl $2, %ecx callq hipMemcpy movss 352(%rsp), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str.7, %edi movb $1, %al callq printf xorl %eax, %eax addq $472, %rsp # imm = 0x1D8 .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 .LBB1_18: .cfi_def_cfa_offset 528 movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $271, %ecx # imm = 0x10F jmp .LBB1_74 .LBB1_20: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $272, %ecx # imm = 0x110 jmp .LBB1_74 .LBB1_22: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $273, %ecx # imm = 0x111 jmp .LBB1_74 .LBB1_24: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $274, %ecx # imm = 0x112 jmp .LBB1_74 .LBB1_26: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $275, %ecx # imm = 0x113 jmp .LBB1_74 .LBB1_28: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $276, %ecx # imm = 0x114 jmp .LBB1_74 .LBB1_30: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $277, %ecx # imm = 0x115 jmp .LBB1_74 .LBB1_32: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $278, %ecx # imm = 0x116 jmp .LBB1_74 .LBB1_34: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $279, %ecx # imm = 0x117 jmp .LBB1_74 .LBB1_36: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $280, %ecx # imm = 0x118 jmp .LBB1_74 .LBB1_38: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $281, %ecx # imm = 0x119 jmp .LBB1_74 .LBB1_40: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $282, %ecx # imm = 0x11A jmp .LBB1_74 .LBB1_42: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $283, %ecx # imm = 0x11B jmp .LBB1_74 .LBB1_44: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $285, %ecx # imm = 0x11D jmp .LBB1_74 .LBB1_46: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $286, %ecx # imm = 0x11E jmp .LBB1_74 .LBB1_48: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $287, %ecx # imm = 0x11F jmp .LBB1_74 .LBB1_50: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $288, %ecx # imm = 0x120 jmp .LBB1_74 .LBB1_52: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $289, %ecx # imm = 0x121 jmp .LBB1_74 .LBB1_54: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $290, %ecx # imm = 0x122 jmp .LBB1_74 .LBB1_56: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $291, %ecx # imm = 0x123 jmp .LBB1_74 .LBB1_58: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $292, %ecx # imm = 0x124 jmp .LBB1_74 .LBB1_60: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $293, %ecx # imm = 0x125 jmp .LBB1_74 .LBB1_62: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $294, %ecx # imm = 0x126 jmp .LBB1_74 .LBB1_64: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $295, %ecx # imm = 0x127 jmp .LBB1_74 .LBB1_66: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $296, %ecx # imm = 0x128 jmp .LBB1_74 .LBB1_68: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi movl $.L.str.3, %edx movq %rbx, %rdi movl $297, %ecx # imm = 0x129 jmp .LBB1_74 .LBB1_72: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.9, %esi jmp .LBB1_73 .LBB1_76: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.10, %esi .LBB1_73: movl $.L.str.3, %edx movq %rbx, %rdi movl $311, %ecx # imm = 0x137 jmp .LBB1_74 .LBB1_78: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.9, %esi jmp .LBB1_79 .LBB1_81: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.10, %esi .LBB1_79: movl $.L.str.3, %edx movq %rbx, %rdi movl $318, %ecx # imm = 0x13E .LBB1_74: movq %rax, %r8 xorl %eax, %eax callq fprintf 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: pushq %rbx .cfi_def_cfa_offset 16 .cfi_offset %rbx, -16 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), %rbx subq $32, %rsp .cfi_adjust_cfa_offset 32 xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction addq $32, %rsp .cfi_adjust_cfa_offset -32 movl $d_Actor_progress, %esi movl $.L__unnamed_2, %edx movl $.L__unnamed_2, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $d_Actor_street, %esi movl $.L__unnamed_3, %edx movl $.L__unnamed_3, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $d_Car_max_velocity, %esi movl $.L__unnamed_4, %edx movl $.L__unnamed_4, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $d_Street_length, %esi movl $.L__unnamed_5, %edx movl $.L__unnamed_5, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $d_Street_max_velocity, %esi movl $.L__unnamed_6, %edx movl $.L__unnamed_6, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $d_Street_neighbors, %esi movl $.L__unnamed_7, %edx movl $.L__unnamed_7, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $d_Array_Street_size, %esi movl $.L__unnamed_8, %edx movl $.L__unnamed_8, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $d_Array_Street_offset, %esi movl $.L__unnamed_9, %edx movl $.L__unnamed_9, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $d_Array_Street_arrays, %esi movl $.L__unnamed_10, %edx movl $.L__unnamed_10, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $d_input_actor_tag, %esi movl $.L__unnamed_11, %edx movl $.L__unnamed_11, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $d_input_actor_id, %esi movl $.L__unnamed_12, %edx movl $.L__unnamed_12, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $d_jobs, %esi movl $.L__unnamed_13, %edx movl $.L__unnamed_13, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $d_randomn, %esi movl $.L__unnamed_14, %edx movl $.L__unnamed_14, %ecx movl $8, %r9d movq %rbx, %rdi xorl %r8d, %r8d pushq $0 .cfi_adjust_cfa_offset 8 pushq $0 .cfi_adjust_cfa_offset 8 callq __hipRegisterVar addq $16, %rsp .cfi_adjust_cfa_offset -16 movl $__hip_module_dtor, %edi popq %rbx .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 d_Actor_progress,@object # @d_Actor_progress .local d_Actor_progress .comm d_Actor_progress,8,8 .type d_Actor_street,@object # @d_Actor_street .local d_Actor_street .comm d_Actor_street,8,8 .type d_Car_max_velocity,@object # @d_Car_max_velocity .local d_Car_max_velocity .comm d_Car_max_velocity,8,8 .type d_Street_length,@object # @d_Street_length .local d_Street_length .comm d_Street_length,8,8 .type d_Street_max_velocity,@object # @d_Street_max_velocity .local d_Street_max_velocity .comm d_Street_max_velocity,8,8 .type d_Street_neighbors,@object # @d_Street_neighbors .local d_Street_neighbors .comm d_Street_neighbors,8,8 .type d_Array_Street_size,@object # @d_Array_Street_size .local d_Array_Street_size .comm d_Array_Street_size,8,8 .type d_Array_Street_offset,@object # @d_Array_Street_offset .local d_Array_Street_offset .comm d_Array_Street_offset,8,8 .type d_Array_Street_arrays,@object # @d_Array_Street_arrays .local d_Array_Street_arrays .comm d_Array_Street_arrays,8,8 .type d_input_actor_tag,@object # @d_input_actor_tag .local d_input_actor_tag .comm d_input_actor_tag,8,8 .type d_input_actor_id,@object # @d_input_actor_id .local d_input_actor_id .comm d_input_actor_id,8,8 .type d_jobs,@object # @d_jobs .local d_jobs .comm d_jobs,8,8 .type d_randomn,@object # @d_randomn .local d_randomn .comm d_randomn,8,8 .type _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_,@object # @_Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .section .rodata,"a",@progbits .globl _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .p2align 3, 0x0 _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_: .quad _Z21__device_stub__kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .size _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_, 8 .type .L.str.3,@object # @.str.3 .section .rodata.str1.1,"aMS",@progbits,1 .L.str.3: .asciz "/home/ubuntu/Datasets/stackv2/train-structured-repos-hip/prg-titech/array2016-paper/master/benchmarks/benchmark_column_reorder.hip" .size .L.str.3, 131 .type .L.str.7,@object # @.str.7 .L.str.7: .asciz "\n\n\nElapsed time millis: %f\n" .size .L.str.7, 28 .type .L.str.8,@object # @.str.8 .L.str.8: .asciz "cudaSafeCall() failed at %s:%i : %s\n" .size .L.str.8, 37 .type .L.str.9,@object # @.str.9 .L.str.9: .asciz "cudaCheckError() failed at %s:%i : %s\n" .size .L.str.9, 39 .type .L.str.10,@object # @.str.10 .L.str.10: .asciz "cudaCheckError() with sync failed at %s:%i : %s\n" .size .L.str.10, 49 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_" .size .L__unnamed_1, 46 .type .L__unnamed_2,@object # @1 .L__unnamed_2: .asciz "d_Actor_progress" .size .L__unnamed_2, 17 .type .L__unnamed_3,@object # @2 .L__unnamed_3: .asciz "d_Actor_street" .size .L__unnamed_3, 15 .type .L__unnamed_4,@object # @3 .L__unnamed_4: .asciz "d_Car_max_velocity" .size .L__unnamed_4, 19 .type .L__unnamed_5,@object # @4 .L__unnamed_5: .asciz "d_Street_length" .size .L__unnamed_5, 16 .type .L__unnamed_6,@object # @5 .L__unnamed_6: .asciz "d_Street_max_velocity" .size .L__unnamed_6, 22 .type .L__unnamed_7,@object # @6 .L__unnamed_7: .asciz "d_Street_neighbors" .size .L__unnamed_7, 19 .type .L__unnamed_8,@object # @7 .L__unnamed_8: .asciz "d_Array_Street_size" .size .L__unnamed_8, 20 .type .L__unnamed_9,@object # @8 .L__unnamed_9: .asciz "d_Array_Street_offset" .size .L__unnamed_9, 22 .type .L__unnamed_10,@object # @9 .L__unnamed_10: .asciz "d_Array_Street_arrays" .size .L__unnamed_10, 22 .type .L__unnamed_11,@object # @10 .L__unnamed_11: .asciz "d_input_actor_tag" .size .L__unnamed_11, 18 .type .L__unnamed_12,@object # @11 .L__unnamed_12: .asciz "d_input_actor_id" .size .L__unnamed_12, 17 .type .L__unnamed_13,@object # @12 .L__unnamed_13: .asciz "d_jobs" .size .L__unnamed_13, 7 .type .L__unnamed_14,@object # @13 .L__unnamed_14: .asciz "d_randomn" .size .L__unnamed_14, 10 .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 "Setting up scenario..." .size .Lstr, 23 .type .Lstr.1,@object # @str.1 .Lstr.1: .asciz "Scenario set up." .size .Lstr.1, 17 .type .Lstr.2,@object # @str.2 .Lstr.2: .asciz "Copying data to GPU..." .size .Lstr.2, 23 .type .Lstr.3,@object # @str.3 .Lstr.3: .asciz "Finished copying data." .size .Lstr.3, 23 .type .Lstr.4,@object # @str.4 .Lstr.4: .asciz "Launching kernel..." .size .Lstr.4, 20 .type .Lstr.5,@object # @str.5 .Lstr.5: .asciz "Kernel finished." .size .Lstr.5, 17 .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__kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym d_Actor_progress .addrsig_sym d_Actor_street .addrsig_sym d_Car_max_velocity .addrsig_sym d_Street_length .addrsig_sym d_Street_max_velocity .addrsig_sym d_Street_neighbors .addrsig_sym d_Array_Street_size .addrsig_sym d_Array_Street_offset .addrsig_sym d_Array_Street_arrays .addrsig_sym d_input_actor_tag .addrsig_sym d_input_actor_id .addrsig_sym d_jobs .addrsig_sym d_randomn .addrsig_sym _Z6kerneliiPfPiS_S_S_S0_S0_S0_S0_S0_S0_S0_S0_ .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 <complex> #include <cmath> #include <iomanip> #include <string> #include <fstream> using namespace std; __global__ void z_funct(double * d_mat_re, double * d_mat_im, int *d_img, int nb_ite) { int i = blockDim.x * blockIdx.x + threadIdx.x; double c_re = d_mat_re[i]; double c_im = d_mat_im[i]; double z_re = 0, z_im =0; double z_re_tmp = 0; int t=0; while ((z_re*z_re+z_im*z_im <= 4) && (t<nb_ite)) { z_re_tmp = z_re*z_re - z_im*z_im + c_re; z_im = 2*z_im*z_re + c_im; z_re = z_re_tmp; t++; } d_img[i] = t; } int main(int argc, char * argv[]) { if (argc == 1) { cout << "Help" << endl << " 8 args :" << endl << endl << "arg1 = min real part | arg2 = min imaginary part " << endl << "arg3 = max real part | arg4 = max imaginary part " << endl << "arg5 = number of points on the real axe | arg6 = number of points on the imaginary axe " << endl << "arg7 = nb of iterations | arg8 = limit convergence" << endl << endl << " 4 args :" << endl << endl << "arg1 = number of points on the real axe | arg2 = number of points on the imaginary axe " << endl << "arg3 = nb of iterations " << endl << endl ; return 1; } double max_re, max_im, min_re, min_im; int nb_pts_re, nb_pts_im, nb_ite; if (argc == 8) { try { min_re = stod(argv[1]); min_im = stod(argv[2]); max_re = stod(argv[3]); max_im = stod(argv[4]); nb_pts_re = stoi(argv[5]); nb_pts_im = stoi(argv[6]); nb_ite = stoi(argv[7]); } catch (...) { cout << "Bad Args : see help (type nameofprogram without args)" << endl << endl; return 1; } } if (argc == 4 ) { min_re = -2; min_im = -1; max_re = 1; max_im = 1; try { nb_pts_re = stoi(argv[1]); nb_pts_im = stoi(argv[2]); nb_ite = stoi(argv[3]); } catch (...) { cout << "Bad Args : see help (type nameofprogram without args)" << endl << endl; return 1; } } cout << "Initializing..." << endl; int size_d = sizeof(double)*nb_pts_re*nb_pts_im; int size_i = sizeof(int)*nb_pts_re*nb_pts_im; double * mat_re = (double *)malloc(size_d); double * mat_im = (double *)malloc(size_d); double re, im; for (int i=0; i<nb_pts_im; i++) { im = max_im - (max_im-min_im)/nb_pts_im*i; for (int j=0; j<nb_pts_re; j++) { re = max_re - (max_re-min_re)/nb_pts_re*j; mat_re[i*nb_pts_re+j] = re; mat_im[i*nb_pts_re+j] = im; } } double *d_mat_re, *d_mat_im; int *d_img; cudaMalloc(&d_mat_re,size_d); cudaMalloc(&d_mat_im,size_d); cudaMalloc(&d_img,size_i); cudaMemcpy(d_mat_re, mat_re, size_d, cudaMemcpyHostToDevice); cudaMemcpy(d_mat_im, mat_im, size_d, cudaMemcpyHostToDevice); cout << "Running on GPU..." << endl; dim3 blockDim = 1024; dim3 gridDim = (nb_pts_re*nb_pts_im)/1024 + 1; z_funct<<<gridDim,blockDim>>>(d_mat_re, d_mat_im, d_img, nb_ite); cout << "Fetching datas..." << endl; int * img = (int*) malloc(size_i); cudaMemcpy(mat_re, d_mat_re, size_d, cudaMemcpyDeviceToHost); cudaMemcpy(mat_im, d_mat_im, size_d, cudaMemcpyDeviceToHost); cudaMemcpy(img, d_img, size_i, cudaMemcpyDeviceToHost); cout << "Writing on the disk..." << endl; ofstream file; file.open("conv"); for (int i=0; i<nb_pts_im; i++) { for (int j=0; j<nb_pts_re; j++) { file << setw(15) << img[i*nb_pts_re+j] ; } file << endl ; } file.close(); /* file.open("val"); for (int i=0; i<nb_pts_im; i++) { for (int j=0; j<nb_pts_re; j++) { file << setw(15) << pow(mat_re[i*nb_pts_re+j],2) + pow(mat_im[i*nb_pts_re+j],2) ; } file << endl ; } file.close(); */ cout << "Done !" << endl; return 0; }
code for sm_80 Function : _Z7z_functPdS_Pii .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 R6, SR_CTAID.X ; /* 0x0000000000067919 */ /* 0x000e220000002500 */ /*0020*/ IMAD.MOV.U32 R0, RZ, RZ, c[0x0][0x178] ; /* 0x00005e00ff007624 */ /* 0x000fe200078e00ff */ /*0030*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*0040*/ HFMA2.MMA R19, -RZ, RZ, 0, 0 ; /* 0x00000000ff137435 */ /* 0x000fe200000001ff */ /*0050*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e240000002100 */ /*0060*/ ISETP.GE.AND P0, PT, R0, 0x1, PT ; /* 0x000000010000780c */ /* 0x000fe20003f06270 */ /*0070*/ IMAD R6, R6, c[0x0][0x0], R3 ; /* 0x0000000006067a24 */ /* 0x001fca00078e0203 */ /*0080*/ SHF.R.S32.HI R3, RZ, 0x1f, R6 ; /* 0x0000001fff037819 */ /* 0x000fe40000011406 */ /*0090*/ LEA R2, P1, R6, c[0x0][0x170], 0x2 ; /* 0x00005c0006027a11 */ /* 0x000fc800078210ff */ /*00a0*/ LEA.HI.X R3, R6, c[0x0][0x174], R3, 0x2, P1 ; /* 0x00005d0006037a11 */ /* 0x000fe200008f1403 */ /*00b0*/ @!P0 BRA 0x250 ; /* 0x0000019000008947 */ /* 0x000fea0003800000 */ /*00c0*/ IMAD.MOV.U32 R7, RZ, RZ, 0x8 ; /* 0x00000008ff077424 */ /* 0x000fc800078e00ff */ /*00d0*/ IMAD.WIDE R4, R6, R7, c[0x0][0x160] ; /* 0x0000580006047625 */ /* 0x000fc800078e0207 */ /*00e0*/ IMAD.WIDE R6, R6, R7, c[0x0][0x168] ; /* 0x00005a0006067625 */ /* 0x000fe400078e0207 */ /*00f0*/ LDG.E.64 R4, [R4.64] ; /* 0x0000000404047981 */ /* 0x000f68000c1e1b00 */ /*0100*/ LDG.E.64 R6, [R6.64] ; /* 0x0000000406067981 */ /* 0x000f62000c1e1b00 */ /*0110*/ BSSY B0, 0x250 ; /* 0x0000013000007945 */ /* 0x000fe20003800000 */ /*0120*/ MOV R19, RZ ; /* 0x000000ff00137202 */ /* 0x000fe20000000f00 */ /*0130*/ CS2R R8, SRZ ; /* 0x0000000000087805 */ /* 0x000fe2000001ff00 */ /*0140*/ CS2R R10, SRZ ; /* 0x00000000000a7805 */ /* 0x000fe2000001ff00 */ /*0150*/ CS2R R12, SRZ ; /* 0x00000000000c7805 */ /* 0x000fe2000001ff00 */ /*0160*/ CS2R R14, SRZ ; /* 0x00000000000e7805 */ /* 0x000fc8000001ff00 */ /*0170*/ DADD R8, R10, -R8 ; /* 0x000000000a087229 */ /* 0x000e220000000808 */ /*0180*/ IADD3 R19, R19, 0x1, RZ ; /* 0x0000000113137810 */ /* 0x000fc60007ffe0ff */ /*0190*/ DADD R12, R12, R12 ; /* 0x000000000c0c7229 */ /* 0x000e62000000000c */ /*01a0*/ ISETP.GE.AND P0, PT, R19, c[0x0][0x178], PT ; /* 0x00005e0013007a0c */ /* 0x000fc60003f06270 */ /*01b0*/ DADD R16, R4, R8 ; /* 0x0000000004107229 */ /* 0x021e080000000008 */ /*01c0*/ DFMA R12, R12, R14, R6 ; /* 0x0000000e0c0c722b */ /* 0x002e480000000006 */ /*01d0*/ DMUL R10, R16, R16 ; /* 0x00000010100a7228 */ /* 0x001fc80000000000 */ /*01e0*/ DMUL R8, R12, R12 ; /* 0x0000000c0c087228 */ /* 0x002e0c0000000000 */ /*01f0*/ DADD R14, R10, R8 ; /* 0x000000000a0e7229 */ /* 0x001e0c0000000008 */ /*0200*/ DSETP.LE.AND P1, PT, R14, 4, PT ; /* 0x401000000e00742a */ /* 0x0010640003f23000 */ /*0210*/ IMAD.MOV.U32 R14, RZ, RZ, R16 ; /* 0x000000ffff0e7224 */ /* 0x001fe200078e0010 */ /*0220*/ MOV R15, R17 ; /* 0x00000011000f7202 */ /* 0x000fd60000000f00 */ /*0230*/ @!P0 BRA P1, 0x170 ; /* 0xffffff3000008947 */ /* 0x002fea000083ffff */ /*0240*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0250*/ STG.E [R2.64], R19 ; /* 0x0000001302007986 */ /* 0x000fe2000c101904 */ /*0260*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0270*/ BRA 0x270; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0280*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0290*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*02a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*02b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*02c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*02d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*02e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*02f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include <iostream> #include <complex> #include <cmath> #include <iomanip> #include <string> #include <fstream> using namespace std; __global__ void z_funct(double * d_mat_re, double * d_mat_im, int *d_img, int nb_ite) { int i = blockDim.x * blockIdx.x + threadIdx.x; double c_re = d_mat_re[i]; double c_im = d_mat_im[i]; double z_re = 0, z_im =0; double z_re_tmp = 0; int t=0; while ((z_re*z_re+z_im*z_im <= 4) && (t<nb_ite)) { z_re_tmp = z_re*z_re - z_im*z_im + c_re; z_im = 2*z_im*z_re + c_im; z_re = z_re_tmp; t++; } d_img[i] = t; } int main(int argc, char * argv[]) { if (argc == 1) { cout << "Help" << endl << " 8 args :" << endl << endl << "arg1 = min real part | arg2 = min imaginary part " << endl << "arg3 = max real part | arg4 = max imaginary part " << endl << "arg5 = number of points on the real axe | arg6 = number of points on the imaginary axe " << endl << "arg7 = nb of iterations | arg8 = limit convergence" << endl << endl << " 4 args :" << endl << endl << "arg1 = number of points on the real axe | arg2 = number of points on the imaginary axe " << endl << "arg3 = nb of iterations " << endl << endl ; return 1; } double max_re, max_im, min_re, min_im; int nb_pts_re, nb_pts_im, nb_ite; if (argc == 8) { try { min_re = stod(argv[1]); min_im = stod(argv[2]); max_re = stod(argv[3]); max_im = stod(argv[4]); nb_pts_re = stoi(argv[5]); nb_pts_im = stoi(argv[6]); nb_ite = stoi(argv[7]); } catch (...) { cout << "Bad Args : see help (type nameofprogram without args)" << endl << endl; return 1; } } if (argc == 4 ) { min_re = -2; min_im = -1; max_re = 1; max_im = 1; try { nb_pts_re = stoi(argv[1]); nb_pts_im = stoi(argv[2]); nb_ite = stoi(argv[3]); } catch (...) { cout << "Bad Args : see help (type nameofprogram without args)" << endl << endl; return 1; } } cout << "Initializing..." << endl; int size_d = sizeof(double)*nb_pts_re*nb_pts_im; int size_i = sizeof(int)*nb_pts_re*nb_pts_im; double * mat_re = (double *)malloc(size_d); double * mat_im = (double *)malloc(size_d); double re, im; for (int i=0; i<nb_pts_im; i++) { im = max_im - (max_im-min_im)/nb_pts_im*i; for (int j=0; j<nb_pts_re; j++) { re = max_re - (max_re-min_re)/nb_pts_re*j; mat_re[i*nb_pts_re+j] = re; mat_im[i*nb_pts_re+j] = im; } } double *d_mat_re, *d_mat_im; int *d_img; cudaMalloc(&d_mat_re,size_d); cudaMalloc(&d_mat_im,size_d); cudaMalloc(&d_img,size_i); cudaMemcpy(d_mat_re, mat_re, size_d, cudaMemcpyHostToDevice); cudaMemcpy(d_mat_im, mat_im, size_d, cudaMemcpyHostToDevice); cout << "Running on GPU..." << endl; dim3 blockDim = 1024; dim3 gridDim = (nb_pts_re*nb_pts_im)/1024 + 1; z_funct<<<gridDim,blockDim>>>(d_mat_re, d_mat_im, d_img, nb_ite); cout << "Fetching datas..." << endl; int * img = (int*) malloc(size_i); cudaMemcpy(mat_re, d_mat_re, size_d, cudaMemcpyDeviceToHost); cudaMemcpy(mat_im, d_mat_im, size_d, cudaMemcpyDeviceToHost); cudaMemcpy(img, d_img, size_i, cudaMemcpyDeviceToHost); cout << "Writing on the disk..." << endl; ofstream file; file.open("conv"); for (int i=0; i<nb_pts_im; i++) { for (int j=0; j<nb_pts_re; j++) { file << setw(15) << img[i*nb_pts_re+j] ; } file << endl ; } file.close(); /* file.open("val"); for (int i=0; i<nb_pts_im; i++) { for (int j=0; j<nb_pts_re; j++) { file << setw(15) << pow(mat_re[i*nb_pts_re+j],2) + pow(mat_im[i*nb_pts_re+j],2) ; } file << endl ; } file.close(); */ cout << "Done !" << endl; return 0; }
.file "tmpxft_0009ed35_00000000-6_cu_mandelbrot.cudafe1.cpp" .text #APP .globl _ZSt21ios_base_library_initv #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB4285: .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 .LFE4285: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z31__device_stub__Z7z_functPdS_PiiPdS_Pii .type _Z31__device_stub__Z7z_functPdS_PiiPdS_Pii, @function _Z31__device_stub__Z7z_functPdS_PiiPdS_Pii: .LFB4307: .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 .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 _Z7z_functPdS_Pii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE4307: .size _Z31__device_stub__Z7z_functPdS_PiiPdS_Pii, .-_Z31__device_stub__Z7z_functPdS_PiiPdS_Pii .globl _Z7z_functPdS_Pii .type _Z7z_functPdS_Pii, @function _Z7z_functPdS_Pii: .LFB4308: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z31__device_stub__Z7z_functPdS_PiiPdS_Pii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE4308: .size _Z7z_functPdS_Pii, .-_Z7z_functPdS_Pii .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "_Z7z_functPdS_Pii" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB4310: .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 _Z7z_functPdS_Pii(%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 .LFE4310: .size _ZL24__sti____cudaRegisterAllv, .-_ZL24__sti____cudaRegisterAllv .section .init_array,"aw" .align 8 .quad _ZL24__sti____cudaRegisterAllv .section .text._ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_,"axG",@progbits,_ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_,comdat .weak _ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .type _ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_, @function _ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_: .LFB4384: .cfi_startproc .cfi_personality 0x9b,DW.ref.__gxx_personality_v0 .cfi_lsda 0x1b,.LLSDA4384 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, %r13 movq %rsi, 8(%rsp) movq %rdx, %rbp movq %rcx, %r12 movl %r8d, %r14d movq %fs:40, %rax movq %rax, 24(%rsp) xorl %eax, %eax call __errno_location@PLT movq %rax, %rbx movl (%rax), %r15d movl $0, (%rax) leaq 16(%rsp), %rsi movl %r14d, %edx movq %rbp, %rdi .LEHB0: call *%r13 movq 16(%rsp), %rcx cmpq %rbp, %rcx je .L27 cmpl $34, (%rbx) je .L16 movl $2147483648, %edx addq %rax, %rdx shrq $32, %rdx jne .L16 testq %r12, %r12 je .L19 subq %rbp, %rcx movq %rcx, (%r12) .L19: cmpl $0, (%rbx) jne .L13 movl %r15d, (%rbx) .L13: movq 24(%rsp), %rdx subq %fs:40, %rdx jne .L28 addq $40, %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 .L27: .cfi_restore_state movq 24(%rsp), %rax subq %fs:40, %rax jne .L29 movq 8(%rsp), %rdi call _ZSt24__throw_invalid_argumentPKc@PLT .L29: call __stack_chk_fail@PLT .L16: movq 24(%rsp), %rax subq %fs:40, %rax jne .L30 movq 8(%rsp), %rdi call _ZSt20__throw_out_of_rangePKc@PLT .LEHE0: .L25: endbr64 movq %rax, %rdi cmpl $0, (%rbx) jne .L22 movl %r15d, (%rbx) .L22: movq 24(%rsp), %rax subq %fs:40, %rax je .L23 call __stack_chk_fail@PLT .L30: call __stack_chk_fail@PLT .L23: .LEHB1: call _Unwind_Resume@PLT .LEHE1: .L28: call __stack_chk_fail@PLT .cfi_endproc .LFE4384: .globl __gxx_personality_v0 .section .gcc_except_table._ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_,"aG",@progbits,_ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_,comdat .LLSDA4384: .byte 0xff .byte 0xff .byte 0x1 .uleb128 .LLSDACSE4384-.LLSDACSB4384 .LLSDACSB4384: .uleb128 .LEHB0-.LFB4384 .uleb128 .LEHE0-.LEHB0 .uleb128 .L25-.LFB4384 .uleb128 0 .uleb128 .LEHB1-.LFB4384 .uleb128 .LEHE1-.LEHB1 .uleb128 0 .uleb128 0 .LLSDACSE4384: .section .text._ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_,"axG",@progbits,_ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_,comdat .size _ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_, .-_ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .section .text._ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_,"axG",@progbits,_ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_,comdat .weak _ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .type _ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_, @function _ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_: .LFB4438: .cfi_startproc .cfi_personality 0x9b,DW.ref.__gxx_personality_v0 .cfi_lsda 0x1b,.LLSDA4438 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 $24, %rsp .cfi_def_cfa_offset 80 movq %rdi, %r15 movq %rsi, %r14 movq %rdx, %rbp movq %rcx, %r12 movq %fs:40, %rax movq %rax, 8(%rsp) xorl %eax, %eax call __errno_location@PLT movq %rax, %rbx movl (%rax), %r13d movl $0, (%rax) movq %rsp, %rsi movq %rbp, %rdi .LEHB2: call *%r15 movq (%rsp), %rax cmpq %rbp, %rax je .L44 cmpl $34, (%rbx) je .L45 testq %r12, %r12 je .L36 subq %rbp, %rax movq %rax, (%r12) .L36: cmpl $0, (%rbx) jne .L31 movl %r13d, (%rbx) .L31: movq 8(%rsp), %rax subq %fs:40, %rax jne .L46 addq $24, %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 .L44: .cfi_restore_state movq 8(%rsp), %rax subq %fs:40, %rax jne .L47 movq %r14, %rdi call _ZSt24__throw_invalid_argumentPKc@PLT .L42: endbr64 movq %rax, %rdi cmpl $0, (%rbx) jne .L39 movl %r13d, (%rbx) .L39: movq 8(%rsp), %rax subq %fs:40, %rax je .L40 call __stack_chk_fail@PLT .L47: call __stack_chk_fail@PLT .L45: movq 8(%rsp), %rax subq %fs:40, %rax jne .L48 movq %r14, %rdi call _ZSt20__throw_out_of_rangePKc@PLT .LEHE2: .L48: call __stack_chk_fail@PLT .L40: .LEHB3: call _Unwind_Resume@PLT .LEHE3: .L46: call __stack_chk_fail@PLT .cfi_endproc .LFE4438: .section .gcc_except_table._ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_,"aG",@progbits,_ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_,comdat .LLSDA4438: .byte 0xff .byte 0xff .byte 0x1 .uleb128 .LLSDACSE4438-.LLSDACSB4438 .LLSDACSB4438: .uleb128 .LEHB2-.LFB4438 .uleb128 .LEHE2-.LEHB2 .uleb128 .L42-.LFB4438 .uleb128 0 .uleb128 .LEHB3-.LFB4438 .uleb128 .LEHE3-.LEHB3 .uleb128 0 .uleb128 0 .LLSDACSE4438: .section .text._ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_,"axG",@progbits,_ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_,comdat .size _ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_, .-_ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .section .rodata._ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC2IS3_EEPKcRKS3_.str1.8,"aMS",@progbits,1 .align 8 .LC1: .string "basic_string: construction from null is not valid" .section .text._ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC2IS3_EEPKcRKS3_,"axG",@progbits,_ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC5IS3_EEPKcRKS3_,comdat .align 2 .weak _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC2IS3_EEPKcRKS3_ .type _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC2IS3_EEPKcRKS3_, @function _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC2IS3_EEPKcRKS3_: .LFB4630: .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 $24, %rsp .cfi_def_cfa_offset 64 movq %fs:40, %rax movq %rax, 8(%rsp) xorl %eax, %eax leaq 16(%rdi), %r12 movq %r12, (%rdi) testq %rsi, %rsi je .L58 movq %rdi, %rbx movq %rsi, %r13 movq %rsi, %rdi call strlen@PLT movq %rax, %rbp movq %rax, (%rsp) cmpq $15, %rax ja .L59 cmpq $1, %rax jne .L54 movzbl 0(%r13), %eax movb %al, 16(%rbx) .L55: movq (%rsp), %rax movq %rax, 8(%rbx) movq (%rbx), %rdx movb $0, (%rdx,%rax) movq 8(%rsp), %rax subq %fs:40, %rax jne .L60 addq $24, %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 .L58: .cfi_restore_state movq 8(%rsp), %rax subq %fs:40, %rax jne .L61 leaq .LC1(%rip), %rdi call _ZSt19__throw_logic_errorPKc@PLT .L61: call __stack_chk_fail@PLT .L59: movq %rsp, %rsi movl $0, %edx movq %rbx, %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE9_M_createERmm@PLT movq %rax, %r12 movq %rax, (%rbx) movq (%rsp), %rax movq %rax, 16(%rbx) .L53: movq %rbp, %rdx movq %r13, %rsi movq %r12, %rdi call memcpy@PLT jmp .L55 .L54: testq %rax, %rax je .L55 jmp .L53 .L60: call __stack_chk_fail@PLT .cfi_endproc .LFE4630: .size _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC2IS3_EEPKcRKS3_, .-_ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC2IS3_EEPKcRKS3_ .weak _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC1IS3_EEPKcRKS3_ .set _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC1IS3_EEPKcRKS3_,_ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC2IS3_EEPKcRKS3_ .section .rodata.str1.1 .LC5: .string "Help" .LC6: .string " 8 args :" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC7: .string "arg1 = min real part | arg2 = min imaginary part " .align 8 .LC8: .string "arg3 = max real part | arg4 = max imaginary part " .align 8 .LC9: .string "arg5 = number of points on the real axe | arg6 = number of points on the imaginary axe " .align 8 .LC10: .string "arg7 = nb of iterations | arg8 = limit convergence" .section .rodata.str1.1 .LC11: .string " 4 args :" .section .rodata.str1.8 .align 8 .LC12: .string "arg1 = number of points on the real axe | arg2 = number of points on the imaginary axe " .section .rodata.str1.1 .LC13: .string "arg3 = nb of iterations " .LC14: .string "stod" .LC15: .string "stoi" .LC16: .string "Initializing..." .LC17: .string "Running on GPU..." .LC18: .string "Fetching datas..." .LC19: .string "Writing on the disk..." .LC20: .string "conv" .LC21: .string "Done !" .section .rodata.str1.8 .align 8 .LC22: .string "Bad Args : see help (type nameofprogram without args)" .text .globl main .type main, @function main: .LFB4282: .cfi_startproc .cfi_personality 0x9b,DW.ref.__gxx_personality_v0 .cfi_lsda 0x1b,.LLSDA4282 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 $632, %rsp .cfi_def_cfa_offset 688 movl %edi, %ebx movq %fs:40, %rax movq %rax, 616(%rsp) xorl %eax, %eax cmpl $1, %edi je .L119 movq %rsi, %r12 cmpl $8, %edi je .L120 cmpl $4, %edi je .L121 .L66: leaq .LC16(%rip), %rsi leaq _ZSt4cout(%rip), %rdi .LEHB4: call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movl 16(%rsp), %r9d movslq %r9d, %r15 movl 20(%rsp), %eax imull %r9d, %eax leal 0(,%rax,4), %ebx sall $3, %eax movslq %eax, %r14 movq %r14, %rdi call malloc@PLT movq %rax, %r12 movq %r14, %rdi call malloc@PLT movq %rax, %r13 movl 20(%rsp), %r10d testl %r10d, %r10d jle .L67 movsd 32(%rsp), %xmm5 movapd %xmm5, %xmm4 subsd 40(%rsp), %xmm4 pxor %xmm0, %xmm0 cvtsi2sdl %r10d, %xmm0 divsd %xmm0, %xmm4 movl 16(%rsp), %r9d movl %r9d, %edi movl $0, %esi movl $0, %ecx .L70: pxor %xmm0, %xmm0 cvtsi2sdl %ecx, %xmm0 mulsd %xmm4, %xmm0 movapd %xmm5, %xmm3 subsd %xmm0, %xmm3 testl %r9d, %r9d jle .L68 pxor %xmm0, %xmm0 cvtsi2sdl %r9d, %xmm0 movsd 8(%rsp), %xmm2 subsd 24(%rsp), %xmm2 divsd %xmm0, %xmm2 movslq %esi, %rdx salq $3, %rdx leaq (%r12,%rdx), %r8 addq %r13, %rdx movl $0, %eax .L69: pxor %xmm0, %xmm0 cvtsi2sdl %eax, %xmm0 mulsd %xmm2, %xmm0 movsd 8(%rsp), %xmm1 subsd %xmm0, %xmm1 movsd %xmm1, (%r8,%rax,8) movsd %xmm3, (%rdx,%rax,8) addq $1, %rax cmpq %rax, %r15 jne .L69 .L68: addl $1, %ecx addl %edi, %esi cmpl %ecx, %r10d jne .L70 .L67: leaq 48(%rsp), %rdi movq %r14, %rsi call cudaMalloc@PLT leaq 56(%rsp), %rdi movq %r14, %rsi call cudaMalloc@PLT movslq %ebx, %rbx leaq 64(%rsp), %rdi movq %rbx, %rsi call cudaMalloc@PLT movl $1, %ecx movq %r14, %rdx movq %r12, %rsi movq 48(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movq %r14, %rdx movq %r13, %rsi movq 56(%rsp), %rdi call cudaMemcpy@PLT leaq .LC17(%rip), %rsi leaq _ZSt4cout(%rip), %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movl $1024, 72(%rsp) movl $1, 76(%rsp) movl $1, 80(%rsp) movl 16(%rsp), %edx movl 20(%rsp), %eax imull %eax, %edx leal 1023(%rdx), %eax testl %edx, %edx cmovns %edx, %eax sarl $10, %eax addl $1, %eax movl %eax, 84(%rsp) movl $1, 88(%rsp) movl $1, 92(%rsp) movl $0, %r9d movl $0, %r8d movq 72(%rsp), %rdx movl $1, %ecx movq 84(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L122 .L71: leaq .LC18(%rip), %rsi leaq _ZSt4cout(%rip), %rbp movq %rbp, %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rbx, %rdi call malloc@PLT movq %rax, 8(%rsp) movl $2, %ecx movq %r14, %rdx movq 48(%rsp), %rsi movq %r12, %rdi call cudaMemcpy@PLT movl $2, %ecx movq %r14, %rdx movq 56(%rsp), %rsi movq %r13, %rdi call cudaMemcpy@PLT movl $2, %ecx movq %rbx, %rdx movq 64(%rsp), %rsi movq 8(%rsp), %rdi call cudaMemcpy@PLT leaq .LC19(%rip), %rsi movq %rbp, %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT leaq 96(%rsp), %rbx movq %rbx, %rdi call _ZNSt14basic_ofstreamIcSt11char_traitsIcEEC1Ev@PLT .LEHE4: movl $16, %edx leaq .LC20(%rip), %rsi movq %rbx, %rdi .LEHB5: call _ZNSt14basic_ofstreamIcSt11char_traitsIcEE4openEPKcSt13_Ios_Openmode@PLT .LEHE5: jmp .L123 .L119: leaq .LC5(%rip), %rsi leaq _ZSt4cout(%rip), %rdi .LEHB6: call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi leaq .LC6(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi leaq .LC7(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi leaq .LC8(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi leaq .LC9(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi leaq .LC10(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi leaq .LC11(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi leaq .LC12(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi leaq .LC13(%rip), %rsi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT .LEHE6: movl %ebx, %eax jmp .L62 .L120: leaq 84(%rsp), %rdx movq 8(%rsi), %rsi leaq 96(%rsp), %rdi .LEHB7: call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC1IS3_EEPKcRKS3_ .LEHE7: movl $0, %ecx movq 96(%rsp), %rdx leaq .LC14(%rip), %rsi movq strtod@GOTPCREL(%rip), %rdi .LEHB8: call _ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .LEHE8: movsd %xmm0, 24(%rsp) leaq 96(%rsp), %rbx movq %rbx, %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT leaq 84(%rsp), %rdx movq 16(%r12), %rsi movq %rbx, %rdi .LEHB9: call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC1IS3_EEPKcRKS3_ .LEHE9: movl $0, %ecx movq 96(%rsp), %rdx leaq .LC14(%rip), %rsi movq strtod@GOTPCREL(%rip), %rdi .LEHB10: call _ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .LEHE10: movsd %xmm0, 40(%rsp) movq %rbx, %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT leaq 84(%rsp), %rdx movq 24(%r12), %rsi movq %rbx, %rdi .LEHB11: call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC1IS3_EEPKcRKS3_ .LEHE11: movl $0, %ecx movq 96(%rsp), %rdx leaq .LC14(%rip), %rsi movq strtod@GOTPCREL(%rip), %rdi .LEHB12: call _ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .LEHE12: movsd %xmm0, 8(%rsp) movq %rbx, %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT leaq 84(%rsp), %rdx movq 32(%r12), %rsi movq %rbx, %rdi .LEHB13: call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC1IS3_EEPKcRKS3_ .LEHE13: movl $0, %ecx movq 96(%rsp), %rdx leaq .LC14(%rip), %rsi movq strtod@GOTPCREL(%rip), %rdi .LEHB14: call _ZN9__gnu_cxx6__stoaIddcJEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .LEHE14: movsd %xmm0, 32(%rsp) movq %rbx, %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT leaq 84(%rsp), %rdx movq 40(%r12), %rsi movq %rbx, %rdi .LEHB15: call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC1IS3_EEPKcRKS3_ .LEHE15: movl $10, %r8d movl $0, %ecx movq 96(%rsp), %rdx leaq .LC15(%rip), %rsi movq __isoc23_strtol@GOTPCREL(%rip), %rdi .LEHB16: call _ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .LEHE16: movl %eax, 16(%rsp) movq %rbx, %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT leaq 84(%rsp), %rdx movq 48(%r12), %rsi movq %rbx, %rdi .LEHB17: call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC1IS3_EEPKcRKS3_ .LEHE17: movl $10, %r8d movl $0, %ecx movq 96(%rsp), %rdx leaq .LC15(%rip), %rsi movq __isoc23_strtol@GOTPCREL(%rip), %rdi .LEHB18: call _ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .LEHE18: movl %eax, 20(%rsp) movq %rbx, %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT leaq 84(%rsp), %rdx movq 56(%r12), %rsi movq %rbx, %rdi .LEHB19: call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC1IS3_EEPKcRKS3_ .LEHE19: movl $10, %r8d movl $0, %ecx movq 96(%rsp), %rdx leaq .LC15(%rip), %rsi movq __isoc23_strtol@GOTPCREL(%rip), %rdi .LEHB20: call _ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .LEHE20: movl %eax, %ebp movq %rbx, %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT jmp .L66 .L121: leaq 84(%rsp), %rdx movq 8(%rsi), %rsi leaq 96(%rsp), %rdi .LEHB21: call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC1IS3_EEPKcRKS3_ .LEHE21: movl $10, %r8d movl $0, %ecx movq 96(%rsp), %rdx leaq .LC15(%rip), %rsi movq __isoc23_strtol@GOTPCREL(%rip), %rdi .LEHB22: call _ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .LEHE22: movl %eax, 16(%rsp) leaq 96(%rsp), %rbx movq %rbx, %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT leaq 84(%rsp), %rdx movq 16(%r12), %rsi movq %rbx, %rdi .LEHB23: call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC1IS3_EEPKcRKS3_ .LEHE23: movl $10, %r8d movl $0, %ecx movq 96(%rsp), %rdx leaq .LC15(%rip), %rsi movq __isoc23_strtol@GOTPCREL(%rip), %rdi .LEHB24: call _ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .LEHE24: movl %eax, 20(%rsp) movq %rbx, %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT leaq 84(%rsp), %rdx movq 24(%r12), %rsi movq %rbx, %rdi .LEHB25: call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEC1IS3_EEPKcRKS3_ .LEHE25: movl $10, %r8d movl $0, %ecx movq 96(%rsp), %rdx leaq .LC15(%rip), %rsi movq __isoc23_strtol@GOTPCREL(%rip), %rdi .LEHB26: call _ZN9__gnu_cxx6__stoaIlicJiEEET0_PFT_PKT1_PPS3_DpT2_EPKcS5_PmS9_ .LEHE26: movl %eax, %ebp movq %rbx, %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT movsd .LC2(%rip), %xmm7 movsd %xmm7, 40(%rsp) movsd .LC3(%rip), %xmm7 movsd %xmm7, 24(%rsp) movsd .LC4(%rip), %xmm7 movsd %xmm7, 32(%rsp) movsd %xmm7, 8(%rsp) jmp .L66 .L122: movl %ebp, %ecx movq 64(%rsp), %rdx movq 56(%rsp), %rsi movq 48(%rsp), %rdi .LEHB27: call _Z31__device_stub__Z7z_functPdS_PiiPdS_Pii .LEHE27: jmp .L71 .L123: cmpl $0, 20(%rsp) jle .L72 movl 16(%rsp), %eax movl %eax, 24(%rsp) movl $0, %r14d movl $0, %r13d movq %rbx, %rbp jmp .L73 .L124: addq $4, %rbx cmpq %r12, %rbx je .L80 .L74: movq 96(%rsp), %rax movq -24(%rax), %rax movq $15, 112(%rsp,%rax) movl (%rbx), %esi movq %rbp, %rdi .LEHB28: call _ZNSolsEi@PLT jmp .L124 .L80: movq 96(%rsp), %rax movq -24(%rax), %rax movq 336(%rsp,%rax), %rbx testq %rbx, %rbx je .L125 cmpb $0, 56(%rbx) je .L77 movzbl 67(%rbx), %esi .L78: movsbl %sil, %esi movq %rbp, %rdi call _ZNSo3putEc@PLT jmp .L126 .L125: movq 616(%rsp), %rax subq %fs:40, %rax jne .L127 call _ZSt16__throw_bad_castv@PLT .L114: endbr64 movq %rax, %rbx leaq 96(%rsp), %rdi call _ZNSt14basic_ofstreamIcSt11char_traitsIcEED1Ev@PLT movq 616(%rsp), %rax subq %fs:40, %rax je .L98 call __stack_chk_fail@PLT .L127: call __stack_chk_fail@PLT .L77: movq %rbx, %rdi call _ZNKSt5ctypeIcE13_M_widen_initEv@PLT movq (%rbx), %rax movl $10, %esi movq %rbx, %rdi call *48(%rax) movl %eax, %esi jmp .L78 .L126: movq %rax, %rdi call _ZNSo5flushEv@PLT addl $1, %r13d movl 24(%rsp), %eax addl %eax, %r14d cmpl %r13d, 20(%rsp) je .L72 .L73: cmpl $0, 16(%rsp) jle .L80 movslq %r14d, %rax movq 8(%rsp), %rcx leaq (%rcx,%rax,4), %rbx addq %r15, %rax leaq (%rcx,%rax,4), %r12 jmp .L74 .L72: leaq 96(%rsp), %rdi call _ZNSt14basic_ofstreamIcSt11char_traitsIcEE5closeEv@PLT leaq .LC21(%rip), %rsi leaq _ZSt4cout(%rip), %rdi call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT .LEHE28: leaq 96(%rsp), %rdi call _ZNSt14basic_ofstreamIcSt11char_traitsIcEED1Ev@PLT movl $0, %eax .L62: movq 616(%rsp), %rdx subq %fs:40, %rdx jne .L128 addq $632, %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 .L100: .cfi_restore_state endbr64 movq %rax, %rbx leaq 96(%rsp), %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT .L82: movq %rbx, %rdi call __cxa_begin_catch@PLT leaq .LC22(%rip), %rsi leaq _ZSt4cout(%rip), %rdi .LEHB29: call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT jmp .L129 .L102: endbr64 movq %rax, %rbx leaq 96(%rsp), %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT jmp .L82 .L103: endbr64 movq %rax, %rbx leaq 96(%rsp), %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT jmp .L82 .L104: endbr64 movq %rax, %rbx leaq 96(%rsp), %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT jmp .L82 .L105: endbr64 movq %rax, %rbx leaq 96(%rsp), %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT jmp .L82 .L106: endbr64 movq %rax, %rbx leaq 96(%rsp), %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT jmp .L82 .L107: endbr64 movq %rax, %rbx leaq 96(%rsp), %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT jmp .L82 .L101: endbr64 movq %rax, %rbx jmp .L82 .L129: movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT .LEHE29: .LEHB30: call __cxa_end_catch@PLT .LEHE30: movl $1, %eax jmp .L62 .L108: endbr64 movq %rax, %rbx call __cxa_end_catch@PLT movq 616(%rsp), %rax subq %fs:40, %rax je .L90 call __stack_chk_fail@PLT .L90: movq %rbx, %rdi .LEHB31: call _Unwind_Resume@PLT .LEHE31: .L109: endbr64 movq %rax, %rbx leaq 96(%rsp), %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT .L92: movq %rbx, %rdi call __cxa_begin_catch@PLT leaq .LC22(%rip), %rsi leaq _ZSt4cout(%rip), %rdi .LEHB32: call _ZStlsISt11char_traitsIcEERSt13basic_ostreamIcT_ES5_PKc@PLT jmp .L130 .L111: endbr64 movq %rax, %rbx leaq 96(%rsp), %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT jmp .L92 .L112: endbr64 movq %rax, %rbx leaq 96(%rsp), %rdi call _ZNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEE10_M_disposeEv@PLT jmp .L92 .L110: endbr64 movq %rax, %rbx jmp .L92 .L130: movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT movq %rax, %rdi call _ZSt4endlIcSt11char_traitsIcEERSt13basic_ostreamIT_T0_ES6_@PLT .LEHE32: .LEHB33: call __cxa_end_catch@PLT .LEHE33: movl $1, %eax jmp .L62 .L113: endbr64 movq %rax, %rbx call __cxa_end_catch@PLT movq 616(%rsp), %rax subq %fs:40, %rax je .L96 call __stack_chk_fail@PLT .L96: movq %rbx, %rdi .LEHB34: call _Unwind_Resume@PLT .L98: movq %rbx, %rdi call _Unwind_Resume@PLT .LEHE34: .L128: call __stack_chk_fail@PLT .cfi_endproc .LFE4282: .section .gcc_except_table,"a",@progbits .align 4 .LLSDA4282: .byte 0xff .byte 0x9b .uleb128 .LLSDATT4282-.LLSDATTD4282 .LLSDATTD4282: .byte 0x1 .uleb128 .LLSDACSE4282-.LLSDACSB4282 .LLSDACSB4282: .uleb128 .LEHB4-.LFB4282 .uleb128 .LEHE4-.LEHB4 .uleb128 0 .uleb128 0 .uleb128 .LEHB5-.LFB4282 .uleb128 .LEHE5-.LEHB5 .uleb128 .L114-.LFB4282 .uleb128 0 .uleb128 .LEHB6-.LFB4282 .uleb128 .LEHE6-.LEHB6 .uleb128 0 .uleb128 0 .uleb128 .LEHB7-.LFB4282 .uleb128 .LEHE7-.LEHB7 .uleb128 .L101-.LFB4282 .uleb128 0x1 .uleb128 .LEHB8-.LFB4282 .uleb128 .LEHE8-.LEHB8 .uleb128 .L100-.LFB4282 .uleb128 0x3 .uleb128 .LEHB9-.LFB4282 .uleb128 .LEHE9-.LEHB9 .uleb128 .L101-.LFB4282 .uleb128 0x1 .uleb128 .LEHB10-.LFB4282 .uleb128 .LEHE10-.LEHB10 .uleb128 .L102-.LFB4282 .uleb128 0x3 .uleb128 .LEHB11-.LFB4282 .uleb128 .LEHE11-.LEHB11 .uleb128 .L101-.LFB4282 .uleb128 0x1 .uleb128 .LEHB12-.LFB4282 .uleb128 .LEHE12-.LEHB12 .uleb128 .L103-.LFB4282 .uleb128 0x3 .uleb128 .LEHB13-.LFB4282 .uleb128 .LEHE13-.LEHB13 .uleb128 .L101-.LFB4282 .uleb128 0x1 .uleb128 .LEHB14-.LFB4282 .uleb128 .LEHE14-.LEHB14 .uleb128 .L104-.LFB4282 .uleb128 0x3 .uleb128 .LEHB15-.LFB4282 .uleb128 .LEHE15-.LEHB15 .uleb128 .L101-.LFB4282 .uleb128 0x1 .uleb128 .LEHB16-.LFB4282 .uleb128 .LEHE16-.LEHB16 .uleb128 .L105-.LFB4282 .uleb128 0x3 .uleb128 .LEHB17-.LFB4282 .uleb128 .LEHE17-.LEHB17 .uleb128 .L101-.LFB4282 .uleb128 0x1 .uleb128 .LEHB18-.LFB4282 .uleb128 .LEHE18-.LEHB18 .uleb128 .L106-.LFB4282 .uleb128 0x3 .uleb128 .LEHB19-.LFB4282 .uleb128 .LEHE19-.LEHB19 .uleb128 .L101-.LFB4282 .uleb128 0x1 .uleb128 .LEHB20-.LFB4282 .uleb128 .LEHE20-.LEHB20 .uleb128 .L107-.LFB4282 .uleb128 0x3 .uleb128 .LEHB21-.LFB4282 .uleb128 .LEHE21-.LEHB21 .uleb128 .L110-.LFB4282 .uleb128 0x1 .uleb128 .LEHB22-.LFB4282 .uleb128 .LEHE22-.LEHB22 .uleb128 .L109-.LFB4282 .uleb128 0x3 .uleb128 .LEHB23-.LFB4282 .uleb128 .LEHE23-.LEHB23 .uleb128 .L110-.LFB4282 .uleb128 0x1 .uleb128 .LEHB24-.LFB4282 .uleb128 .LEHE24-.LEHB24 .uleb128 .L111-.LFB4282 .uleb128 0x3 .uleb128 .LEHB25-.LFB4282 .uleb128 .LEHE25-.LEHB25 .uleb128 .L110-.LFB4282 .uleb128 0x1 .uleb128 .LEHB26-.LFB4282 .uleb128 .LEHE26-.LEHB26 .uleb128 .L112-.LFB4282 .uleb128 0x3 .uleb128 .LEHB27-.LFB4282 .uleb128 .LEHE27-.LEHB27 .uleb128 0 .uleb128 0 .uleb128 .LEHB28-.LFB4282 .uleb128 .LEHE28-.LEHB28 .uleb128 .L114-.LFB4282 .uleb128 0 .uleb128 .LEHB29-.LFB4282 .uleb128 .LEHE29-.LEHB29 .uleb128 .L108-.LFB4282 .uleb128 0 .uleb128 .LEHB30-.LFB4282 .uleb128 .LEHE30-.LEHB30 .uleb128 0 .uleb128 0 .uleb128 .LEHB31-.LFB4282 .uleb128 .LEHE31-.LEHB31 .uleb128 0 .uleb128 0 .uleb128 .LEHB32-.LFB4282 .uleb128 .LEHE32-.LEHB32 .uleb128 .L113-.LFB4282 .uleb128 0 .uleb128 .LEHB33-.LFB4282 .uleb128 .LEHE33-.LEHB33 .uleb128 0 .uleb128 0 .uleb128 .LEHB34-.LFB4282 .uleb128 .LEHE34-.LEHB34 .uleb128 0 .uleb128 0 .LLSDACSE4282: .byte 0x1 .byte 0 .byte 0 .byte 0x7d .align 4 .long 0 .LLSDATT4282: .text .size main, .-main .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 .LC2: .long 0 .long -1074790400 .align 8 .LC3: .long 0 .long -1073741824 .align 8 .LC4: .long 0 .long 1072693248 .hidden DW.ref.__gxx_personality_v0 .weak DW.ref.__gxx_personality_v0 .section .data.rel.local.DW.ref.__gxx_personality_v0,"awG",@progbits,DW.ref.__gxx_personality_v0,comdat .align 8 .type DW.ref.__gxx_personality_v0, @object .size DW.ref.__gxx_personality_v0, 8 DW.ref.__gxx_personality_v0: .quad __gxx_personality_v0 .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 <complex> #include <cmath> #include <iomanip> #include <string> #include <fstream> using namespace std; __global__ void z_funct(double * d_mat_re, double * d_mat_im, int *d_img, int nb_ite) { int i = blockDim.x * blockIdx.x + threadIdx.x; double c_re = d_mat_re[i]; double c_im = d_mat_im[i]; double z_re = 0, z_im =0; double z_re_tmp = 0; int t=0; while ((z_re*z_re+z_im*z_im <= 4) && (t<nb_ite)) { z_re_tmp = z_re*z_re - z_im*z_im + c_re; z_im = 2*z_im*z_re + c_im; z_re = z_re_tmp; t++; } d_img[i] = t; } int main(int argc, char * argv[]) { if (argc == 1) { cout << "Help" << endl << " 8 args :" << endl << endl << "arg1 = min real part | arg2 = min imaginary part " << endl << "arg3 = max real part | arg4 = max imaginary part " << endl << "arg5 = number of points on the real axe | arg6 = number of points on the imaginary axe " << endl << "arg7 = nb of iterations | arg8 = limit convergence" << endl << endl << " 4 args :" << endl << endl << "arg1 = number of points on the real axe | arg2 = number of points on the imaginary axe " << endl << "arg3 = nb of iterations " << endl << endl ; return 1; } double max_re, max_im, min_re, min_im; int nb_pts_re, nb_pts_im, nb_ite; if (argc == 8) { try { min_re = stod(argv[1]); min_im = stod(argv[2]); max_re = stod(argv[3]); max_im = stod(argv[4]); nb_pts_re = stoi(argv[5]); nb_pts_im = stoi(argv[6]); nb_ite = stoi(argv[7]); } catch (...) { cout << "Bad Args : see help (type nameofprogram without args)" << endl << endl; return 1; } } if (argc == 4 ) { min_re = -2; min_im = -1; max_re = 1; max_im = 1; try { nb_pts_re = stoi(argv[1]); nb_pts_im = stoi(argv[2]); nb_ite = stoi(argv[3]); } catch (...) { cout << "Bad Args : see help (type nameofprogram without args)" << endl << endl; return 1; } } cout << "Initializing..." << endl; int size_d = sizeof(double)*nb_pts_re*nb_pts_im; int size_i = sizeof(int)*nb_pts_re*nb_pts_im; double * mat_re = (double *)malloc(size_d); double * mat_im = (double *)malloc(size_d); double re, im; for (int i=0; i<nb_pts_im; i++) { im = max_im - (max_im-min_im)/nb_pts_im*i; for (int j=0; j<nb_pts_re; j++) { re = max_re - (max_re-min_re)/nb_pts_re*j; mat_re[i*nb_pts_re+j] = re; mat_im[i*nb_pts_re+j] = im; } } double *d_mat_re, *d_mat_im; int *d_img; cudaMalloc(&d_mat_re,size_d); cudaMalloc(&d_mat_im,size_d); cudaMalloc(&d_img,size_i); cudaMemcpy(d_mat_re, mat_re, size_d, cudaMemcpyHostToDevice); cudaMemcpy(d_mat_im, mat_im, size_d, cudaMemcpyHostToDevice); cout << "Running on GPU..." << endl; dim3 blockDim = 1024; dim3 gridDim = (nb_pts_re*nb_pts_im)/1024 + 1; z_funct<<<gridDim,blockDim>>>(d_mat_re, d_mat_im, d_img, nb_ite); cout << "Fetching datas..." << endl; int * img = (int*) malloc(size_i); cudaMemcpy(mat_re, d_mat_re, size_d, cudaMemcpyDeviceToHost); cudaMemcpy(mat_im, d_mat_im, size_d, cudaMemcpyDeviceToHost); cudaMemcpy(img, d_img, size_i, cudaMemcpyDeviceToHost); cout << "Writing on the disk..." << endl; ofstream file; file.open("conv"); for (int i=0; i<nb_pts_im; i++) { for (int j=0; j<nb_pts_re; j++) { file << setw(15) << img[i*nb_pts_re+j] ; } file << endl ; } file.close(); /* file.open("val"); for (int i=0; i<nb_pts_im; i++) { for (int j=0; j<nb_pts_re; j++) { file << setw(15) << pow(mat_re[i*nb_pts_re+j],2) + pow(mat_im[i*nb_pts_re+j],2) ; } file << endl ; } file.close(); */ cout << "Done !" << endl; return 0; }
#include <hip/hip_runtime.h> #include <iostream> #include <complex> #include <cmath> #include <iomanip> #include <string> #include <fstream> using namespace std; __global__ void z_funct(double * d_mat_re, double * d_mat_im, int *d_img, int nb_ite) { int i = blockDim.x * blockIdx.x + threadIdx.x; double c_re = d_mat_re[i]; double c_im = d_mat_im[i]; double z_re = 0, z_im =0; double z_re_tmp = 0; int t=0; while ((z_re*z_re+z_im*z_im <= 4) && (t<nb_ite)) { z_re_tmp = z_re*z_re - z_im*z_im + c_re; z_im = 2*z_im*z_re + c_im; z_re = z_re_tmp; t++; } d_img[i] = t; } int main(int argc, char * argv[]) { if (argc == 1) { cout << "Help" << endl << " 8 args :" << endl << endl << "arg1 = min real part | arg2 = min imaginary part " << endl << "arg3 = max real part | arg4 = max imaginary part " << endl << "arg5 = number of points on the real axe | arg6 = number of points on the imaginary axe " << endl << "arg7 = nb of iterations | arg8 = limit convergence" << endl << endl << " 4 args :" << endl << endl << "arg1 = number of points on the real axe | arg2 = number of points on the imaginary axe " << endl << "arg3 = nb of iterations " << endl << endl ; return 1; } double max_re, max_im, min_re, min_im; int nb_pts_re, nb_pts_im, nb_ite; if (argc == 8) { try { min_re = stod(argv[1]); min_im = stod(argv[2]); max_re = stod(argv[3]); max_im = stod(argv[4]); nb_pts_re = stoi(argv[5]); nb_pts_im = stoi(argv[6]); nb_ite = stoi(argv[7]); } catch (...) { cout << "Bad Args : see help (type nameofprogram without args)" << endl << endl; return 1; } } if (argc == 4 ) { min_re = -2; min_im = -1; max_re = 1; max_im = 1; try { nb_pts_re = stoi(argv[1]); nb_pts_im = stoi(argv[2]); nb_ite = stoi(argv[3]); } catch (...) { cout << "Bad Args : see help (type nameofprogram without args)" << endl << endl; return 1; } } cout << "Initializing..." << endl; int size_d = sizeof(double)*nb_pts_re*nb_pts_im; int size_i = sizeof(int)*nb_pts_re*nb_pts_im; double * mat_re = (double *)malloc(size_d); double * mat_im = (double *)malloc(size_d); double re, im; for (int i=0; i<nb_pts_im; i++) { im = max_im - (max_im-min_im)/nb_pts_im*i; for (int j=0; j<nb_pts_re; j++) { re = max_re - (max_re-min_re)/nb_pts_re*j; mat_re[i*nb_pts_re+j] = re; mat_im[i*nb_pts_re+j] = im; } } double *d_mat_re, *d_mat_im; int *d_img; hipMalloc(&d_mat_re,size_d); hipMalloc(&d_mat_im,size_d); hipMalloc(&d_img,size_i); hipMemcpy(d_mat_re, mat_re, size_d, hipMemcpyHostToDevice); hipMemcpy(d_mat_im, mat_im, size_d, hipMemcpyHostToDevice); cout << "Running on GPU..." << endl; dim3 blockDim = 1024; dim3 gridDim = (nb_pts_re*nb_pts_im)/1024 + 1; z_funct<<<gridDim,blockDim>>>(d_mat_re, d_mat_im, d_img, nb_ite); cout << "Fetching datas..." << endl; int * img = (int*) malloc(size_i); hipMemcpy(mat_re, d_mat_re, size_d, hipMemcpyDeviceToHost); hipMemcpy(mat_im, d_mat_im, size_d, hipMemcpyDeviceToHost); hipMemcpy(img, d_img, size_i, hipMemcpyDeviceToHost); cout << "Writing on the disk..." << endl; ofstream file; file.open("conv"); for (int i=0; i<nb_pts_im; i++) { for (int j=0; j<nb_pts_re; j++) { file << setw(15) << img[i*nb_pts_re+j] ; } file << endl ; } file.close(); /* file.open("val"); for (int i=0; i<nb_pts_im; i++) { for (int j=0; j<nb_pts_re; j++) { file << setw(15) << pow(mat_re[i*nb_pts_re+j],2) + pow(mat_im[i*nb_pts_re+j],2) ; } file << endl ; } file.close(); */ cout << "Done !" << endl; return 0; }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include <iostream> #include <complex> #include <cmath> #include <iomanip> #include <string> #include <fstream> using namespace std; __global__ void z_funct(double * d_mat_re, double * d_mat_im, int *d_img, int nb_ite) { int i = blockDim.x * blockIdx.x + threadIdx.x; double c_re = d_mat_re[i]; double c_im = d_mat_im[i]; double z_re = 0, z_im =0; double z_re_tmp = 0; int t=0; while ((z_re*z_re+z_im*z_im <= 4) && (t<nb_ite)) { z_re_tmp = z_re*z_re - z_im*z_im + c_re; z_im = 2*z_im*z_re + c_im; z_re = z_re_tmp; t++; } d_img[i] = t; } int main(int argc, char * argv[]) { if (argc == 1) { cout << "Help" << endl << " 8 args :" << endl << endl << "arg1 = min real part | arg2 = min imaginary part " << endl << "arg3 = max real part | arg4 = max imaginary part " << endl << "arg5 = number of points on the real axe | arg6 = number of points on the imaginary axe " << endl << "arg7 = nb of iterations | arg8 = limit convergence" << endl << endl << " 4 args :" << endl << endl << "arg1 = number of points on the real axe | arg2 = number of points on the imaginary axe " << endl << "arg3 = nb of iterations " << endl << endl ; return 1; } double max_re, max_im, min_re, min_im; int nb_pts_re, nb_pts_im, nb_ite; if (argc == 8) { try { min_re = stod(argv[1]); min_im = stod(argv[2]); max_re = stod(argv[3]); max_im = stod(argv[4]); nb_pts_re = stoi(argv[5]); nb_pts_im = stoi(argv[6]); nb_ite = stoi(argv[7]); } catch (...) { cout << "Bad Args : see help (type nameofprogram without args)" << endl << endl; return 1; } } if (argc == 4 ) { min_re = -2; min_im = -1; max_re = 1; max_im = 1; try { nb_pts_re = stoi(argv[1]); nb_pts_im = stoi(argv[2]); nb_ite = stoi(argv[3]); } catch (...) { cout << "Bad Args : see help (type nameofprogram without args)" << endl << endl; return 1; } } cout << "Initializing..." << endl; int size_d = sizeof(double)*nb_pts_re*nb_pts_im; int size_i = sizeof(int)*nb_pts_re*nb_pts_im; double * mat_re = (double *)malloc(size_d); double * mat_im = (double *)malloc(size_d); double re, im; for (int i=0; i<nb_pts_im; i++) { im = max_im - (max_im-min_im)/nb_pts_im*i; for (int j=0; j<nb_pts_re; j++) { re = max_re - (max_re-min_re)/nb_pts_re*j; mat_re[i*nb_pts_re+j] = re; mat_im[i*nb_pts_re+j] = im; } } double *d_mat_re, *d_mat_im; int *d_img; hipMalloc(&d_mat_re,size_d); hipMalloc(&d_mat_im,size_d); hipMalloc(&d_img,size_i); hipMemcpy(d_mat_re, mat_re, size_d, hipMemcpyHostToDevice); hipMemcpy(d_mat_im, mat_im, size_d, hipMemcpyHostToDevice); cout << "Running on GPU..." << endl; dim3 blockDim = 1024; dim3 gridDim = (nb_pts_re*nb_pts_im)/1024 + 1; z_funct<<<gridDim,blockDim>>>(d_mat_re, d_mat_im, d_img, nb_ite); cout << "Fetching datas..." << endl; int * img = (int*) malloc(size_i); hipMemcpy(mat_re, d_mat_re, size_d, hipMemcpyDeviceToHost); hipMemcpy(mat_im, d_mat_im, size_d, hipMemcpyDeviceToHost); hipMemcpy(img, d_img, size_i, hipMemcpyDeviceToHost); cout << "Writing on the disk..." << endl; ofstream file; file.open("conv"); for (int i=0; i<nb_pts_im; i++) { for (int j=0; j<nb_pts_re; j++) { file << setw(15) << img[i*nb_pts_re+j] ; } file << endl ; } file.close(); /* file.open("val"); for (int i=0; i<nb_pts_im; i++) { for (int j=0; j<nb_pts_re; j++) { file << setw(15) << pow(mat_re[i*nb_pts_re+j],2) + pow(mat_im[i*nb_pts_re+j],2) ; } file << endl ; } file.close(); */ cout << "Done !" << endl; return 0; }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z7z_functPdS_Pii .globl _Z7z_functPdS_Pii .p2align 8 .type _Z7z_functPdS_Pii,@function _Z7z_functPdS_Pii: s_clause 0x1 s_load_b32 s3, s[0:1], 0x2c s_load_b32 s2, s[0:1], 0x18 s_waitcnt lgkmcnt(0) s_and_b32 s3, s3, 0xffff s_cmp_lt_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_scc1 .LBB0_4 s_load_b128 s[4:7], s[0:1], 0x0 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_1) v_lshlrev_b64 v[3:4], 3, v[1:2] s_mov_b32 s3, 0 s_waitcnt lgkmcnt(0) v_add_co_u32 v5, vcc_lo, s4, v3 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v6, vcc_lo, s5, v4, vcc_lo v_add_co_u32 v7, vcc_lo, s6, v3 v_add_co_ci_u32_e32 v8, vcc_lo, s7, v4, vcc_lo s_mov_b32 s4, 0 global_load_b64 v[3:4], v[5:6], off global_load_b64 v[5:6], v[7:8], off v_mov_b32_e32 v7, 0 v_mov_b32_e32 v8, 0 s_delay_alu instid0(VALU_DEP_1) v_dual_mov_b32 v12, v8 :: v_dual_mov_b32 v11, v7 v_dual_mov_b32 v10, v8 :: v_dual_mov_b32 v9, v7 v_dual_mov_b32 v14, v8 :: v_dual_mov_b32 v13, v7 .p2align 6 .LBB0_2: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[9:10], v[9:10], v[9:10] v_add_f64 v[7:8], v[11:12], -v[7:8] s_add_i32 s4, s4, 1 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_3) | instid1(VALU_DEP_3) s_cmp_ge_i32 s4, s2 v_mov_b32_e32 v0, s4 s_cselect_b32 s5, -1, 0 s_waitcnt vmcnt(0) v_fma_f64 v[9:10], v[13:14], v[9:10], v[5:6] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[13:14], v[3:4], v[7:8] v_mul_f64 v[7:8], v[9:10], v[9:10] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[11:12], v[13:14], v[13:14], v[7:8] v_cmp_nge_f64_e32 vcc_lo, 4.0, v[11:12] v_mul_f64 v[11:12], v[13:14], v[13:14] s_or_b32 s5, vcc_lo, s5 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_and_b32 s5, exec_lo, s5 s_or_b32 s3, s5, s3 s_delay_alu instid0(SALU_CYCLE_1) s_and_not1_b32 exec_lo, exec_lo, s3 s_cbranch_execnz .LBB0_2 s_or_b32 exec_lo, exec_lo, s3 s_branch .LBB0_5 .LBB0_4: v_mov_b32_e32 v0, 0 .LBB0_5: s_load_b64 s[0:1], s[0:1], 0x10 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[1:2], 2, v[1:2] s_waitcnt lgkmcnt(0) v_add_co_u32 v1, vcc_lo, s0, v1 s_delay_alu instid0(VALU_DEP_2) 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 _Z7z_functPdS_Pii .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 _Z7z_functPdS_Pii, .Lfunc_end0-_Z7z_functPdS_Pii .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: _Z7z_functPdS_Pii .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z7z_functPdS_Pii.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 device assembly to AMD device assembly.
code for sm_80 Function : _Z7z_functPdS_Pii .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 R6, SR_CTAID.X ; /* 0x0000000000067919 */ /* 0x000e220000002500 */ /*0020*/ IMAD.MOV.U32 R0, RZ, RZ, c[0x0][0x178] ; /* 0x00005e00ff007624 */ /* 0x000fe200078e00ff */ /*0030*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*0040*/ HFMA2.MMA R19, -RZ, RZ, 0, 0 ; /* 0x00000000ff137435 */ /* 0x000fe200000001ff */ /*0050*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e240000002100 */ /*0060*/ ISETP.GE.AND P0, PT, R0, 0x1, PT ; /* 0x000000010000780c */ /* 0x000fe20003f06270 */ /*0070*/ IMAD R6, R6, c[0x0][0x0], R3 ; /* 0x0000000006067a24 */ /* 0x001fca00078e0203 */ /*0080*/ SHF.R.S32.HI R3, RZ, 0x1f, R6 ; /* 0x0000001fff037819 */ /* 0x000fe40000011406 */ /*0090*/ LEA R2, P1, R6, c[0x0][0x170], 0x2 ; /* 0x00005c0006027a11 */ /* 0x000fc800078210ff */ /*00a0*/ LEA.HI.X R3, R6, c[0x0][0x174], R3, 0x2, P1 ; /* 0x00005d0006037a11 */ /* 0x000fe200008f1403 */ /*00b0*/ @!P0 BRA 0x250 ; /* 0x0000019000008947 */ /* 0x000fea0003800000 */ /*00c0*/ IMAD.MOV.U32 R7, RZ, RZ, 0x8 ; /* 0x00000008ff077424 */ /* 0x000fc800078e00ff */ /*00d0*/ IMAD.WIDE R4, R6, R7, c[0x0][0x160] ; /* 0x0000580006047625 */ /* 0x000fc800078e0207 */ /*00e0*/ IMAD.WIDE R6, R6, R7, c[0x0][0x168] ; /* 0x00005a0006067625 */ /* 0x000fe400078e0207 */ /*00f0*/ LDG.E.64 R4, [R4.64] ; /* 0x0000000404047981 */ /* 0x000f68000c1e1b00 */ /*0100*/ LDG.E.64 R6, [R6.64] ; /* 0x0000000406067981 */ /* 0x000f62000c1e1b00 */ /*0110*/ BSSY B0, 0x250 ; /* 0x0000013000007945 */ /* 0x000fe20003800000 */ /*0120*/ MOV R19, RZ ; /* 0x000000ff00137202 */ /* 0x000fe20000000f00 */ /*0130*/ CS2R R8, SRZ ; /* 0x0000000000087805 */ /* 0x000fe2000001ff00 */ /*0140*/ CS2R R10, SRZ ; /* 0x00000000000a7805 */ /* 0x000fe2000001ff00 */ /*0150*/ CS2R R12, SRZ ; /* 0x00000000000c7805 */ /* 0x000fe2000001ff00 */ /*0160*/ CS2R R14, SRZ ; /* 0x00000000000e7805 */ /* 0x000fc8000001ff00 */ /*0170*/ DADD R8, R10, -R8 ; /* 0x000000000a087229 */ /* 0x000e220000000808 */ /*0180*/ IADD3 R19, R19, 0x1, RZ ; /* 0x0000000113137810 */ /* 0x000fc60007ffe0ff */ /*0190*/ DADD R12, R12, R12 ; /* 0x000000000c0c7229 */ /* 0x000e62000000000c */ /*01a0*/ ISETP.GE.AND P0, PT, R19, c[0x0][0x178], PT ; /* 0x00005e0013007a0c */ /* 0x000fc60003f06270 */ /*01b0*/ DADD R16, R4, R8 ; /* 0x0000000004107229 */ /* 0x021e080000000008 */ /*01c0*/ DFMA R12, R12, R14, R6 ; /* 0x0000000e0c0c722b */ /* 0x002e480000000006 */ /*01d0*/ DMUL R10, R16, R16 ; /* 0x00000010100a7228 */ /* 0x001fc80000000000 */ /*01e0*/ DMUL R8, R12, R12 ; /* 0x0000000c0c087228 */ /* 0x002e0c0000000000 */ /*01f0*/ DADD R14, R10, R8 ; /* 0x000000000a0e7229 */ /* 0x001e0c0000000008 */ /*0200*/ DSETP.LE.AND P1, PT, R14, 4, PT ; /* 0x401000000e00742a */ /* 0x0010640003f23000 */ /*0210*/ IMAD.MOV.U32 R14, RZ, RZ, R16 ; /* 0x000000ffff0e7224 */ /* 0x001fe200078e0010 */ /*0220*/ MOV R15, R17 ; /* 0x00000011000f7202 */ /* 0x000fd60000000f00 */ /*0230*/ @!P0 BRA P1, 0x170 ; /* 0xffffff3000008947 */ /* 0x002fea000083ffff */ /*0240*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0250*/ STG.E [R2.64], R19 ; /* 0x0000001302007986 */ /* 0x000fe2000c101904 */ /*0260*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0270*/ BRA 0x270; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0280*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0290*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*02a0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*02b0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*02c0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*02d0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*02e0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*02f0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z7z_functPdS_Pii .globl _Z7z_functPdS_Pii .p2align 8 .type _Z7z_functPdS_Pii,@function _Z7z_functPdS_Pii: s_clause 0x1 s_load_b32 s3, s[0:1], 0x2c s_load_b32 s2, s[0:1], 0x18 s_waitcnt lgkmcnt(0) s_and_b32 s3, s3, 0xffff s_cmp_lt_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_scc1 .LBB0_4 s_load_b128 s[4:7], s[0:1], 0x0 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_1) v_lshlrev_b64 v[3:4], 3, v[1:2] s_mov_b32 s3, 0 s_waitcnt lgkmcnt(0) v_add_co_u32 v5, vcc_lo, s4, v3 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v6, vcc_lo, s5, v4, vcc_lo v_add_co_u32 v7, vcc_lo, s6, v3 v_add_co_ci_u32_e32 v8, vcc_lo, s7, v4, vcc_lo s_mov_b32 s4, 0 global_load_b64 v[3:4], v[5:6], off global_load_b64 v[5:6], v[7:8], off v_mov_b32_e32 v7, 0 v_mov_b32_e32 v8, 0 s_delay_alu instid0(VALU_DEP_1) v_dual_mov_b32 v12, v8 :: v_dual_mov_b32 v11, v7 v_dual_mov_b32 v10, v8 :: v_dual_mov_b32 v9, v7 v_dual_mov_b32 v14, v8 :: v_dual_mov_b32 v13, v7 .p2align 6 .LBB0_2: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[9:10], v[9:10], v[9:10] v_add_f64 v[7:8], v[11:12], -v[7:8] s_add_i32 s4, s4, 1 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_3) | instid1(VALU_DEP_3) s_cmp_ge_i32 s4, s2 v_mov_b32_e32 v0, s4 s_cselect_b32 s5, -1, 0 s_waitcnt vmcnt(0) v_fma_f64 v[9:10], v[13:14], v[9:10], v[5:6] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f64 v[13:14], v[3:4], v[7:8] v_mul_f64 v[7:8], v[9:10], v[9:10] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[11:12], v[13:14], v[13:14], v[7:8] v_cmp_nge_f64_e32 vcc_lo, 4.0, v[11:12] v_mul_f64 v[11:12], v[13:14], v[13:14] s_or_b32 s5, vcc_lo, s5 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_and_b32 s5, exec_lo, s5 s_or_b32 s3, s5, s3 s_delay_alu instid0(SALU_CYCLE_1) s_and_not1_b32 exec_lo, exec_lo, s3 s_cbranch_execnz .LBB0_2 s_or_b32 exec_lo, exec_lo, s3 s_branch .LBB0_5 .LBB0_4: v_mov_b32_e32 v0, 0 .LBB0_5: s_load_b64 s[0:1], s[0:1], 0x10 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[1:2], 2, v[1:2] s_waitcnt lgkmcnt(0) v_add_co_u32 v1, vcc_lo, s0, v1 s_delay_alu instid0(VALU_DEP_2) 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 _Z7z_functPdS_Pii .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 _Z7z_functPdS_Pii, .Lfunc_end0-_Z7z_functPdS_Pii .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: _Z7z_functPdS_Pii .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z7z_functPdS_Pii.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 <math.h> void __global__ kernel_linearInterpolation(float* array_device, int* rowArray, int rowArrayLength, int col, int colArrayLength, int totalCols, int totalRows, float* originalTimes, int originalTimesLength, float* newTimes, int newTimesLength, float* results_device) { // where original times and new times should probably be unix time? but technically won't matter probably int n = blockIdx.x * blockDim.x + threadIdx.x; //int m = blockIdx.y * blockDim.y + threadIdx.y; int m = col; if (n < newTimesLength && m < colArrayLength) { // so we created a row-wise thread per element in the newTimes, with m being number of columns float newTime = newTimes[n]; int resultsInd = n*colArrayLength + m; // need to potentially handle the first and last points in a special way ,,, if (n==0 && newTimes[0] < originalTimes[0]) { // lets say if newTimes[0] is prior to originalTimes[0] less than the distance from originalTimes[1] to originalTimes[2], we'll accept it // otherwise call it a NaN if ( originalTimes[0] - newTimes[0] < originalTimes[1] - originalTimes[0] ) { // continue slope from points 0 to 1? // ie assume there's a previous point at same slope and same sample rate int arrayInd1 = (0)*totalCols + m; int arrayInd2 = (1)*totalCols + m; float valueChange = array_device[arrayInd2] - array_device[arrayInd1]; float timeDiff = originalTimes[1] - originalTimes[0]; float pseudoPointTime = originalTimes[0] - timeDiff; float pseudoPointValue = array_device[arrayInd1] - valueChange; float perc = (newTime - pseudoPointTime) / timeDiff; results_device[resultsInd] = pseudoPointValue + perc*valueChange; } else results_device[resultsInd] = NAN; return; } // ================================================ // ================================================ // ===== DO THIS SAME LOGIC FOR LAST POINT? ======= // ================================================ // ================================================ // we now need to find the points before and after ... // do we need to iterate through the entire originalTimes vector or is there some way we can approximate the starting location? //float origSampleRate = (originalTimes[originalTimesLength-1] - originalTimes[0]) / originalTimesLength; //float newSampleRate = (newTimes[newTimesLength-1] - newTimes[0]) / newTimesLength; //int originalTimeLoc = (newTime - newTimes[0]) / (originalTimes[originalTimesLength-1] - originalTimes) float newTimePerc = (newTime - newTimes[0]) / (newTimes[newTimesLength-1] - newTimes[0]); int estimation = newTimePerc * originalTimesLength; // start at approximately the correct point int iterationsToFindMatch = 0; int originalTimeLoc; if (originalTimes[estimation] > newTime) { // time we guessed is greater than the time we want ... so walk backwards? for (originalTimeLoc=estimation; originalTimeLoc>0; originalTimeLoc--) { if (originalTimes[originalTimeLoc] < newTime) break; iterationsToFindMatch++; } } else { // the time we guessed is greater than the time we want ... so walk forwards? for (originalTimeLoc=estimation; originalTimeLoc<originalTimesLength; originalTimeLoc++) { if (originalTimes[originalTimeLoc] > newTime) { originalTimeLoc--; // want the left side point break; } iterationsToFindMatch++; } } printf("Loc %i needed %i iterations, starting from %i -> %i\n", n, iterationsToFindMatch, estimation, originalTimeLoc); int arrayInd1 = originalTimeLoc*totalCols + m; int arrayInd2 = (originalTimeLoc+1)*totalCols + m; float valueChange = (array_device[arrayInd2] - array_device[arrayInd1]); // we need to figure out what percentage the newTime is between the two points float origTime1 = originalTimes[originalTimeLoc]; float origTime2 = originalTimes[originalTimeLoc+1]; float perc = (newTime - origTime1) / (origTime2 - origTime1); results_device[resultsInd] = array_device[arrayInd1] + perc*valueChange; } }
code for sm_80 Function : _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .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 */ /* 0x000fc600078e00ff */ /*0010*/ S2R R16, SR_CTAID.X ; /* 0x0000000000107919 */ /* 0x000e220000002500 */ /*0020*/ IMAD.MOV.U32 R7, RZ, RZ, c[0x0][0x174] ; /* 0x00005d00ff077624 */ /* 0x000fe200078e00ff */ /*0030*/ IADD3 R1, R1, -0x10, RZ ; /* 0xfffffff001017810 */ /* 0x000fe40007ffe0ff */ /*0040*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e240000002100 */ /*0050*/ IMAD R16, R16, c[0x0][0x0], R3 ; /* 0x0000000010107a24 */ /* 0x001fca00078e0203 */ /*0060*/ ISETP.GE.AND P0, PT, R16, c[0x0][0x1a0], PT ; /* 0x0000680010007a0c */ /* 0x000fc80003f06270 */ /*0070*/ ISETP.GE.OR P0, PT, R7, c[0x0][0x178], P0 ; /* 0x00005e0007007a0c */ /* 0x000fda0000706670 */ /*0080*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0090*/ IMAD.MOV.U32 R9, RZ, RZ, 0x4 ; /* 0x00000004ff097424 */ /* 0x000fe200078e00ff */ /*00a0*/ MOV R3, c[0x0][0x19c] ; /* 0x0000670000037a02 */ /* 0x000fe20000000f00 */ /*00b0*/ IMAD.MOV.U32 R2, RZ, RZ, c[0x0][0x198] ; /* 0x00006600ff027624 */ /* 0x000fe200078e00ff */ /*00c0*/ ULDC.64 UR36, c[0x0][0x118] ; /* 0x0000460000247ab9 */ /* 0x000fe20000000a00 */ /*00d0*/ IMAD.WIDE R24, R16, R9, c[0x0][0x198] ; /* 0x0000660010187625 */ /* 0x000fc600078e0209 */ /*00e0*/ LDG.E R5, [R2.64] ; /* 0x0000002402057981 */ /* 0x000168000c1e1900 */ /*00f0*/ LDG.E R24, [R24.64] ; /* 0x0000002418187981 */ /* 0x000162000c1e1900 */ /*0100*/ ISETP.NE.AND P0, PT, R16.reuse, RZ, PT ; /* 0x000000ff1000720c */ /* 0x040fe20003f05270 */ /*0110*/ IMAD R22, R16, c[0x0][0x178], R7 ; /* 0x00005e0010167a24 */ /* 0x000fe200078e0207 */ /*0120*/ BSSY B0, 0x1b0 ; /* 0x0000008000007945 */ /* 0x000fe60003800000 */ /*0130*/ IMAD.WIDE R22, R22, R9, c[0x0][0x1a8] ; /* 0x00006a0016167625 */ /* 0x000fd000078e0209 */ /*0140*/ @P0 BRA 0x1a0 ; /* 0x0000005000000947 */ /* 0x000fea0003800000 */ /*0150*/ IMAD.MOV.U32 R2, RZ, RZ, c[0x0][0x188] ; /* 0x00006200ff027624 */ /* 0x001fe400078e00ff */ /*0160*/ IMAD.MOV.U32 R3, RZ, RZ, c[0x0][0x18c] ; /* 0x00006300ff037624 */ /* 0x000fca00078e00ff */ /*0170*/ LDG.E R0, [R2.64] ; /* 0x0000002402007981 */ /* 0x000ea4000c1e1900 */ /*0180*/ FSETP.GEU.AND P0, PT, R5, R0, PT ; /* 0x000000000500720b */ /* 0x024fda0003f0e000 */ /*0190*/ @!P0 BRA 0x890 ; /* 0x000006f000008947 */ /* 0x000fea0003800000 */ /*01a0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x001fea0003800000 */ /*01b0*/ IMAD.MOV.U32 R2, RZ, RZ, c[0x0][0x1a0] ; /* 0x00006800ff027624 */ /* 0x000fca00078e00ff */ /*01c0*/ IADD3 R2, R2, -0x1, RZ ; /* 0xffffffff02027810 */ /* 0x000fca0007ffe0ff */ /*01d0*/ IMAD.WIDE R2, R2, R9, c[0x0][0x198] ; /* 0x0000660002027625 */ /* 0x000fcc00078e0209 */ /*01e0*/ LDG.E R2, [R2.64] ; /* 0x0000002402027981 */ /* 0x000ea2000c1e1900 */ /*01f0*/ BSSY B0, 0x300 ; /* 0x0000010000007945 */ /* 0x000fe20003800000 */ /*0200*/ FADD R0, -R5, R2 ; /* 0x0000000205007221 */ /* 0x024fe40000000100 */ /*0210*/ FADD R5, R24, -R5 ; /* 0x8000000518057221 */ /* 0x000fe40000000000 */ /*0220*/ MUFU.RCP R7, R0 ; /* 0x0000000000077308 */ /* 0x000e300000001000 */ /*0230*/ FCHK P0, R5, R0 ; /* 0x0000000005007302 */ /* 0x000e620000000000 */ /*0240*/ FFMA R4, -R0, R7, 1 ; /* 0x3f80000000047423 */ /* 0x001fc80000000107 */ /*0250*/ FFMA R4, R7, R4, R7 ; /* 0x0000000407047223 */ /* 0x000fc80000000007 */ /*0260*/ FFMA R7, R5, R4, RZ ; /* 0x0000000405077223 */ /* 0x000fc800000000ff */ /*0270*/ FFMA R6, -R0, R7, R5 ; /* 0x0000000700067223 */ /* 0x000fc80000000105 */ /*0280*/ FFMA R4, R4, R6, R7 ; /* 0x0000000604047223 */ /* 0x000fe20000000007 */ /*0290*/ @!P0 BRA 0x2f0 ; /* 0x0000005000008947 */ /* 0x002fea0003800000 */ /*02a0*/ IMAD.MOV.U32 R8, RZ, RZ, R5 ; /* 0x000000ffff087224 */ /* 0x000fe200078e0005 */ /*02b0*/ MOV R7, R0 ; /* 0x0000000000077202 */ /* 0x000fe40000000f00 */ /*02c0*/ MOV R2, 0x2e0 ; /* 0x000002e000027802 */ /* 0x000fe40000000f00 */ /*02d0*/ CALL.REL.NOINC 0xab0 ; /* 0x000007d000007944 */ /* 0x000fea0003c00000 */ /*02e0*/ IMAD.MOV.U32 R4, RZ, RZ, R8 ; /* 0x000000ffff047224 */ /* 0x000fe400078e0008 */ /*02f0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0300*/ I2F R3, c[0x0][0x190] ; /* 0x0000640000037b06 */ /* 0x000e240000201400 */ /*0310*/ FMUL R4, R3, R4 ; /* 0x0000000403047220 */ /* 0x001fe40000400000 */ /*0320*/ IMAD.MOV.U32 R3, RZ, RZ, 0x4 ; /* 0x00000004ff037424 */ /* 0x000fc800078e00ff */ /*0330*/ F2I.TRUNC.NTZ R18, R4 ; /* 0x0000000400127305 */ /* 0x000e24000020f100 */ /*0340*/ IMAD.WIDE R2, R18, R3, c[0x0][0x188] ; /* 0x0000620012027625 */ /* 0x001fcc00078e0203 */ /*0350*/ LDG.E R3, [R2.64] ; /* 0x0000002402037981 */ /* 0x000ea2000c1e1900 */ /*0360*/ IADD3 R6, P1, R1, c[0x0][0x20], RZ ; /* 0x0000080001067a10 */ /* 0x000fe20007f3e0ff */ /*0370*/ BSSY B0, 0x5d0 ; /* 0x0000025000007945 */ /* 0x000fe80003800000 */ /*0380*/ IMAD.X R7, RZ, RZ, c[0x0][0x24], P1 ; /* 0x00000900ff077624 */ /* 0x000fe200008e06ff */ /*0390*/ FSETP.GT.AND P0, PT, R3, R24, PT ; /* 0x000000180300720b */ /* 0x004fda0003f04000 */ /*03a0*/ @P0 BRA 0x4d0 ; /* 0x0000012000000947 */ /* 0x000fea0003800000 */ /*03b0*/ ISETP.GE.AND P0, PT, R18, c[0x0][0x190], PT ; /* 0x0000640012007a0c */ /* 0x000fe20003f06270 */ /*03c0*/ IMAD.MOV.U32 R17, RZ, RZ, RZ ; /* 0x000000ffff117224 */ /* 0x000fe200078e00ff */ /*03d0*/ MOV R19, R18 ; /* 0x0000001200137202 */ /* 0x000fd60000000f00 */ /*03e0*/ @P0 BRA 0x5c0 ; /* 0x000001d000000947 */ /* 0x000fea0003800000 */ /*03f0*/ IMAD.MOV.U32 R17, RZ, RZ, RZ ; /* 0x000000ffff117224 */ /* 0x000fe400078e00ff */ /*0400*/ IMAD.MOV.U32 R19, RZ, RZ, R18 ; /* 0x000000ffff137224 */ /* 0x000fe400078e0012 */ /*0410*/ IMAD.MOV.U32 R2, RZ, RZ, 0x4 ; /* 0x00000004ff027424 */ /* 0x000fc800078e00ff */ /*0420*/ IMAD.WIDE R2, R19, R2, c[0x0][0x188] ; /* 0x0000620013027625 */ /* 0x000fcc00078e0202 */ /*0430*/ LDG.E R3, [R2.64] ; /* 0x0000002402037981 */ /* 0x000ea4000c1e1900 */ /*0440*/ FSETP.GT.AND P0, PT, R3, R24, PT ; /* 0x000000180300720b */ /* 0x004fda0003f04000 */ /*0450*/ @P0 BRA 0x4b0 ; /* 0x0000005000000947 */ /* 0x000fea0003800000 */ /*0460*/ IADD3 R19, R19, 0x1, RZ ; /* 0x0000000113137810 */ /* 0x000fe40007ffe0ff */ /*0470*/ IADD3 R17, R17, 0x1, RZ ; /* 0x0000000111117810 */ /* 0x000fe40007ffe0ff */ /*0480*/ ISETP.GE.AND P0, PT, R19, c[0x0][0x190], PT ; /* 0x0000640013007a0c */ /* 0x000fda0003f06270 */ /*0490*/ @!P0 BRA 0x410 ; /* 0xffffff7000008947 */ /* 0x000fea000383ffff */ /*04a0*/ BRA 0x5c0 ; /* 0x0000011000007947 */ /* 0x000fea0003800000 */ /*04b0*/ IADD3 R19, R19, -0x1, RZ ; /* 0xffffffff13137810 */ /* 0x000fe20007ffe0ff */ /*04c0*/ BRA 0x5c0 ; /* 0x000000f000007947 */ /* 0x000fea0003800000 */ /*04d0*/ ISETP.GE.AND P0, PT, R18, 0x1, PT ; /* 0x000000011200780c */ /* 0x000fe20003f06270 */ /*04e0*/ IMAD.MOV.U32 R17, RZ, RZ, RZ ; /* 0x000000ffff117224 */ /* 0x000fe200078e00ff */ /*04f0*/ MOV R19, R18 ; /* 0x0000001200137202 */ /* 0x000fd60000000f00 */ /*0500*/ @!P0 BRA 0x5c0 ; /* 0x000000b000008947 */ /* 0x000fea0003800000 */ /*0510*/ IMAD.MOV.U32 R17, RZ, RZ, RZ ; /* 0x000000ffff117224 */ /* 0x000fe400078e00ff */ /*0520*/ IMAD.MOV.U32 R19, RZ, RZ, R18 ; /* 0x000000ffff137224 */ /* 0x000fe400078e0012 */ /*0530*/ IMAD.MOV.U32 R2, RZ, RZ, 0x4 ; /* 0x00000004ff027424 */ /* 0x000fc800078e00ff */ /*0540*/ IMAD.WIDE R2, R19, R2, c[0x0][0x188] ; /* 0x0000620013027625 */ /* 0x000fcc00078e0202 */ /*0550*/ LDG.E R3, [R2.64] ; /* 0x0000002402037981 */ /* 0x000ea4000c1e1900 */ /*0560*/ FSETP.GEU.AND P0, PT, R3, R24, PT ; /* 0x000000180300720b */ /* 0x004fda0003f0e000 */ /*0570*/ @!P0 BRA 0x5c0 ; /* 0x0000004000008947 */ /* 0x000fea0003800000 */ /*0580*/ ISETP.GT.AND P0, PT, R19.reuse, 0x1, PT ; /* 0x000000011300780c */ /* 0x040fe40003f04270 */ /*0590*/ IADD3 R19, R19, -0x1, RZ ; /* 0xffffffff13137810 */ /* 0x000fe40007ffe0ff */ /*05a0*/ IADD3 R17, R17, 0x1, RZ ; /* 0x0000000111117810 */ /* 0x000fd20007ffe0ff */ /*05b0*/ @P0 BRA 0x530 ; /* 0xffffff7000000947 */ /* 0x000fea000383ffff */ /*05c0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*05d0*/ MOV R2, 0x0 ; /* 0x0000000000027802 */ /* 0x000fe20000000f00 */ /*05e0*/ STL.128 [R1], R16 ; /* 0x0000001001007387 */ /* 0x0001e20000100c00 */ /*05f0*/ MOV R4, c[0x4][0x8] ; /* 0x0100020000047a02 */ /* 0x000fe20000000f00 */ /*0600*/ IMAD.MOV.U32 R5, RZ, RZ, c[0x4][0xc] ; /* 0x01000300ff057624 */ /* 0x000fc600078e00ff */ /*0610*/ LDC.64 R2, c[0x4][R2] ; /* 0x0100000002027b82 */ /* 0x000e680000000a00 */ /*0620*/ LEPC R8 ; /* 0x000000000008734e */ /* 0x000fe20000000000 */ /*0630*/ MOV R11, 0x6a0 ; /* 0x000006a0000b7802 */ /* 0x000fe40000000f00 */ /*0640*/ MOV R20, 0x620 ; /* 0x0000062000147802 */ /* 0x000fe40000000f00 */ /*0650*/ MOV R21, 0x0 ; /* 0x0000000000157802 */ /* 0x000fe40000000f00 */ /*0660*/ MOV R0, 0x0 ; /* 0x0000000000007802 */ /* 0x000fe40000000f00 */ /*0670*/ IADD3 R20, P0, P1, -R20, R11, R8 ; /* 0x0000000b14147210 */ /* 0x000fc8000791e108 */ /*0680*/ IADD3.X R21, ~R0, R21, R9, P0, P1 ; /* 0x0000001500157210 */ /* 0x000fc800007e2509 */ /*0690*/ CALL.ABS.NOINC R2 ; /* 0x0000000002007343 */ /* 0x003fea0003c00000 */ /*06a0*/ IMAD.MOV.U32 R5, RZ, RZ, 0x4 ; /* 0x00000004ff057424 */ /* 0x000fc800078e00ff */ /*06b0*/ IMAD.WIDE R2, R19, R5, c[0x0][0x188] ; /* 0x0000620013027625 */ /* 0x000fca00078e0205 */ /*06c0*/ LDG.E R9, [R2.64] ; /* 0x0000002402097981 */ /* 0x000ea8000c1e1900 */ /*06d0*/ LDG.E R8, [R2.64+0x4] ; /* 0x0000042402087981 */ /* 0x000ea2000c1e1900 */ /*06e0*/ IMAD.MOV.U32 R6, RZ, RZ, c[0x0][0x17c] ; /* 0x00005f00ff067624 */ /* 0x000fc800078e00ff */ /*06f0*/ IMAD R6, R19, R6, c[0x0][0x174] ; /* 0x00005d0013067624 */ /* 0x000fca00078e0206 */ /*0700*/ IADD3 R4, R6.reuse, c[0x0][0x17c], RZ ; /* 0x00005f0006047a10 */ /* 0x040fe20007ffe0ff */ /*0710*/ IMAD.WIDE R6, R6, R5, c[0x0][0x160] ; /* 0x0000580006067625 */ /* 0x000fc800078e0205 */ /*0720*/ IMAD.WIDE R4, R4, R5, c[0x0][0x160] ; /* 0x0000580004047625 */ /* 0x000fe200078e0205 */ /*0730*/ LDG.E R0, [R6.64] ; /* 0x0000002406007981 */ /* 0x0004ea000c1e1900 */ /*0740*/ LDG.E R5, [R4.64] ; /* 0x0000002404057981 */ /* 0x000ee2000c1e1900 */ /*0750*/ BSSY B0, 0x860 ; /* 0x0000010000007945 */ /* 0x000fe20003800000 */ /*0760*/ FADD R7, -R9, R8 ; /* 0x0000000809077221 */ /* 0x004fe40000000100 */ /*0770*/ FADD R9, R24, -R9 ; /* 0x8000000918097221 */ /* 0x000fc40000000000 */ /*0780*/ MUFU.RCP R8, R7 ; /* 0x0000000700087308 */ /* 0x000e300000001000 */ /*0790*/ FCHK P0, R9, R7 ; /* 0x0000000709007302 */ /* 0x000e620000000000 */ /*07a0*/ FFMA R11, -R7, R8, 1 ; /* 0x3f800000070b7423 */ /* 0x001fe40000000108 */ /*07b0*/ FADD R3, -R0, R5 ; /* 0x0000000500037221 */ /* 0x008fc40000000100 */ /*07c0*/ FFMA R11, R8, R11, R8 ; /* 0x0000000b080b7223 */ /* 0x000fc80000000008 */ /*07d0*/ FFMA R2, R9, R11, RZ ; /* 0x0000000b09027223 */ /* 0x000fc800000000ff */ /*07e0*/ FFMA R8, -R7, R2, R9 ; /* 0x0000000207087223 */ /* 0x000fc80000000109 */ /*07f0*/ FFMA R2, R11, R8, R2 ; /* 0x000000080b027223 */ /* 0x000fe20000000002 */ /*0800*/ @!P0 BRA 0x850 ; /* 0x0000004000008947 */ /* 0x002fea0003800000 */ /*0810*/ MOV R8, R9 ; /* 0x0000000900087202 */ /* 0x000fe40000000f00 */ /*0820*/ MOV R2, 0x840 ; /* 0x0000084000027802 */ /* 0x000fe40000000f00 */ /*0830*/ CALL.REL.NOINC 0xab0 ; /* 0x0000027000007944 */ /* 0x000fea0003c00000 */ /*0840*/ IMAD.MOV.U32 R2, RZ, RZ, R8 ; /* 0x000000ffff027224 */ /* 0x000fe400078e0008 */ /*0850*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0860*/ FFMA R3, R3, R2, R0 ; /* 0x0000000203037223 */ /* 0x000fca0000000000 */ /*0870*/ STG.E [R22.64], R3 ; /* 0x0000000316007986 */ /* 0x000fe2000c101924 */ /*0880*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0890*/ LDG.E R3, [R2.64+0x4] ; /* 0x0000042402037981 */ /* 0x000ea2000c1e1900 */ /*08a0*/ FADD R5, -R5, R0 ; /* 0x0000000005057221 */ /* 0x000fe40000000100 */ /*08b0*/ FADD R6, -R0, R3 ; /* 0x0000000300067221 */ /* 0x004fca0000000100 */ /*08c0*/ FSETP.GEU.AND P0, PT, R5, R6, PT ; /* 0x000000060500720b */ /* 0x000fda0003f0e000 */ /*08d0*/ @P0 IMAD.MOV.U32 R5, RZ, RZ, 0x7fc00000 ; /* 0x7fc00000ff050424 */ /* 0x000fca00078e00ff */ /*08e0*/ @P0 STG.E [R22.64], R5 ; /* 0x0000000516000986 */ /* 0x0001e2000c101924 */ /*08f0*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0900*/ IMAD.MOV.U32 R2, RZ, RZ, c[0x0][0x17c] ; /* 0x00005f00ff027624 */ /* 0x000fe400078e00ff */ /*0910*/ IMAD.WIDE R4, R7, R9, c[0x0][0x160] ; /* 0x0000580007047625 */ /* 0x001fc600078e0209 */ /*0920*/ IADD3 R2, R2, c[0x0][0x174], RZ ; /* 0x00005d0002027a10 */ /* 0x000fc60007ffe0ff */ /*0930*/ LDG.E R5, [R4.64] ; /* 0x0000002404057981 */ /* 0x000ea4000c1e1900 */ /*0940*/ IMAD.WIDE R2, R2, R9, c[0x0][0x160] ; /* 0x0000580002027625 */ /* 0x000fcc00078e0209 */ /*0950*/ LDG.E R2, [R2.64] ; /* 0x0000002402027981 */ /* 0x000ea2000c1e1900 */ /*0960*/ MUFU.RCP R9, R6 ; /* 0x0000000600097308 */ /* 0x000e220000001000 */ /*0970*/ FADD R7, R0, -R6 ; /* 0x8000000600077221 */ /* 0x000fc80000000000 */ /*0980*/ FADD R7, R24, -R7 ; /* 0x8000000718077221 */ /* 0x000fc80000000000 */ /*0990*/ FCHK P0, R7, R6 ; /* 0x0000000607007302 */ /* 0x000e620000000000 */ /*09a0*/ FFMA R0, -R6, R9, 1 ; /* 0x3f80000006007423 */ /* 0x001fc80000000109 */ /*09b0*/ FFMA R8, R9, R0, R9 ; /* 0x0000000009087223 */ /* 0x000fc80000000009 */ /*09c0*/ FFMA R9, R7, R8, RZ ; /* 0x0000000807097223 */ /* 0x000fe200000000ff */ /*09d0*/ BSSY B1, 0xa80 ; /* 0x000000a000017945 */ /* 0x000fe60003800000 */ /*09e0*/ FFMA R10, -R6, R9, R7 ; /* 0x00000009060a7223 */ /* 0x000fc80000000107 */ /*09f0*/ FFMA R8, R8, R10, R9 ; /* 0x0000000a08087223 */ /* 0x000fe40000000009 */ /*0a00*/ FADD R0, -R5, R2 ; /* 0x0000000205007221 */ /* 0x004fc80000000100 */ /*0a10*/ FADD R3, R5, -R0 ; /* 0x8000000005037221 */ /* 0x000fe20000000000 */ /*0a20*/ @!P0 BRA 0xa70 ; /* 0x0000004000008947 */ /* 0x002fea0003800000 */ /*0a30*/ MOV R8, R7 ; /* 0x0000000700087202 */ /* 0x000fe20000000f00 */ /*0a40*/ IMAD.MOV.U32 R7, RZ, RZ, R6 ; /* 0x000000ffff077224 */ /* 0x000fe200078e0006 */ /*0a50*/ MOV R2, 0xa70 ; /* 0x00000a7000027802 */ /* 0x000fe40000000f00 */ /*0a60*/ CALL.REL.NOINC 0xab0 ; /* 0x0000004000007944 */ /* 0x000fea0003c00000 */ /*0a70*/ BSYNC B1 ; /* 0x0000000000017941 */ /* 0x000fea0003800000 */ /*0a80*/ FFMA R3, R0, R8, R3 ; /* 0x0000000800037223 */ /* 0x000fca0000000003 */ /*0a90*/ STG.E [R22.64], R3 ; /* 0x0000000316007986 */ /* 0x000fe2000c101924 */ /*0aa0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0ab0*/ SHF.R.U32.HI R6, RZ, 0x17, R7 ; /* 0x00000017ff067819 */ /* 0x000fe20000011607 */ /*0ac0*/ BSSY B2, 0x1100 ; /* 0x0000063000027945 */ /* 0x000fe20003800000 */ /*0ad0*/ SHF.R.U32.HI R4, RZ, 0x17, R8.reuse ; /* 0x00000017ff047819 */ /* 0x100fe40000011608 */ /*0ae0*/ LOP3.LUT R6, R6, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff06067812 */ /* 0x000fe400078ec0ff */ /*0af0*/ LOP3.LUT R10, R4, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff040a7812 */ /* 0x000fe200078ec0ff */ /*0b00*/ IMAD.MOV.U32 R4, RZ, RZ, R8 ; /* 0x000000ffff047224 */ /* 0x000fe200078e0008 */ /*0b10*/ IADD3 R11, R6, -0x1, RZ ; /* 0xffffffff060b7810 */ /* 0x000fe40007ffe0ff */ /*0b20*/ IADD3 R9, R10, -0x1, RZ ; /* 0xffffffff0a097810 */ /* 0x000fc40007ffe0ff */ /*0b30*/ ISETP.GT.U32.AND P0, PT, R11, 0xfd, PT ; /* 0x000000fd0b00780c */ /* 0x000fc80003f04070 */ /*0b40*/ ISETP.GT.U32.OR P0, PT, R9, 0xfd, P0 ; /* 0x000000fd0900780c */ /* 0x000fda0000704470 */ /*0b50*/ @!P0 IMAD.MOV.U32 R5, RZ, RZ, RZ ; /* 0x000000ffff058224 */ /* 0x000fe200078e00ff */ /*0b60*/ @!P0 BRA 0xce0 ; /* 0x0000017000008947 */ /* 0x000fea0003800000 */ /*0b70*/ FSETP.GTU.FTZ.AND P0, PT, |R8|, +INF , PT ; /* 0x7f8000000800780b */ /* 0x000fe40003f1c200 */ /*0b80*/ FSETP.GTU.FTZ.AND P1, PT, |R7|, +INF , PT ; /* 0x7f8000000700780b */ /* 0x000fc80003f3c200 */ /*0b90*/ PLOP3.LUT P0, PT, P0, P1, PT, 0xa8, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000703570 */ /*0ba0*/ @P0 BRA 0x10e0 ; /* 0x0000053000000947 */ /* 0x000fea0003800000 */ /*0bb0*/ LOP3.LUT P0, RZ, R7, 0x7fffffff, R4, 0xc8, !PT ; /* 0x7fffffff07ff7812 */ /* 0x000fda000780c804 */ /*0bc0*/ @!P0 BRA 0x10c0 ; /* 0x000004f000008947 */ /* 0x000fea0003800000 */ /*0bd0*/ FSETP.NEU.FTZ.AND P2, PT, |R8|.reuse, +INF , PT ; /* 0x7f8000000800780b */ /* 0x040fe40003f5d200 */ /*0be0*/ FSETP.NEU.FTZ.AND P1, PT, |R7|, +INF , PT ; /* 0x7f8000000700780b */ /* 0x000fe40003f3d200 */ /*0bf0*/ FSETP.NEU.FTZ.AND P0, PT, |R8|, +INF , PT ; /* 0x7f8000000800780b */ /* 0x000fd60003f1d200 */ /*0c00*/ @!P1 BRA !P2, 0x10c0 ; /* 0x000004b000009947 */ /* 0x000fea0005000000 */ /*0c10*/ LOP3.LUT P2, RZ, R4, 0x7fffffff, RZ, 0xc0, !PT ; /* 0x7fffffff04ff7812 */ /* 0x000fc8000784c0ff */ /*0c20*/ PLOP3.LUT P1, PT, P1, P2, PT, 0x2a, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000f24572 */ /*0c30*/ @P1 BRA 0x10a0 ; /* 0x0000046000001947 */ /* 0x000fea0003800000 */ /*0c40*/ LOP3.LUT P1, RZ, R7, 0x7fffffff, RZ, 0xc0, !PT ; /* 0x7fffffff07ff7812 */ /* 0x000fc8000782c0ff */ /*0c50*/ PLOP3.LUT P0, PT, P0, P1, PT, 0x2a, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000702572 */ /*0c60*/ @P0 BRA 0x1070 ; /* 0x0000040000000947 */ /* 0x000fea0003800000 */ /*0c70*/ ISETP.GE.AND P0, PT, R9, RZ, PT ; /* 0x000000ff0900720c */ /* 0x000fe40003f06270 */ /*0c80*/ ISETP.GE.AND P1, PT, R11, RZ, PT ; /* 0x000000ff0b00720c */ /* 0x000fd60003f26270 */ /*0c90*/ @P0 MOV R5, RZ ; /* 0x000000ff00050202 */ /* 0x000fe20000000f00 */ /*0ca0*/ @!P0 IMAD.MOV.U32 R5, RZ, RZ, -0x40 ; /* 0xffffffc0ff058424 */ /* 0x000fe400078e00ff */ /*0cb0*/ @!P0 FFMA R4, R8, 1.84467440737095516160e+19, RZ ; /* 0x5f80000008048823 */ /* 0x000fe400000000ff */ /*0cc0*/ @!P1 FFMA R7, R7, 1.84467440737095516160e+19, RZ ; /* 0x5f80000007079823 */ /* 0x000fe200000000ff */ /*0cd0*/ @!P1 IADD3 R5, R5, 0x40, RZ ; /* 0x0000004005059810 */ /* 0x000fe40007ffe0ff */ /*0ce0*/ LEA R8, R6, 0xc0800000, 0x17 ; /* 0xc080000006087811 */ /* 0x000fe200078eb8ff */ /*0cf0*/ BSSY B3, 0x1060 ; /* 0x0000036000037945 */ /* 0x000fe80003800000 */ /*0d00*/ IMAD.IADD R8, R7, 0x1, -R8 ; /* 0x0000000107087824 */ /* 0x000fe200078e0a08 */ /*0d10*/ IADD3 R7, R10, -0x7f, RZ ; /* 0xffffff810a077810 */ /* 0x000fc60007ffe0ff */ /*0d20*/ MUFU.RCP R9, R8 ; /* 0x0000000800097308 */ /* 0x0000620000001000 */ /*0d30*/ FADD.FTZ R11, -R8, -RZ ; /* 0x800000ff080b7221 */ /* 0x000fe40000010100 */ /*0d40*/ IMAD R4, R7.reuse, -0x800000, R4 ; /* 0xff80000007047824 */ /* 0x040fe200078e0204 */ /*0d50*/ IADD3 R8, R7, 0x7f, -R6 ; /* 0x0000007f07087810 */ /* 0x001fca0007ffe806 */ /*0d60*/ IMAD.IADD R5, R8, 0x1, R5 ; /* 0x0000000108057824 */ /* 0x000fe400078e0205 */ /*0d70*/ FFMA R10, R9, R11, 1 ; /* 0x3f800000090a7423 */ /* 0x002fc8000000000b */ /*0d80*/ FFMA R13, R9, R10, R9 ; /* 0x0000000a090d7223 */ /* 0x000fc80000000009 */ /*0d90*/ FFMA R9, R4, R13, RZ ; /* 0x0000000d04097223 */ /* 0x000fc800000000ff */ /*0da0*/ FFMA R10, R11, R9, R4 ; /* 0x000000090b0a7223 */ /* 0x000fc80000000004 */ /*0db0*/ FFMA R10, R13, R10, R9 ; /* 0x0000000a0d0a7223 */ /* 0x000fc80000000009 */ /*0dc0*/ FFMA R11, R11, R10, R4 ; /* 0x0000000a0b0b7223 */ /* 0x000fc80000000004 */ /*0dd0*/ FFMA R4, R13, R11, R10 ; /* 0x0000000b0d047223 */ /* 0x000fca000000000a */ /*0de0*/ SHF.R.U32.HI R6, RZ, 0x17, R4 ; /* 0x00000017ff067819 */ /* 0x000fc80000011604 */ /*0df0*/ LOP3.LUT R6, R6, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff06067812 */ /* 0x000fc800078ec0ff */ /*0e00*/ IADD3 R9, R6, R5, RZ ; /* 0x0000000506097210 */ /* 0x000fc80007ffe0ff */ /*0e10*/ IADD3 R6, R9, -0x1, RZ ; /* 0xffffffff09067810 */ /* 0x000fc80007ffe0ff */ /*0e20*/ ISETP.GE.U32.AND P0, PT, R6, 0xfe, PT ; /* 0x000000fe0600780c */ /* 0x000fda0003f06070 */ /*0e30*/ @!P0 BRA 0x1040 ; /* 0x0000020000008947 */ /* 0x000fea0003800000 */ /*0e40*/ ISETP.GT.AND P0, PT, R9, 0xfe, PT ; /* 0x000000fe0900780c */ /* 0x000fda0003f04270 */ /*0e50*/ @P0 BRA 0x1010 ; /* 0x000001b000000947 */ /* 0x000fea0003800000 */ /*0e60*/ ISETP.GE.AND P0, PT, R9, 0x1, PT ; /* 0x000000010900780c */ /* 0x000fda0003f06270 */ /*0e70*/ @P0 BRA 0x1050 ; /* 0x000001d000000947 */ /* 0x000fea0003800000 */ /*0e80*/ ISETP.GE.AND P0, PT, R9, -0x18, PT ; /* 0xffffffe80900780c */ /* 0x000fe40003f06270 */ /*0e90*/ LOP3.LUT R4, R4, 0x80000000, RZ, 0xc0, !PT ; /* 0x8000000004047812 */ /* 0x000fd600078ec0ff */ /*0ea0*/ @!P0 BRA 0x1050 ; /* 0x000001a000008947 */ /* 0x000fea0003800000 */ /*0eb0*/ FFMA.RZ R5, R13, R11.reuse, R10.reuse ; /* 0x0000000b0d057223 */ /* 0x180fe2000000c00a */ /*0ec0*/ IADD3 R8, R9, 0x20, RZ ; /* 0x0000002009087810 */ /* 0x000fe20007ffe0ff */ /*0ed0*/ FFMA.RM R6, R13, R11.reuse, R10.reuse ; /* 0x0000000b0d067223 */ /* 0x180fe2000000400a */ /*0ee0*/ ISETP.NE.AND P2, PT, R9, RZ, PT ; /* 0x000000ff0900720c */ /* 0x000fe40003f45270 */ /*0ef0*/ LOP3.LUT R7, R5, 0x7fffff, RZ, 0xc0, !PT ; /* 0x007fffff05077812 */ /* 0x000fe200078ec0ff */ /*0f00*/ FFMA.RP R5, R13, R11, R10 ; /* 0x0000000b0d057223 */ /* 0x000fe2000000800a */ /*0f10*/ ISETP.NE.AND P1, PT, R9, RZ, PT ; /* 0x000000ff0900720c */ /* 0x000fe20003f25270 */ /*0f20*/ IMAD.MOV R9, RZ, RZ, -R9 ; /* 0x000000ffff097224 */ /* 0x000fe200078e0a09 */ /*0f30*/ LOP3.LUT R7, R7, 0x800000, RZ, 0xfc, !PT ; /* 0x0080000007077812 */ /* 0x000fe400078efcff */ /*0f40*/ FSETP.NEU.FTZ.AND P0, PT, R5, R6, PT ; /* 0x000000060500720b */ /* 0x000fc40003f1d000 */ /*0f50*/ SHF.L.U32 R8, R7, R8, RZ ; /* 0x0000000807087219 */ /* 0x000fe400000006ff */ /*0f60*/ SEL R6, R9, RZ, P2 ; /* 0x000000ff09067207 */ /* 0x000fe40001000000 */ /*0f70*/ ISETP.NE.AND P1, PT, R8, RZ, P1 ; /* 0x000000ff0800720c */ /* 0x000fe40000f25270 */ /*0f80*/ SHF.R.U32.HI R6, RZ, R6, R7 ; /* 0x00000006ff067219 */ /* 0x000fe40000011607 */ /*0f90*/ PLOP3.LUT P0, PT, P0, P1, PT, 0xa8, 0x0 ; /* 0x000000000000781c */ /* 0x000fe40000703570 */ /*0fa0*/ SHF.R.U32.HI R8, RZ, 0x1, R6 ; /* 0x00000001ff087819 */ /* 0x000fc40000011606 */ /*0fb0*/ SEL R5, RZ, 0x1, !P0 ; /* 0x00000001ff057807 */ /* 0x000fc80004000000 */ /*0fc0*/ LOP3.LUT R5, R5, 0x1, R8, 0xf8, !PT ; /* 0x0000000105057812 */ /* 0x000fc800078ef808 */ /*0fd0*/ LOP3.LUT R5, R5, R6, RZ, 0xc0, !PT ; /* 0x0000000605057212 */ /* 0x000fca00078ec0ff */ /*0fe0*/ IMAD.IADD R5, R8, 0x1, R5 ; /* 0x0000000108057824 */ /* 0x000fca00078e0205 */ /*0ff0*/ LOP3.LUT R4, R5, R4, RZ, 0xfc, !PT ; /* 0x0000000405047212 */ /* 0x000fe200078efcff */ /*1000*/ BRA 0x1050 ; /* 0x0000004000007947 */ /* 0x000fea0003800000 */ /*1010*/ LOP3.LUT R4, R4, 0x80000000, RZ, 0xc0, !PT ; /* 0x8000000004047812 */ /* 0x000fc800078ec0ff */ /*1020*/ LOP3.LUT R4, R4, 0x7f800000, RZ, 0xfc, !PT ; /* 0x7f80000004047812 */ /* 0x000fe200078efcff */ /*1030*/ BRA 0x1050 ; /* 0x0000001000007947 */ /* 0x000fea0003800000 */ /*1040*/ IMAD R4, R5, 0x800000, R4 ; /* 0x0080000005047824 */ /* 0x000fe400078e0204 */ /*1050*/ BSYNC B3 ; /* 0x0000000000037941 */ /* 0x000fea0003800000 */ /*1060*/ BRA 0x10f0 ; /* 0x0000008000007947 */ /* 0x000fea0003800000 */ /*1070*/ LOP3.LUT R4, R7, 0x80000000, R4, 0x48, !PT ; /* 0x8000000007047812 */ /* 0x000fc800078e4804 */ /*1080*/ LOP3.LUT R4, R4, 0x7f800000, RZ, 0xfc, !PT ; /* 0x7f80000004047812 */ /* 0x000fe200078efcff */ /*1090*/ BRA 0x10f0 ; /* 0x0000005000007947 */ /* 0x000fea0003800000 */ /*10a0*/ LOP3.LUT R4, R7, 0x80000000, R4, 0x48, !PT ; /* 0x8000000007047812 */ /* 0x000fe200078e4804 */ /*10b0*/ BRA 0x10f0 ; /* 0x0000003000007947 */ /* 0x000fea0003800000 */ /*10c0*/ MUFU.RSQ R4, -QNAN ; /* 0xffc0000000047908 */ /* 0x000e220000001400 */ /*10d0*/ BRA 0x10f0 ; /* 0x0000001000007947 */ /* 0x000fea0003800000 */ /*10e0*/ FADD.FTZ R4, R8, R7 ; /* 0x0000000708047221 */ /* 0x000fe40000010000 */ /*10f0*/ BSYNC B2 ; /* 0x0000000000027941 */ /* 0x000fea0003800000 */ /*1100*/ MOV R8, R4 ; /* 0x0000000400087202 */ /* 0x001fe20000000f00 */ /*1110*/ IMAD.MOV.U32 R4, RZ, RZ, R2 ; /* 0x000000ffff047224 */ /* 0x000fe400078e0002 */ /*1120*/ IMAD.MOV.U32 R5, RZ, RZ, 0x0 ; /* 0x00000000ff057424 */ /* 0x000fc800078e00ff */ /*1130*/ RET.REL.NODEC R4 0x0 ; /* 0xffffeec004007950 */ /* 0x000fea0003c3ffff */ /*1140*/ BRA 0x1140; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*1150*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1160*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1170*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*1180*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*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 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include <stdio.h> #include <math.h> void __global__ kernel_linearInterpolation(float* array_device, int* rowArray, int rowArrayLength, int col, int colArrayLength, int totalCols, int totalRows, float* originalTimes, int originalTimesLength, float* newTimes, int newTimesLength, float* results_device) { // where original times and new times should probably be unix time? but technically won't matter probably int n = blockIdx.x * blockDim.x + threadIdx.x; //int m = blockIdx.y * blockDim.y + threadIdx.y; int m = col; if (n < newTimesLength && m < colArrayLength) { // so we created a row-wise thread per element in the newTimes, with m being number of columns float newTime = newTimes[n]; int resultsInd = n*colArrayLength + m; // need to potentially handle the first and last points in a special way ,,, if (n==0 && newTimes[0] < originalTimes[0]) { // lets say if newTimes[0] is prior to originalTimes[0] less than the distance from originalTimes[1] to originalTimes[2], we'll accept it // otherwise call it a NaN if ( originalTimes[0] - newTimes[0] < originalTimes[1] - originalTimes[0] ) { // continue slope from points 0 to 1? // ie assume there's a previous point at same slope and same sample rate int arrayInd1 = (0)*totalCols + m; int arrayInd2 = (1)*totalCols + m; float valueChange = array_device[arrayInd2] - array_device[arrayInd1]; float timeDiff = originalTimes[1] - originalTimes[0]; float pseudoPointTime = originalTimes[0] - timeDiff; float pseudoPointValue = array_device[arrayInd1] - valueChange; float perc = (newTime - pseudoPointTime) / timeDiff; results_device[resultsInd] = pseudoPointValue + perc*valueChange; } else results_device[resultsInd] = NAN; return; } // ================================================ // ================================================ // ===== DO THIS SAME LOGIC FOR LAST POINT? ======= // ================================================ // ================================================ // we now need to find the points before and after ... // do we need to iterate through the entire originalTimes vector or is there some way we can approximate the starting location? //float origSampleRate = (originalTimes[originalTimesLength-1] - originalTimes[0]) / originalTimesLength; //float newSampleRate = (newTimes[newTimesLength-1] - newTimes[0]) / newTimesLength; //int originalTimeLoc = (newTime - newTimes[0]) / (originalTimes[originalTimesLength-1] - originalTimes) float newTimePerc = (newTime - newTimes[0]) / (newTimes[newTimesLength-1] - newTimes[0]); int estimation = newTimePerc * originalTimesLength; // start at approximately the correct point int iterationsToFindMatch = 0; int originalTimeLoc; if (originalTimes[estimation] > newTime) { // time we guessed is greater than the time we want ... so walk backwards? for (originalTimeLoc=estimation; originalTimeLoc>0; originalTimeLoc--) { if (originalTimes[originalTimeLoc] < newTime) break; iterationsToFindMatch++; } } else { // the time we guessed is greater than the time we want ... so walk forwards? for (originalTimeLoc=estimation; originalTimeLoc<originalTimesLength; originalTimeLoc++) { if (originalTimes[originalTimeLoc] > newTime) { originalTimeLoc--; // want the left side point break; } iterationsToFindMatch++; } } printf("Loc %i needed %i iterations, starting from %i -> %i\n", n, iterationsToFindMatch, estimation, originalTimeLoc); int arrayInd1 = originalTimeLoc*totalCols + m; int arrayInd2 = (originalTimeLoc+1)*totalCols + m; float valueChange = (array_device[arrayInd2] - array_device[arrayInd1]); // we need to figure out what percentage the newTime is between the two points float origTime1 = originalTimes[originalTimeLoc]; float origTime2 = originalTimes[originalTimeLoc+1]; float perc = (newTime - origTime1) / (origTime2 - origTime1); results_device[resultsInd] = array_device[arrayInd1] + perc*valueChange; } }
.file "tmpxft_0012dc61_00000000-6_kernel_interpolate.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2059: .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 .LFE2059: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z61__device_stub__Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_PfPiiiiiiS_iS_iS_ .type _Z61__device_stub__Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_PfPiiiiiiS_iS_iS_, @function _Z61__device_stub__Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_PfPiiiiiiS_iS_iS_: .LFB2081: .cfi_startproc endbr64 subq $248, %rsp .cfi_def_cfa_offset 256 movq %rdi, 56(%rsp) movq %rsi, 48(%rsp) movl %edx, 44(%rsp) movl %ecx, 40(%rsp) movl %r8d, 36(%rsp) movl %r9d, 32(%rsp) movq 264(%rsp), %rax movq %rax, 24(%rsp) movq 280(%rsp), %rax movq %rax, 16(%rsp) movq 296(%rsp), %rax movq %rax, 8(%rsp) movq %fs:40, %rax movq %rax, 232(%rsp) xorl %eax, %eax leaq 56(%rsp), %rax movq %rax, 128(%rsp) leaq 48(%rsp), %rax movq %rax, 136(%rsp) leaq 44(%rsp), %rax movq %rax, 144(%rsp) leaq 40(%rsp), %rax movq %rax, 152(%rsp) leaq 36(%rsp), %rax movq %rax, 160(%rsp) leaq 32(%rsp), %rax movq %rax, 168(%rsp) leaq 256(%rsp), %rax movq %rax, 176(%rsp) leaq 24(%rsp), %rax movq %rax, 184(%rsp) leaq 272(%rsp), %rax movq %rax, 192(%rsp) leaq 16(%rsp), %rax movq %rax, 200(%rsp) leaq 288(%rsp), %rax movq %rax, 208(%rsp) leaq 8(%rsp), %rax movq %rax, 216(%rsp) movl $1, 80(%rsp) movl $1, 84(%rsp) movl $1, 88(%rsp) movl $1, 92(%rsp) movl $1, 96(%rsp) movl $1, 100(%rsp) leaq 72(%rsp), %rcx leaq 64(%rsp), %rdx leaq 92(%rsp), %rsi leaq 80(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 232(%rsp), %rax subq %fs:40, %rax jne .L8 addq $248, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 72(%rsp) .cfi_def_cfa_offset 264 pushq 72(%rsp) .cfi_def_cfa_offset 272 leaq 144(%rsp), %r9 movq 108(%rsp), %rcx movl 116(%rsp), %r8d movq 96(%rsp), %rsi movl 104(%rsp), %edx leaq _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 256 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2081: .size _Z61__device_stub__Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_PfPiiiiiiS_iS_iS_, .-_Z61__device_stub__Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_PfPiiiiiiS_iS_iS_ .globl _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .type _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_, @function _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_: .LFB2082: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 pushq 56(%rsp) .cfi_def_cfa_offset 24 movl 56(%rsp), %eax pushq %rax .cfi_def_cfa_offset 32 pushq 56(%rsp) .cfi_def_cfa_offset 40 movl 56(%rsp), %eax pushq %rax .cfi_def_cfa_offset 48 pushq 56(%rsp) .cfi_def_cfa_offset 56 movl 56(%rsp), %eax pushq %rax .cfi_def_cfa_offset 64 call _Z61__device_stub__Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_PfPiiiiiiS_iS_iS_ addq $56, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2082: .size _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_, .-_Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "_Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2084: .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 _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_(%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 .LFE2084: .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 <math.h> void __global__ kernel_linearInterpolation(float* array_device, int* rowArray, int rowArrayLength, int col, int colArrayLength, int totalCols, int totalRows, float* originalTimes, int originalTimesLength, float* newTimes, int newTimesLength, float* results_device) { // where original times and new times should probably be unix time? but technically won't matter probably int n = blockIdx.x * blockDim.x + threadIdx.x; //int m = blockIdx.y * blockDim.y + threadIdx.y; int m = col; if (n < newTimesLength && m < colArrayLength) { // so we created a row-wise thread per element in the newTimes, with m being number of columns float newTime = newTimes[n]; int resultsInd = n*colArrayLength + m; // need to potentially handle the first and last points in a special way ,,, if (n==0 && newTimes[0] < originalTimes[0]) { // lets say if newTimes[0] is prior to originalTimes[0] less than the distance from originalTimes[1] to originalTimes[2], we'll accept it // otherwise call it a NaN if ( originalTimes[0] - newTimes[0] < originalTimes[1] - originalTimes[0] ) { // continue slope from points 0 to 1? // ie assume there's a previous point at same slope and same sample rate int arrayInd1 = (0)*totalCols + m; int arrayInd2 = (1)*totalCols + m; float valueChange = array_device[arrayInd2] - array_device[arrayInd1]; float timeDiff = originalTimes[1] - originalTimes[0]; float pseudoPointTime = originalTimes[0] - timeDiff; float pseudoPointValue = array_device[arrayInd1] - valueChange; float perc = (newTime - pseudoPointTime) / timeDiff; results_device[resultsInd] = pseudoPointValue + perc*valueChange; } else results_device[resultsInd] = NAN; return; } // ================================================ // ================================================ // ===== DO THIS SAME LOGIC FOR LAST POINT? ======= // ================================================ // ================================================ // we now need to find the points before and after ... // do we need to iterate through the entire originalTimes vector or is there some way we can approximate the starting location? //float origSampleRate = (originalTimes[originalTimesLength-1] - originalTimes[0]) / originalTimesLength; //float newSampleRate = (newTimes[newTimesLength-1] - newTimes[0]) / newTimesLength; //int originalTimeLoc = (newTime - newTimes[0]) / (originalTimes[originalTimesLength-1] - originalTimes) float newTimePerc = (newTime - newTimes[0]) / (newTimes[newTimesLength-1] - newTimes[0]); int estimation = newTimePerc * originalTimesLength; // start at approximately the correct point int iterationsToFindMatch = 0; int originalTimeLoc; if (originalTimes[estimation] > newTime) { // time we guessed is greater than the time we want ... so walk backwards? for (originalTimeLoc=estimation; originalTimeLoc>0; originalTimeLoc--) { if (originalTimes[originalTimeLoc] < newTime) break; iterationsToFindMatch++; } } else { // the time we guessed is greater than the time we want ... so walk forwards? for (originalTimeLoc=estimation; originalTimeLoc<originalTimesLength; originalTimeLoc++) { if (originalTimes[originalTimeLoc] > newTime) { originalTimeLoc--; // want the left side point break; } iterationsToFindMatch++; } } printf("Loc %i needed %i iterations, starting from %i -> %i\n", n, iterationsToFindMatch, estimation, originalTimeLoc); int arrayInd1 = originalTimeLoc*totalCols + m; int arrayInd2 = (originalTimeLoc+1)*totalCols + m; float valueChange = (array_device[arrayInd2] - array_device[arrayInd1]); // we need to figure out what percentage the newTime is between the two points float origTime1 = originalTimes[originalTimeLoc]; float origTime2 = originalTimes[originalTimeLoc+1]; float perc = (newTime - origTime1) / (origTime2 - origTime1); results_device[resultsInd] = array_device[arrayInd1] + perc*valueChange; } }
#include <hip/hip_runtime.h> #include <stdio.h> #include <math.h> void __global__ kernel_linearInterpolation(float* array_device, int* rowArray, int rowArrayLength, int col, int colArrayLength, int totalCols, int totalRows, float* originalTimes, int originalTimesLength, float* newTimes, int newTimesLength, float* results_device) { // where original times and new times should probably be unix time? but technically won't matter probably int n = blockIdx.x * blockDim.x + threadIdx.x; //int m = blockIdx.y * blockDim.y + threadIdx.y; int m = col; if (n < newTimesLength && m < colArrayLength) { // so we created a row-wise thread per element in the newTimes, with m being number of columns float newTime = newTimes[n]; int resultsInd = n*colArrayLength + m; // need to potentially handle the first and last points in a special way ,,, if (n==0 && newTimes[0] < originalTimes[0]) { // lets say if newTimes[0] is prior to originalTimes[0] less than the distance from originalTimes[1] to originalTimes[2], we'll accept it // otherwise call it a NaN if ( originalTimes[0] - newTimes[0] < originalTimes[1] - originalTimes[0] ) { // continue slope from points 0 to 1? // ie assume there's a previous point at same slope and same sample rate int arrayInd1 = (0)*totalCols + m; int arrayInd2 = (1)*totalCols + m; float valueChange = array_device[arrayInd2] - array_device[arrayInd1]; float timeDiff = originalTimes[1] - originalTimes[0]; float pseudoPointTime = originalTimes[0] - timeDiff; float pseudoPointValue = array_device[arrayInd1] - valueChange; float perc = (newTime - pseudoPointTime) / timeDiff; results_device[resultsInd] = pseudoPointValue + perc*valueChange; } else results_device[resultsInd] = NAN; return; } // ================================================ // ================================================ // ===== DO THIS SAME LOGIC FOR LAST POINT? ======= // ================================================ // ================================================ // we now need to find the points before and after ... // do we need to iterate through the entire originalTimes vector or is there some way we can approximate the starting location? //float origSampleRate = (originalTimes[originalTimesLength-1] - originalTimes[0]) / originalTimesLength; //float newSampleRate = (newTimes[newTimesLength-1] - newTimes[0]) / newTimesLength; //int originalTimeLoc = (newTime - newTimes[0]) / (originalTimes[originalTimesLength-1] - originalTimes) float newTimePerc = (newTime - newTimes[0]) / (newTimes[newTimesLength-1] - newTimes[0]); int estimation = newTimePerc * originalTimesLength; // start at approximately the correct point int iterationsToFindMatch = 0; int originalTimeLoc; if (originalTimes[estimation] > newTime) { // time we guessed is greater than the time we want ... so walk backwards? for (originalTimeLoc=estimation; originalTimeLoc>0; originalTimeLoc--) { if (originalTimes[originalTimeLoc] < newTime) break; iterationsToFindMatch++; } } else { // the time we guessed is greater than the time we want ... so walk forwards? for (originalTimeLoc=estimation; originalTimeLoc<originalTimesLength; originalTimeLoc++) { if (originalTimes[originalTimeLoc] > newTime) { originalTimeLoc--; // want the left side point break; } iterationsToFindMatch++; } } printf("Loc %i needed %i iterations, starting from %i -> %i\n", n, iterationsToFindMatch, estimation, originalTimeLoc); int arrayInd1 = originalTimeLoc*totalCols + m; int arrayInd2 = (originalTimeLoc+1)*totalCols + m; float valueChange = (array_device[arrayInd2] - array_device[arrayInd1]); // we need to figure out what percentage the newTime is between the two points float origTime1 = originalTimes[originalTimeLoc]; float origTime2 = originalTimes[originalTimeLoc+1]; float perc = (newTime - origTime1) / (origTime2 - origTime1); results_device[resultsInd] = array_device[arrayInd1] + perc*valueChange; } }
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 <math.h> void __global__ kernel_linearInterpolation(float* array_device, int* rowArray, int rowArrayLength, int col, int colArrayLength, int totalCols, int totalRows, float* originalTimes, int originalTimesLength, float* newTimes, int newTimesLength, float* results_device) { // where original times and new times should probably be unix time? but technically won't matter probably int n = blockIdx.x * blockDim.x + threadIdx.x; //int m = blockIdx.y * blockDim.y + threadIdx.y; int m = col; if (n < newTimesLength && m < colArrayLength) { // so we created a row-wise thread per element in the newTimes, with m being number of columns float newTime = newTimes[n]; int resultsInd = n*colArrayLength + m; // need to potentially handle the first and last points in a special way ,,, if (n==0 && newTimes[0] < originalTimes[0]) { // lets say if newTimes[0] is prior to originalTimes[0] less than the distance from originalTimes[1] to originalTimes[2], we'll accept it // otherwise call it a NaN if ( originalTimes[0] - newTimes[0] < originalTimes[1] - originalTimes[0] ) { // continue slope from points 0 to 1? // ie assume there's a previous point at same slope and same sample rate int arrayInd1 = (0)*totalCols + m; int arrayInd2 = (1)*totalCols + m; float valueChange = array_device[arrayInd2] - array_device[arrayInd1]; float timeDiff = originalTimes[1] - originalTimes[0]; float pseudoPointTime = originalTimes[0] - timeDiff; float pseudoPointValue = array_device[arrayInd1] - valueChange; float perc = (newTime - pseudoPointTime) / timeDiff; results_device[resultsInd] = pseudoPointValue + perc*valueChange; } else results_device[resultsInd] = NAN; return; } // ================================================ // ================================================ // ===== DO THIS SAME LOGIC FOR LAST POINT? ======= // ================================================ // ================================================ // we now need to find the points before and after ... // do we need to iterate through the entire originalTimes vector or is there some way we can approximate the starting location? //float origSampleRate = (originalTimes[originalTimesLength-1] - originalTimes[0]) / originalTimesLength; //float newSampleRate = (newTimes[newTimesLength-1] - newTimes[0]) / newTimesLength; //int originalTimeLoc = (newTime - newTimes[0]) / (originalTimes[originalTimesLength-1] - originalTimes) float newTimePerc = (newTime - newTimes[0]) / (newTimes[newTimesLength-1] - newTimes[0]); int estimation = newTimePerc * originalTimesLength; // start at approximately the correct point int iterationsToFindMatch = 0; int originalTimeLoc; if (originalTimes[estimation] > newTime) { // time we guessed is greater than the time we want ... so walk backwards? for (originalTimeLoc=estimation; originalTimeLoc>0; originalTimeLoc--) { if (originalTimes[originalTimeLoc] < newTime) break; iterationsToFindMatch++; } } else { // the time we guessed is greater than the time we want ... so walk forwards? for (originalTimeLoc=estimation; originalTimeLoc<originalTimesLength; originalTimeLoc++) { if (originalTimes[originalTimeLoc] > newTime) { originalTimeLoc--; // want the left side point break; } iterationsToFindMatch++; } } printf("Loc %i needed %i iterations, starting from %i -> %i\n", n, iterationsToFindMatch, estimation, originalTimeLoc); int arrayInd1 = originalTimeLoc*totalCols + m; int arrayInd2 = (originalTimeLoc+1)*totalCols + m; float valueChange = (array_device[arrayInd2] - array_device[arrayInd1]); // we need to figure out what percentage the newTime is between the two points float origTime1 = originalTimes[originalTimeLoc]; float origTime2 = originalTimes[originalTimeLoc+1]; float perc = (newTime - origTime1) / (origTime2 - origTime1); results_device[resultsInd] = array_device[arrayInd1] + perc*valueChange; } }
.text .file "kernel_interpolate.hip" .globl _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_ # -- Begin function _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .p2align 4, 0x90 .type _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_,@function _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_: # @_Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .cfi_startproc # %bb.0: subq $184, %rsp .cfi_def_cfa_offset 192 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 192(%rsp), %rax movq %rax, 128(%rsp) leaq 200(%rsp), %rax movq %rax, 136(%rsp) leaq 208(%rsp), %rax movq %rax, 144(%rsp) leaq 216(%rsp), %rax movq %rax, 152(%rsp) leaq 224(%rsp), %rax movq %rax, 160(%rsp) leaq 232(%rsp), %rax movq %rax, 168(%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 $_Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $200, %rsp .cfi_adjust_cfa_offset -200 retq .Lfunc_end0: .size _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_, .Lfunc_end0-_Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .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 $_Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_, %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 _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_,@object # @_Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .section .rodata,"a",@progbits .globl _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .p2align 3, 0x0 _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_: .quad _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .size _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_" .size .L__unnamed_1, 48 .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 _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .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_0012dc61_00000000-6_kernel_interpolate.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2059: .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 .LFE2059: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z61__device_stub__Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_PfPiiiiiiS_iS_iS_ .type _Z61__device_stub__Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_PfPiiiiiiS_iS_iS_, @function _Z61__device_stub__Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_PfPiiiiiiS_iS_iS_: .LFB2081: .cfi_startproc endbr64 subq $248, %rsp .cfi_def_cfa_offset 256 movq %rdi, 56(%rsp) movq %rsi, 48(%rsp) movl %edx, 44(%rsp) movl %ecx, 40(%rsp) movl %r8d, 36(%rsp) movl %r9d, 32(%rsp) movq 264(%rsp), %rax movq %rax, 24(%rsp) movq 280(%rsp), %rax movq %rax, 16(%rsp) movq 296(%rsp), %rax movq %rax, 8(%rsp) movq %fs:40, %rax movq %rax, 232(%rsp) xorl %eax, %eax leaq 56(%rsp), %rax movq %rax, 128(%rsp) leaq 48(%rsp), %rax movq %rax, 136(%rsp) leaq 44(%rsp), %rax movq %rax, 144(%rsp) leaq 40(%rsp), %rax movq %rax, 152(%rsp) leaq 36(%rsp), %rax movq %rax, 160(%rsp) leaq 32(%rsp), %rax movq %rax, 168(%rsp) leaq 256(%rsp), %rax movq %rax, 176(%rsp) leaq 24(%rsp), %rax movq %rax, 184(%rsp) leaq 272(%rsp), %rax movq %rax, 192(%rsp) leaq 16(%rsp), %rax movq %rax, 200(%rsp) leaq 288(%rsp), %rax movq %rax, 208(%rsp) leaq 8(%rsp), %rax movq %rax, 216(%rsp) movl $1, 80(%rsp) movl $1, 84(%rsp) movl $1, 88(%rsp) movl $1, 92(%rsp) movl $1, 96(%rsp) movl $1, 100(%rsp) leaq 72(%rsp), %rcx leaq 64(%rsp), %rdx leaq 92(%rsp), %rsi leaq 80(%rsp), %rdi call __cudaPopCallConfiguration@PLT testl %eax, %eax je .L7 .L3: movq 232(%rsp), %rax subq %fs:40, %rax jne .L8 addq $248, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L7: .cfi_restore_state pushq 72(%rsp) .cfi_def_cfa_offset 264 pushq 72(%rsp) .cfi_def_cfa_offset 272 leaq 144(%rsp), %r9 movq 108(%rsp), %rcx movl 116(%rsp), %r8d movq 96(%rsp), %rsi movl 104(%rsp), %edx leaq _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 256 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2081: .size _Z61__device_stub__Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_PfPiiiiiiS_iS_iS_, .-_Z61__device_stub__Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_PfPiiiiiiS_iS_iS_ .globl _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .type _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_, @function _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_: .LFB2082: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 pushq 56(%rsp) .cfi_def_cfa_offset 24 movl 56(%rsp), %eax pushq %rax .cfi_def_cfa_offset 32 pushq 56(%rsp) .cfi_def_cfa_offset 40 movl 56(%rsp), %eax pushq %rax .cfi_def_cfa_offset 48 pushq 56(%rsp) .cfi_def_cfa_offset 56 movl 56(%rsp), %eax pushq %rax .cfi_def_cfa_offset 64 call _Z61__device_stub__Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_PfPiiiiiiS_iS_iS_ addq $56, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2082: .size _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_, .-_Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "_Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2084: .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 _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_(%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 .LFE2084: .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 "kernel_interpolate.hip" .globl _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_ # -- Begin function _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .p2align 4, 0x90 .type _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_,@function _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_: # @_Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .cfi_startproc # %bb.0: subq $184, %rsp .cfi_def_cfa_offset 192 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 192(%rsp), %rax movq %rax, 128(%rsp) leaq 200(%rsp), %rax movq %rax, 136(%rsp) leaq 208(%rsp), %rax movq %rax, 144(%rsp) leaq 216(%rsp), %rax movq %rax, 152(%rsp) leaq 224(%rsp), %rax movq %rax, 160(%rsp) leaq 232(%rsp), %rax movq %rax, 168(%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 $_Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $200, %rsp .cfi_adjust_cfa_offset -200 retq .Lfunc_end0: .size _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_, .Lfunc_end0-_Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .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 $_Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_, %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 _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_,@object # @_Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .section .rodata,"a",@progbits .globl _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .p2align 3, 0x0 _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_: .quad _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .size _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_" .size .L__unnamed_1, 48 .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 _Z41__device_stub__kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z26kernel_linearInterpolationPfPiiiiiiS_iS_iS_ .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.
// Some useful utilities // system includes #include <stdio.h> #include <assert.h> #include <cuda.h> // External function definitions void checkCUDAError(const char *msg) { cudaError_t err = cudaGetLastError(); if( cudaSuccess != err) { fprintf(stderr, "Cuda error: %s: %s.\n", msg, cudaGetErrorString( err) ); exit(EXIT_FAILURE); } } // When there is more than one device, // select the fastest device and report characteristics void selectAndReport(int *major, int *minor) { int best_gpu = 0; // Use the device with the most cores int number_of_devices; cudaGetDeviceCount(&number_of_devices); checkCUDAError("Get Count"); printf("\n%d Devices\n",number_of_devices); if (number_of_devices > 1) { int max_cores = 0; int device_number; for (device_number = 0; device_number < number_of_devices; device_number++) { cudaDeviceProp device_properties; cudaGetDeviceProperties(&device_properties, device_number); printf("Device # %d has %d cores\n",device_number, device_properties.multiProcessorCount); double gb = 1024*1024*1024; printf("Device # %d has %f GB global memory\n",device_number, ((double)device_properties.totalGlobalMem)/gb); if (max_cores < device_properties.multiProcessorCount) { max_cores = device_properties.multiProcessorCount; best_gpu = device_number; } } printf("\n *** Best GPU is: %d\n",best_gpu); cudaSetDevice(best_gpu); checkCUDAError("Can't set device\n"); } printf("\n"); // get number of SMs on this GPU int devID; cudaGetDevice(&devID); cudaDeviceProp deviceProp; cudaGetDeviceProperties(&deviceProp, devID); printf("Device %d has %d cores\n", best_gpu, deviceProp.multiProcessorCount); double gb = 1024*1024*1024; printf("Device # %d has %f GB global memory\n",best_gpu, ((double)deviceProp.totalGlobalMem)/gb); if (deviceProp.major == 9999 && deviceProp.minor == 9999) { printf("There is no device supporting CUDA.\n"); cudaThreadExit(); } else { printf("Device is a %s, capability: %d.%d\n", deviceProp.name, deviceProp.major, deviceProp.minor); *major = deviceProp.major; *minor = deviceProp.minor; } int driverVersion, runtimeVersion; assert(cudaSuccess == cudaDriverGetVersion(&driverVersion)); assert(cudaSuccess == cudaRuntimeGetVersion(&runtimeVersion)); printf("CUDA Driver version: %d, runtime version: %d\n\n", driverVersion, runtimeVersion); }
code for sm_80
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
// Some useful utilities // system includes #include <stdio.h> #include <assert.h> #include <cuda.h> // External function definitions void checkCUDAError(const char *msg) { cudaError_t err = cudaGetLastError(); if( cudaSuccess != err) { fprintf(stderr, "Cuda error: %s: %s.\n", msg, cudaGetErrorString( err) ); exit(EXIT_FAILURE); } } // When there is more than one device, // select the fastest device and report characteristics void selectAndReport(int *major, int *minor) { int best_gpu = 0; // Use the device with the most cores int number_of_devices; cudaGetDeviceCount(&number_of_devices); checkCUDAError("Get Count"); printf("\n%d Devices\n",number_of_devices); if (number_of_devices > 1) { int max_cores = 0; int device_number; for (device_number = 0; device_number < number_of_devices; device_number++) { cudaDeviceProp device_properties; cudaGetDeviceProperties(&device_properties, device_number); printf("Device # %d has %d cores\n",device_number, device_properties.multiProcessorCount); double gb = 1024*1024*1024; printf("Device # %d has %f GB global memory\n",device_number, ((double)device_properties.totalGlobalMem)/gb); if (max_cores < device_properties.multiProcessorCount) { max_cores = device_properties.multiProcessorCount; best_gpu = device_number; } } printf("\n *** Best GPU is: %d\n",best_gpu); cudaSetDevice(best_gpu); checkCUDAError("Can't set device\n"); } printf("\n"); // get number of SMs on this GPU int devID; cudaGetDevice(&devID); cudaDeviceProp deviceProp; cudaGetDeviceProperties(&deviceProp, devID); printf("Device %d has %d cores\n", best_gpu, deviceProp.multiProcessorCount); double gb = 1024*1024*1024; printf("Device # %d has %f GB global memory\n",best_gpu, ((double)deviceProp.totalGlobalMem)/gb); if (deviceProp.major == 9999 && deviceProp.minor == 9999) { printf("There is no device supporting CUDA.\n"); cudaThreadExit(); } else { printf("Device is a %s, capability: %d.%d\n", deviceProp.name, deviceProp.major, deviceProp.minor); *major = deviceProp.major; *minor = deviceProp.minor; } int driverVersion, runtimeVersion; assert(cudaSuccess == cudaDriverGetVersion(&driverVersion)); assert(cudaSuccess == cudaRuntimeGetVersion(&runtimeVersion)); printf("CUDA Driver version: %d, runtime version: %d\n\n", driverVersion, runtimeVersion); }
.file "tmpxft_00122abd_00000000-6_utils.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 .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "Cuda error: %s: %s.\n" .text .globl _Z14checkCUDAErrorPKc .type _Z14checkCUDAErrorPKc, @function _Z14checkCUDAErrorPKc: .LFB2057: .cfi_startproc endbr64 pushq %rbx .cfi_def_cfa_offset 16 .cfi_offset 3, -16 movq %rdi, %rbx call cudaGetLastError@PLT testl %eax, %eax jne .L6 popq %rbx .cfi_remember_state .cfi_def_cfa_offset 8 ret .L6: .cfi_restore_state movl %eax, %edi call cudaGetErrorString@PLT movq %rax, %r8 movq %rbx, %rcx leaq .LC0(%rip), %rdx movl $2, %esi movq stderr(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movl $1, %edi call exit@PLT .cfi_endproc .LFE2057: .size _Z14checkCUDAErrorPKc, .-_Z14checkCUDAErrorPKc .section .rodata.str1.1 .LC1: .string "Get Count" .LC2: .string "\n%d Devices\n" .LC3: .string "Device # %d has %d cores\n" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC5: .string "Device # %d has %f GB global memory\n" .section .rodata.str1.1 .LC6: .string "\n *** Best GPU is: %d\n" .LC7: .string "Can't set device\n" .LC8: .string "\n" .LC9: .string "Device %d has %d cores\n" .section .rodata.str1.8 .align 8 .LC10: .string "There is no device supporting CUDA.\n" .align 8 .LC11: .string "Device is a %s, capability: %d.%d\n" .align 8 .LC12: .string "CUDA Driver version: %d, runtime version: %d\n\n" .text .globl _Z15selectAndReportPiS_ .type _Z15selectAndReportPiS_, @function _Z15selectAndReportPiS_: .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 $1080, %rsp .cfi_def_cfa_offset 1136 movq %rdi, (%rsp) movq %rsi, 8(%rsp) movq %fs:40, %rax movq %rax, 1064(%rsp) xorl %eax, %eax leaq 24(%rsp), %rdi call cudaGetDeviceCount@PLT leaq .LC1(%rip), %rdi call _Z14checkCUDAErrorPKc movl 24(%rsp), %edx leaq .LC2(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $0, %r12d cmpl $1, 24(%rsp) jle .L8 movl $0, %ebx movl $0, %ebp leaq 32(%rsp), %r15 leaq .LC3(%rip), %r14 leaq .LC5(%rip), %r13 jmp .L12 .L9: movq %rax, %rdx shrq %rdx andl $1, %eax orq %rax, %rdx pxor %xmm0, %xmm0 cvtsi2sdq %rdx, %xmm0 addsd %xmm0, %xmm0 jmp .L10 .L11: addl $1, %ebx cmpl %ebx, 24(%rsp) jle .L22 .L12: movl %ebx, %esi movq %r15, %rdi call cudaGetDeviceProperties_v2@PLT movl 420(%rsp), %ecx movl %ebx, %edx movq %r14, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 320(%rsp), %rax testq %rax, %rax js .L9 pxor %xmm0, %xmm0 cvtsi2sdq %rax, %xmm0 .L10: mulsd .LC4(%rip), %xmm0 movl %ebx, %edx movq %r13, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movl 420(%rsp), %eax cmpl %ebp, %eax jle .L11 movl %eax, %ebp movl %ebx, %r12d jmp .L11 .L22: movl %r12d, %edx leaq .LC6(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl %r12d, %edi call cudaSetDevice@PLT leaq .LC7(%rip), %rdi call _Z14checkCUDAErrorPKc .L8: leaq .LC8(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT leaq 28(%rsp), %rdi call cudaGetDevice@PLT leaq 32(%rsp), %rdi movl 28(%rsp), %esi call cudaGetDeviceProperties_v2@PLT movl 420(%rsp), %ecx movl %r12d, %edx leaq .LC9(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 320(%rsp), %rax testq %rax, %rax js .L13 pxor %xmm0, %xmm0 cvtsi2sdq %rax, %xmm0 .L14: mulsd .LC4(%rip), %xmm0 movl %r12d, %edx leaq .LC5(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movabsq $42945378002703, %rax cmpq %rax, 392(%rsp) je .L23 leaq 32(%rsp), %rdx movl 396(%rsp), %r8d movl 392(%rsp), %ecx leaq .LC11(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 392(%rsp), %eax movq (%rsp), %rcx movl %eax, (%rcx) movl 396(%rsp), %eax movq 8(%rsp), %rcx movl %eax, (%rcx) .L16: movl $0, %ecx movl $0, %edx leaq .LC12(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 1064(%rsp), %rax subq %fs:40, %rax jne .L24 addq $1080, %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 .L13: .cfi_restore_state movq %rax, %rdx shrq %rdx andl $1, %eax orq %rax, %rdx pxor %xmm0, %xmm0 cvtsi2sdq %rdx, %xmm0 addsd %xmm0, %xmm0 jmp .L14 .L23: leaq .LC10(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT call cudaThreadExit@PLT jmp .L16 .L24: call __stack_chk_fail@PLT .cfi_endproc .LFE2058: .size _Z15selectAndReportPiS_, .-_Z15selectAndReportPiS_ .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2084: .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 .LFE2084: .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 .LC4: .long 0 .long 1041235968 .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.
// Some useful utilities // system includes #include <stdio.h> #include <assert.h> #include <cuda.h> // External function definitions void checkCUDAError(const char *msg) { cudaError_t err = cudaGetLastError(); if( cudaSuccess != err) { fprintf(stderr, "Cuda error: %s: %s.\n", msg, cudaGetErrorString( err) ); exit(EXIT_FAILURE); } } // When there is more than one device, // select the fastest device and report characteristics void selectAndReport(int *major, int *minor) { int best_gpu = 0; // Use the device with the most cores int number_of_devices; cudaGetDeviceCount(&number_of_devices); checkCUDAError("Get Count"); printf("\n%d Devices\n",number_of_devices); if (number_of_devices > 1) { int max_cores = 0; int device_number; for (device_number = 0; device_number < number_of_devices; device_number++) { cudaDeviceProp device_properties; cudaGetDeviceProperties(&device_properties, device_number); printf("Device # %d has %d cores\n",device_number, device_properties.multiProcessorCount); double gb = 1024*1024*1024; printf("Device # %d has %f GB global memory\n",device_number, ((double)device_properties.totalGlobalMem)/gb); if (max_cores < device_properties.multiProcessorCount) { max_cores = device_properties.multiProcessorCount; best_gpu = device_number; } } printf("\n *** Best GPU is: %d\n",best_gpu); cudaSetDevice(best_gpu); checkCUDAError("Can't set device\n"); } printf("\n"); // get number of SMs on this GPU int devID; cudaGetDevice(&devID); cudaDeviceProp deviceProp; cudaGetDeviceProperties(&deviceProp, devID); printf("Device %d has %d cores\n", best_gpu, deviceProp.multiProcessorCount); double gb = 1024*1024*1024; printf("Device # %d has %f GB global memory\n",best_gpu, ((double)deviceProp.totalGlobalMem)/gb); if (deviceProp.major == 9999 && deviceProp.minor == 9999) { printf("There is no device supporting CUDA.\n"); cudaThreadExit(); } else { printf("Device is a %s, capability: %d.%d\n", deviceProp.name, deviceProp.major, deviceProp.minor); *major = deviceProp.major; *minor = deviceProp.minor; } int driverVersion, runtimeVersion; assert(cudaSuccess == cudaDriverGetVersion(&driverVersion)); assert(cudaSuccess == cudaRuntimeGetVersion(&runtimeVersion)); printf("CUDA Driver version: %d, runtime version: %d\n\n", driverVersion, runtimeVersion); }
// Some useful utilities // system includes #include <stdio.h> #include <assert.h> #include <hip/hip_runtime.h> // External function definitions void checkCUDAError(const char *msg) { hipError_t err = hipGetLastError(); if( hipSuccess != err) { fprintf(stderr, "Cuda error: %s: %s.\n", msg, hipGetErrorString( err) ); exit(EXIT_FAILURE); } } // When there is more than one device, // select the fastest device and report characteristics void selectAndReport(int *major, int *minor) { int best_gpu = 0; // Use the device with the most cores int number_of_devices; hipGetDeviceCount(&number_of_devices); checkCUDAError("Get Count"); printf("\n%d Devices\n",number_of_devices); if (number_of_devices > 1) { int max_cores = 0; int device_number; for (device_number = 0; device_number < number_of_devices; device_number++) { hipDeviceProp_t device_properties; hipGetDeviceProperties(&device_properties, device_number); printf("Device # %d has %d cores\n",device_number, device_properties.multiProcessorCount); double gb = 1024*1024*1024; printf("Device # %d has %f GB global memory\n",device_number, ((double)device_properties.totalGlobalMem)/gb); if (max_cores < device_properties.multiProcessorCount) { max_cores = device_properties.multiProcessorCount; best_gpu = device_number; } } printf("\n *** Best GPU is: %d\n",best_gpu); hipSetDevice(best_gpu); checkCUDAError("Can't set device\n"); } printf("\n"); // get number of SMs on this GPU int devID; hipGetDevice(&devID); hipDeviceProp_t deviceProp; hipGetDeviceProperties(&deviceProp, devID); printf("Device %d has %d cores\n", best_gpu, deviceProp.multiProcessorCount); double gb = 1024*1024*1024; printf("Device # %d has %f GB global memory\n",best_gpu, ((double)deviceProp.totalGlobalMem)/gb); if (deviceProp.major == 9999 && deviceProp.minor == 9999) { printf("There is no device supporting CUDA.\n"); hipDeviceReset(); } else { printf("Device is a %s, capability: %d.%d\n", deviceProp.name, deviceProp.major, deviceProp.minor); *major = deviceProp.major; *minor = deviceProp.minor; } int driverVersion, runtimeVersion; assert(hipSuccess == hipDriverGetVersion(&driverVersion)); assert(hipSuccess == hipRuntimeGetVersion(&runtimeVersion)); printf("CUDA Driver version: %d, runtime version: %d\n\n", driverVersion, runtimeVersion); }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
// Some useful utilities // system includes #include <stdio.h> #include <assert.h> #include <hip/hip_runtime.h> // External function definitions void checkCUDAError(const char *msg) { hipError_t err = hipGetLastError(); if( hipSuccess != err) { fprintf(stderr, "Cuda error: %s: %s.\n", msg, hipGetErrorString( err) ); exit(EXIT_FAILURE); } } // When there is more than one device, // select the fastest device and report characteristics void selectAndReport(int *major, int *minor) { int best_gpu = 0; // Use the device with the most cores int number_of_devices; hipGetDeviceCount(&number_of_devices); checkCUDAError("Get Count"); printf("\n%d Devices\n",number_of_devices); if (number_of_devices > 1) { int max_cores = 0; int device_number; for (device_number = 0; device_number < number_of_devices; device_number++) { hipDeviceProp_t device_properties; hipGetDeviceProperties(&device_properties, device_number); printf("Device # %d has %d cores\n",device_number, device_properties.multiProcessorCount); double gb = 1024*1024*1024; printf("Device # %d has %f GB global memory\n",device_number, ((double)device_properties.totalGlobalMem)/gb); if (max_cores < device_properties.multiProcessorCount) { max_cores = device_properties.multiProcessorCount; best_gpu = device_number; } } printf("\n *** Best GPU is: %d\n",best_gpu); hipSetDevice(best_gpu); checkCUDAError("Can't set device\n"); } printf("\n"); // get number of SMs on this GPU int devID; hipGetDevice(&devID); hipDeviceProp_t deviceProp; hipGetDeviceProperties(&deviceProp, devID); printf("Device %d has %d cores\n", best_gpu, deviceProp.multiProcessorCount); double gb = 1024*1024*1024; printf("Device # %d has %f GB global memory\n",best_gpu, ((double)deviceProp.totalGlobalMem)/gb); if (deviceProp.major == 9999 && deviceProp.minor == 9999) { printf("There is no device supporting CUDA.\n"); hipDeviceReset(); } else { printf("Device is a %s, capability: %d.%d\n", deviceProp.name, deviceProp.major, deviceProp.minor); *major = deviceProp.major; *minor = deviceProp.minor; } int driverVersion, runtimeVersion; assert(hipSuccess == hipDriverGetVersion(&driverVersion)); assert(hipSuccess == hipRuntimeGetVersion(&runtimeVersion)); printf("CUDA Driver version: %d, runtime version: %d\n\n", driverVersion, runtimeVersion); }
.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.
// Some useful utilities // system includes #include <stdio.h> #include <assert.h> #include <hip/hip_runtime.h> // External function definitions void checkCUDAError(const char *msg) { hipError_t err = hipGetLastError(); if( hipSuccess != err) { fprintf(stderr, "Cuda error: %s: %s.\n", msg, hipGetErrorString( err) ); exit(EXIT_FAILURE); } } // When there is more than one device, // select the fastest device and report characteristics void selectAndReport(int *major, int *minor) { int best_gpu = 0; // Use the device with the most cores int number_of_devices; hipGetDeviceCount(&number_of_devices); checkCUDAError("Get Count"); printf("\n%d Devices\n",number_of_devices); if (number_of_devices > 1) { int max_cores = 0; int device_number; for (device_number = 0; device_number < number_of_devices; device_number++) { hipDeviceProp_t device_properties; hipGetDeviceProperties(&device_properties, device_number); printf("Device # %d has %d cores\n",device_number, device_properties.multiProcessorCount); double gb = 1024*1024*1024; printf("Device # %d has %f GB global memory\n",device_number, ((double)device_properties.totalGlobalMem)/gb); if (max_cores < device_properties.multiProcessorCount) { max_cores = device_properties.multiProcessorCount; best_gpu = device_number; } } printf("\n *** Best GPU is: %d\n",best_gpu); hipSetDevice(best_gpu); checkCUDAError("Can't set device\n"); } printf("\n"); // get number of SMs on this GPU int devID; hipGetDevice(&devID); hipDeviceProp_t deviceProp; hipGetDeviceProperties(&deviceProp, devID); printf("Device %d has %d cores\n", best_gpu, deviceProp.multiProcessorCount); double gb = 1024*1024*1024; printf("Device # %d has %f GB global memory\n",best_gpu, ((double)deviceProp.totalGlobalMem)/gb); if (deviceProp.major == 9999 && deviceProp.minor == 9999) { printf("There is no device supporting CUDA.\n"); hipDeviceReset(); } else { printf("Device is a %s, capability: %d.%d\n", deviceProp.name, deviceProp.major, deviceProp.minor); *major = deviceProp.major; *minor = deviceProp.minor; } int driverVersion, runtimeVersion; assert(hipSuccess == hipDriverGetVersion(&driverVersion)); assert(hipSuccess == hipRuntimeGetVersion(&runtimeVersion)); printf("CUDA Driver version: %d, runtime version: %d\n\n", driverVersion, runtimeVersion); }
.text .file "utils.hip" .globl _Z14checkCUDAErrorPKc # -- Begin function _Z14checkCUDAErrorPKc .p2align 4, 0x90 .type _Z14checkCUDAErrorPKc,@function _Z14checkCUDAErrorPKc: # @_Z14checkCUDAErrorPKc .cfi_startproc # %bb.0: pushq %r14 .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 %r14, -16 movq %rdi, %rbx callq hipGetLastError testl %eax, %eax jne .LBB0_2 # %bb.1: addq $8, %rsp .cfi_def_cfa_offset 24 popq %rbx .cfi_def_cfa_offset 16 popq %r14 .cfi_def_cfa_offset 8 retq .LBB0_2: .cfi_def_cfa_offset 32 movq stderr(%rip), %r14 movl %eax, %edi callq hipGetErrorString movl $.L.str, %esi movq %r14, %rdi movq %rbx, %rdx movq %rax, %rcx xorl %eax, %eax callq fprintf movl $1, %edi callq exit .Lfunc_end0: .size _Z14checkCUDAErrorPKc, .Lfunc_end0-_Z14checkCUDAErrorPKc .cfi_endproc # -- End function .section .rodata.cst16,"aM",@progbits,16 .p2align 4, 0x0 # -- Begin function _Z15selectAndReportPiS_ .LCPI1_0: .long 1127219200 # 0x43300000 .long 1160773632 # 0x45300000 .long 0 # 0x0 .long 0 # 0x0 .LCPI1_1: .quad 0x4330000000000000 # double 4503599627370496 .quad 0x4530000000000000 # double 1.9342813113834067E+25 .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 .LCPI1_2: .quad 0x3e10000000000000 # double 9.3132257461547852E-10 .text .globl _Z15selectAndReportPiS_ .p2align 4, 0x90 .type _Z15selectAndReportPiS_,@function _Z15selectAndReportPiS_: # @_Z15selectAndReportPiS_ .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 $1480, %rsp # imm = 0x5C8 .cfi_def_cfa_offset 1536 .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 movq %rdi, %r14 movq %rsp, %rdi callq hipGetDeviceCount callq hipGetLastError testl %eax, %eax jne .LBB1_1 # %bb.3: # %_Z14checkCUDAErrorPKc.exit movl (%rsp), %esi xorl %ebp, %ebp movl $.L.str.2, %edi xorl %eax, %eax callq printf cmpl $2, (%rsp) jl .LBB1_10 # %bb.4: # %.preheader cmpl $0, (%rsp) jle .LBB1_5 # %bb.6: # %.lr.ph leaq 8(%rsp), %r15 xorl %ebp, %ebp xorl %r13d, %r13d xorl %r12d, %r12d .p2align 4, 0x90 .LBB1_7: # =>This Inner Loop Header: Depth=1 movq %r15, %rdi movl %r12d, %esi callq hipGetDevicePropertiesR0600 movl 396(%rsp), %edx movl $.L.str.3, %edi movl %r12d, %esi xorl %eax, %eax callq printf movsd 296(%rsp), %xmm1 # xmm1 = mem[0],zero unpcklps .LCPI1_0(%rip), %xmm1 # xmm1 = xmm1[0],mem[0],xmm1[1],mem[1] subpd .LCPI1_1(%rip), %xmm1 movapd %xmm1, %xmm0 unpckhpd %xmm1, %xmm0 # xmm0 = xmm0[1],xmm1[1] addsd %xmm1, %xmm0 mulsd .LCPI1_2(%rip), %xmm0 movl $.L.str.4, %edi movl %r12d, %esi movb $1, %al callq printf movl 396(%rsp), %eax cmpl %eax, %r13d cmovlel %eax, %r13d cmovll %r12d, %ebp incl %r12d cmpl (%rsp), %r12d jl .LBB1_7 jmp .LBB1_8 .LBB1_5: xorl %ebp, %ebp .LBB1_8: # %._crit_edge movl $.L.str.5, %edi movl %ebp, %esi xorl %eax, %eax callq printf movl %ebp, %edi callq hipSetDevice callq hipGetLastError testl %eax, %eax jne .LBB1_9 .LBB1_10: # %_Z14checkCUDAErrorPKc.exit24 movl $10, %edi callq putchar@PLT leaq 4(%rsp), %rdi callq hipGetDevice movl 4(%rsp), %esi leaq 8(%rsp), %rdi callq hipGetDevicePropertiesR0600 movl 396(%rsp), %edx movl $.L.str.8, %edi movl %ebp, %esi xorl %eax, %eax callq printf movsd 296(%rsp), %xmm1 # xmm1 = mem[0],zero unpcklps .LCPI1_0(%rip), %xmm1 # xmm1 = xmm1[0],mem[0],xmm1[1],mem[1] subpd .LCPI1_1(%rip), %xmm1 movapd %xmm1, %xmm0 unpckhpd %xmm1, %xmm0 # xmm0 = xmm0[1],xmm1[1] addsd %xmm1, %xmm0 mulsd .LCPI1_2(%rip), %xmm0 movl $.L.str.4, %edi movl %ebp, %esi movb $1, %al callq printf movl 368(%rsp), %edx movl 372(%rsp), %ecx cmpl $9999, %edx # imm = 0x270F jne .LBB1_13 # %bb.11: # %_Z14checkCUDAErrorPKc.exit24 cmpl $9999, %ecx # imm = 0x270F jne .LBB1_13 # %bb.12: movl $.Lstr, %edi callq puts@PLT callq hipDeviceReset jmp .LBB1_14 .LBB1_13: leaq 8(%rsp), %rsi movl $.L.str.10, %edi xorl %eax, %eax callq printf movl 368(%rsp), %eax movl %eax, (%r14) movl 372(%rsp), %eax movl %eax, (%rbx) .LBB1_14: movl $.L.str.11, %edi xorl %eax, %eax callq printf addq $1480, %rsp # imm = 0x5C8 .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 .LBB1_1: .cfi_def_cfa_offset 1536 movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str, %esi movl $.L.str.1, %edx jmp .LBB1_2 .LBB1_9: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str, %esi movl $.L.str.6, %edx .LBB1_2: movq %rbx, %rdi movq %rax, %rcx xorl %eax, %eax callq fprintf movl $1, %edi callq exit .Lfunc_end1: .size _Z15selectAndReportPiS_, .Lfunc_end1-_Z15selectAndReportPiS_ .cfi_endproc # -- End function .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "Cuda error: %s: %s.\n" .size .L.str, 21 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "Get Count" .size .L.str.1, 10 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "\n%d Devices\n" .size .L.str.2, 13 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "Device # %d has %d cores\n" .size .L.str.3, 26 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "Device # %d has %f GB global memory\n" .size .L.str.4, 37 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "\n *** Best GPU is: %d\n" .size .L.str.5, 23 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz "Can't set device\n" .size .L.str.6, 18 .type .L.str.8,@object # @.str.8 .L.str.8: .asciz "Device %d has %d cores\n" .size .L.str.8, 24 .type .L.str.10,@object # @.str.10 .L.str.10: .asciz "Device is a %s, capability: %d.%d\n" .size .L.str.10, 35 .type .L.str.11,@object # @.str.11 .L.str.11: .asciz "CUDA Driver version: %d, runtime version: %d\n\n" .size .L.str.11, 47 .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 "There is no device supporting CUDA." .size .Lstr, 36 .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_00122abd_00000000-6_utils.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 .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "Cuda error: %s: %s.\n" .text .globl _Z14checkCUDAErrorPKc .type _Z14checkCUDAErrorPKc, @function _Z14checkCUDAErrorPKc: .LFB2057: .cfi_startproc endbr64 pushq %rbx .cfi_def_cfa_offset 16 .cfi_offset 3, -16 movq %rdi, %rbx call cudaGetLastError@PLT testl %eax, %eax jne .L6 popq %rbx .cfi_remember_state .cfi_def_cfa_offset 8 ret .L6: .cfi_restore_state movl %eax, %edi call cudaGetErrorString@PLT movq %rax, %r8 movq %rbx, %rcx leaq .LC0(%rip), %rdx movl $2, %esi movq stderr(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movl $1, %edi call exit@PLT .cfi_endproc .LFE2057: .size _Z14checkCUDAErrorPKc, .-_Z14checkCUDAErrorPKc .section .rodata.str1.1 .LC1: .string "Get Count" .LC2: .string "\n%d Devices\n" .LC3: .string "Device # %d has %d cores\n" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC5: .string "Device # %d has %f GB global memory\n" .section .rodata.str1.1 .LC6: .string "\n *** Best GPU is: %d\n" .LC7: .string "Can't set device\n" .LC8: .string "\n" .LC9: .string "Device %d has %d cores\n" .section .rodata.str1.8 .align 8 .LC10: .string "There is no device supporting CUDA.\n" .align 8 .LC11: .string "Device is a %s, capability: %d.%d\n" .align 8 .LC12: .string "CUDA Driver version: %d, runtime version: %d\n\n" .text .globl _Z15selectAndReportPiS_ .type _Z15selectAndReportPiS_, @function _Z15selectAndReportPiS_: .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 $1080, %rsp .cfi_def_cfa_offset 1136 movq %rdi, (%rsp) movq %rsi, 8(%rsp) movq %fs:40, %rax movq %rax, 1064(%rsp) xorl %eax, %eax leaq 24(%rsp), %rdi call cudaGetDeviceCount@PLT leaq .LC1(%rip), %rdi call _Z14checkCUDAErrorPKc movl 24(%rsp), %edx leaq .LC2(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $0, %r12d cmpl $1, 24(%rsp) jle .L8 movl $0, %ebx movl $0, %ebp leaq 32(%rsp), %r15 leaq .LC3(%rip), %r14 leaq .LC5(%rip), %r13 jmp .L12 .L9: movq %rax, %rdx shrq %rdx andl $1, %eax orq %rax, %rdx pxor %xmm0, %xmm0 cvtsi2sdq %rdx, %xmm0 addsd %xmm0, %xmm0 jmp .L10 .L11: addl $1, %ebx cmpl %ebx, 24(%rsp) jle .L22 .L12: movl %ebx, %esi movq %r15, %rdi call cudaGetDeviceProperties_v2@PLT movl 420(%rsp), %ecx movl %ebx, %edx movq %r14, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 320(%rsp), %rax testq %rax, %rax js .L9 pxor %xmm0, %xmm0 cvtsi2sdq %rax, %xmm0 .L10: mulsd .LC4(%rip), %xmm0 movl %ebx, %edx movq %r13, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movl 420(%rsp), %eax cmpl %ebp, %eax jle .L11 movl %eax, %ebp movl %ebx, %r12d jmp .L11 .L22: movl %r12d, %edx leaq .LC6(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl %r12d, %edi call cudaSetDevice@PLT leaq .LC7(%rip), %rdi call _Z14checkCUDAErrorPKc .L8: leaq .LC8(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT leaq 28(%rsp), %rdi call cudaGetDevice@PLT leaq 32(%rsp), %rdi movl 28(%rsp), %esi call cudaGetDeviceProperties_v2@PLT movl 420(%rsp), %ecx movl %r12d, %edx leaq .LC9(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 320(%rsp), %rax testq %rax, %rax js .L13 pxor %xmm0, %xmm0 cvtsi2sdq %rax, %xmm0 .L14: mulsd .LC4(%rip), %xmm0 movl %r12d, %edx leaq .LC5(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movabsq $42945378002703, %rax cmpq %rax, 392(%rsp) je .L23 leaq 32(%rsp), %rdx movl 396(%rsp), %r8d movl 392(%rsp), %ecx leaq .LC11(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl 392(%rsp), %eax movq (%rsp), %rcx movl %eax, (%rcx) movl 396(%rsp), %eax movq 8(%rsp), %rcx movl %eax, (%rcx) .L16: movl $0, %ecx movl $0, %edx leaq .LC12(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movq 1064(%rsp), %rax subq %fs:40, %rax jne .L24 addq $1080, %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 .L13: .cfi_restore_state movq %rax, %rdx shrq %rdx andl $1, %eax orq %rax, %rdx pxor %xmm0, %xmm0 cvtsi2sdq %rdx, %xmm0 addsd %xmm0, %xmm0 jmp .L14 .L23: leaq .LC10(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT call cudaThreadExit@PLT jmp .L16 .L24: call __stack_chk_fail@PLT .cfi_endproc .LFE2058: .size _Z15selectAndReportPiS_, .-_Z15selectAndReportPiS_ .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2084: .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 .LFE2084: .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 .LC4: .long 0 .long 1041235968 .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 "utils.hip" .globl _Z14checkCUDAErrorPKc # -- Begin function _Z14checkCUDAErrorPKc .p2align 4, 0x90 .type _Z14checkCUDAErrorPKc,@function _Z14checkCUDAErrorPKc: # @_Z14checkCUDAErrorPKc .cfi_startproc # %bb.0: pushq %r14 .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 %r14, -16 movq %rdi, %rbx callq hipGetLastError testl %eax, %eax jne .LBB0_2 # %bb.1: addq $8, %rsp .cfi_def_cfa_offset 24 popq %rbx .cfi_def_cfa_offset 16 popq %r14 .cfi_def_cfa_offset 8 retq .LBB0_2: .cfi_def_cfa_offset 32 movq stderr(%rip), %r14 movl %eax, %edi callq hipGetErrorString movl $.L.str, %esi movq %r14, %rdi movq %rbx, %rdx movq %rax, %rcx xorl %eax, %eax callq fprintf movl $1, %edi callq exit .Lfunc_end0: .size _Z14checkCUDAErrorPKc, .Lfunc_end0-_Z14checkCUDAErrorPKc .cfi_endproc # -- End function .section .rodata.cst16,"aM",@progbits,16 .p2align 4, 0x0 # -- Begin function _Z15selectAndReportPiS_ .LCPI1_0: .long 1127219200 # 0x43300000 .long 1160773632 # 0x45300000 .long 0 # 0x0 .long 0 # 0x0 .LCPI1_1: .quad 0x4330000000000000 # double 4503599627370496 .quad 0x4530000000000000 # double 1.9342813113834067E+25 .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 .LCPI1_2: .quad 0x3e10000000000000 # double 9.3132257461547852E-10 .text .globl _Z15selectAndReportPiS_ .p2align 4, 0x90 .type _Z15selectAndReportPiS_,@function _Z15selectAndReportPiS_: # @_Z15selectAndReportPiS_ .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 $1480, %rsp # imm = 0x5C8 .cfi_def_cfa_offset 1536 .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 movq %rdi, %r14 movq %rsp, %rdi callq hipGetDeviceCount callq hipGetLastError testl %eax, %eax jne .LBB1_1 # %bb.3: # %_Z14checkCUDAErrorPKc.exit movl (%rsp), %esi xorl %ebp, %ebp movl $.L.str.2, %edi xorl %eax, %eax callq printf cmpl $2, (%rsp) jl .LBB1_10 # %bb.4: # %.preheader cmpl $0, (%rsp) jle .LBB1_5 # %bb.6: # %.lr.ph leaq 8(%rsp), %r15 xorl %ebp, %ebp xorl %r13d, %r13d xorl %r12d, %r12d .p2align 4, 0x90 .LBB1_7: # =>This Inner Loop Header: Depth=1 movq %r15, %rdi movl %r12d, %esi callq hipGetDevicePropertiesR0600 movl 396(%rsp), %edx movl $.L.str.3, %edi movl %r12d, %esi xorl %eax, %eax callq printf movsd 296(%rsp), %xmm1 # xmm1 = mem[0],zero unpcklps .LCPI1_0(%rip), %xmm1 # xmm1 = xmm1[0],mem[0],xmm1[1],mem[1] subpd .LCPI1_1(%rip), %xmm1 movapd %xmm1, %xmm0 unpckhpd %xmm1, %xmm0 # xmm0 = xmm0[1],xmm1[1] addsd %xmm1, %xmm0 mulsd .LCPI1_2(%rip), %xmm0 movl $.L.str.4, %edi movl %r12d, %esi movb $1, %al callq printf movl 396(%rsp), %eax cmpl %eax, %r13d cmovlel %eax, %r13d cmovll %r12d, %ebp incl %r12d cmpl (%rsp), %r12d jl .LBB1_7 jmp .LBB1_8 .LBB1_5: xorl %ebp, %ebp .LBB1_8: # %._crit_edge movl $.L.str.5, %edi movl %ebp, %esi xorl %eax, %eax callq printf movl %ebp, %edi callq hipSetDevice callq hipGetLastError testl %eax, %eax jne .LBB1_9 .LBB1_10: # %_Z14checkCUDAErrorPKc.exit24 movl $10, %edi callq putchar@PLT leaq 4(%rsp), %rdi callq hipGetDevice movl 4(%rsp), %esi leaq 8(%rsp), %rdi callq hipGetDevicePropertiesR0600 movl 396(%rsp), %edx movl $.L.str.8, %edi movl %ebp, %esi xorl %eax, %eax callq printf movsd 296(%rsp), %xmm1 # xmm1 = mem[0],zero unpcklps .LCPI1_0(%rip), %xmm1 # xmm1 = xmm1[0],mem[0],xmm1[1],mem[1] subpd .LCPI1_1(%rip), %xmm1 movapd %xmm1, %xmm0 unpckhpd %xmm1, %xmm0 # xmm0 = xmm0[1],xmm1[1] addsd %xmm1, %xmm0 mulsd .LCPI1_2(%rip), %xmm0 movl $.L.str.4, %edi movl %ebp, %esi movb $1, %al callq printf movl 368(%rsp), %edx movl 372(%rsp), %ecx cmpl $9999, %edx # imm = 0x270F jne .LBB1_13 # %bb.11: # %_Z14checkCUDAErrorPKc.exit24 cmpl $9999, %ecx # imm = 0x270F jne .LBB1_13 # %bb.12: movl $.Lstr, %edi callq puts@PLT callq hipDeviceReset jmp .LBB1_14 .LBB1_13: leaq 8(%rsp), %rsi movl $.L.str.10, %edi xorl %eax, %eax callq printf movl 368(%rsp), %eax movl %eax, (%r14) movl 372(%rsp), %eax movl %eax, (%rbx) .LBB1_14: movl $.L.str.11, %edi xorl %eax, %eax callq printf addq $1480, %rsp # imm = 0x5C8 .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 .LBB1_1: .cfi_def_cfa_offset 1536 movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str, %esi movl $.L.str.1, %edx jmp .LBB1_2 .LBB1_9: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str, %esi movl $.L.str.6, %edx .LBB1_2: movq %rbx, %rdi movq %rax, %rcx xorl %eax, %eax callq fprintf movl $1, %edi callq exit .Lfunc_end1: .size _Z15selectAndReportPiS_, .Lfunc_end1-_Z15selectAndReportPiS_ .cfi_endproc # -- End function .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "Cuda error: %s: %s.\n" .size .L.str, 21 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "Get Count" .size .L.str.1, 10 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "\n%d Devices\n" .size .L.str.2, 13 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "Device # %d has %d cores\n" .size .L.str.3, 26 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "Device # %d has %f GB global memory\n" .size .L.str.4, 37 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "\n *** Best GPU is: %d\n" .size .L.str.5, 23 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz "Can't set device\n" .size .L.str.6, 18 .type .L.str.8,@object # @.str.8 .L.str.8: .asciz "Device %d has %d cores\n" .size .L.str.8, 24 .type .L.str.10,@object # @.str.10 .L.str.10: .asciz "Device is a %s, capability: %d.%d\n" .size .L.str.10, 35 .type .L.str.11,@object # @.str.11 .L.str.11: .asciz "CUDA Driver version: %d, runtime version: %d\n\n" .size .L.str.11, 47 .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 "There is no device supporting CUDA." .size .Lstr, 36 .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 CUDA host assembly.
#include "includes.h" __device__ float sigmoid(float x) { return 1 / (1 + expf(-x)); } __global__ void produceState2(const float* arguments, const int argsSize, const float* weights, const int* topology, const int topSize, float* outStates) { const int tid = threadIdx.x; const int dim = argsSize + topSize; extern __shared__ float s[]; float* states = s; bool* ready = (bool*)&states[dim]; __shared__ int counter[1]; int r = tid; while(r < dim) { ready[r] = false; r += blockDim.x; } if (tid == 0) { counter[tid] = argsSize; } if (tid < argsSize) { states[tid] = arguments[tid]; ready[tid] = true; } __syncthreads(); while(counter[0] < dim) { const int index = counter[0] + tid; const int topIndex = index - argsSize; if (topIndex < topSize) { const int leftBorder = topology[topIndex*3]; const int rightBorder = topology[topIndex*3 + 1]; const int weightsStart = topology[topIndex*3 + 2]; if (rightBorder <= counter[0]) { float sum = 0; for (int i = leftBorder; i < rightBorder; i++) { sum += states[i] * weights[weightsStart + i - leftBorder]; } states[index] = sigmoid(sum); ready[index] = true; } } __syncthreads(); if (tid == 0) { int total = counter[0]; for (int i = total; i < total + blockDim.x && i < dim; i++) { if (ready[i]) { counter[0]++; } } } __syncthreads(); } int n = tid; while(n < dim) { outStates[n] = states[n]; n += blockDim.x; } }
.file "tmpxft_000a7ab7_00000000-6_produceState2.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2030: .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 .LFE2030: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z7sigmoidf .type _Z7sigmoidf, @function _Z7sigmoidf: .LFB2027: .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 .LFE2027: .size _Z7sigmoidf, .-_Z7sigmoidf .globl _Z44__device_stub__Z13produceState2PKfiS0_PKiiPfPKfiS0_PKiiPf .type _Z44__device_stub__Z13produceState2PKfiS0_PKiiPfPKfiS0_PKiiPf, @function _Z44__device_stub__Z13produceState2PKfiS0_PKiiPfPKfiS0_PKiiPf: .LFB2052: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movq %rdi, 40(%rsp) movl %esi, 36(%rsp) movq %rdx, 24(%rsp) movq %rcx, 16(%rsp) movl %r8d, 32(%rsp) movq %r9, 8(%rsp) movq %fs:40, %rax movq %rax, 168(%rsp) xorl %eax, %eax leaq 40(%rsp), %rax movq %rax, 112(%rsp) leaq 36(%rsp), %rax movq %rax, 120(%rsp) leaq 24(%rsp), %rax movq %rax, 128(%rsp) leaq 16(%rsp), %rax movq %rax, 136(%rsp) leaq 32(%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 .L9 .L5: movq 168(%rsp), %rax subq %fs:40, %rax jne .L10 addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L9: .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 _Z13produceState2PKfiS0_PKiiPf(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 192 jmp .L5 .L10: call __stack_chk_fail@PLT .cfi_endproc .LFE2052: .size _Z44__device_stub__Z13produceState2PKfiS0_PKiiPfPKfiS0_PKiiPf, .-_Z44__device_stub__Z13produceState2PKfiS0_PKiiPfPKfiS0_PKiiPf .globl _Z13produceState2PKfiS0_PKiiPf .type _Z13produceState2PKfiS0_PKiiPf, @function _Z13produceState2PKfiS0_PKiiPf: .LFB2053: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z44__device_stub__Z13produceState2PKfiS0_PKiiPfPKfiS0_PKiiPf addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2053: .size _Z13produceState2PKfiS0_PKiiPf, .-_Z13produceState2PKfiS0_PKiiPf .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "_Z13produceState2PKfiS0_PKiiPf" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2055: .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 _Z13produceState2PKfiS0_PKiiPf(%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 .LFE2055: .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" __device__ float sigmoid(float x) { return 1 / (1 + expf(-x)); } __global__ void produceState2(const float* arguments, const int argsSize, const float* weights, const int* topology, const int topSize, float* outStates) { const int tid = threadIdx.x; const int dim = argsSize + topSize; extern __shared__ float s[]; float* states = s; bool* ready = (bool*)&states[dim]; __shared__ int counter[1]; int r = tid; while(r < dim) { ready[r] = false; r += blockDim.x; } if (tid == 0) { counter[tid] = argsSize; } if (tid < argsSize) { states[tid] = arguments[tid]; ready[tid] = true; } __syncthreads(); while(counter[0] < dim) { const int index = counter[0] + tid; const int topIndex = index - argsSize; if (topIndex < topSize) { const int leftBorder = topology[topIndex*3]; const int rightBorder = topology[topIndex*3 + 1]; const int weightsStart = topology[topIndex*3 + 2]; if (rightBorder <= counter[0]) { float sum = 0; for (int i = leftBorder; i < rightBorder; i++) { sum += states[i] * weights[weightsStart + i - leftBorder]; } states[index] = sigmoid(sum); ready[index] = true; } } __syncthreads(); if (tid == 0) { int total = counter[0]; for (int i = total; i < total + blockDim.x && i < dim; i++) { if (ready[i]) { counter[0]++; } } } __syncthreads(); } int n = tid; while(n < dim) { outStates[n] = states[n]; n += blockDim.x; } }
#include <hip/hip_runtime.h> #include "includes.h" __device__ float sigmoid(float x) { return 1 / (1 + expf(-x)); } __global__ void produceState2(const float* arguments, const int argsSize, const float* weights, const int* topology, const int topSize, float* outStates) { const int tid = threadIdx.x; const int dim = argsSize + topSize; extern __shared__ float s[]; float* states = s; bool* ready = (bool*)&states[dim]; __shared__ int counter[1]; int r = tid; while(r < dim) { ready[r] = false; r += blockDim.x; } if (tid == 0) { counter[tid] = argsSize; } if (tid < argsSize) { states[tid] = arguments[tid]; ready[tid] = true; } __syncthreads(); while(counter[0] < dim) { const int index = counter[0] + tid; const int topIndex = index - argsSize; if (topIndex < topSize) { const int leftBorder = topology[topIndex*3]; const int rightBorder = topology[topIndex*3 + 1]; const int weightsStart = topology[topIndex*3 + 2]; if (rightBorder <= counter[0]) { float sum = 0; for (int i = leftBorder; i < rightBorder; i++) { sum += states[i] * weights[weightsStart + i - leftBorder]; } states[index] = sigmoid(sum); ready[index] = true; } } __syncthreads(); if (tid == 0) { int total = counter[0]; for (int i = total; i < total + blockDim.x && i < dim; i++) { if (ready[i]) { counter[0]++; } } } __syncthreads(); } int n = tid; while(n < dim) { outStates[n] = states[n]; n += blockDim.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" __device__ float sigmoid(float x) { return 1 / (1 + expf(-x)); } __global__ void produceState2(const float* arguments, const int argsSize, const float* weights, const int* topology, const int topSize, float* outStates) { const int tid = threadIdx.x; const int dim = argsSize + topSize; extern __shared__ float s[]; float* states = s; bool* ready = (bool*)&states[dim]; __shared__ int counter[1]; int r = tid; while(r < dim) { ready[r] = false; r += blockDim.x; } if (tid == 0) { counter[tid] = argsSize; } if (tid < argsSize) { states[tid] = arguments[tid]; ready[tid] = true; } __syncthreads(); while(counter[0] < dim) { const int index = counter[0] + tid; const int topIndex = index - argsSize; if (topIndex < topSize) { const int leftBorder = topology[topIndex*3]; const int rightBorder = topology[topIndex*3 + 1]; const int weightsStart = topology[topIndex*3 + 2]; if (rightBorder <= counter[0]) { float sum = 0; for (int i = leftBorder; i < rightBorder; i++) { sum += states[i] * weights[weightsStart + i - leftBorder]; } states[index] = sigmoid(sum); ready[index] = true; } } __syncthreads(); if (tid == 0) { int total = counter[0]; for (int i = total; i < total + blockDim.x && i < dim; i++) { if (ready[i]) { counter[0]++; } } } __syncthreads(); } int n = tid; while(n < dim) { outStates[n] = states[n]; n += blockDim.x; } }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z13produceState2PKfiS0_PKiiPf .globl _Z13produceState2PKfiS0_PKiiPf .p2align 8 .type _Z13produceState2PKfiS0_PKiiPf,@function _Z13produceState2PKfiS0_PKiiPf: s_clause 0x1 s_load_b32 s11, s[0:1], 0x8 s_load_b32 s12, s[0:1], 0x20 s_waitcnt lgkmcnt(0) s_add_i32 s10, s12, s11 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_lshl_b32 s2, s10, 2 s_add_i32 s13, s2, 4 s_mov_b32 s2, exec_lo v_cmpx_gt_i32_e64 s10, v0 s_cbranch_execz .LBB0_3 s_load_b32 s3, s[0:1], 0x3c v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, v0 s_mov_b32 s4, 0 s_waitcnt lgkmcnt(0) s_and_b32 s3, s3, 0xffff .LBB0_2: s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_4) | instid1(SALU_CYCLE_1) v_add_nc_u32_e32 v3, s13, v2 v_add_nc_u32_e32 v2, s3, v2 ds_store_b8 v3, v1 v_cmp_le_i32_e32 vcc_lo, s10, v2 s_or_b32 s4, vcc_lo, s4 s_and_not1_b32 exec_lo, exec_lo, s4 s_cbranch_execnz .LBB0_2 .LBB0_3: s_or_b32 exec_lo, exec_lo, s2 v_cmp_eq_u32_e64 s2, 0, v0 s_delay_alu instid0(VALU_DEP_1) s_and_saveexec_b32 s3, s2 s_cbranch_execz .LBB0_5 v_dual_mov_b32 v1, 0 :: v_dual_mov_b32 v2, s11 ds_store_b32 v1, v2 .LBB0_5: s_or_b32 exec_lo, exec_lo, s3 s_delay_alu instid0(SALU_CYCLE_1) s_mov_b32 s3, exec_lo v_cmpx_gt_i32_e64 s11, v0 s_cbranch_execz .LBB0_7 s_load_b64 s[4:5], s[0:1], 0x0 v_dual_mov_b32 v4, 1 :: v_dual_lshlrev_b32 v1, 2, v0 v_add_nc_u32_e32 v3, s13, v0 s_waitcnt lgkmcnt(0) global_load_b32 v2, v1, s[4:5] v_add_nc_u32_e32 v1, 4, v1 s_waitcnt vmcnt(0) ds_store_b32 v1, v2 ds_store_b8 v3, v4 .LBB0_7: s_or_b32 exec_lo, exec_lo, s3 v_mov_b32_e32 v3, 0 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv ds_load_b32 v5, v3 s_waitcnt lgkmcnt(0) v_cmp_le_i32_e32 vcc_lo, s10, v5 s_cbranch_vccnz .LBB0_22 s_load_b128 s[4:7], s[0:1], 0x10 s_add_u32 s8, s0, 48 s_addc_u32 s9, s1, 0 s_branch .LBB0_10 .LBB0_9: s_or_b32 exec_lo, exec_lo, s14 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv ds_load_b32 v5, v3 s_waitcnt lgkmcnt(0) v_cmp_le_i32_e32 vcc_lo, s10, v5 s_cbranch_vccnz .LBB0_22 .LBB0_10: v_add_nc_u32_e32 v4, v5, v0 s_mov_b32 s3, exec_lo s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_subrev_nc_u32_e32 v1, s11, v4 v_cmpx_gt_i32_e64 s12, v1 s_cbranch_execz .LBB0_17 v_lshl_add_u32 v1, v1, 1, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v2, 31, v1 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, s6, v1 v_add_co_ci_u32_e32 v2, vcc_lo, s7, v2, vcc_lo global_load_b32 v6, v[1:2], off offset:4 s_waitcnt vmcnt(0) v_cmp_le_i32_e32 vcc_lo, v6, v5 s_and_b32 exec_lo, exec_lo, vcc_lo s_cbranch_execz .LBB0_17 global_load_b32 v7, v[1:2], off v_mov_b32_e32 v5, 0 s_mov_b32 s14, exec_lo s_waitcnt vmcnt(0) v_cmpx_lt_i32_e64 v7, v6 s_cbranch_execz .LBB0_16 global_load_b32 v1, v[1:2], off offset:8 v_lshl_add_u32 v8, v7, 2, 4 v_mov_b32_e32 v5, 0 s_mov_b32 s15, 0 .p2align 6 .LBB0_14: s_waitcnt vmcnt(0) v_ashrrev_i32_e32 v2, 31, v1 v_add_nc_u32_e32 v7, 1, v7 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_lshlrev_b64 v[9:10], 2, v[1:2] v_add_nc_u32_e32 v1, 1, v1 v_add_co_u32 v9, vcc_lo, s4, v9 s_delay_alu instid0(VALU_DEP_3) v_add_co_ci_u32_e32 v10, vcc_lo, s5, v10, vcc_lo v_cmp_ge_i32_e32 vcc_lo, v7, v6 global_load_b32 v2, v[9:10], off ds_load_b32 v9, v8 v_add_nc_u32_e32 v8, 4, v8 s_or_b32 s15, vcc_lo, s15 s_waitcnt vmcnt(0) lgkmcnt(0) v_fmac_f32_e32 v5, v9, v2 s_and_not1_b32 exec_lo, exec_lo, s15 s_cbranch_execnz .LBB0_14 s_or_b32 exec_lo, exec_lo, s15 .LBB0_16: s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_2) | instid1(VALU_DEP_2) s_or_b32 exec_lo, exec_lo, s14 v_mul_f32_e32 v1, 0xbfb8aa3b, v5 v_cmp_nlt_f32_e32 vcc_lo, 0x42ce8ed0, v5 v_rndne_f32_e32 v2, v1 v_fma_f32 v6, v5, 0xbfb8aa3b, -v1 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_dual_sub_f32 v1, v1, v2 :: v_dual_fmac_f32 v6, 0xb2a5705f, v5 v_cvt_i32_f32_e32 v2, v2 v_add_f32_e32 v1, v1, v6 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_1) v_exp_f32_e32 v1, v1 s_waitcnt_depctr 0xfff v_ldexp_f32 v1, v1, v2 v_cndmask_b32_e32 v1, 0, v1, vcc_lo v_cmp_ngt_f32_e32 vcc_lo, 0xc2b17218, v5 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_cndmask_b32_e32 v1, 0x7f800000, v1, vcc_lo v_add_f32_e32 v1, 1.0, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_div_scale_f32 v2, null, v1, v1, 1.0 v_div_scale_f32 v7, vcc_lo, 1.0, v1, 1.0 v_rcp_f32_e32 v5, v2 s_waitcnt_depctr 0xfff v_fma_f32 v6, -v2, v5, 1.0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmac_f32_e32 v5, v6, v5 v_mul_f32_e32 v6, v7, v5 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f32 v8, -v2, v6, v7 v_fmac_f32_e32 v6, v8, v5 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f32 v2, -v2, v6, v7 v_div_fmas_f32 v2, v2, v5, v6 v_lshl_add_u32 v5, v4, 2, 4 v_add_nc_u32_e32 v4, s13, v4 s_delay_alu instid0(VALU_DEP_3) v_div_fixup_f32 v1, v2, v1, 1.0 v_mov_b32_e32 v2, 1 ds_store_b32 v5, v1 ds_store_b8 v4, v2 .LBB0_17: s_or_b32 exec_lo, exec_lo, s3 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_and_saveexec_b32 s14, s2 s_cbranch_execz .LBB0_9 s_load_b32 s3, s[8:9], 0xc ds_load_b32 v1, v3 s_waitcnt lgkmcnt(0) s_and_b32 s3, s3, 0xffff v_cmp_le_i32_e32 vcc_lo, s10, v1 v_add_nc_u32_e32 v2, s3, v1 v_mov_b32_e32 v4, v1 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_cmp_ge_u32_e64 s3, v1, v2 s_or_b32 s3, vcc_lo, s3 s_delay_alu instid0(SALU_CYCLE_1) s_and_b32 vcc_lo, exec_lo, s3 s_cbranch_vccz .LBB0_20 s_branch .LBB0_9 .p2align 6 .LBB0_19: v_add_nc_u32_e32 v1, 1, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_cmp_lt_u32_e32 vcc_lo, v1, v2 v_cmp_gt_i32_e64 s3, s10, v1 s_and_b32 s3, s3, vcc_lo s_delay_alu instid0(SALU_CYCLE_1) s_and_b32 vcc_lo, exec_lo, s3 s_cbranch_vccz .LBB0_9 .LBB0_20: v_add_nc_u32_e32 v5, s13, v1 ds_load_u8 v5, v5 s_waitcnt lgkmcnt(0) v_cmp_eq_u16_e32 vcc_lo, 0, v5 s_cbranch_vccnz .LBB0_19 v_add_nc_u32_e32 v4, 1, v4 ds_store_b32 v3, v4 s_branch .LBB0_19 .LBB0_22: s_mov_b32 s2, exec_lo v_cmpx_gt_i32_e64 s10, v0 s_cbranch_execz .LBB0_25 s_clause 0x1 s_load_b32 s4, s[0:1], 0x3c s_load_b64 s[2:3], s[0:1], 0x28 v_lshl_add_u32 v2, v0, 2, 4 s_mov_b32 s5, 0 s_waitcnt lgkmcnt(0) s_and_b32 s1, s4, 0xffff s_delay_alu instid0(SALU_CYCLE_1) s_lshl_b32 s4, s1, 2 .p2align 6 .LBB0_24: ds_load_b32 v5, v2 v_ashrrev_i32_e32 v1, 31, v0 v_add_nc_u32_e32 v2, s4, v2 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[3:4], 2, v[0:1] v_add_nc_u32_e32 v0, s1, v0 v_cmp_le_i32_e32 vcc_lo, s10, v0 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v3, s0, s2, v3 v_add_co_ci_u32_e64 v4, s0, s3, v4, s0 s_or_b32 s5, vcc_lo, s5 s_waitcnt lgkmcnt(0) global_store_b32 v[3:4], v5, off s_and_not1_b32 exec_lo, exec_lo, s5 s_cbranch_execnz .LBB0_24 .LBB0_25: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z13produceState2PKfiS0_PKiiPf .amdhsa_group_segment_fixed_size 4 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 304 .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 _Z13produceState2PKfiS0_PKiiPf, .Lfunc_end0-_Z13produceState2PKfiS0_PKiiPf .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 - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 24 .size: 8 .value_kind: global_buffer - .offset: 32 .size: 4 .value_kind: by_value - .address_space: global .offset: 40 .size: 8 .value_kind: global_buffer - .offset: 48 .size: 4 .value_kind: hidden_block_count_x - .offset: 52 .size: 4 .value_kind: hidden_block_count_y - .offset: 56 .size: 4 .value_kind: hidden_block_count_z - .offset: 60 .size: 2 .value_kind: hidden_group_size_x - .offset: 62 .size: 2 .value_kind: hidden_group_size_y - .offset: 64 .size: 2 .value_kind: hidden_group_size_z - .offset: 66 .size: 2 .value_kind: hidden_remainder_x - .offset: 68 .size: 2 .value_kind: hidden_remainder_y - .offset: 70 .size: 2 .value_kind: hidden_remainder_z - .offset: 88 .size: 8 .value_kind: hidden_global_offset_x - .offset: 96 .size: 8 .value_kind: hidden_global_offset_y - .offset: 104 .size: 8 .value_kind: hidden_global_offset_z - .offset: 112 .size: 2 .value_kind: hidden_grid_dims - .offset: 168 .size: 4 .value_kind: hidden_dynamic_lds_size .group_segment_fixed_size: 4 .kernarg_segment_align: 8 .kernarg_segment_size: 304 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z13produceState2PKfiS0_PKiiPf .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z13produceState2PKfiS0_PKiiPf.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" __device__ float sigmoid(float x) { return 1 / (1 + expf(-x)); } __global__ void produceState2(const float* arguments, const int argsSize, const float* weights, const int* topology, const int topSize, float* outStates) { const int tid = threadIdx.x; const int dim = argsSize + topSize; extern __shared__ float s[]; float* states = s; bool* ready = (bool*)&states[dim]; __shared__ int counter[1]; int r = tid; while(r < dim) { ready[r] = false; r += blockDim.x; } if (tid == 0) { counter[tid] = argsSize; } if (tid < argsSize) { states[tid] = arguments[tid]; ready[tid] = true; } __syncthreads(); while(counter[0] < dim) { const int index = counter[0] + tid; const int topIndex = index - argsSize; if (topIndex < topSize) { const int leftBorder = topology[topIndex*3]; const int rightBorder = topology[topIndex*3 + 1]; const int weightsStart = topology[topIndex*3 + 2]; if (rightBorder <= counter[0]) { float sum = 0; for (int i = leftBorder; i < rightBorder; i++) { sum += states[i] * weights[weightsStart + i - leftBorder]; } states[index] = sigmoid(sum); ready[index] = true; } } __syncthreads(); if (tid == 0) { int total = counter[0]; for (int i = total; i < total + blockDim.x && i < dim; i++) { if (ready[i]) { counter[0]++; } } } __syncthreads(); } int n = tid; while(n < dim) { outStates[n] = states[n]; n += blockDim.x; } }
.text .file "produceState2.hip" .globl _Z28__device_stub__produceState2PKfiS0_PKiiPf # -- Begin function _Z28__device_stub__produceState2PKfiS0_PKiiPf .p2align 4, 0x90 .type _Z28__device_stub__produceState2PKfiS0_PKiiPf,@function _Z28__device_stub__produceState2PKfiS0_PKiiPf: # @_Z28__device_stub__produceState2PKfiS0_PKiiPf .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 88(%rsp) movl %esi, 12(%rsp) movq %rdx, 80(%rsp) movq %rcx, 72(%rsp) movl %r8d, 8(%rsp) movq %r9, 64(%rsp) leaq 88(%rsp), %rax movq %rax, 96(%rsp) leaq 12(%rsp), %rax movq %rax, 104(%rsp) leaq 80(%rsp), %rax movq %rax, 112(%rsp) leaq 72(%rsp), %rax movq %rax, 120(%rsp) leaq 8(%rsp), %rax movq %rax, 128(%rsp) leaq 64(%rsp), %rax movq %rax, 136(%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 96(%rsp), %r9 movl $_Z13produceState2PKfiS0_PKiiPf, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $168, %rsp .cfi_adjust_cfa_offset -168 retq .Lfunc_end0: .size _Z28__device_stub__produceState2PKfiS0_PKiiPf, .Lfunc_end0-_Z28__device_stub__produceState2PKfiS0_PKiiPf .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 $_Z13produceState2PKfiS0_PKiiPf, %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 _Z13produceState2PKfiS0_PKiiPf,@object # @_Z13produceState2PKfiS0_PKiiPf .section .rodata,"a",@progbits .globl _Z13produceState2PKfiS0_PKiiPf .p2align 3, 0x0 _Z13produceState2PKfiS0_PKiiPf: .quad _Z28__device_stub__produceState2PKfiS0_PKiiPf .size _Z13produceState2PKfiS0_PKiiPf, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z13produceState2PKfiS0_PKiiPf" .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 _Z28__device_stub__produceState2PKfiS0_PKiiPf .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z13produceState2PKfiS0_PKiiPf .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_000a7ab7_00000000-6_produceState2.cudafe1.cpp" .text #APP #NO_APP .type _ZL26__cudaUnregisterBinaryUtilv, @function _ZL26__cudaUnregisterBinaryUtilv: .LFB2030: .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 .LFE2030: .size _ZL26__cudaUnregisterBinaryUtilv, .-_ZL26__cudaUnregisterBinaryUtilv .globl _Z7sigmoidf .type _Z7sigmoidf, @function _Z7sigmoidf: .LFB2027: .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 .LFE2027: .size _Z7sigmoidf, .-_Z7sigmoidf .globl _Z44__device_stub__Z13produceState2PKfiS0_PKiiPfPKfiS0_PKiiPf .type _Z44__device_stub__Z13produceState2PKfiS0_PKiiPfPKfiS0_PKiiPf, @function _Z44__device_stub__Z13produceState2PKfiS0_PKiiPfPKfiS0_PKiiPf: .LFB2052: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movq %rdi, 40(%rsp) movl %esi, 36(%rsp) movq %rdx, 24(%rsp) movq %rcx, 16(%rsp) movl %r8d, 32(%rsp) movq %r9, 8(%rsp) movq %fs:40, %rax movq %rax, 168(%rsp) xorl %eax, %eax leaq 40(%rsp), %rax movq %rax, 112(%rsp) leaq 36(%rsp), %rax movq %rax, 120(%rsp) leaq 24(%rsp), %rax movq %rax, 128(%rsp) leaq 16(%rsp), %rax movq %rax, 136(%rsp) leaq 32(%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 .L9 .L5: movq 168(%rsp), %rax subq %fs:40, %rax jne .L10 addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L9: .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 _Z13produceState2PKfiS0_PKiiPf(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 192 jmp .L5 .L10: call __stack_chk_fail@PLT .cfi_endproc .LFE2052: .size _Z44__device_stub__Z13produceState2PKfiS0_PKiiPfPKfiS0_PKiiPf, .-_Z44__device_stub__Z13produceState2PKfiS0_PKiiPfPKfiS0_PKiiPf .globl _Z13produceState2PKfiS0_PKiiPf .type _Z13produceState2PKfiS0_PKiiPf, @function _Z13produceState2PKfiS0_PKiiPf: .LFB2053: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z44__device_stub__Z13produceState2PKfiS0_PKiiPfPKfiS0_PKiiPf addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2053: .size _Z13produceState2PKfiS0_PKiiPf, .-_Z13produceState2PKfiS0_PKiiPf .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "_Z13produceState2PKfiS0_PKiiPf" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2055: .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 _Z13produceState2PKfiS0_PKiiPf(%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 .LFE2055: .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 "produceState2.hip" .globl _Z28__device_stub__produceState2PKfiS0_PKiiPf # -- Begin function _Z28__device_stub__produceState2PKfiS0_PKiiPf .p2align 4, 0x90 .type _Z28__device_stub__produceState2PKfiS0_PKiiPf,@function _Z28__device_stub__produceState2PKfiS0_PKiiPf: # @_Z28__device_stub__produceState2PKfiS0_PKiiPf .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 88(%rsp) movl %esi, 12(%rsp) movq %rdx, 80(%rsp) movq %rcx, 72(%rsp) movl %r8d, 8(%rsp) movq %r9, 64(%rsp) leaq 88(%rsp), %rax movq %rax, 96(%rsp) leaq 12(%rsp), %rax movq %rax, 104(%rsp) leaq 80(%rsp), %rax movq %rax, 112(%rsp) leaq 72(%rsp), %rax movq %rax, 120(%rsp) leaq 8(%rsp), %rax movq %rax, 128(%rsp) leaq 64(%rsp), %rax movq %rax, 136(%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 96(%rsp), %r9 movl $_Z13produceState2PKfiS0_PKiiPf, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $168, %rsp .cfi_adjust_cfa_offset -168 retq .Lfunc_end0: .size _Z28__device_stub__produceState2PKfiS0_PKiiPf, .Lfunc_end0-_Z28__device_stub__produceState2PKfiS0_PKiiPf .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 $_Z13produceState2PKfiS0_PKiiPf, %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 _Z13produceState2PKfiS0_PKiiPf,@object # @_Z13produceState2PKfiS0_PKiiPf .section .rodata,"a",@progbits .globl _Z13produceState2PKfiS0_PKiiPf .p2align 3, 0x0 _Z13produceState2PKfiS0_PKiiPf: .quad _Z28__device_stub__produceState2PKfiS0_PKiiPf .size _Z13produceState2PKfiS0_PKiiPf, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z13produceState2PKfiS0_PKiiPf" .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 _Z28__device_stub__produceState2PKfiS0_PKiiPf .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z13produceState2PKfiS0_PKiiPf .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" # define MAX(a, b) ((a) > (b) ? (a) : (b)) # define GAUSSIAN_KERNEL_SIZE 3 # define SOBEL_KERNEL_SIZE 5 # define TILE_WIDTH 32 # define SMEM_SIZE 128 __global__ void initializeSobel(float *d_sobelKernelX, float *d_sobelKernelY) { int ix = threadIdx.x; int iy = threadIdx.y; int weight = SOBEL_KERNEL_SIZE / 2; if (ix < SOBEL_KERNEL_SIZE && iy < SOBEL_KERNEL_SIZE) { int index = iy * SOBEL_KERNEL_SIZE + ix; float sx = ix - SOBEL_KERNEL_SIZE / 2; float sy = iy - SOBEL_KERNEL_SIZE / 2; float norm = sx * sx + sy *sy; if (norm == 0.0f) { d_sobelKernelX[index] = 0.0f; d_sobelKernelY[index] = 0.0f; } else { d_sobelKernelX[index] = sx * weight / norm; d_sobelKernelY[index] = sy * weight / norm; } } }
code for sm_80 Function : _Z15initializeSobelPfS_ .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 R3, SR_TID.Y ; /* 0x0000000000037919 */ /* 0x000e280000002200 */ /*0020*/ S2R R4, SR_TID.X ; /* 0x0000000000047919 */ /* 0x000e620000002100 */ /*0030*/ ISETP.GT.AND P0, PT, R3, 0x4, PT ; /* 0x000000040300780c */ /* 0x001fc80003f04270 */ /*0040*/ ISETP.GT.OR P0, PT, R4, 0x4, P0 ; /* 0x000000040400780c */ /* 0x002fda0000704670 */ /*0050*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0060*/ IADD3 R6, R3.reuse, -0x2, RZ ; /* 0xfffffffe03067810 */ /* 0x040fe20007ffe0ff */ /*0070*/ IMAD.MOV.U32 R5, RZ, RZ, 0x4 ; /* 0x00000004ff057424 */ /* 0x000fe200078e00ff */ /*0080*/ IADD3 R7, R4, -0x2, RZ ; /* 0xfffffffe04077810 */ /* 0x000fe20007ffe0ff */ /*0090*/ IMAD R4, R3, 0x5, R4 ; /* 0x0000000503047824 */ /* 0x000fe200078e0204 */ /*00a0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe40000000a00 */ /*00b0*/ I2F R6, R6 ; /* 0x0000000600067306 */ /* 0x000e220000201400 */ /*00c0*/ IMAD.WIDE R2, R4, R5, c[0x0][0x160] ; /* 0x0000580004027625 */ /* 0x000fc800078e0205 */ /*00d0*/ IMAD.WIDE R4, R4, R5, c[0x0][0x168] ; /* 0x00005a0004047625 */ /* 0x000fc600078e0205 */ /*00e0*/ I2F R7, R7 ; /* 0x0000000700077306 */ /* 0x000e620000201400 */ /*00f0*/ FMUL R0, R6, R6 ; /* 0x0000000606007220 */ /* 0x001fc80000400000 */ /*0100*/ FFMA R0, R7, R7, R0 ; /* 0x0000000707007223 */ /* 0x002fca0000000000 */ /*0110*/ FSETP.NEU.AND P0, PT, R0, RZ, PT ; /* 0x000000ff0000720b */ /* 0x000fda0003f0d000 */ /*0120*/ @!P0 BRA 0x340 ; /* 0x0000021000008947 */ /* 0x000fea0003800000 */ /*0130*/ MUFU.RCP R9, R0 ; /* 0x0000000000097308 */ /* 0x000e220000001000 */ /*0140*/ FADD R7, R7, R7 ; /* 0x0000000707077221 */ /* 0x000fe20000000000 */ /*0150*/ BSSY B0, 0x220 ; /* 0x000000c000007945 */ /* 0x000fec0003800000 */ /*0160*/ FCHK P0, R7, R0 ; /* 0x0000000007007302 */ /* 0x000e620000000000 */ /*0170*/ FFMA R8, -R0, R9, 1 ; /* 0x3f80000000087423 */ /* 0x001fc80000000109 */ /*0180*/ FFMA R8, R9, R8, R9 ; /* 0x0000000809087223 */ /* 0x000fc80000000009 */ /*0190*/ FFMA R9, R7, R8, RZ ; /* 0x0000000807097223 */ /* 0x000fc800000000ff */ /*01a0*/ FFMA R10, -R0, R9, R7 ; /* 0x00000009000a7223 */ /* 0x000fc80000000107 */ /*01b0*/ FFMA R9, R8, R10, R9 ; /* 0x0000000a08097223 */ /* 0x000fe20000000009 */ /*01c0*/ @!P0 BRA 0x210 ; /* 0x0000004000008947 */ /* 0x002fea0003800000 */ /*01d0*/ IMAD.MOV.U32 R11, RZ, RZ, R7 ; /* 0x000000ffff0b7224 */ /* 0x000fe200078e0007 */ /*01e0*/ MOV R8, 0x200 ; /* 0x0000020000087802 */ /* 0x000fe40000000f00 */ /*01f0*/ CALL.REL.NOINC 0x370 ; /* 0x0000017000007944 */ /* 0x000fea0003c00000 */ /*0200*/ IMAD.MOV.U32 R9, RZ, RZ, R7 ; /* 0x000000ffff097224 */ /* 0x001fe400078e0007 */ /*0210*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0220*/ MUFU.RCP R11, R0 ; /* 0x00000000000b7308 */ /* 0x000e220000001000 */ /*0230*/ FADD R7, R6, R6 ; /* 0x0000000606077221 */ /* 0x000fe20000000000 */ /*0240*/ STG.E [R2.64], R9 ; /* 0x0000000902007986 */ /* 0x0003e2000c101904 */ /*0250*/ BSSY B0, 0x320 ; /* 0x000000c000007945 */ /* 0x000fea0003800000 */ /*0260*/ FCHK P0, R7, R0 ; /* 0x0000000007007302 */ /* 0x000ea20000000000 */ /*0270*/ FFMA R8, -R0, R11, 1 ; /* 0x3f80000000087423 */ /* 0x001fc8000000010b */ /*0280*/ FFMA R8, R11, R8, R11 ; /* 0x000000080b087223 */ /* 0x000fc8000000000b */ /*0290*/ FFMA R6, R8, R7, RZ ; /* 0x0000000708067223 */ /* 0x000fc800000000ff */ /*02a0*/ FFMA R11, -R0, R6, R7 ; /* 0x00000006000b7223 */ /* 0x000fc80000000107 */ /*02b0*/ FFMA R11, R8, R11, R6 ; /* 0x0000000b080b7223 */ /* 0x000fe20000000006 */ /*02c0*/ @!P0 BRA 0x310 ; /* 0x0000004000008947 */ /* 0x004fea0003800000 */ /*02d0*/ IMAD.MOV.U32 R11, RZ, RZ, R7 ; /* 0x000000ffff0b7224 */ /* 0x002fe200078e0007 */ /*02e0*/ MOV R8, 0x300 ; /* 0x0000030000087802 */ /* 0x000fe40000000f00 */ /*02f0*/ CALL.REL.NOINC 0x370 ; /* 0x0000007000007944 */ /* 0x000fea0003c00000 */ /*0300*/ IMAD.MOV.U32 R11, RZ, RZ, R7 ; /* 0x000000ffff0b7224 */ /* 0x001fe400078e0007 */ /*0310*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x002fea0003800000 */ /*0320*/ STG.E [R4.64], R11 ; /* 0x0000000b04007986 */ /* 0x000fe2000c101904 */ /*0330*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0340*/ STG.E [R2.64], RZ ; /* 0x000000ff02007986 */ /* 0x000fe8000c101904 */ /*0350*/ STG.E [R4.64], RZ ; /* 0x000000ff04007986 */ /* 0x000fe2000c101904 */ /*0360*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0370*/ SHF.R.U32.HI R10, RZ, 0x17, R0.reuse ; /* 0x00000017ff0a7819 */ /* 0x100fe20000011600 */ /*0380*/ BSSY B1, 0x9d0 ; /* 0x0000064000017945 */ /* 0x000fe20003800000 */ /*0390*/ SHF.R.U32.HI R7, RZ, 0x17, R11 ; /* 0x00000017ff077819 */ /* 0x000fe2000001160b */ /*03a0*/ IMAD.MOV.U32 R12, RZ, RZ, R0 ; /* 0x000000ffff0c7224 */ /* 0x000fe200078e0000 */ /*03b0*/ LOP3.LUT R10, R10, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff0a0a7812 */ /* 0x000fc400078ec0ff */ /*03c0*/ LOP3.LUT R15, R7, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff070f7812 */ /* 0x000fe200078ec0ff */ /*03d0*/ IMAD.MOV.U32 R7, RZ, RZ, R11 ; /* 0x000000ffff077224 */ /* 0x000fe200078e000b */ /*03e0*/ IADD3 R14, R10, -0x1, RZ ; /* 0xffffffff0a0e7810 */ /* 0x000fe40007ffe0ff */ /*03f0*/ IADD3 R13, R15, -0x1, RZ ; /* 0xffffffff0f0d7810 */ /* 0x000fe40007ffe0ff */ /*0400*/ ISETP.GT.U32.AND P0, PT, R14, 0xfd, PT ; /* 0x000000fd0e00780c */ /* 0x000fc80003f04070 */ /*0410*/ ISETP.GT.U32.OR P0, PT, R13, 0xfd, P0 ; /* 0x000000fd0d00780c */ /* 0x000fda0000704470 */ /*0420*/ @!P0 IMAD.MOV.U32 R9, RZ, RZ, RZ ; /* 0x000000ffff098224 */ /* 0x000fe200078e00ff */ /*0430*/ @!P0 BRA 0x5b0 ; /* 0x0000017000008947 */ /* 0x000fea0003800000 */ /*0440*/ FSETP.GTU.FTZ.AND P0, PT, |R11|, +INF , PT ; /* 0x7f8000000b00780b */ /* 0x000fe40003f1c200 */ /*0450*/ FSETP.GTU.FTZ.AND P1, PT, |R0|, +INF , PT ; /* 0x7f8000000000780b */ /* 0x000fc80003f3c200 */ /*0460*/ PLOP3.LUT P0, PT, P0, P1, PT, 0xa8, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000703570 */ /*0470*/ @P0 BRA 0x9b0 ; /* 0x0000053000000947 */ /* 0x000fea0003800000 */ /*0480*/ LOP3.LUT P0, RZ, R12, 0x7fffffff, R7, 0xc8, !PT ; /* 0x7fffffff0cff7812 */ /* 0x000fda000780c807 */ /*0490*/ @!P0 BRA 0x990 ; /* 0x000004f000008947 */ /* 0x000fea0003800000 */ /*04a0*/ FSETP.NEU.FTZ.AND P2, PT, |R11|.reuse, +INF , PT ; /* 0x7f8000000b00780b */ /* 0x040fe40003f5d200 */ /*04b0*/ FSETP.NEU.FTZ.AND P1, PT, |R0|, +INF , PT ; /* 0x7f8000000000780b */ /* 0x000fe40003f3d200 */ /*04c0*/ FSETP.NEU.FTZ.AND P0, PT, |R11|, +INF , PT ; /* 0x7f8000000b00780b */ /* 0x000fd60003f1d200 */ /*04d0*/ @!P1 BRA !P2, 0x990 ; /* 0x000004b000009947 */ /* 0x000fea0005000000 */ /*04e0*/ LOP3.LUT P2, RZ, R7, 0x7fffffff, RZ, 0xc0, !PT ; /* 0x7fffffff07ff7812 */ /* 0x000fc8000784c0ff */ /*04f0*/ PLOP3.LUT P1, PT, P1, P2, PT, 0x2a, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000f24572 */ /*0500*/ @P1 BRA 0x970 ; /* 0x0000046000001947 */ /* 0x000fea0003800000 */ /*0510*/ LOP3.LUT P1, RZ, R12, 0x7fffffff, RZ, 0xc0, !PT ; /* 0x7fffffff0cff7812 */ /* 0x000fc8000782c0ff */ /*0520*/ PLOP3.LUT P0, PT, P0, P1, PT, 0x2a, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000702572 */ /*0530*/ @P0 BRA 0x940 ; /* 0x0000040000000947 */ /* 0x000fea0003800000 */ /*0540*/ ISETP.GE.AND P0, PT, R13, RZ, PT ; /* 0x000000ff0d00720c */ /* 0x000fe40003f06270 */ /*0550*/ ISETP.GE.AND P1, PT, R14, RZ, PT ; /* 0x000000ff0e00720c */ /* 0x000fd60003f26270 */ /*0560*/ @P0 IMAD.MOV.U32 R9, RZ, RZ, RZ ; /* 0x000000ffff090224 */ /* 0x000fe400078e00ff */ /*0570*/ @!P0 IMAD.MOV.U32 R9, RZ, RZ, -0x40 ; /* 0xffffffc0ff098424 */ /* 0x000fe400078e00ff */ /*0580*/ @!P0 FFMA R7, R11, 1.84467440737095516160e+19, RZ ; /* 0x5f8000000b078823 */ /* 0x000fe400000000ff */ /*0590*/ @!P1 FFMA R12, R0, 1.84467440737095516160e+19, RZ ; /* 0x5f800000000c9823 */ /* 0x000fe200000000ff */ /*05a0*/ @!P1 IADD3 R9, R9, 0x40, RZ ; /* 0x0000004009099810 */ /* 0x000fe40007ffe0ff */ /*05b0*/ LEA R11, R10, 0xc0800000, 0x17 ; /* 0xc08000000a0b7811 */ /* 0x000fe200078eb8ff */ /*05c0*/ BSSY B2, 0x930 ; /* 0x0000036000027945 */ /* 0x000fe80003800000 */ /*05d0*/ IMAD.IADD R12, R12, 0x1, -R11 ; /* 0x000000010c0c7824 */ /* 0x000fe200078e0a0b */ /*05e0*/ IADD3 R11, R15, -0x7f, RZ ; /* 0xffffff810f0b7810 */ /* 0x000fc60007ffe0ff */ /*05f0*/ MUFU.RCP R13, R12 ; /* 0x0000000c000d7308 */ /* 0x0000620000001000 */ /*0600*/ FADD.FTZ R14, -R12, -RZ ; /* 0x800000ff0c0e7221 */ /* 0x000fe40000010100 */ /*0610*/ IMAD R7, R11.reuse, -0x800000, R7 ; /* 0xff8000000b077824 */ /* 0x040fe200078e0207 */ /*0620*/ IADD3 R12, R11, 0x7f, -R10 ; /* 0x0000007f0b0c7810 */ /* 0x001fca0007ffe80a */ /*0630*/ IMAD.IADD R12, R12, 0x1, R9 ; /* 0x000000010c0c7824 */ /* 0x000fe400078e0209 */ /*0640*/ FFMA R16, R13, R14, 1 ; /* 0x3f8000000d107423 */ /* 0x002fc8000000000e */ /*0650*/ FFMA R18, R13, R16, R13 ; /* 0x000000100d127223 */ /* 0x000fc8000000000d */ /*0660*/ FFMA R13, R7, R18, RZ ; /* 0x00000012070d7223 */ /* 0x000fc800000000ff */ /*0670*/ FFMA R16, R14, R13, R7 ; /* 0x0000000d0e107223 */ /* 0x000fc80000000007 */ /*0680*/ FFMA R13, R18, R16, R13 ; /* 0x00000010120d7223 */ /* 0x000fc8000000000d */ /*0690*/ FFMA R14, R14, R13, R7 ; /* 0x0000000d0e0e7223 */ /* 0x000fc80000000007 */ /*06a0*/ FFMA R7, R18, R14, R13 ; /* 0x0000000e12077223 */ /* 0x000fca000000000d */ /*06b0*/ SHF.R.U32.HI R10, RZ, 0x17, R7 ; /* 0x00000017ff0a7819 */ /* 0x000fc80000011607 */ /*06c0*/ LOP3.LUT R10, R10, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff0a0a7812 */ /* 0x000fca00078ec0ff */ /*06d0*/ IMAD.IADD R15, R10, 0x1, R12 ; /* 0x000000010a0f7824 */ /* 0x000fca00078e020c */ /*06e0*/ IADD3 R9, R15, -0x1, RZ ; /* 0xffffffff0f097810 */ /* 0x000fc80007ffe0ff */ /*06f0*/ ISETP.GE.U32.AND P0, PT, R9, 0xfe, PT ; /* 0x000000fe0900780c */ /* 0x000fda0003f06070 */ /*0700*/ @!P0 BRA 0x910 ; /* 0x0000020000008947 */ /* 0x000fea0003800000 */ /*0710*/ ISETP.GT.AND P0, PT, R15, 0xfe, PT ; /* 0x000000fe0f00780c */ /* 0x000fda0003f04270 */ /*0720*/ @P0 BRA 0x8e0 ; /* 0x000001b000000947 */ /* 0x000fea0003800000 */ /*0730*/ ISETP.GE.AND P0, PT, R15, 0x1, PT ; /* 0x000000010f00780c */ /* 0x000fda0003f06270 */ /*0740*/ @P0 BRA 0x920 ; /* 0x000001d000000947 */ /* 0x000fea0003800000 */ /*0750*/ ISETP.GE.AND P0, PT, R15, -0x18, PT ; /* 0xffffffe80f00780c */ /* 0x000fe40003f06270 */ /*0760*/ LOP3.LUT R7, R7, 0x80000000, RZ, 0xc0, !PT ; /* 0x8000000007077812 */ /* 0x000fd600078ec0ff */ /*0770*/ @!P0 BRA 0x920 ; /* 0x000001a000008947 */ /* 0x000fea0003800000 */ /*0780*/ FFMA.RZ R9, R18.reuse, R14.reuse, R13.reuse ; /* 0x0000000e12097223 */ /* 0x1c0fe2000000c00d */ /*0790*/ IADD3 R12, R15.reuse, 0x20, RZ ; /* 0x000000200f0c7810 */ /* 0x040fe20007ffe0ff */ /*07a0*/ FFMA.RM R10, R18.reuse, R14.reuse, R13.reuse ; /* 0x0000000e120a7223 */ /* 0x1c0fe2000000400d */ /*07b0*/ ISETP.NE.AND P2, PT, R15, RZ, PT ; /* 0x000000ff0f00720c */ /* 0x000fe40003f45270 */ /*07c0*/ LOP3.LUT R11, R9, 0x7fffff, RZ, 0xc0, !PT ; /* 0x007fffff090b7812 */ /* 0x000fe200078ec0ff */ /*07d0*/ FFMA.RP R9, R18, R14, R13 ; /* 0x0000000e12097223 */ /* 0x000fe2000000800d */ /*07e0*/ ISETP.NE.AND P1, PT, R15, RZ, PT ; /* 0x000000ff0f00720c */ /* 0x000fe20003f25270 */ /*07f0*/ IMAD.MOV R13, RZ, RZ, -R15 ; /* 0x000000ffff0d7224 */ /* 0x000fe200078e0a0f */ /*0800*/ LOP3.LUT R11, R11, 0x800000, RZ, 0xfc, !PT ; /* 0x008000000b0b7812 */ /* 0x000fe400078efcff */ /*0810*/ FSETP.NEU.FTZ.AND P0, PT, R9, R10, PT ; /* 0x0000000a0900720b */ /* 0x000fc40003f1d000 */ /*0820*/ SHF.L.U32 R12, R11, R12, RZ ; /* 0x0000000c0b0c7219 */ /* 0x000fe400000006ff */ /*0830*/ SEL R10, R13, RZ, P2 ; /* 0x000000ff0d0a7207 */ /* 0x000fe40001000000 */ /*0840*/ ISETP.NE.AND P1, PT, R12, RZ, P1 ; /* 0x000000ff0c00720c */ /* 0x000fe40000f25270 */ /*0850*/ SHF.R.U32.HI R10, RZ, R10, R11 ; /* 0x0000000aff0a7219 */ /* 0x000fe4000001160b */ /*0860*/ PLOP3.LUT P0, PT, P0, P1, PT, 0xa8, 0x0 ; /* 0x000000000000781c */ /* 0x000fe40000703570 */ /*0870*/ SHF.R.U32.HI R12, RZ, 0x1, R10 ; /* 0x00000001ff0c7819 */ /* 0x000fc4000001160a */ /*0880*/ SEL R9, RZ, 0x1, !P0 ; /* 0x00000001ff097807 */ /* 0x000fc80004000000 */ /*0890*/ LOP3.LUT R9, R9, 0x1, R12, 0xf8, !PT ; /* 0x0000000109097812 */ /* 0x000fc800078ef80c */ /*08a0*/ LOP3.LUT R9, R9, R10, RZ, 0xc0, !PT ; /* 0x0000000a09097212 */ /* 0x000fca00078ec0ff */ /*08b0*/ IMAD.IADD R12, R12, 0x1, R9 ; /* 0x000000010c0c7824 */ /* 0x000fca00078e0209 */ /*08c0*/ LOP3.LUT R7, R12, R7, RZ, 0xfc, !PT ; /* 0x000000070c077212 */ /* 0x000fe200078efcff */ /*08d0*/ BRA 0x920 ; /* 0x0000004000007947 */ /* 0x000fea0003800000 */ /*08e0*/ LOP3.LUT R7, R7, 0x80000000, RZ, 0xc0, !PT ; /* 0x8000000007077812 */ /* 0x000fc800078ec0ff */ /*08f0*/ LOP3.LUT R7, R7, 0x7f800000, RZ, 0xfc, !PT ; /* 0x7f80000007077812 */ /* 0x000fe200078efcff */ /*0900*/ BRA 0x920 ; /* 0x0000001000007947 */ /* 0x000fea0003800000 */ /*0910*/ IMAD R7, R12, 0x800000, R7 ; /* 0x008000000c077824 */ /* 0x000fe400078e0207 */ /*0920*/ BSYNC B2 ; /* 0x0000000000027941 */ /* 0x000fea0003800000 */ /*0930*/ BRA 0x9c0 ; /* 0x0000008000007947 */ /* 0x000fea0003800000 */ /*0940*/ LOP3.LUT R7, R12, 0x80000000, R7, 0x48, !PT ; /* 0x800000000c077812 */ /* 0x000fc800078e4807 */ /*0950*/ LOP3.LUT R7, R7, 0x7f800000, RZ, 0xfc, !PT ; /* 0x7f80000007077812 */ /* 0x000fe200078efcff */ /*0960*/ BRA 0x9c0 ; /* 0x0000005000007947 */ /* 0x000fea0003800000 */ /*0970*/ LOP3.LUT R7, R12, 0x80000000, R7, 0x48, !PT ; /* 0x800000000c077812 */ /* 0x000fe200078e4807 */ /*0980*/ BRA 0x9c0 ; /* 0x0000003000007947 */ /* 0x000fea0003800000 */ /*0990*/ MUFU.RSQ R7, -QNAN ; /* 0xffc0000000077908 */ /* 0x000e220000001400 */ /*09a0*/ BRA 0x9c0 ; /* 0x0000001000007947 */ /* 0x000fea0003800000 */ /*09b0*/ FADD.FTZ R7, R11, R0 ; /* 0x000000000b077221 */ /* 0x000fe40000010000 */ /*09c0*/ BSYNC B1 ; /* 0x0000000000017941 */ /* 0x000fea0003800000 */ /*09d0*/ IMAD.MOV.U32 R9, RZ, RZ, 0x0 ; /* 0x00000000ff097424 */ /* 0x000fc800078e00ff */ /*09e0*/ RET.REL.NODEC R8 0x0 ; /* 0xfffff61008007950 */ /* 0x000fea0003c3ffff */ /*09f0*/ BRA 0x9f0; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0a00*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a10*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a20*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a30*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a40*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a50*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a60*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a70*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include "includes.h" # define MAX(a, b) ((a) > (b) ? (a) : (b)) # define GAUSSIAN_KERNEL_SIZE 3 # define SOBEL_KERNEL_SIZE 5 # define TILE_WIDTH 32 # define SMEM_SIZE 128 __global__ void initializeSobel(float *d_sobelKernelX, float *d_sobelKernelY) { int ix = threadIdx.x; int iy = threadIdx.y; int weight = SOBEL_KERNEL_SIZE / 2; if (ix < SOBEL_KERNEL_SIZE && iy < SOBEL_KERNEL_SIZE) { int index = iy * SOBEL_KERNEL_SIZE + ix; float sx = ix - SOBEL_KERNEL_SIZE / 2; float sy = iy - SOBEL_KERNEL_SIZE / 2; float norm = sx * sx + sy *sy; if (norm == 0.0f) { d_sobelKernelX[index] = 0.0f; d_sobelKernelY[index] = 0.0f; } else { d_sobelKernelX[index] = sx * weight / norm; d_sobelKernelY[index] = sy * weight / norm; } } }
.file "tmpxft_0000b160_00000000-6_initializeSobel.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 _Z37__device_stub__Z15initializeSobelPfS_PfS_ .type _Z37__device_stub__Z15initializeSobelPfS_PfS_, @function _Z37__device_stub__Z15initializeSobelPfS_PfS_: .LFB2051: .cfi_startproc endbr64 subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 8(%rsp) movq %rsi, (%rsp) movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax leaq 8(%rsp), %rax movq %rax, 80(%rsp) movq %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 _Z15initializeSobelPfS_(%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 _Z37__device_stub__Z15initializeSobelPfS_PfS_, .-_Z37__device_stub__Z15initializeSobelPfS_PfS_ .globl _Z15initializeSobelPfS_ .type _Z15initializeSobelPfS_, @function _Z15initializeSobelPfS_: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z37__device_stub__Z15initializeSobelPfS_PfS_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z15initializeSobelPfS_, .-_Z15initializeSobelPfS_ .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "_Z15initializeSobelPfS_" .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 _Z15initializeSobelPfS_(%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 MAX(a, b) ((a) > (b) ? (a) : (b)) # define GAUSSIAN_KERNEL_SIZE 3 # define SOBEL_KERNEL_SIZE 5 # define TILE_WIDTH 32 # define SMEM_SIZE 128 __global__ void initializeSobel(float *d_sobelKernelX, float *d_sobelKernelY) { int ix = threadIdx.x; int iy = threadIdx.y; int weight = SOBEL_KERNEL_SIZE / 2; if (ix < SOBEL_KERNEL_SIZE && iy < SOBEL_KERNEL_SIZE) { int index = iy * SOBEL_KERNEL_SIZE + ix; float sx = ix - SOBEL_KERNEL_SIZE / 2; float sy = iy - SOBEL_KERNEL_SIZE / 2; float norm = sx * sx + sy *sy; if (norm == 0.0f) { d_sobelKernelX[index] = 0.0f; d_sobelKernelY[index] = 0.0f; } else { d_sobelKernelX[index] = sx * weight / norm; d_sobelKernelY[index] = sy * weight / norm; } } }
#include <hip/hip_runtime.h> #include "includes.h" # define MAX(a, b) ((a) > (b) ? (a) : (b)) # define GAUSSIAN_KERNEL_SIZE 3 # define SOBEL_KERNEL_SIZE 5 # define TILE_WIDTH 32 # define SMEM_SIZE 128 __global__ void initializeSobel(float *d_sobelKernelX, float *d_sobelKernelY) { int ix = threadIdx.x; int iy = threadIdx.y; int weight = SOBEL_KERNEL_SIZE / 2; if (ix < SOBEL_KERNEL_SIZE && iy < SOBEL_KERNEL_SIZE) { int index = iy * SOBEL_KERNEL_SIZE + ix; float sx = ix - SOBEL_KERNEL_SIZE / 2; float sy = iy - SOBEL_KERNEL_SIZE / 2; float norm = sx * sx + sy *sy; if (norm == 0.0f) { d_sobelKernelX[index] = 0.0f; d_sobelKernelY[index] = 0.0f; } else { d_sobelKernelX[index] = sx * weight / norm; d_sobelKernelY[index] = sy * weight / norm; } } }
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 MAX(a, b) ((a) > (b) ? (a) : (b)) # define GAUSSIAN_KERNEL_SIZE 3 # define SOBEL_KERNEL_SIZE 5 # define TILE_WIDTH 32 # define SMEM_SIZE 128 __global__ void initializeSobel(float *d_sobelKernelX, float *d_sobelKernelY) { int ix = threadIdx.x; int iy = threadIdx.y; int weight = SOBEL_KERNEL_SIZE / 2; if (ix < SOBEL_KERNEL_SIZE && iy < SOBEL_KERNEL_SIZE) { int index = iy * SOBEL_KERNEL_SIZE + ix; float sx = ix - SOBEL_KERNEL_SIZE / 2; float sy = iy - SOBEL_KERNEL_SIZE / 2; float norm = sx * sx + sy *sy; if (norm == 0.0f) { d_sobelKernelX[index] = 0.0f; d_sobelKernelY[index] = 0.0f; } else { d_sobelKernelX[index] = sx * weight / norm; d_sobelKernelY[index] = sy * weight / norm; } } }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z15initializeSobelPfS_ .globl _Z15initializeSobelPfS_ .p2align 8 .type _Z15initializeSobelPfS_,@function _Z15initializeSobelPfS_: v_and_b32_e32 v3, 0x3ff, v0 v_bfe_u32 v0, v0, 10, 10 s_mov_b32 s2, exec_lo s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_max_u32_e32 v1, v3, v0 v_cmpx_gt_u32_e32 5, v1 s_cbranch_execz .LBB0_6 v_add_nc_u32_e32 v1, -2, v0 s_load_b64 s[4:5], s[0:1], 0x0 v_add_nc_u32_e32 v2, -2, v3 v_mad_u32_u24 v0, v0, 5, v3 s_mov_b32 s2, exec_lo v_cvt_f32_i32_e32 v1, v1 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_cvt_f32_i32_e32 v4, v2 v_mul_f32_e32 v2, v1, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmac_f32_e32 v2, v4, v4 v_cmpx_neq_f32_e32 0, v2 s_xor_b32 s3, exec_lo, s2 s_cbranch_execz .LBB0_3 v_dual_add_f32 v3, v4, v4 :: v_dual_add_f32 v4, v1, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_div_scale_f32 v1, null, v2, v2, v3 v_div_scale_f32 v5, null, v2, v2, v4 v_div_scale_f32 v10, vcc_lo, v3, v2, v3 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_rcp_f32_e32 v6, v1 v_rcp_f32_e32 v7, v5 s_waitcnt_depctr 0xfff v_fma_f32 v8, -v1, v6, 1.0 v_fma_f32 v9, -v5, v7, 1.0 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_dual_fmac_f32 v6, v8, v6 :: v_dual_fmac_f32 v7, v9, v7 v_div_scale_f32 v8, s2, v4, v2, v4 v_mul_f32_e32 v9, v10, v6 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_mul_f32_e32 v11, v8, v7 v_fma_f32 v12, -v1, v9, v10 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f32 v13, -v5, v11, v8 v_fmac_f32_e32 v9, v12, v6 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fmac_f32_e32 v11, v13, v7 v_fma_f32 v1, -v1, v9, v10 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f32 v5, -v5, v11, v8 v_div_fmas_f32 v1, v1, v6, v9 s_mov_b32 vcc_lo, s2 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_3) v_div_fmas_f32 v5, v5, v7, v11 v_lshlrev_b32_e32 v7, 2, v0 v_div_fixup_f32 v6, v1, v2, v3 v_mov_b32_e32 v1, 0 s_delay_alu instid0(VALU_DEP_4) v_div_fixup_f32 v3, v5, v2, v4 s_waitcnt lgkmcnt(0) global_store_b32 v7, v6, s[4:5] .LBB0_3: s_and_not1_saveexec_b32 s2, s3 s_cbranch_execz .LBB0_5 v_dual_mov_b32 v1, 0 :: v_dual_lshlrev_b32 v2, 2, v0 s_delay_alu instid0(VALU_DEP_1) v_mov_b32_e32 v3, v1 s_waitcnt lgkmcnt(0) global_store_b32 v2, v1, s[4:5] .LBB0_5: s_or_b32 exec_lo, exec_lo, s2 s_load_b64 s[0:1], s[0:1], 0x8 v_lshlrev_b64 v[0:1], 2, v[0:1] 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_6: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z15initializeSobelPfS_ .amdhsa_group_segment_fixed_size 0 .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 1 .amdhsa_next_free_vgpr 14 .amdhsa_next_free_sgpr 6 .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 _Z15initializeSobelPfS_, .Lfunc_end0-_Z15initializeSobelPfS_ .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 .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 16 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z15initializeSobelPfS_ .private_segment_fixed_size: 0 .sgpr_count: 8 .sgpr_spill_count: 0 .symbol: _Z15initializeSobelPfS_.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" # define MAX(a, b) ((a) > (b) ? (a) : (b)) # define GAUSSIAN_KERNEL_SIZE 3 # define SOBEL_KERNEL_SIZE 5 # define TILE_WIDTH 32 # define SMEM_SIZE 128 __global__ void initializeSobel(float *d_sobelKernelX, float *d_sobelKernelY) { int ix = threadIdx.x; int iy = threadIdx.y; int weight = SOBEL_KERNEL_SIZE / 2; if (ix < SOBEL_KERNEL_SIZE && iy < SOBEL_KERNEL_SIZE) { int index = iy * SOBEL_KERNEL_SIZE + ix; float sx = ix - SOBEL_KERNEL_SIZE / 2; float sy = iy - SOBEL_KERNEL_SIZE / 2; float norm = sx * sx + sy *sy; if (norm == 0.0f) { d_sobelKernelX[index] = 0.0f; d_sobelKernelY[index] = 0.0f; } else { d_sobelKernelX[index] = sx * weight / norm; d_sobelKernelY[index] = sy * weight / norm; } } }
.text .file "initializeSobel.hip" .globl _Z30__device_stub__initializeSobelPfS_ # -- Begin function _Z30__device_stub__initializeSobelPfS_ .p2align 4, 0x90 .type _Z30__device_stub__initializeSobelPfS_,@function _Z30__device_stub__initializeSobelPfS_: # @_Z30__device_stub__initializeSobelPfS_ .cfi_startproc # %bb.0: subq $88, %rsp .cfi_def_cfa_offset 96 movq %rdi, 56(%rsp) movq %rsi, 48(%rsp) leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 48(%rsp), %rax movq %rax, 72(%rsp) 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 64(%rsp), %r9 movl $_Z15initializeSobelPfS_, %edi pushq (%rsp) .cfi_adjust_cfa_offset 8 pushq 16(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $104, %rsp .cfi_adjust_cfa_offset -104 retq .Lfunc_end0: .size _Z30__device_stub__initializeSobelPfS_, .Lfunc_end0-_Z30__device_stub__initializeSobelPfS_ .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 $_Z15initializeSobelPfS_, %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 _Z15initializeSobelPfS_,@object # @_Z15initializeSobelPfS_ .section .rodata,"a",@progbits .globl _Z15initializeSobelPfS_ .p2align 3, 0x0 _Z15initializeSobelPfS_: .quad _Z30__device_stub__initializeSobelPfS_ .size _Z15initializeSobelPfS_, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z15initializeSobelPfS_" .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 _Z30__device_stub__initializeSobelPfS_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z15initializeSobelPfS_ .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 : _Z15initializeSobelPfS_ .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 R3, SR_TID.Y ; /* 0x0000000000037919 */ /* 0x000e280000002200 */ /*0020*/ S2R R4, SR_TID.X ; /* 0x0000000000047919 */ /* 0x000e620000002100 */ /*0030*/ ISETP.GT.AND P0, PT, R3, 0x4, PT ; /* 0x000000040300780c */ /* 0x001fc80003f04270 */ /*0040*/ ISETP.GT.OR P0, PT, R4, 0x4, P0 ; /* 0x000000040400780c */ /* 0x002fda0000704670 */ /*0050*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0060*/ IADD3 R6, R3.reuse, -0x2, RZ ; /* 0xfffffffe03067810 */ /* 0x040fe20007ffe0ff */ /*0070*/ IMAD.MOV.U32 R5, RZ, RZ, 0x4 ; /* 0x00000004ff057424 */ /* 0x000fe200078e00ff */ /*0080*/ IADD3 R7, R4, -0x2, RZ ; /* 0xfffffffe04077810 */ /* 0x000fe20007ffe0ff */ /*0090*/ IMAD R4, R3, 0x5, R4 ; /* 0x0000000503047824 */ /* 0x000fe200078e0204 */ /*00a0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe40000000a00 */ /*00b0*/ I2F R6, R6 ; /* 0x0000000600067306 */ /* 0x000e220000201400 */ /*00c0*/ IMAD.WIDE R2, R4, R5, c[0x0][0x160] ; /* 0x0000580004027625 */ /* 0x000fc800078e0205 */ /*00d0*/ IMAD.WIDE R4, R4, R5, c[0x0][0x168] ; /* 0x00005a0004047625 */ /* 0x000fc600078e0205 */ /*00e0*/ I2F R7, R7 ; /* 0x0000000700077306 */ /* 0x000e620000201400 */ /*00f0*/ FMUL R0, R6, R6 ; /* 0x0000000606007220 */ /* 0x001fc80000400000 */ /*0100*/ FFMA R0, R7, R7, R0 ; /* 0x0000000707007223 */ /* 0x002fca0000000000 */ /*0110*/ FSETP.NEU.AND P0, PT, R0, RZ, PT ; /* 0x000000ff0000720b */ /* 0x000fda0003f0d000 */ /*0120*/ @!P0 BRA 0x340 ; /* 0x0000021000008947 */ /* 0x000fea0003800000 */ /*0130*/ MUFU.RCP R9, R0 ; /* 0x0000000000097308 */ /* 0x000e220000001000 */ /*0140*/ FADD R7, R7, R7 ; /* 0x0000000707077221 */ /* 0x000fe20000000000 */ /*0150*/ BSSY B0, 0x220 ; /* 0x000000c000007945 */ /* 0x000fec0003800000 */ /*0160*/ FCHK P0, R7, R0 ; /* 0x0000000007007302 */ /* 0x000e620000000000 */ /*0170*/ FFMA R8, -R0, R9, 1 ; /* 0x3f80000000087423 */ /* 0x001fc80000000109 */ /*0180*/ FFMA R8, R9, R8, R9 ; /* 0x0000000809087223 */ /* 0x000fc80000000009 */ /*0190*/ FFMA R9, R7, R8, RZ ; /* 0x0000000807097223 */ /* 0x000fc800000000ff */ /*01a0*/ FFMA R10, -R0, R9, R7 ; /* 0x00000009000a7223 */ /* 0x000fc80000000107 */ /*01b0*/ FFMA R9, R8, R10, R9 ; /* 0x0000000a08097223 */ /* 0x000fe20000000009 */ /*01c0*/ @!P0 BRA 0x210 ; /* 0x0000004000008947 */ /* 0x002fea0003800000 */ /*01d0*/ IMAD.MOV.U32 R11, RZ, RZ, R7 ; /* 0x000000ffff0b7224 */ /* 0x000fe200078e0007 */ /*01e0*/ MOV R8, 0x200 ; /* 0x0000020000087802 */ /* 0x000fe40000000f00 */ /*01f0*/ CALL.REL.NOINC 0x370 ; /* 0x0000017000007944 */ /* 0x000fea0003c00000 */ /*0200*/ IMAD.MOV.U32 R9, RZ, RZ, R7 ; /* 0x000000ffff097224 */ /* 0x001fe400078e0007 */ /*0210*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0220*/ MUFU.RCP R11, R0 ; /* 0x00000000000b7308 */ /* 0x000e220000001000 */ /*0230*/ FADD R7, R6, R6 ; /* 0x0000000606077221 */ /* 0x000fe20000000000 */ /*0240*/ STG.E [R2.64], R9 ; /* 0x0000000902007986 */ /* 0x0003e2000c101904 */ /*0250*/ BSSY B0, 0x320 ; /* 0x000000c000007945 */ /* 0x000fea0003800000 */ /*0260*/ FCHK P0, R7, R0 ; /* 0x0000000007007302 */ /* 0x000ea20000000000 */ /*0270*/ FFMA R8, -R0, R11, 1 ; /* 0x3f80000000087423 */ /* 0x001fc8000000010b */ /*0280*/ FFMA R8, R11, R8, R11 ; /* 0x000000080b087223 */ /* 0x000fc8000000000b */ /*0290*/ FFMA R6, R8, R7, RZ ; /* 0x0000000708067223 */ /* 0x000fc800000000ff */ /*02a0*/ FFMA R11, -R0, R6, R7 ; /* 0x00000006000b7223 */ /* 0x000fc80000000107 */ /*02b0*/ FFMA R11, R8, R11, R6 ; /* 0x0000000b080b7223 */ /* 0x000fe20000000006 */ /*02c0*/ @!P0 BRA 0x310 ; /* 0x0000004000008947 */ /* 0x004fea0003800000 */ /*02d0*/ IMAD.MOV.U32 R11, RZ, RZ, R7 ; /* 0x000000ffff0b7224 */ /* 0x002fe200078e0007 */ /*02e0*/ MOV R8, 0x300 ; /* 0x0000030000087802 */ /* 0x000fe40000000f00 */ /*02f0*/ CALL.REL.NOINC 0x370 ; /* 0x0000007000007944 */ /* 0x000fea0003c00000 */ /*0300*/ IMAD.MOV.U32 R11, RZ, RZ, R7 ; /* 0x000000ffff0b7224 */ /* 0x001fe400078e0007 */ /*0310*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x002fea0003800000 */ /*0320*/ STG.E [R4.64], R11 ; /* 0x0000000b04007986 */ /* 0x000fe2000c101904 */ /*0330*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0340*/ STG.E [R2.64], RZ ; /* 0x000000ff02007986 */ /* 0x000fe8000c101904 */ /*0350*/ STG.E [R4.64], RZ ; /* 0x000000ff04007986 */ /* 0x000fe2000c101904 */ /*0360*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0370*/ SHF.R.U32.HI R10, RZ, 0x17, R0.reuse ; /* 0x00000017ff0a7819 */ /* 0x100fe20000011600 */ /*0380*/ BSSY B1, 0x9d0 ; /* 0x0000064000017945 */ /* 0x000fe20003800000 */ /*0390*/ SHF.R.U32.HI R7, RZ, 0x17, R11 ; /* 0x00000017ff077819 */ /* 0x000fe2000001160b */ /*03a0*/ IMAD.MOV.U32 R12, RZ, RZ, R0 ; /* 0x000000ffff0c7224 */ /* 0x000fe200078e0000 */ /*03b0*/ LOP3.LUT R10, R10, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff0a0a7812 */ /* 0x000fc400078ec0ff */ /*03c0*/ LOP3.LUT R15, R7, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff070f7812 */ /* 0x000fe200078ec0ff */ /*03d0*/ IMAD.MOV.U32 R7, RZ, RZ, R11 ; /* 0x000000ffff077224 */ /* 0x000fe200078e000b */ /*03e0*/ IADD3 R14, R10, -0x1, RZ ; /* 0xffffffff0a0e7810 */ /* 0x000fe40007ffe0ff */ /*03f0*/ IADD3 R13, R15, -0x1, RZ ; /* 0xffffffff0f0d7810 */ /* 0x000fe40007ffe0ff */ /*0400*/ ISETP.GT.U32.AND P0, PT, R14, 0xfd, PT ; /* 0x000000fd0e00780c */ /* 0x000fc80003f04070 */ /*0410*/ ISETP.GT.U32.OR P0, PT, R13, 0xfd, P0 ; /* 0x000000fd0d00780c */ /* 0x000fda0000704470 */ /*0420*/ @!P0 IMAD.MOV.U32 R9, RZ, RZ, RZ ; /* 0x000000ffff098224 */ /* 0x000fe200078e00ff */ /*0430*/ @!P0 BRA 0x5b0 ; /* 0x0000017000008947 */ /* 0x000fea0003800000 */ /*0440*/ FSETP.GTU.FTZ.AND P0, PT, |R11|, +INF , PT ; /* 0x7f8000000b00780b */ /* 0x000fe40003f1c200 */ /*0450*/ FSETP.GTU.FTZ.AND P1, PT, |R0|, +INF , PT ; /* 0x7f8000000000780b */ /* 0x000fc80003f3c200 */ /*0460*/ PLOP3.LUT P0, PT, P0, P1, PT, 0xa8, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000703570 */ /*0470*/ @P0 BRA 0x9b0 ; /* 0x0000053000000947 */ /* 0x000fea0003800000 */ /*0480*/ LOP3.LUT P0, RZ, R12, 0x7fffffff, R7, 0xc8, !PT ; /* 0x7fffffff0cff7812 */ /* 0x000fda000780c807 */ /*0490*/ @!P0 BRA 0x990 ; /* 0x000004f000008947 */ /* 0x000fea0003800000 */ /*04a0*/ FSETP.NEU.FTZ.AND P2, PT, |R11|.reuse, +INF , PT ; /* 0x7f8000000b00780b */ /* 0x040fe40003f5d200 */ /*04b0*/ FSETP.NEU.FTZ.AND P1, PT, |R0|, +INF , PT ; /* 0x7f8000000000780b */ /* 0x000fe40003f3d200 */ /*04c0*/ FSETP.NEU.FTZ.AND P0, PT, |R11|, +INF , PT ; /* 0x7f8000000b00780b */ /* 0x000fd60003f1d200 */ /*04d0*/ @!P1 BRA !P2, 0x990 ; /* 0x000004b000009947 */ /* 0x000fea0005000000 */ /*04e0*/ LOP3.LUT P2, RZ, R7, 0x7fffffff, RZ, 0xc0, !PT ; /* 0x7fffffff07ff7812 */ /* 0x000fc8000784c0ff */ /*04f0*/ PLOP3.LUT P1, PT, P1, P2, PT, 0x2a, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000f24572 */ /*0500*/ @P1 BRA 0x970 ; /* 0x0000046000001947 */ /* 0x000fea0003800000 */ /*0510*/ LOP3.LUT P1, RZ, R12, 0x7fffffff, RZ, 0xc0, !PT ; /* 0x7fffffff0cff7812 */ /* 0x000fc8000782c0ff */ /*0520*/ PLOP3.LUT P0, PT, P0, P1, PT, 0x2a, 0x0 ; /* 0x000000000000781c */ /* 0x000fda0000702572 */ /*0530*/ @P0 BRA 0x940 ; /* 0x0000040000000947 */ /* 0x000fea0003800000 */ /*0540*/ ISETP.GE.AND P0, PT, R13, RZ, PT ; /* 0x000000ff0d00720c */ /* 0x000fe40003f06270 */ /*0550*/ ISETP.GE.AND P1, PT, R14, RZ, PT ; /* 0x000000ff0e00720c */ /* 0x000fd60003f26270 */ /*0560*/ @P0 IMAD.MOV.U32 R9, RZ, RZ, RZ ; /* 0x000000ffff090224 */ /* 0x000fe400078e00ff */ /*0570*/ @!P0 IMAD.MOV.U32 R9, RZ, RZ, -0x40 ; /* 0xffffffc0ff098424 */ /* 0x000fe400078e00ff */ /*0580*/ @!P0 FFMA R7, R11, 1.84467440737095516160e+19, RZ ; /* 0x5f8000000b078823 */ /* 0x000fe400000000ff */ /*0590*/ @!P1 FFMA R12, R0, 1.84467440737095516160e+19, RZ ; /* 0x5f800000000c9823 */ /* 0x000fe200000000ff */ /*05a0*/ @!P1 IADD3 R9, R9, 0x40, RZ ; /* 0x0000004009099810 */ /* 0x000fe40007ffe0ff */ /*05b0*/ LEA R11, R10, 0xc0800000, 0x17 ; /* 0xc08000000a0b7811 */ /* 0x000fe200078eb8ff */ /*05c0*/ BSSY B2, 0x930 ; /* 0x0000036000027945 */ /* 0x000fe80003800000 */ /*05d0*/ IMAD.IADD R12, R12, 0x1, -R11 ; /* 0x000000010c0c7824 */ /* 0x000fe200078e0a0b */ /*05e0*/ IADD3 R11, R15, -0x7f, RZ ; /* 0xffffff810f0b7810 */ /* 0x000fc60007ffe0ff */ /*05f0*/ MUFU.RCP R13, R12 ; /* 0x0000000c000d7308 */ /* 0x0000620000001000 */ /*0600*/ FADD.FTZ R14, -R12, -RZ ; /* 0x800000ff0c0e7221 */ /* 0x000fe40000010100 */ /*0610*/ IMAD R7, R11.reuse, -0x800000, R7 ; /* 0xff8000000b077824 */ /* 0x040fe200078e0207 */ /*0620*/ IADD3 R12, R11, 0x7f, -R10 ; /* 0x0000007f0b0c7810 */ /* 0x001fca0007ffe80a */ /*0630*/ IMAD.IADD R12, R12, 0x1, R9 ; /* 0x000000010c0c7824 */ /* 0x000fe400078e0209 */ /*0640*/ FFMA R16, R13, R14, 1 ; /* 0x3f8000000d107423 */ /* 0x002fc8000000000e */ /*0650*/ FFMA R18, R13, R16, R13 ; /* 0x000000100d127223 */ /* 0x000fc8000000000d */ /*0660*/ FFMA R13, R7, R18, RZ ; /* 0x00000012070d7223 */ /* 0x000fc800000000ff */ /*0670*/ FFMA R16, R14, R13, R7 ; /* 0x0000000d0e107223 */ /* 0x000fc80000000007 */ /*0680*/ FFMA R13, R18, R16, R13 ; /* 0x00000010120d7223 */ /* 0x000fc8000000000d */ /*0690*/ FFMA R14, R14, R13, R7 ; /* 0x0000000d0e0e7223 */ /* 0x000fc80000000007 */ /*06a0*/ FFMA R7, R18, R14, R13 ; /* 0x0000000e12077223 */ /* 0x000fca000000000d */ /*06b0*/ SHF.R.U32.HI R10, RZ, 0x17, R7 ; /* 0x00000017ff0a7819 */ /* 0x000fc80000011607 */ /*06c0*/ LOP3.LUT R10, R10, 0xff, RZ, 0xc0, !PT ; /* 0x000000ff0a0a7812 */ /* 0x000fca00078ec0ff */ /*06d0*/ IMAD.IADD R15, R10, 0x1, R12 ; /* 0x000000010a0f7824 */ /* 0x000fca00078e020c */ /*06e0*/ IADD3 R9, R15, -0x1, RZ ; /* 0xffffffff0f097810 */ /* 0x000fc80007ffe0ff */ /*06f0*/ ISETP.GE.U32.AND P0, PT, R9, 0xfe, PT ; /* 0x000000fe0900780c */ /* 0x000fda0003f06070 */ /*0700*/ @!P0 BRA 0x910 ; /* 0x0000020000008947 */ /* 0x000fea0003800000 */ /*0710*/ ISETP.GT.AND P0, PT, R15, 0xfe, PT ; /* 0x000000fe0f00780c */ /* 0x000fda0003f04270 */ /*0720*/ @P0 BRA 0x8e0 ; /* 0x000001b000000947 */ /* 0x000fea0003800000 */ /*0730*/ ISETP.GE.AND P0, PT, R15, 0x1, PT ; /* 0x000000010f00780c */ /* 0x000fda0003f06270 */ /*0740*/ @P0 BRA 0x920 ; /* 0x000001d000000947 */ /* 0x000fea0003800000 */ /*0750*/ ISETP.GE.AND P0, PT, R15, -0x18, PT ; /* 0xffffffe80f00780c */ /* 0x000fe40003f06270 */ /*0760*/ LOP3.LUT R7, R7, 0x80000000, RZ, 0xc0, !PT ; /* 0x8000000007077812 */ /* 0x000fd600078ec0ff */ /*0770*/ @!P0 BRA 0x920 ; /* 0x000001a000008947 */ /* 0x000fea0003800000 */ /*0780*/ FFMA.RZ R9, R18.reuse, R14.reuse, R13.reuse ; /* 0x0000000e12097223 */ /* 0x1c0fe2000000c00d */ /*0790*/ IADD3 R12, R15.reuse, 0x20, RZ ; /* 0x000000200f0c7810 */ /* 0x040fe20007ffe0ff */ /*07a0*/ FFMA.RM R10, R18.reuse, R14.reuse, R13.reuse ; /* 0x0000000e120a7223 */ /* 0x1c0fe2000000400d */ /*07b0*/ ISETP.NE.AND P2, PT, R15, RZ, PT ; /* 0x000000ff0f00720c */ /* 0x000fe40003f45270 */ /*07c0*/ LOP3.LUT R11, R9, 0x7fffff, RZ, 0xc0, !PT ; /* 0x007fffff090b7812 */ /* 0x000fe200078ec0ff */ /*07d0*/ FFMA.RP R9, R18, R14, R13 ; /* 0x0000000e12097223 */ /* 0x000fe2000000800d */ /*07e0*/ ISETP.NE.AND P1, PT, R15, RZ, PT ; /* 0x000000ff0f00720c */ /* 0x000fe20003f25270 */ /*07f0*/ IMAD.MOV R13, RZ, RZ, -R15 ; /* 0x000000ffff0d7224 */ /* 0x000fe200078e0a0f */ /*0800*/ LOP3.LUT R11, R11, 0x800000, RZ, 0xfc, !PT ; /* 0x008000000b0b7812 */ /* 0x000fe400078efcff */ /*0810*/ FSETP.NEU.FTZ.AND P0, PT, R9, R10, PT ; /* 0x0000000a0900720b */ /* 0x000fc40003f1d000 */ /*0820*/ SHF.L.U32 R12, R11, R12, RZ ; /* 0x0000000c0b0c7219 */ /* 0x000fe400000006ff */ /*0830*/ SEL R10, R13, RZ, P2 ; /* 0x000000ff0d0a7207 */ /* 0x000fe40001000000 */ /*0840*/ ISETP.NE.AND P1, PT, R12, RZ, P1 ; /* 0x000000ff0c00720c */ /* 0x000fe40000f25270 */ /*0850*/ SHF.R.U32.HI R10, RZ, R10, R11 ; /* 0x0000000aff0a7219 */ /* 0x000fe4000001160b */ /*0860*/ PLOP3.LUT P0, PT, P0, P1, PT, 0xa8, 0x0 ; /* 0x000000000000781c */ /* 0x000fe40000703570 */ /*0870*/ SHF.R.U32.HI R12, RZ, 0x1, R10 ; /* 0x00000001ff0c7819 */ /* 0x000fc4000001160a */ /*0880*/ SEL R9, RZ, 0x1, !P0 ; /* 0x00000001ff097807 */ /* 0x000fc80004000000 */ /*0890*/ LOP3.LUT R9, R9, 0x1, R12, 0xf8, !PT ; /* 0x0000000109097812 */ /* 0x000fc800078ef80c */ /*08a0*/ LOP3.LUT R9, R9, R10, RZ, 0xc0, !PT ; /* 0x0000000a09097212 */ /* 0x000fca00078ec0ff */ /*08b0*/ IMAD.IADD R12, R12, 0x1, R9 ; /* 0x000000010c0c7824 */ /* 0x000fca00078e0209 */ /*08c0*/ LOP3.LUT R7, R12, R7, RZ, 0xfc, !PT ; /* 0x000000070c077212 */ /* 0x000fe200078efcff */ /*08d0*/ BRA 0x920 ; /* 0x0000004000007947 */ /* 0x000fea0003800000 */ /*08e0*/ LOP3.LUT R7, R7, 0x80000000, RZ, 0xc0, !PT ; /* 0x8000000007077812 */ /* 0x000fc800078ec0ff */ /*08f0*/ LOP3.LUT R7, R7, 0x7f800000, RZ, 0xfc, !PT ; /* 0x7f80000007077812 */ /* 0x000fe200078efcff */ /*0900*/ BRA 0x920 ; /* 0x0000001000007947 */ /* 0x000fea0003800000 */ /*0910*/ IMAD R7, R12, 0x800000, R7 ; /* 0x008000000c077824 */ /* 0x000fe400078e0207 */ /*0920*/ BSYNC B2 ; /* 0x0000000000027941 */ /* 0x000fea0003800000 */ /*0930*/ BRA 0x9c0 ; /* 0x0000008000007947 */ /* 0x000fea0003800000 */ /*0940*/ LOP3.LUT R7, R12, 0x80000000, R7, 0x48, !PT ; /* 0x800000000c077812 */ /* 0x000fc800078e4807 */ /*0950*/ LOP3.LUT R7, R7, 0x7f800000, RZ, 0xfc, !PT ; /* 0x7f80000007077812 */ /* 0x000fe200078efcff */ /*0960*/ BRA 0x9c0 ; /* 0x0000005000007947 */ /* 0x000fea0003800000 */ /*0970*/ LOP3.LUT R7, R12, 0x80000000, R7, 0x48, !PT ; /* 0x800000000c077812 */ /* 0x000fe200078e4807 */ /*0980*/ BRA 0x9c0 ; /* 0x0000003000007947 */ /* 0x000fea0003800000 */ /*0990*/ MUFU.RSQ R7, -QNAN ; /* 0xffc0000000077908 */ /* 0x000e220000001400 */ /*09a0*/ BRA 0x9c0 ; /* 0x0000001000007947 */ /* 0x000fea0003800000 */ /*09b0*/ FADD.FTZ R7, R11, R0 ; /* 0x000000000b077221 */ /* 0x000fe40000010000 */ /*09c0*/ BSYNC B1 ; /* 0x0000000000017941 */ /* 0x000fea0003800000 */ /*09d0*/ IMAD.MOV.U32 R9, RZ, RZ, 0x0 ; /* 0x00000000ff097424 */ /* 0x000fc800078e00ff */ /*09e0*/ RET.REL.NODEC R8 0x0 ; /* 0xfffff61008007950 */ /* 0x000fea0003c3ffff */ /*09f0*/ BRA 0x9f0; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0a00*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a10*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a20*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a30*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a40*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a50*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a60*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0a70*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z15initializeSobelPfS_ .globl _Z15initializeSobelPfS_ .p2align 8 .type _Z15initializeSobelPfS_,@function _Z15initializeSobelPfS_: v_and_b32_e32 v3, 0x3ff, v0 v_bfe_u32 v0, v0, 10, 10 s_mov_b32 s2, exec_lo s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_max_u32_e32 v1, v3, v0 v_cmpx_gt_u32_e32 5, v1 s_cbranch_execz .LBB0_6 v_add_nc_u32_e32 v1, -2, v0 s_load_b64 s[4:5], s[0:1], 0x0 v_add_nc_u32_e32 v2, -2, v3 v_mad_u32_u24 v0, v0, 5, v3 s_mov_b32 s2, exec_lo v_cvt_f32_i32_e32 v1, v1 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_cvt_f32_i32_e32 v4, v2 v_mul_f32_e32 v2, v1, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmac_f32_e32 v2, v4, v4 v_cmpx_neq_f32_e32 0, v2 s_xor_b32 s3, exec_lo, s2 s_cbranch_execz .LBB0_3 v_dual_add_f32 v3, v4, v4 :: v_dual_add_f32 v4, v1, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_div_scale_f32 v1, null, v2, v2, v3 v_div_scale_f32 v5, null, v2, v2, v4 v_div_scale_f32 v10, vcc_lo, v3, v2, v3 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_rcp_f32_e32 v6, v1 v_rcp_f32_e32 v7, v5 s_waitcnt_depctr 0xfff v_fma_f32 v8, -v1, v6, 1.0 v_fma_f32 v9, -v5, v7, 1.0 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_dual_fmac_f32 v6, v8, v6 :: v_dual_fmac_f32 v7, v9, v7 v_div_scale_f32 v8, s2, v4, v2, v4 v_mul_f32_e32 v9, v10, v6 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_mul_f32_e32 v11, v8, v7 v_fma_f32 v12, -v1, v9, v10 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f32 v13, -v5, v11, v8 v_fmac_f32_e32 v9, v12, v6 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fmac_f32_e32 v11, v13, v7 v_fma_f32 v1, -v1, v9, v10 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f32 v5, -v5, v11, v8 v_div_fmas_f32 v1, v1, v6, v9 s_mov_b32 vcc_lo, s2 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_3) v_div_fmas_f32 v5, v5, v7, v11 v_lshlrev_b32_e32 v7, 2, v0 v_div_fixup_f32 v6, v1, v2, v3 v_mov_b32_e32 v1, 0 s_delay_alu instid0(VALU_DEP_4) v_div_fixup_f32 v3, v5, v2, v4 s_waitcnt lgkmcnt(0) global_store_b32 v7, v6, s[4:5] .LBB0_3: s_and_not1_saveexec_b32 s2, s3 s_cbranch_execz .LBB0_5 v_dual_mov_b32 v1, 0 :: v_dual_lshlrev_b32 v2, 2, v0 s_delay_alu instid0(VALU_DEP_1) v_mov_b32_e32 v3, v1 s_waitcnt lgkmcnt(0) global_store_b32 v2, v1, s[4:5] .LBB0_5: s_or_b32 exec_lo, exec_lo, s2 s_load_b64 s[0:1], s[0:1], 0x8 v_lshlrev_b64 v[0:1], 2, v[0:1] 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_6: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z15initializeSobelPfS_ .amdhsa_group_segment_fixed_size 0 .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 1 .amdhsa_next_free_vgpr 14 .amdhsa_next_free_sgpr 6 .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 _Z15initializeSobelPfS_, .Lfunc_end0-_Z15initializeSobelPfS_ .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 .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 16 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z15initializeSobelPfS_ .private_segment_fixed_size: 0 .sgpr_count: 8 .sgpr_spill_count: 0 .symbol: _Z15initializeSobelPfS_.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 from CUDA host assembly to AMD host assembly.
.file "tmpxft_0000b160_00000000-6_initializeSobel.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 _Z37__device_stub__Z15initializeSobelPfS_PfS_ .type _Z37__device_stub__Z15initializeSobelPfS_PfS_, @function _Z37__device_stub__Z15initializeSobelPfS_PfS_: .LFB2051: .cfi_startproc endbr64 subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 8(%rsp) movq %rsi, (%rsp) movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax leaq 8(%rsp), %rax movq %rax, 80(%rsp) movq %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 _Z15initializeSobelPfS_(%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 _Z37__device_stub__Z15initializeSobelPfS_PfS_, .-_Z37__device_stub__Z15initializeSobelPfS_PfS_ .globl _Z15initializeSobelPfS_ .type _Z15initializeSobelPfS_, @function _Z15initializeSobelPfS_: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z37__device_stub__Z15initializeSobelPfS_PfS_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z15initializeSobelPfS_, .-_Z15initializeSobelPfS_ .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "_Z15initializeSobelPfS_" .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 _Z15initializeSobelPfS_(%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 "initializeSobel.hip" .globl _Z30__device_stub__initializeSobelPfS_ # -- Begin function _Z30__device_stub__initializeSobelPfS_ .p2align 4, 0x90 .type _Z30__device_stub__initializeSobelPfS_,@function _Z30__device_stub__initializeSobelPfS_: # @_Z30__device_stub__initializeSobelPfS_ .cfi_startproc # %bb.0: subq $88, %rsp .cfi_def_cfa_offset 96 movq %rdi, 56(%rsp) movq %rsi, 48(%rsp) leaq 56(%rsp), %rax movq %rax, 64(%rsp) leaq 48(%rsp), %rax movq %rax, 72(%rsp) 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 64(%rsp), %r9 movl $_Z15initializeSobelPfS_, %edi pushq (%rsp) .cfi_adjust_cfa_offset 8 pushq 16(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $104, %rsp .cfi_adjust_cfa_offset -104 retq .Lfunc_end0: .size _Z30__device_stub__initializeSobelPfS_, .Lfunc_end0-_Z30__device_stub__initializeSobelPfS_ .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 $_Z15initializeSobelPfS_, %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 _Z15initializeSobelPfS_,@object # @_Z15initializeSobelPfS_ .section .rodata,"a",@progbits .globl _Z15initializeSobelPfS_ .p2align 3, 0x0 _Z15initializeSobelPfS_: .quad _Z30__device_stub__initializeSobelPfS_ .size _Z15initializeSobelPfS_, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z15initializeSobelPfS_" .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 _Z30__device_stub__initializeSobelPfS_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z15initializeSobelPfS_ .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 <cuda.h> #include <stdio.h> #include <math.h> #define NUM_BANKS 8 #define LOG_NUM_BANKS 3 #define CONFLICT_FREE_OFFSET(n) \ ((n) >> NUM_BANKS + (n) >> (2 * LOG_NUM_BANKS)) // Print device properties void printDevProp(cudaDeviceProp devProp) { printf("Major revision number: %d\n", devProp.major); printf("Minor revision number: %d\n", devProp.minor); printf("Name: %s\n", devProp.name); printf("Total global memory: %u\n", devProp.totalGlobalMem); printf("Total shared memory per block: %u\n", devProp.sharedMemPerBlock); printf("Total registers per block: %d\n", devProp.regsPerBlock); printf("Warp size: %d\n", devProp.warpSize); printf("Maximum memory pitch: %u\n", devProp.memPitch); printf("Maximum threads per block: %d\n", devProp.maxThreadsPerBlock); for (int i = 0; i < 3; ++i) printf("Maximum dimension %d of block: %d\n", i, devProp.maxThreadsDim[i]); for (int i = 0; i < 3; ++i) printf("Maximum dimension %d of grid: %d\n", i, devProp.maxGridSize[i]); printf("Clock rate: %d\n", devProp.clockRate); printf("Total constant memory: %u\n", devProp.totalConstMem); printf("Texture alignment: %u\n", devProp.textureAlignment); printf("Concurrent copy and execution: %s\n", (devProp.deviceOverlap ? "Yes" : "No")); printf("Number of multiprocessors: %d\n", devProp.multiProcessorCount); printf("Kernel execution timeout: %s\n", (devProp.kernelExecTimeoutEnabled ? "Yes" : "No")); return; } void proprietati() { int devCount; cudaGetDeviceCount(&devCount); printf("Avem %d device-uri.\n", devCount); for (int i = 0; i < devCount; ++i) { // Get device properties printf("\nCUDA Device #%d\n", i); cudaDeviceProp devProp; cudaGetDeviceProperties(&devProp, i); printDevProp(devProp); } } __global__ void vecDiv(float *a, float *b, float *c, unsigned int n) { // Get our global thread ID int id = blockIdx.x*blockDim.x+threadIdx.x; // Make sure we do not go out of bounds if (id < n) c[id] = a[id] / b[id]; } __global__ void vecProd(float *a, float *b, float *c, unsigned int n) { // Get our global thread ID int id = blockIdx.x*blockDim.x+threadIdx.x; // Make sure we do not go out of bounds if (id < n) c[id] = a[id] * b[id]; } __global__ void vecAdd(float *a, float *b, float *c, unsigned int n) { // Get our global thread ID int id = blockIdx.x*blockDim.x+threadIdx.x; // Make sure we do not go out of bounds if (id < n) c[id] = a[id] + b[id]; } __global__ void vecSub(float *a, float *b, float *c, unsigned int n) { // Get our global thread ID int id = blockIdx.x*blockDim.x+threadIdx.x; // Make sure we do not go out of bounds if (id < n) c[id] = a[id] - b[id]; } __global__ void eulVector(float *a, float *b,float *c, unsigned int n) { int id = blockIdx.x*blockDim.x+threadIdx.x; if(id < n){ c[id] = a[id] - b[id]; c[id]*= c[id]; } } __global__ void smallSum(unsigned char *out, unsigned char *in, int n) { int id = blockIdx.x*blockDim.x+threadIdx.x; if(threadIdx.x == 1023) { out[id/1024] = id; } } // reduction sum __global__ void reductionSum(float *in_array, float *out_array, unsigned long size) { extern __shared__ float shared[]; unsigned int tid = threadIdx.x; unsigned int i = blockIdx.x*blockDim.x + threadIdx.x; shared[tid] = in_array[i]; __syncthreads(); for(unsigned int s=blockDim.x/2; s>0; s>>=1) { if (tid < s) { shared[tid] += shared[tid + s]; } __syncthreads(); } if(tid==0) //first thread, last sum { out_array[blockIdx.x]=shared[0]; } } __global__ void cumSum(float *in_array, float *out_array, unsigned long size) { extern __shared__ float shared[]; unsigned int tid = threadIdx.x; unsigned int b_offset = blockIdx.x * blockDim.x; unsigned int offset = 1; int i = tid; int j = tid + blockDim.x / 2; int offset_i = CONFLICT_FREE_OFFSET(i); int offset_j = CONFLICT_FREE_OFFSET(j); shared[i + offset_i] = in_array[i + b_offset]; shared[j + offset_j] = in_array[j + b_offset]; // scan up for (int s = (blockDim.x >> 1); s > 0; s >>= 1) { __syncthreads(); if (tid < s) { int i = offset * (2 * tid + 1) - 1; int j = offset * (2 * tid + 2) - 1; i += CONFLICT_FREE_OFFSET(i); j += CONFLICT_FREE_OFFSET(j); shared[j] += shared[i]; } offset <<= 1; } if (tid == 0) { shared[blockDim.x - 1 + CONFLICT_FREE_OFFSET(blockDim.x - 1)] = 0; } // scan down for (int s = 1; s < blockDim.x; s <<= 1) { offset >>= 1; __syncthreads(); if (tid < s) { int i = offset * (2 * tid + 1) - 1; int j = offset * (2 * tid + 2) - 1; i += CONFLICT_FREE_OFFSET(i); j += CONFLICT_FREE_OFFSET(j); float tmp = shared[i]; shared[i] = shared[j]; shared[j] += tmp; } } __syncthreads(); out_array[i] = shared[i + offset_i]; out_array[j] = shared[j + offset_j]; } void test1(float *a, float *b, unsigned long n) { float *d_a, *d_b, *d_rez; float *rez = (float*) malloc(n*sizeof(float)); cudaMalloc((void **)&d_a,n*sizeof(float)); cudaMalloc((void **)&d_b,n*sizeof(float)); cudaMalloc((void **)&d_rez,n*sizeof(float)); cudaMemcpy(d_a,a,n*sizeof(float),cudaMemcpyHostToDevice); cudaMemcpy(d_b,b,n*sizeof(float),cudaMemcpyHostToDevice); vecAdd<<<161,64>>>(d_a,d_b,d_rez,n); cudaMemcpy(rez,d_rez,10304*sizeof(float),cudaMemcpyDeviceToHost); printf("\n REZULTAT TEST: %f \n",rez[0]); cudaFree(d_a); cudaFree(d_b); cudaFree(d_rez); free(rez); } // a imagini // b imagine void euclideanNormAsync(float *a, float *b, unsigned long n, float *result, unsigned long nrPoze) { float *d_a,*d_b,*d_rez,*d_temp; float *sum; cudaStream_t stream; cudaStreamCreate(&stream); cudaHostRegister((void **)&a, n*nrPoze*sizeof(float), cudaHostRegisterPortable); cudaHostAlloc((void **)&sum, nrPoze*sizeof(float), cudaHostAllocDefault); cudaMalloc((void **)&d_a,n*sizeof(float)); cudaMalloc((void **)&d_b,n*sizeof(float)); cudaMalloc((void **)&d_temp,n*sizeof(float)); cudaMalloc((void **)&d_rez,(n/64+(256-n/64))*sizeof(float)); //cudaMemcpy(d_a,a,n*sizeof(float),cudaMemcpyHostToDevice); cudaMemcpy(d_b,b,n*sizeof(float),cudaMemcpyHostToDevice); for(unsigned long int i=0, k=0; i<n, k< nrPoze; i+= n, k++) { cudaMemcpyAsync(d_a,a+i,n*sizeof(float),cudaMemcpyHostToDevice, stream); eulVector<<<n/64, 64, n*sizeof(float), stream>>>(d_a,d_b,d_temp, n); reductionSum<<<n/64, 64, n*sizeof(float), stream>>>(d_temp,d_rez, n); cudaMemsetAsync(d_rez+n/64,0,95*sizeof(float),stream); reductionSum<<<1,256, n*sizeof(float), stream>>>(d_rez,d_temp,n); //cumSum<<<n/64, n/64, n*sizeof(float)>>>(d_rez,d_temp, n/64); cudaMemcpyAsync(sum+k,d_temp,sizeof(float),cudaMemcpyDeviceToHost, stream); } cudaStreamSynchronize(stream); for(int i=0;i<nrPoze;i++) { result[i] = sum[i]; } cudaFree(d_a); cudaFree(d_b); cudaFree(d_temp); cudaFree(d_rez); cudaFreeHost(sum); cudaHostUnregister(a); cudaStreamDestroy(stream); } // a imagini // b imagine void cityblockNormAsync(float *a, float *b, unsigned long n, float *result, unsigned long nrPoze) { float *d_a,*d_b,*d_rez,*d_temp; float *sum; cudaStream_t stream; cudaStreamCreate(&stream); cudaHostRegister((void **)&a, n*nrPoze*sizeof(float), cudaHostRegisterPortable); cudaHostAlloc((void **)&sum, nrPoze*sizeof(float), cudaHostAllocDefault); cudaMalloc((void **)&d_a,n*sizeof(float)); cudaMalloc((void **)&d_b,n*sizeof(float)); cudaMalloc((void **)&d_temp,n*sizeof(float)); cudaMalloc((void **)&d_rez,(n/64+(256-n/64))*sizeof(float)); //cudaMemcpy(d_a,a,n*sizeof(float),cudaMemcpyHostToDevice); cudaMemcpy(d_b,b,n*sizeof(float),cudaMemcpyHostToDevice); for(unsigned long int i=0, k=0; i<n, k< nrPoze; i+= n, k++) { cudaMemcpyAsync(d_a,a+i,n*sizeof(float),cudaMemcpyHostToDevice, stream); vecSub<<<n/64, 64, n*sizeof(float), stream>>>(d_a,d_b,d_temp, n); reductionSum<<<n/64, 64, n*sizeof(float), stream>>>(d_temp,d_rez, n); cudaMemsetAsync(d_rez+n/64,0,95*sizeof(float),stream); reductionSum<<<1,256, n*sizeof(float), stream>>>(d_rez,d_temp,n); //cumSum<<<n/64, n/64, n*sizeof(float)>>>(d_rez,d_temp, n/64); cudaMemcpyAsync(sum+k,d_temp,sizeof(float),cudaMemcpyDeviceToHost, stream); } cudaStreamSynchronize(stream); for(int i=0;i<nrPoze;i++) { result[i] = sum[i]; } //printf("%f ", min); cudaFree(d_a); cudaFree(d_b); cudaFree(d_temp); cudaFree(d_rez); cudaFreeHost(sum); cudaHostUnregister(a); cudaStreamDestroy(stream); } // a imagini // b imagine void cosNormAsync(float *a, float *b, unsigned long n, float *result, unsigned long nrPoze) { float *d_a,*d_b,*d_rez,*d_temp; float *d_a1,*d_b1,*d_rez1, *d_temp1; float *d_a2,*d_b2,*d_rez2, *d_temp2; float *dotProd, *normA, *normB; cudaStream_t stream, stream1, stream2; cudaStreamCreate(&stream); cudaStreamCreate(&stream1); cudaStreamCreate(&stream2); cudaHostRegister((void **)&a, n*nrPoze*sizeof(float), cudaHostRegisterPortable); cudaHostAlloc((void **)&dotProd, nrPoze*sizeof(float), cudaHostAllocDefault); cudaHostAlloc((void **)&normA, nrPoze*sizeof(float), cudaHostAllocDefault); cudaHostAlloc((void **)&normB, nrPoze*sizeof(float), cudaHostAllocDefault); cudaMalloc((void **)&d_a,n*sizeof(float)); cudaMalloc((void **)&d_b,n*sizeof(float)); cudaMalloc((void **)&d_temp,n*sizeof(float)); cudaMalloc((void **)&d_rez,(n/64+(256-n/64))*sizeof(float)); cudaMalloc((void **)&d_a1,n*sizeof(float)); cudaMalloc((void **)&d_b1,n*sizeof(float)); cudaMalloc((void **)&d_temp1,n*sizeof(float)); cudaMalloc((void **)&d_rez1,(n/64+(256-n/64))*sizeof(float)); cudaMalloc((void **)&d_a2,n*sizeof(float)); cudaMalloc((void **)&d_b2,n*sizeof(float)); cudaMalloc((void **)&d_temp2,n*sizeof(float)); cudaMalloc((void **)&d_rez2,(n/64+(256-n/64))*sizeof(float)); //cudaMemcpy(d_a,a,n*sizeof(float),cudaMemcpyHostToDevice); cudaMemcpy(d_b,b,n*sizeof(float),cudaMemcpyHostToDevice); cudaMemcpy(d_b1,b,n*sizeof(float),cudaMemcpyHostToDevice); cudaMemcpy(d_b2,b,n*sizeof(float),cudaMemcpyHostToDevice); for(unsigned long int i=0, k=0; i<n, k< nrPoze; i+= n, k++) { cudaMemcpyAsync(d_a,a+i,n*sizeof(float),cudaMemcpyHostToDevice, stream); cudaMemcpyAsync(d_a1,a+i,n*sizeof(float),cudaMemcpyHostToDevice, stream1); cudaMemcpyAsync(d_a2,a+i,n*sizeof(float),cudaMemcpyHostToDevice, stream1); vecProd<<<n/64, 64, n*sizeof(float), stream>>>(d_a,d_b,d_temp, n); vecProd<<<n/64, 64, n*sizeof(float), stream1>>>(d_a1,d_a2,d_temp1, n); if(i==0) vecProd<<<n/64, 64, n*sizeof(float), stream2>>>(d_b1,d_b2,d_temp2, n); reductionSum<<<n/64, 64, n*sizeof(float), stream>>>(d_temp,d_rez, n); reductionSum<<<n/64, 64, n*sizeof(float), stream1>>>(d_temp1,d_rez1, n); if(i==0) reductionSum<<<n/64, 64, n*sizeof(float), stream2>>>(d_temp2,d_rez2, n); cudaMemsetAsync(d_rez+n/64,0,95*sizeof(float),stream); cudaMemsetAsync(d_rez1+n/64,0,95*sizeof(float),stream1); if(i==0) cudaMemsetAsync(d_rez2+n/64,0,95*sizeof(float),stream2); reductionSum<<<1,256, n*sizeof(float), stream>>>(d_rez,d_temp,n); reductionSum<<<1,256, n*sizeof(float), stream1>>>(d_rez1,d_temp1,n); if(i==0) reductionSum<<<1,256, n*sizeof(float), stream2>>>(d_rez2,d_temp2,n); // we now have in d_temp - dot product, d_temp1 - norm of a, d_temp2 norm of b cudaMemcpyAsync(dotProd+k,d_temp,sizeof(float),cudaMemcpyDeviceToHost, stream); cudaMemcpyAsync(normA+k,d_temp1,sizeof(float),cudaMemcpyDeviceToHost, stream1); cudaMemcpyAsync(normB+k,d_temp2,sizeof(float),cudaMemcpyDeviceToHost, stream2); } cudaStreamSynchronize(stream); cudaStreamSynchronize(stream1); cudaStreamSynchronize(stream2); for(int i=0;i<nrPoze;i++) { float calc = 1 - dotProd[i]/(sqrtf(normA[i])*sqrtf(normB[i])); result[i] = isnan(calc)?1:calc; } //printf("%f ",result[0]); cudaFree(d_a); cudaFree(d_b); cudaFree(d_temp); cudaFree(d_rez); cudaFree(d_a1); cudaFree(d_b2); cudaFree(d_temp1); cudaFree(d_rez1); cudaFree(d_a2); cudaFree(d_b2); cudaFree(d_temp2); cudaFree(d_rez2); cudaFreeHost(dotProd); cudaFreeHost(normA); cudaFreeHost(normB); cudaHostUnregister(a); cudaStreamDestroy(stream); cudaStreamDestroy(stream1); cudaStreamDestroy(stream2); }
#include <hip/hip_runtime.h> #include <stdio.h> #include <math.h> #define NUM_BANKS 8 #define LOG_NUM_BANKS 3 #define CONFLICT_FREE_OFFSET(n) \ ((n) >> NUM_BANKS + (n) >> (2 * LOG_NUM_BANKS)) // Print device properties void printDevProp(hipDeviceProp_t devProp) { printf("Major revision number: %d\n", devProp.major); printf("Minor revision number: %d\n", devProp.minor); printf("Name: %s\n", devProp.name); printf("Total global memory: %u\n", devProp.totalGlobalMem); printf("Total shared memory per block: %u\n", devProp.sharedMemPerBlock); printf("Total registers per block: %d\n", devProp.regsPerBlock); printf("Warp size: %d\n", devProp.warpSize); printf("Maximum memory pitch: %u\n", devProp.memPitch); printf("Maximum threads per block: %d\n", devProp.maxThreadsPerBlock); for (int i = 0; i < 3; ++i) printf("Maximum dimension %d of block: %d\n", i, devProp.maxThreadsDim[i]); for (int i = 0; i < 3; ++i) printf("Maximum dimension %d of grid: %d\n", i, devProp.maxGridSize[i]); printf("Clock rate: %d\n", devProp.clockRate); printf("Total constant memory: %u\n", devProp.totalConstMem); printf("Texture alignment: %u\n", devProp.textureAlignment); printf("Concurrent copy and execution: %s\n", (devProp.deviceOverlap ? "Yes" : "No")); printf("Number of multiprocessors: %d\n", devProp.multiProcessorCount); printf("Kernel execution timeout: %s\n", (devProp.kernelExecTimeoutEnabled ? "Yes" : "No")); return; } void proprietati() { int devCount; hipGetDeviceCount(&devCount); printf("Avem %d device-uri.\n", devCount); for (int i = 0; i < devCount; ++i) { // Get device properties printf("\nCUDA Device #%d\n", i); hipDeviceProp_t devProp; hipGetDeviceProperties(&devProp, i); printDevProp(devProp); } } __global__ void vecDiv(float *a, float *b, float *c, unsigned int n) { // Get our global thread ID int id = blockIdx.x*blockDim.x+threadIdx.x; // Make sure we do not go out of bounds if (id < n) c[id] = a[id] / b[id]; } __global__ void vecProd(float *a, float *b, float *c, unsigned int n) { // Get our global thread ID int id = blockIdx.x*blockDim.x+threadIdx.x; // Make sure we do not go out of bounds if (id < n) c[id] = a[id] * b[id]; } __global__ void vecAdd(float *a, float *b, float *c, unsigned int n) { // Get our global thread ID int id = blockIdx.x*blockDim.x+threadIdx.x; // Make sure we do not go out of bounds if (id < n) c[id] = a[id] + b[id]; } __global__ void vecSub(float *a, float *b, float *c, unsigned int n) { // Get our global thread ID int id = blockIdx.x*blockDim.x+threadIdx.x; // Make sure we do not go out of bounds if (id < n) c[id] = a[id] - b[id]; } __global__ void eulVector(float *a, float *b,float *c, unsigned int n) { int id = blockIdx.x*blockDim.x+threadIdx.x; if(id < n){ c[id] = a[id] - b[id]; c[id]*= c[id]; } } __global__ void smallSum(unsigned char *out, unsigned char *in, int n) { int id = blockIdx.x*blockDim.x+threadIdx.x; if(threadIdx.x == 1023) { out[id/1024] = id; } } // reduction sum __global__ void reductionSum(float *in_array, float *out_array, unsigned long size) { extern __shared__ float shared[]; unsigned int tid = threadIdx.x; unsigned int i = blockIdx.x*blockDim.x + threadIdx.x; shared[tid] = in_array[i]; __syncthreads(); for(unsigned int s=blockDim.x/2; s>0; s>>=1) { if (tid < s) { shared[tid] += shared[tid + s]; } __syncthreads(); } if(tid==0) //first thread, last sum { out_array[blockIdx.x]=shared[0]; } } __global__ void cumSum(float *in_array, float *out_array, unsigned long size) { extern __shared__ float shared[]; unsigned int tid = threadIdx.x; unsigned int b_offset = blockIdx.x * blockDim.x; unsigned int offset = 1; int i = tid; int j = tid + blockDim.x / 2; int offset_i = CONFLICT_FREE_OFFSET(i); int offset_j = CONFLICT_FREE_OFFSET(j); shared[i + offset_i] = in_array[i + b_offset]; shared[j + offset_j] = in_array[j + b_offset]; // scan up for (int s = (blockDim.x >> 1); s > 0; s >>= 1) { __syncthreads(); if (tid < s) { int i = offset * (2 * tid + 1) - 1; int j = offset * (2 * tid + 2) - 1; i += CONFLICT_FREE_OFFSET(i); j += CONFLICT_FREE_OFFSET(j); shared[j] += shared[i]; } offset <<= 1; } if (tid == 0) { shared[blockDim.x - 1 + CONFLICT_FREE_OFFSET(blockDim.x - 1)] = 0; } // scan down for (int s = 1; s < blockDim.x; s <<= 1) { offset >>= 1; __syncthreads(); if (tid < s) { int i = offset * (2 * tid + 1) - 1; int j = offset * (2 * tid + 2) - 1; i += CONFLICT_FREE_OFFSET(i); j += CONFLICT_FREE_OFFSET(j); float tmp = shared[i]; shared[i] = shared[j]; shared[j] += tmp; } } __syncthreads(); out_array[i] = shared[i + offset_i]; out_array[j] = shared[j + offset_j]; } void test1(float *a, float *b, unsigned long n) { float *d_a, *d_b, *d_rez; float *rez = (float*) malloc(n*sizeof(float)); hipMalloc((void **)&d_a,n*sizeof(float)); hipMalloc((void **)&d_b,n*sizeof(float)); hipMalloc((void **)&d_rez,n*sizeof(float)); hipMemcpy(d_a,a,n*sizeof(float),hipMemcpyHostToDevice); hipMemcpy(d_b,b,n*sizeof(float),hipMemcpyHostToDevice); vecAdd<<<161,64>>>(d_a,d_b,d_rez,n); hipMemcpy(rez,d_rez,10304*sizeof(float),hipMemcpyDeviceToHost); printf("\n REZULTAT TEST: %f \n",rez[0]); hipFree(d_a); hipFree(d_b); hipFree(d_rez); free(rez); } // a imagini // b imagine void euclideanNormAsync(float *a, float *b, unsigned long n, float *result, unsigned long nrPoze) { float *d_a,*d_b,*d_rez,*d_temp; float *sum; hipStream_t stream; hipStreamCreate(&stream); hipHostRegister((void **)&a, n*nrPoze*sizeof(float), hipHostRegisterPortable); hipHostAlloc((void **)&sum, nrPoze*sizeof(float), hipHostMallocDefault); hipMalloc((void **)&d_a,n*sizeof(float)); hipMalloc((void **)&d_b,n*sizeof(float)); hipMalloc((void **)&d_temp,n*sizeof(float)); hipMalloc((void **)&d_rez,(n/64+(256-n/64))*sizeof(float)); //cudaMemcpy(d_a,a,n*sizeof(float),cudaMemcpyHostToDevice); hipMemcpy(d_b,b,n*sizeof(float),hipMemcpyHostToDevice); for(unsigned long int i=0, k=0; i<n, k< nrPoze; i+= n, k++) { hipMemcpyAsync(d_a,a+i,n*sizeof(float),hipMemcpyHostToDevice, stream); eulVector<<<n/64, 64, n*sizeof(float), stream>>>(d_a,d_b,d_temp, n); reductionSum<<<n/64, 64, n*sizeof(float), stream>>>(d_temp,d_rez, n); hipMemsetAsync(d_rez+n/64,0,95*sizeof(float),stream); reductionSum<<<1,256, n*sizeof(float), stream>>>(d_rez,d_temp,n); //cumSum<<<n/64, n/64, n*sizeof(float)>>>(d_rez,d_temp, n/64); hipMemcpyAsync(sum+k,d_temp,sizeof(float),hipMemcpyDeviceToHost, stream); } hipStreamSynchronize(stream); for(int i=0;i<nrPoze;i++) { result[i] = sum[i]; } hipFree(d_a); hipFree(d_b); hipFree(d_temp); hipFree(d_rez); hipHostFree(sum); hipHostUnregister(a); hipStreamDestroy(stream); } // a imagini // b imagine void cityblockNormAsync(float *a, float *b, unsigned long n, float *result, unsigned long nrPoze) { float *d_a,*d_b,*d_rez,*d_temp; float *sum; hipStream_t stream; hipStreamCreate(&stream); hipHostRegister((void **)&a, n*nrPoze*sizeof(float), hipHostRegisterPortable); hipHostAlloc((void **)&sum, nrPoze*sizeof(float), hipHostMallocDefault); hipMalloc((void **)&d_a,n*sizeof(float)); hipMalloc((void **)&d_b,n*sizeof(float)); hipMalloc((void **)&d_temp,n*sizeof(float)); hipMalloc((void **)&d_rez,(n/64+(256-n/64))*sizeof(float)); //cudaMemcpy(d_a,a,n*sizeof(float),cudaMemcpyHostToDevice); hipMemcpy(d_b,b,n*sizeof(float),hipMemcpyHostToDevice); for(unsigned long int i=0, k=0; i<n, k< nrPoze; i+= n, k++) { hipMemcpyAsync(d_a,a+i,n*sizeof(float),hipMemcpyHostToDevice, stream); vecSub<<<n/64, 64, n*sizeof(float), stream>>>(d_a,d_b,d_temp, n); reductionSum<<<n/64, 64, n*sizeof(float), stream>>>(d_temp,d_rez, n); hipMemsetAsync(d_rez+n/64,0,95*sizeof(float),stream); reductionSum<<<1,256, n*sizeof(float), stream>>>(d_rez,d_temp,n); //cumSum<<<n/64, n/64, n*sizeof(float)>>>(d_rez,d_temp, n/64); hipMemcpyAsync(sum+k,d_temp,sizeof(float),hipMemcpyDeviceToHost, stream); } hipStreamSynchronize(stream); for(int i=0;i<nrPoze;i++) { result[i] = sum[i]; } //printf("%f ", min); hipFree(d_a); hipFree(d_b); hipFree(d_temp); hipFree(d_rez); hipHostFree(sum); hipHostUnregister(a); hipStreamDestroy(stream); } // a imagini // b imagine void cosNormAsync(float *a, float *b, unsigned long n, float *result, unsigned long nrPoze) { float *d_a,*d_b,*d_rez,*d_temp; float *d_a1,*d_b1,*d_rez1, *d_temp1; float *d_a2,*d_b2,*d_rez2, *d_temp2; float *dotProd, *normA, *normB; hipStream_t stream, stream1, stream2; hipStreamCreate(&stream); hipStreamCreate(&stream1); hipStreamCreate(&stream2); hipHostRegister((void **)&a, n*nrPoze*sizeof(float), hipHostRegisterPortable); hipHostAlloc((void **)&dotProd, nrPoze*sizeof(float), hipHostMallocDefault); hipHostAlloc((void **)&normA, nrPoze*sizeof(float), hipHostMallocDefault); hipHostAlloc((void **)&normB, nrPoze*sizeof(float), hipHostMallocDefault); hipMalloc((void **)&d_a,n*sizeof(float)); hipMalloc((void **)&d_b,n*sizeof(float)); hipMalloc((void **)&d_temp,n*sizeof(float)); hipMalloc((void **)&d_rez,(n/64+(256-n/64))*sizeof(float)); hipMalloc((void **)&d_a1,n*sizeof(float)); hipMalloc((void **)&d_b1,n*sizeof(float)); hipMalloc((void **)&d_temp1,n*sizeof(float)); hipMalloc((void **)&d_rez1,(n/64+(256-n/64))*sizeof(float)); hipMalloc((void **)&d_a2,n*sizeof(float)); hipMalloc((void **)&d_b2,n*sizeof(float)); hipMalloc((void **)&d_temp2,n*sizeof(float)); hipMalloc((void **)&d_rez2,(n/64+(256-n/64))*sizeof(float)); //cudaMemcpy(d_a,a,n*sizeof(float),cudaMemcpyHostToDevice); hipMemcpy(d_b,b,n*sizeof(float),hipMemcpyHostToDevice); hipMemcpy(d_b1,b,n*sizeof(float),hipMemcpyHostToDevice); hipMemcpy(d_b2,b,n*sizeof(float),hipMemcpyHostToDevice); for(unsigned long int i=0, k=0; i<n, k< nrPoze; i+= n, k++) { hipMemcpyAsync(d_a,a+i,n*sizeof(float),hipMemcpyHostToDevice, stream); hipMemcpyAsync(d_a1,a+i,n*sizeof(float),hipMemcpyHostToDevice, stream1); hipMemcpyAsync(d_a2,a+i,n*sizeof(float),hipMemcpyHostToDevice, stream1); vecProd<<<n/64, 64, n*sizeof(float), stream>>>(d_a,d_b,d_temp, n); vecProd<<<n/64, 64, n*sizeof(float), stream1>>>(d_a1,d_a2,d_temp1, n); if(i==0) vecProd<<<n/64, 64, n*sizeof(float), stream2>>>(d_b1,d_b2,d_temp2, n); reductionSum<<<n/64, 64, n*sizeof(float), stream>>>(d_temp,d_rez, n); reductionSum<<<n/64, 64, n*sizeof(float), stream1>>>(d_temp1,d_rez1, n); if(i==0) reductionSum<<<n/64, 64, n*sizeof(float), stream2>>>(d_temp2,d_rez2, n); hipMemsetAsync(d_rez+n/64,0,95*sizeof(float),stream); hipMemsetAsync(d_rez1+n/64,0,95*sizeof(float),stream1); if(i==0) hipMemsetAsync(d_rez2+n/64,0,95*sizeof(float),stream2); reductionSum<<<1,256, n*sizeof(float), stream>>>(d_rez,d_temp,n); reductionSum<<<1,256, n*sizeof(float), stream1>>>(d_rez1,d_temp1,n); if(i==0) reductionSum<<<1,256, n*sizeof(float), stream2>>>(d_rez2,d_temp2,n); // we now have in d_temp - dot product, d_temp1 - norm of a, d_temp2 norm of b hipMemcpyAsync(dotProd+k,d_temp,sizeof(float),hipMemcpyDeviceToHost, stream); hipMemcpyAsync(normA+k,d_temp1,sizeof(float),hipMemcpyDeviceToHost, stream1); hipMemcpyAsync(normB+k,d_temp2,sizeof(float),hipMemcpyDeviceToHost, stream2); } hipStreamSynchronize(stream); hipStreamSynchronize(stream1); hipStreamSynchronize(stream2); for(int i=0;i<nrPoze;i++) { float calc = 1 - dotProd[i]/(sqrtf(normA[i])*sqrtf(normB[i])); result[i] = isnan(calc)?1:calc; } //printf("%f ",result[0]); hipFree(d_a); hipFree(d_b); hipFree(d_temp); hipFree(d_rez); hipFree(d_a1); hipFree(d_b2); hipFree(d_temp1); hipFree(d_rez1); hipFree(d_a2); hipFree(d_b2); hipFree(d_temp2); hipFree(d_rez2); hipHostFree(dotProd); hipHostFree(normA); hipHostFree(normB); hipHostUnregister(a); hipStreamDestroy(stream); hipStreamDestroy(stream1); hipStreamDestroy(stream2); }
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 <math.h> #define NUM_BANKS 8 #define LOG_NUM_BANKS 3 #define CONFLICT_FREE_OFFSET(n) \ ((n) >> NUM_BANKS + (n) >> (2 * LOG_NUM_BANKS)) // Print device properties void printDevProp(hipDeviceProp_t devProp) { printf("Major revision number: %d\n", devProp.major); printf("Minor revision number: %d\n", devProp.minor); printf("Name: %s\n", devProp.name); printf("Total global memory: %u\n", devProp.totalGlobalMem); printf("Total shared memory per block: %u\n", devProp.sharedMemPerBlock); printf("Total registers per block: %d\n", devProp.regsPerBlock); printf("Warp size: %d\n", devProp.warpSize); printf("Maximum memory pitch: %u\n", devProp.memPitch); printf("Maximum threads per block: %d\n", devProp.maxThreadsPerBlock); for (int i = 0; i < 3; ++i) printf("Maximum dimension %d of block: %d\n", i, devProp.maxThreadsDim[i]); for (int i = 0; i < 3; ++i) printf("Maximum dimension %d of grid: %d\n", i, devProp.maxGridSize[i]); printf("Clock rate: %d\n", devProp.clockRate); printf("Total constant memory: %u\n", devProp.totalConstMem); printf("Texture alignment: %u\n", devProp.textureAlignment); printf("Concurrent copy and execution: %s\n", (devProp.deviceOverlap ? "Yes" : "No")); printf("Number of multiprocessors: %d\n", devProp.multiProcessorCount); printf("Kernel execution timeout: %s\n", (devProp.kernelExecTimeoutEnabled ? "Yes" : "No")); return; } void proprietati() { int devCount; hipGetDeviceCount(&devCount); printf("Avem %d device-uri.\n", devCount); for (int i = 0; i < devCount; ++i) { // Get device properties printf("\nCUDA Device #%d\n", i); hipDeviceProp_t devProp; hipGetDeviceProperties(&devProp, i); printDevProp(devProp); } } __global__ void vecDiv(float *a, float *b, float *c, unsigned int n) { // Get our global thread ID int id = blockIdx.x*blockDim.x+threadIdx.x; // Make sure we do not go out of bounds if (id < n) c[id] = a[id] / b[id]; } __global__ void vecProd(float *a, float *b, float *c, unsigned int n) { // Get our global thread ID int id = blockIdx.x*blockDim.x+threadIdx.x; // Make sure we do not go out of bounds if (id < n) c[id] = a[id] * b[id]; } __global__ void vecAdd(float *a, float *b, float *c, unsigned int n) { // Get our global thread ID int id = blockIdx.x*blockDim.x+threadIdx.x; // Make sure we do not go out of bounds if (id < n) c[id] = a[id] + b[id]; } __global__ void vecSub(float *a, float *b, float *c, unsigned int n) { // Get our global thread ID int id = blockIdx.x*blockDim.x+threadIdx.x; // Make sure we do not go out of bounds if (id < n) c[id] = a[id] - b[id]; } __global__ void eulVector(float *a, float *b,float *c, unsigned int n) { int id = blockIdx.x*blockDim.x+threadIdx.x; if(id < n){ c[id] = a[id] - b[id]; c[id]*= c[id]; } } __global__ void smallSum(unsigned char *out, unsigned char *in, int n) { int id = blockIdx.x*blockDim.x+threadIdx.x; if(threadIdx.x == 1023) { out[id/1024] = id; } } // reduction sum __global__ void reductionSum(float *in_array, float *out_array, unsigned long size) { extern __shared__ float shared[]; unsigned int tid = threadIdx.x; unsigned int i = blockIdx.x*blockDim.x + threadIdx.x; shared[tid] = in_array[i]; __syncthreads(); for(unsigned int s=blockDim.x/2; s>0; s>>=1) { if (tid < s) { shared[tid] += shared[tid + s]; } __syncthreads(); } if(tid==0) //first thread, last sum { out_array[blockIdx.x]=shared[0]; } } __global__ void cumSum(float *in_array, float *out_array, unsigned long size) { extern __shared__ float shared[]; unsigned int tid = threadIdx.x; unsigned int b_offset = blockIdx.x * blockDim.x; unsigned int offset = 1; int i = tid; int j = tid + blockDim.x / 2; int offset_i = CONFLICT_FREE_OFFSET(i); int offset_j = CONFLICT_FREE_OFFSET(j); shared[i + offset_i] = in_array[i + b_offset]; shared[j + offset_j] = in_array[j + b_offset]; // scan up for (int s = (blockDim.x >> 1); s > 0; s >>= 1) { __syncthreads(); if (tid < s) { int i = offset * (2 * tid + 1) - 1; int j = offset * (2 * tid + 2) - 1; i += CONFLICT_FREE_OFFSET(i); j += CONFLICT_FREE_OFFSET(j); shared[j] += shared[i]; } offset <<= 1; } if (tid == 0) { shared[blockDim.x - 1 + CONFLICT_FREE_OFFSET(blockDim.x - 1)] = 0; } // scan down for (int s = 1; s < blockDim.x; s <<= 1) { offset >>= 1; __syncthreads(); if (tid < s) { int i = offset * (2 * tid + 1) - 1; int j = offset * (2 * tid + 2) - 1; i += CONFLICT_FREE_OFFSET(i); j += CONFLICT_FREE_OFFSET(j); float tmp = shared[i]; shared[i] = shared[j]; shared[j] += tmp; } } __syncthreads(); out_array[i] = shared[i + offset_i]; out_array[j] = shared[j + offset_j]; } void test1(float *a, float *b, unsigned long n) { float *d_a, *d_b, *d_rez; float *rez = (float*) malloc(n*sizeof(float)); hipMalloc((void **)&d_a,n*sizeof(float)); hipMalloc((void **)&d_b,n*sizeof(float)); hipMalloc((void **)&d_rez,n*sizeof(float)); hipMemcpy(d_a,a,n*sizeof(float),hipMemcpyHostToDevice); hipMemcpy(d_b,b,n*sizeof(float),hipMemcpyHostToDevice); vecAdd<<<161,64>>>(d_a,d_b,d_rez,n); hipMemcpy(rez,d_rez,10304*sizeof(float),hipMemcpyDeviceToHost); printf("\n REZULTAT TEST: %f \n",rez[0]); hipFree(d_a); hipFree(d_b); hipFree(d_rez); free(rez); } // a imagini // b imagine void euclideanNormAsync(float *a, float *b, unsigned long n, float *result, unsigned long nrPoze) { float *d_a,*d_b,*d_rez,*d_temp; float *sum; hipStream_t stream; hipStreamCreate(&stream); hipHostRegister((void **)&a, n*nrPoze*sizeof(float), hipHostRegisterPortable); hipHostAlloc((void **)&sum, nrPoze*sizeof(float), hipHostMallocDefault); hipMalloc((void **)&d_a,n*sizeof(float)); hipMalloc((void **)&d_b,n*sizeof(float)); hipMalloc((void **)&d_temp,n*sizeof(float)); hipMalloc((void **)&d_rez,(n/64+(256-n/64))*sizeof(float)); //cudaMemcpy(d_a,a,n*sizeof(float),cudaMemcpyHostToDevice); hipMemcpy(d_b,b,n*sizeof(float),hipMemcpyHostToDevice); for(unsigned long int i=0, k=0; i<n, k< nrPoze; i+= n, k++) { hipMemcpyAsync(d_a,a+i,n*sizeof(float),hipMemcpyHostToDevice, stream); eulVector<<<n/64, 64, n*sizeof(float), stream>>>(d_a,d_b,d_temp, n); reductionSum<<<n/64, 64, n*sizeof(float), stream>>>(d_temp,d_rez, n); hipMemsetAsync(d_rez+n/64,0,95*sizeof(float),stream); reductionSum<<<1,256, n*sizeof(float), stream>>>(d_rez,d_temp,n); //cumSum<<<n/64, n/64, n*sizeof(float)>>>(d_rez,d_temp, n/64); hipMemcpyAsync(sum+k,d_temp,sizeof(float),hipMemcpyDeviceToHost, stream); } hipStreamSynchronize(stream); for(int i=0;i<nrPoze;i++) { result[i] = sum[i]; } hipFree(d_a); hipFree(d_b); hipFree(d_temp); hipFree(d_rez); hipHostFree(sum); hipHostUnregister(a); hipStreamDestroy(stream); } // a imagini // b imagine void cityblockNormAsync(float *a, float *b, unsigned long n, float *result, unsigned long nrPoze) { float *d_a,*d_b,*d_rez,*d_temp; float *sum; hipStream_t stream; hipStreamCreate(&stream); hipHostRegister((void **)&a, n*nrPoze*sizeof(float), hipHostRegisterPortable); hipHostAlloc((void **)&sum, nrPoze*sizeof(float), hipHostMallocDefault); hipMalloc((void **)&d_a,n*sizeof(float)); hipMalloc((void **)&d_b,n*sizeof(float)); hipMalloc((void **)&d_temp,n*sizeof(float)); hipMalloc((void **)&d_rez,(n/64+(256-n/64))*sizeof(float)); //cudaMemcpy(d_a,a,n*sizeof(float),cudaMemcpyHostToDevice); hipMemcpy(d_b,b,n*sizeof(float),hipMemcpyHostToDevice); for(unsigned long int i=0, k=0; i<n, k< nrPoze; i+= n, k++) { hipMemcpyAsync(d_a,a+i,n*sizeof(float),hipMemcpyHostToDevice, stream); vecSub<<<n/64, 64, n*sizeof(float), stream>>>(d_a,d_b,d_temp, n); reductionSum<<<n/64, 64, n*sizeof(float), stream>>>(d_temp,d_rez, n); hipMemsetAsync(d_rez+n/64,0,95*sizeof(float),stream); reductionSum<<<1,256, n*sizeof(float), stream>>>(d_rez,d_temp,n); //cumSum<<<n/64, n/64, n*sizeof(float)>>>(d_rez,d_temp, n/64); hipMemcpyAsync(sum+k,d_temp,sizeof(float),hipMemcpyDeviceToHost, stream); } hipStreamSynchronize(stream); for(int i=0;i<nrPoze;i++) { result[i] = sum[i]; } //printf("%f ", min); hipFree(d_a); hipFree(d_b); hipFree(d_temp); hipFree(d_rez); hipHostFree(sum); hipHostUnregister(a); hipStreamDestroy(stream); } // a imagini // b imagine void cosNormAsync(float *a, float *b, unsigned long n, float *result, unsigned long nrPoze) { float *d_a,*d_b,*d_rez,*d_temp; float *d_a1,*d_b1,*d_rez1, *d_temp1; float *d_a2,*d_b2,*d_rez2, *d_temp2; float *dotProd, *normA, *normB; hipStream_t stream, stream1, stream2; hipStreamCreate(&stream); hipStreamCreate(&stream1); hipStreamCreate(&stream2); hipHostRegister((void **)&a, n*nrPoze*sizeof(float), hipHostRegisterPortable); hipHostAlloc((void **)&dotProd, nrPoze*sizeof(float), hipHostMallocDefault); hipHostAlloc((void **)&normA, nrPoze*sizeof(float), hipHostMallocDefault); hipHostAlloc((void **)&normB, nrPoze*sizeof(float), hipHostMallocDefault); hipMalloc((void **)&d_a,n*sizeof(float)); hipMalloc((void **)&d_b,n*sizeof(float)); hipMalloc((void **)&d_temp,n*sizeof(float)); hipMalloc((void **)&d_rez,(n/64+(256-n/64))*sizeof(float)); hipMalloc((void **)&d_a1,n*sizeof(float)); hipMalloc((void **)&d_b1,n*sizeof(float)); hipMalloc((void **)&d_temp1,n*sizeof(float)); hipMalloc((void **)&d_rez1,(n/64+(256-n/64))*sizeof(float)); hipMalloc((void **)&d_a2,n*sizeof(float)); hipMalloc((void **)&d_b2,n*sizeof(float)); hipMalloc((void **)&d_temp2,n*sizeof(float)); hipMalloc((void **)&d_rez2,(n/64+(256-n/64))*sizeof(float)); //cudaMemcpy(d_a,a,n*sizeof(float),cudaMemcpyHostToDevice); hipMemcpy(d_b,b,n*sizeof(float),hipMemcpyHostToDevice); hipMemcpy(d_b1,b,n*sizeof(float),hipMemcpyHostToDevice); hipMemcpy(d_b2,b,n*sizeof(float),hipMemcpyHostToDevice); for(unsigned long int i=0, k=0; i<n, k< nrPoze; i+= n, k++) { hipMemcpyAsync(d_a,a+i,n*sizeof(float),hipMemcpyHostToDevice, stream); hipMemcpyAsync(d_a1,a+i,n*sizeof(float),hipMemcpyHostToDevice, stream1); hipMemcpyAsync(d_a2,a+i,n*sizeof(float),hipMemcpyHostToDevice, stream1); vecProd<<<n/64, 64, n*sizeof(float), stream>>>(d_a,d_b,d_temp, n); vecProd<<<n/64, 64, n*sizeof(float), stream1>>>(d_a1,d_a2,d_temp1, n); if(i==0) vecProd<<<n/64, 64, n*sizeof(float), stream2>>>(d_b1,d_b2,d_temp2, n); reductionSum<<<n/64, 64, n*sizeof(float), stream>>>(d_temp,d_rez, n); reductionSum<<<n/64, 64, n*sizeof(float), stream1>>>(d_temp1,d_rez1, n); if(i==0) reductionSum<<<n/64, 64, n*sizeof(float), stream2>>>(d_temp2,d_rez2, n); hipMemsetAsync(d_rez+n/64,0,95*sizeof(float),stream); hipMemsetAsync(d_rez1+n/64,0,95*sizeof(float),stream1); if(i==0) hipMemsetAsync(d_rez2+n/64,0,95*sizeof(float),stream2); reductionSum<<<1,256, n*sizeof(float), stream>>>(d_rez,d_temp,n); reductionSum<<<1,256, n*sizeof(float), stream1>>>(d_rez1,d_temp1,n); if(i==0) reductionSum<<<1,256, n*sizeof(float), stream2>>>(d_rez2,d_temp2,n); // we now have in d_temp - dot product, d_temp1 - norm of a, d_temp2 norm of b hipMemcpyAsync(dotProd+k,d_temp,sizeof(float),hipMemcpyDeviceToHost, stream); hipMemcpyAsync(normA+k,d_temp1,sizeof(float),hipMemcpyDeviceToHost, stream1); hipMemcpyAsync(normB+k,d_temp2,sizeof(float),hipMemcpyDeviceToHost, stream2); } hipStreamSynchronize(stream); hipStreamSynchronize(stream1); hipStreamSynchronize(stream2); for(int i=0;i<nrPoze;i++) { float calc = 1 - dotProd[i]/(sqrtf(normA[i])*sqrtf(normB[i])); result[i] = isnan(calc)?1:calc; } //printf("%f ",result[0]); hipFree(d_a); hipFree(d_b); hipFree(d_temp); hipFree(d_rez); hipFree(d_a1); hipFree(d_b2); hipFree(d_temp1); hipFree(d_rez1); hipFree(d_a2); hipFree(d_b2); hipFree(d_temp2); hipFree(d_rez2); hipHostFree(dotProd); hipHostFree(normA); hipHostFree(normB); hipHostUnregister(a); hipStreamDestroy(stream); hipStreamDestroy(stream1); hipStreamDestroy(stream2); }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z6vecDivPfS_S_j .globl _Z6vecDivPfS_S_j .p2align 8 .type _Z6vecDivPfS_S_j,@function _Z6vecDivPfS_S_j: s_clause 0x1 s_load_b32 s2, s[0:1], 0x2c s_load_b32 s3, s[0:1], 0x18 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_u32_e64 s3, v1 s_cbranch_execz .LBB0_2 s_load_b128 s[4:7], s[0:1], 0x0 v_ashrrev_i32_e32 v2, 31, v1 s_load_b64 s[0:1], s[0:1], 0x10 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[0:1], 2, v[1:2] s_waitcnt lgkmcnt(0) v_add_co_u32 v2, vcc_lo, s4, v0 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v3, vcc_lo, s5, v1, vcc_lo v_add_co_u32 v4, vcc_lo, s6, v0 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v1, vcc_lo global_load_b32 v2, v[2:3], off global_load_b32 v3, v[4:5], off s_waitcnt vmcnt(0) v_div_scale_f32 v4, null, v3, v3, v2 v_div_scale_f32 v7, vcc_lo, v2, v3, v2 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_2) | instid1(VALU_DEP_1) v_rcp_f32_e32 v5, v4 s_waitcnt_depctr 0xfff v_fma_f32 v6, -v4, v5, 1.0 v_fmac_f32_e32 v5, v6, v5 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mul_f32_e32 v6, v7, v5 v_fma_f32 v8, -v4, v6, v7 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmac_f32_e32 v6, v8, v5 v_fma_f32 v4, -v4, v6, v7 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_3) v_div_fmas_f32 v4, v4, v5, v6 v_add_co_u32 v0, vcc_lo, s0, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo v_div_fixup_f32 v2, v4, v3, v2 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 _Z6vecDivPfS_S_j .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 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 _Z6vecDivPfS_S_j, .Lfunc_end0-_Z6vecDivPfS_S_j .section .AMDGPU.csdata,"",@progbits .text .protected _Z7vecProdPfS_S_j .globl _Z7vecProdPfS_S_j .p2align 8 .type _Z7vecProdPfS_S_j,@function _Z7vecProdPfS_S_j: s_clause 0x1 s_load_b32 s2, s[0:1], 0x2c s_load_b32 s3, s[0:1], 0x18 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_u32_e64 s3, v1 s_cbranch_execz .LBB1_2 s_load_b128 s[4:7], s[0:1], 0x0 v_ashrrev_i32_e32 v2, 31, v1 s_load_b64 s[0:1], s[0:1], 0x10 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[0:1], 2, v[1:2] s_waitcnt lgkmcnt(0) v_add_co_u32 v2, vcc_lo, s4, v0 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v3, vcc_lo, s5, v1, vcc_lo v_add_co_u32 v4, vcc_lo, s6, v0 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v1, vcc_lo v_add_co_u32 v0, vcc_lo, s0, v0 global_load_b32 v2, v[2:3], off global_load_b32 v3, v[4:5], off v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo s_waitcnt vmcnt(0) v_mul_f32_e32 v2, v2, v3 global_store_b32 v[0:1], v2, off .LBB1_2: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z7vecProdPfS_S_j .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 6 .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_end1: .size _Z7vecProdPfS_S_j, .Lfunc_end1-_Z7vecProdPfS_S_j .section .AMDGPU.csdata,"",@progbits .text .protected _Z6vecAddPfS_S_j .globl _Z6vecAddPfS_S_j .p2align 8 .type _Z6vecAddPfS_S_j,@function _Z6vecAddPfS_S_j: s_clause 0x1 s_load_b32 s2, s[0:1], 0x2c s_load_b32 s3, s[0:1], 0x18 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_u32_e64 s3, v1 s_cbranch_execz .LBB2_2 s_load_b128 s[4:7], s[0:1], 0x0 v_ashrrev_i32_e32 v2, 31, v1 s_load_b64 s[0:1], s[0:1], 0x10 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[0:1], 2, v[1:2] s_waitcnt lgkmcnt(0) v_add_co_u32 v2, vcc_lo, s4, v0 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v3, vcc_lo, s5, v1, vcc_lo v_add_co_u32 v4, vcc_lo, s6, v0 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v1, vcc_lo v_add_co_u32 v0, vcc_lo, s0, v0 global_load_b32 v2, v[2:3], off global_load_b32 v3, v[4:5], off v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo s_waitcnt vmcnt(0) v_add_f32_e32 v2, v2, v3 global_store_b32 v[0:1], v2, off .LBB2_2: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z6vecAddPfS_S_j .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 6 .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_end2: .size _Z6vecAddPfS_S_j, .Lfunc_end2-_Z6vecAddPfS_S_j .section .AMDGPU.csdata,"",@progbits .text .protected _Z6vecSubPfS_S_j .globl _Z6vecSubPfS_S_j .p2align 8 .type _Z6vecSubPfS_S_j,@function _Z6vecSubPfS_S_j: s_clause 0x1 s_load_b32 s2, s[0:1], 0x2c s_load_b32 s3, s[0:1], 0x18 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_u32_e64 s3, v1 s_cbranch_execz .LBB3_2 s_load_b128 s[4:7], s[0:1], 0x0 v_ashrrev_i32_e32 v2, 31, v1 s_load_b64 s[0:1], s[0:1], 0x10 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[0:1], 2, v[1:2] s_waitcnt lgkmcnt(0) v_add_co_u32 v2, vcc_lo, s4, v0 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v3, vcc_lo, s5, v1, vcc_lo v_add_co_u32 v4, vcc_lo, s6, v0 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v1, vcc_lo v_add_co_u32 v0, vcc_lo, s0, v0 global_load_b32 v2, v[2:3], off global_load_b32 v3, v[4:5], off v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo s_waitcnt vmcnt(0) v_sub_f32_e32 v2, v2, v3 global_store_b32 v[0:1], v2, off .LBB3_2: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z6vecSubPfS_S_j .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 6 .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_end3: .size _Z6vecSubPfS_S_j, .Lfunc_end3-_Z6vecSubPfS_S_j .section .AMDGPU.csdata,"",@progbits .text .protected _Z9eulVectorPfS_S_j .globl _Z9eulVectorPfS_S_j .p2align 8 .type _Z9eulVectorPfS_S_j,@function _Z9eulVectorPfS_S_j: s_clause 0x1 s_load_b32 s2, s[0:1], 0x2c s_load_b32 s3, s[0:1], 0x18 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_u32_e64 s3, v1 s_cbranch_execz .LBB4_2 s_load_b128 s[4:7], s[0:1], 0x0 v_ashrrev_i32_e32 v2, 31, v1 s_load_b64 s[0:1], s[0:1], 0x10 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[0:1], 2, v[1:2] s_waitcnt lgkmcnt(0) v_add_co_u32 v2, vcc_lo, s4, v0 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v3, vcc_lo, s5, v1, vcc_lo v_add_co_u32 v4, vcc_lo, s6, v0 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v1, vcc_lo v_add_co_u32 v0, vcc_lo, s0, v0 global_load_b32 v2, v[2:3], off global_load_b32 v3, v[4:5], off v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo s_waitcnt vmcnt(0) v_sub_f32_e32 v2, v2, v3 s_delay_alu instid0(VALU_DEP_1) v_mul_f32_e32 v2, v2, v2 global_store_b32 v[0:1], v2, off .LBB4_2: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z9eulVectorPfS_S_j .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 6 .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_end4: .size _Z9eulVectorPfS_S_j, .Lfunc_end4-_Z9eulVectorPfS_S_j .section .AMDGPU.csdata,"",@progbits .text .protected _Z8smallSumPhS_i .globl _Z8smallSumPhS_i .p2align 8 .type _Z8smallSumPhS_i,@function _Z8smallSumPhS_i: s_mov_b32 s2, exec_lo v_cmpx_eq_u32_e32 0x3ff, v0 s_cbranch_execz .LBB5_2 s_clause 0x1 s_load_b32 s2, s[0:1], 0x24 s_load_b64 s[0:1], s[0:1], 0x0 s_waitcnt lgkmcnt(0) s_and_b32 s2, s2, 0xffff s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mad_u64_u32 v[1:2], null, s15, s2, v[0:1] v_ashrrev_i32_e32 v0, 31, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshrrev_b32_e32 v0, 22, v0 v_add_nc_u32_e32 v0, v1, v0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v0, 10, v0 v_ashrrev_i32_e32 v3, 31, v0 v_add_co_u32 v2, vcc_lo, s0, v0 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v3, vcc_lo, s1, v3, vcc_lo global_store_b8 v[2:3], v1, off .LBB5_2: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z8smallSumPhS_i .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 280 .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 4 .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_end5: .size _Z8smallSumPhS_i, .Lfunc_end5-_Z8smallSumPhS_i .section .AMDGPU.csdata,"",@progbits .text .protected _Z12reductionSumPfS_m .globl _Z12reductionSumPfS_m .p2align 8 .type _Z12reductionSumPfS_m,@function _Z12reductionSumPfS_m: s_clause 0x1 s_load_b32 s3, s[0:1], 0x24 s_load_b64 s[4:5], s[0:1], 0x0 s_mov_b32 s2, s15 s_waitcnt lgkmcnt(0) s_and_b32 s3, s3, 0xffff s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_2) | instid1(VALU_DEP_1) v_mad_u64_u32 v[1:2], null, s2, s3, v[0:1] v_mov_b32_e32 v2, 0 s_cmp_lt_u32 s3, 2 v_lshlrev_b64 v[1:2], 2, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v1, vcc_lo, s4, v1 v_add_co_ci_u32_e32 v2, vcc_lo, s5, v2, vcc_lo global_load_b32 v2, v[1:2], off v_lshl_add_u32 v1, v0, 2, 0 s_waitcnt vmcnt(0) ds_store_b32 v1, v2 s_waitcnt lgkmcnt(0) s_barrier s_branch .LBB6_2 .p2align 6 .LBB6_1: s_or_b32 exec_lo, exec_lo, s5 s_waitcnt lgkmcnt(0) s_barrier s_cmp_lt_u32 s3, 4 s_mov_b32 s3, s4 .LBB6_2: buffer_gl0_inv s_cbranch_scc1 .LBB6_5 s_lshr_b32 s4, s3, 1 s_mov_b32 s5, exec_lo v_cmpx_gt_u32_e64 s4, v0 s_cbranch_execz .LBB6_1 v_add_nc_u32_e32 v2, s4, v0 s_delay_alu instid0(VALU_DEP_1) v_lshl_add_u32 v2, v2, 2, 0 ds_load_b32 v2, v2 ds_load_b32 v3, v1 s_waitcnt lgkmcnt(0) v_add_f32_e32 v2, v2, v3 ds_store_b32 v1, v2 s_branch .LBB6_1 .LBB6_5: s_mov_b32 s3, exec_lo v_cmpx_eq_u32_e32 0, v0 s_cbranch_execz .LBB6_7 v_dual_mov_b32 v0, 0 :: v_dual_mov_b32 v1, 0 s_load_b64 s[0:1], s[0:1], 0x8 s_mov_b32 s3, 0 s_delay_alu instid0(SALU_CYCLE_1) s_lshl_b64 s[2:3], s[2:3], 2 ds_load_b32 v0, v0 s_waitcnt lgkmcnt(0) s_add_u32 s0, s0, s2 s_addc_u32 s1, s1, s3 global_store_b32 v1, v0, s[0:1] .LBB6_7: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z12reductionSumPfS_m .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 280 .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 4 .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_end6: .size _Z12reductionSumPfS_m, .Lfunc_end6-_Z12reductionSumPfS_m .section .AMDGPU.csdata,"",@progbits .text .protected _Z6cumSumPfS_m .globl _Z6cumSumPfS_m .p2align 8 .type _Z6cumSumPfS_m,@function _Z6cumSumPfS_m: s_clause 0x1 s_load_b32 s3, s[0:1], 0x24 s_load_b64 s[4:5], s[0:1], 0x0 s_waitcnt lgkmcnt(0) s_and_b32 s2, s3, 0xffff s_bfe_u32 s3, s3, 0xf0001 s_mul_i32 s15, s15, s2 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_3) | instid1(VALU_DEP_2) v_dual_mov_b32 v3, 0 :: v_dual_add_nc_u32 v2, s15, v0 v_add_nc_u32_e32 v1, s3, v0 s_cmp_lt_u32 s2, 2 s_mov_b32 s3, 1 v_lshlrev_b64 v[4:5], 2, v[2:3] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_nc_u32_e32 v2, s15, v1 v_lshlrev_b64 v[2:3], 2, v[2:3] s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_u32 v4, vcc_lo, s4, v4 v_add_co_ci_u32_e32 v5, vcc_lo, s5, v5, vcc_lo s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_u32 v2, vcc_lo, s4, v2 v_add_co_ci_u32_e32 v3, vcc_lo, s5, v3, vcc_lo s_clause 0x1 global_load_b32 v4, v[4:5], off global_load_b32 v5, v[2:3], off v_add_nc_u32_e32 v2, 8, v1 v_lshl_add_u32 v3, v0, 2, 0 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshrrev_b32_e32 v2, v2, v1 v_lshrrev_b32_e32 v2, 6, v2 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_nc_u32_e32 v2, v2, v1 v_lshl_add_u32 v2, v2, 2, 0 s_waitcnt vmcnt(1) ds_store_b32 v3, v4 s_waitcnt vmcnt(0) ds_store_b32 v2, v5 s_cbranch_scc1 .LBB7_6 v_lshlrev_b32_e32 v5, 1, v0 s_mov_b32 s4, s2 s_delay_alu instid0(VALU_DEP_1) v_or_b32_e32 v4, 1, v5 v_add_nc_u32_e32 v5, 2, v5 s_set_inst_prefetch_distance 0x1 .p2align 6 .LBB7_2: s_lshr_b32 s5, s4, 1 s_mov_b32 s6, exec_lo s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv v_cmpx_gt_u32_e64 s5, v0 s_cbranch_execz .LBB7_4 v_mul_lo_u32 v6, s3, v4 v_mul_lo_u32 v7, s3, v5 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_3) v_add_nc_u32_e32 v8, -1, v6 v_add_nc_u32_e32 v6, 7, v6 v_add_nc_u32_e32 v9, -1, v7 v_add_nc_u32_e32 v7, 7, v7 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_ashrrev_i32_e32 v6, v6, v8 v_ashrrev_i32_e32 v7, v7, v9 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_ashrrev_i32_e32 v6, 6, v6 v_ashrrev_i32_e32 v7, 6, v7 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_nc_u32_e32 v6, v6, v8 v_add_nc_u32_e32 v7, v7, v9 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_lshl_add_u32 v6, v6, 2, 0 v_lshl_add_u32 v7, v7, 2, 0 ds_load_b32 v6, v6 ds_load_b32 v8, v7 s_waitcnt lgkmcnt(0) v_add_f32_e32 v6, v6, v8 ds_store_b32 v7, v6 .LBB7_4: s_or_b32 exec_lo, exec_lo, s6 s_lshl_b32 s3, s3, 1 s_cmp_lt_u32 s4, 4 s_cbranch_scc1 .LBB7_6 s_mov_b32 s4, s5 s_branch .LBB7_2 .LBB7_6: s_set_inst_prefetch_distance 0x2 s_mov_b32 s4, exec_lo v_cmpx_eq_u32_e32 0, v0 s_cbranch_execz .LBB7_8 s_add_i32 s5, s2, -1 s_add_i32 s6, s2, 7 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_lshr_b32 s6, s5, s6 s_lshr_b32 s6, s6, 6 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_add_i32 s6, s6, s5 s_lshl_b32 s5, s6, 2 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_add_i32 s5, s5, 0 v_dual_mov_b32 v4, 0 :: v_dual_mov_b32 v5, s5 ds_store_b32 v5, v4 .LBB7_8: s_or_b32 exec_lo, exec_lo, s4 s_cmp_lt_u32 s2, 2 s_cbranch_scc1 .LBB7_13 v_lshlrev_b32_e32 v5, 1, v0 s_mov_b32 s4, 1 s_delay_alu instid0(VALU_DEP_1) v_or_b32_e32 v4, 1, v5 v_add_nc_u32_e32 v5, 2, v5 s_set_inst_prefetch_distance 0x1 s_branch .LBB7_11 .p2align 6 .LBB7_10: s_or_b32 exec_lo, exec_lo, s5 s_lshl_b32 s4, s4, 1 s_delay_alu instid0(SALU_CYCLE_1) s_cmp_ge_u32 s4, s2 s_cbranch_scc1 .LBB7_13 .LBB7_11: s_lshr_b32 s3, s3, 1 s_mov_b32 s5, exec_lo s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv v_cmpx_gt_u32_e64 s4, v0 s_cbranch_execz .LBB7_10 v_mul_lo_u32 v6, s3, v5 v_mul_lo_u32 v8, s3, v4 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_add_nc_u32_e32 v7, -1, v6 v_add_nc_u32_e32 v6, 7, v6 v_ashrrev_i32_e32 v6, v6, v7 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v6, 6, v6 v_add_nc_u32_e32 v6, v6, v7 v_add_nc_u32_e32 v7, -1, v8 v_add_nc_u32_e32 v8, 7, v8 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_2) v_lshl_add_u32 v6, v6, 2, 0 v_ashrrev_i32_e32 v8, v8, v7 ds_load_b32 v9, v6 v_ashrrev_i32_e32 v8, 6, v8 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_nc_u32_e32 v7, v8, v7 v_lshl_add_u32 v7, v7, 2, 0 ds_load_b32 v8, v7 s_waitcnt lgkmcnt(1) ds_store_b32 v7, v9 ds_load_b32 v7, v6 s_waitcnt lgkmcnt(0) v_add_f32_e32 v7, v8, v7 ds_store_b32 v6, v7 s_branch .LBB7_10 .LBB7_13: s_set_inst_prefetch_distance 0x2 s_load_b64 s[0:1], s[0:1], 0x8 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv ds_load_b32 v3, v3 ds_load_b32 v2, v2 v_lshlrev_b32_e32 v0, 2, v0 v_lshlrev_b32_e32 v1, 2, v1 s_waitcnt lgkmcnt(1) global_store_b32 v0, v3, s[0:1] s_waitcnt lgkmcnt(0) global_store_b32 v1, v2, s[0:1] s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z6cumSumPfS_m .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 280 .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_end7: .size _Z6cumSumPfS_m, .Lfunc_end7-_Z6cumSumPfS_m .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: _Z6vecDivPfS_S_j .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z6vecDivPfS_S_j.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 9 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .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: _Z7vecProdPfS_S_j .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z7vecProdPfS_S_j.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 6 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .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: _Z6vecAddPfS_S_j .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z6vecAddPfS_S_j.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 6 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .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: _Z6vecSubPfS_S_j .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z6vecSubPfS_S_j.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 6 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .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: _Z9eulVectorPfS_S_j .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z9eulVectorPfS_S_j.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 6 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .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: 24 .size: 4 .value_kind: hidden_block_count_x - .offset: 28 .size: 4 .value_kind: hidden_block_count_y - .offset: 32 .size: 4 .value_kind: hidden_block_count_z - .offset: 36 .size: 2 .value_kind: hidden_group_size_x - .offset: 38 .size: 2 .value_kind: hidden_group_size_y - .offset: 40 .size: 2 .value_kind: hidden_group_size_z - .offset: 42 .size: 2 .value_kind: hidden_remainder_x - .offset: 44 .size: 2 .value_kind: hidden_remainder_y - .offset: 46 .size: 2 .value_kind: hidden_remainder_z - .offset: 64 .size: 8 .value_kind: hidden_global_offset_x - .offset: 72 .size: 8 .value_kind: hidden_global_offset_y - .offset: 80 .size: 8 .value_kind: hidden_global_offset_z - .offset: 88 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 280 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z8smallSumPhS_i .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z8smallSumPhS_i.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 4 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .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: 8 .value_kind: by_value - .offset: 24 .size: 4 .value_kind: hidden_block_count_x - .offset: 28 .size: 4 .value_kind: hidden_block_count_y - .offset: 32 .size: 4 .value_kind: hidden_block_count_z - .offset: 36 .size: 2 .value_kind: hidden_group_size_x - .offset: 38 .size: 2 .value_kind: hidden_group_size_y - .offset: 40 .size: 2 .value_kind: hidden_group_size_z - .offset: 42 .size: 2 .value_kind: hidden_remainder_x - .offset: 44 .size: 2 .value_kind: hidden_remainder_y - .offset: 46 .size: 2 .value_kind: hidden_remainder_z - .offset: 64 .size: 8 .value_kind: hidden_global_offset_x - .offset: 72 .size: 8 .value_kind: hidden_global_offset_y - .offset: 80 .size: 8 .value_kind: hidden_global_offset_z - .offset: 88 .size: 2 .value_kind: hidden_grid_dims - .offset: 144 .size: 4 .value_kind: hidden_dynamic_lds_size .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 280 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z12reductionSumPfS_m .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z12reductionSumPfS_m.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 4 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .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: 8 .value_kind: by_value - .offset: 24 .size: 4 .value_kind: hidden_block_count_x - .offset: 28 .size: 4 .value_kind: hidden_block_count_y - .offset: 32 .size: 4 .value_kind: hidden_block_count_z - .offset: 36 .size: 2 .value_kind: hidden_group_size_x - .offset: 38 .size: 2 .value_kind: hidden_group_size_y - .offset: 40 .size: 2 .value_kind: hidden_group_size_z - .offset: 42 .size: 2 .value_kind: hidden_remainder_x - .offset: 44 .size: 2 .value_kind: hidden_remainder_y - .offset: 46 .size: 2 .value_kind: hidden_remainder_z - .offset: 64 .size: 8 .value_kind: hidden_global_offset_x - .offset: 72 .size: 8 .value_kind: hidden_global_offset_y - .offset: 80 .size: 8 .value_kind: hidden_global_offset_z - .offset: 88 .size: 2 .value_kind: hidden_grid_dims - .offset: 144 .size: 4 .value_kind: hidden_dynamic_lds_size .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 280 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z6cumSumPfS_m .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z6cumSumPfS_m.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 CUDA source code to HIP/AMD source code.
/** * CUDA implementation of a fully-connected feed forward neural network * * Authors: Jevgenija Aksjonova (jevaks@kth.se) * Beatrice Ionascu (bionascu@kth.se) * * Last changed: 04/30/2017 */ #include <stdio.h> #include <assert.h> #include <stdlib.h> #include <string.h> #include <math.h> #define TOL 0.00000001 #define BATCH_SIZE 100 #define TRAIN_SET_SIZE 60000 #define TEST_SET_SIZE 10000 #define FUNCTION_TYPE 1 // 0 - tanh, 1 - relu /* Read labels data */ void read_labels(const char filename[], int *array, int data_size) { memset(array, 0, sizeof(int) * data_size); FILE *file = fopen(filename, "r"); if (file) { size_t i; /* Read each line from the file. */ for (i = 0; i < data_size && !feof(file); i++) { fscanf(file, "%d", array + i); } assert(i == data_size); fclose(file); } else { perror(filename); } } /* Read image data */ void read_images(const char *filename, double *array, int data_size, int batch_size, int batch_ind) { FILE *file = fopen(filename, "r"); if (file) { size_t i, j; char buffer[data_size * 4], *ptr; /* Read each line from the file. */ for (i = 0; fgets(buffer, sizeof buffer, file); ++i) { /* Only parse data from desired batch. */ if (i >= batch_ind * batch_size && i < (batch_ind + 1) * batch_size ) { /* Parse the comma-separated values from each line into 'array'. */ for (j = 0, ptr = buffer; j < data_size; j++, ptr++) { array[(i - batch_ind * batch_size) * data_size + j] = strtol(ptr, &ptr, 10) / 256.0; } } } fclose(file); } else { perror(filename); } } /* Compute local number of parameters (includes weights and biases) */ int param_size(int *layer_size, int num_layers) { int size = 0; size_t i; for (i = 1; i < num_layers; i++) { size = size + layer_size[i] * (layer_size[i-1] +1); } return size; } /* Compute total amount of data (includes the input data features and the layer outputs) */ int data_size(int *layer_size, int num_layers) { int size = 0; size_t i; for (i = 0; i < num_layers; i++) { size += layer_size[i]; } return size; } /* Compute local starting index of the parameters corresponding to layer */ int param_ind(int layer, int *layer_size) { int ind = 0; size_t i; for (i = 1; i < layer; i++) { ind += layer_size[i] * (layer_size[i - 1] + 1); } return ind; } /* Compute local starting index of the data corresponding to layer */ int data_ind(int layer, int *layer_size) { int ind = 0; size_t i; for (i = 0; i < layer; i++) { ind += layer_size[i]; } return ind; } /* Forward pass */ __global__ void forward(double *input, int inputSize, double *output, int outputSize, double *param, int fun) { size_t i, j; double l_output; i = blockIdx.x * blockDim.x + threadIdx.x; if (i < outputSize ) { /* Initialize output with the bias term */ l_output = param[i]; for (j = 0; j < inputSize; j++) { // Add weighted inputs l_output += param[(j+1)*outputSize + i] * input[j]; } /* Activation */ if (fun > 0) { if (FUNCTION_TYPE == 0) { l_output = tanh(l_output); } else if (FUNCTION_TYPE == 1) { // relu function if (l_output <= 0) { l_output = 0.0; } } } output[i] = l_output; } } /* Backward pass (backpropagate gradients from current_layer to prev_layer) */ __global__ void backward_part1(double *current_layer, int layer_size, double *prev_layer, int prev_layer_size, double *param, double *grad_param, int layer,double lambda) { size_t i, j; double l_current_layer; i = blockIdx.x * blockDim.x + threadIdx.x; if (i < layer_size) { l_current_layer = current_layer[i]; // Update parameters grad_param[i] += l_current_layer; for (j = 0; j < prev_layer_size; j++) { grad_param[(j+1) * layer_size + i ] += l_current_layer * prev_layer[j] + 2.0 * lambda * param[(j+1) * layer_size + i]; // regularization } } } __global__ void backward_part2(double *current_layer, int layer_size, double *prev_layer, int prev_layer_size, double *param, double *grad_param, int layer,double lambda) { size_t i, k; double l_prev_layer; i = blockIdx.x * blockDim.x + threadIdx.x; if (i < prev_layer_size ) { l_prev_layer = prev_layer[i]; // backprapogate partial derivatives if (FUNCTION_TYPE == 0) { double localGrad = (1-pow(l_prev_layer,2)); // tanh derivative l_prev_layer = 0.0; for (k = 0; k < layer_size; k++) { l_prev_layer += current_layer[k] * param[(i+1) * layer_size + k]; } l_prev_layer *= localGrad; } else if (FUNCTION_TYPE == 1) { if (l_prev_layer >0.0) { l_prev_layer = 0.0; for (k = 0; k < layer_size; k++) { l_prev_layer += current_layer[k] * param[(i+1) * layer_size + k]; } } else { l_prev_layer = 0.0; } } prev_layer[i] = l_prev_layer; } } __global__ void update_param(double *param, double *grad_param, int num_param, double learning_rate, int batch_size) { int i = blockIdx.x * blockDim.x + threadIdx.x; if (i < num_param) { param[i] = param[i] - learning_rate * grad_param[i] / (float)batch_size; } } /* Train the network */ void train(const char filename[], int* label, double *d_param, double *d_grad_param, int *layer_size, int num_layers, int epochs, double lambda, double learning_rate) { size_t img, ep = 0; int i, layer, batch_index; int input_pointer, output_pointer, param_pointer; double sum, global_loss = 1; int image_size = layer_size[0]; int num_param = param_size(layer_size, num_layers); /* Allocate image data array */ double *images; images = (double *) malloc(BATCH_SIZE * image_size * sizeof(double)); /* Allocate data array */ double *d_data ; cudaMalloc( (void**)&d_data, data_size(layer_size, num_layers) * sizeof(double)); /* Loop over epochs (one epoch = loop over all images) */ while (ep++ < epochs && global_loss > TOL) { /* Loop over batches */ for (batch_index = 0; batch_index < TRAIN_SET_SIZE / BATCH_SIZE; batch_index++) { /* Initialize loss */ global_loss = 0.0; /* Reset gradients to 0 */ cudaMemset(d_grad_param, 0.0, num_param * sizeof(double)); /* Load images in batch */ read_images(filename, images, image_size, BATCH_SIZE, batch_index); /* Loop over images in batch */ for (img = 0; img < BATCH_SIZE; img++) { /* Copy one image to the 1st layer of the data */ cudaMemcpy(d_data, images + img * image_size, image_size * sizeof(double), cudaMemcpyHostToDevice ); /* Forward pass */ for (layer = 1; layer < num_layers; layer++) { input_pointer = data_ind(layer - 1, layer_size); output_pointer = data_ind(layer, layer_size); param_pointer = param_ind(layer, layer_size); dim3 dimBlock( 128,1,1 ); dim3 dimGrid(ceil((float)layer_size[layer]/128.0),1,1); forward<<<dimGrid, dimBlock>>>(d_data + input_pointer, layer_size[layer - 1], d_data + output_pointer, layer_size[layer], d_param + param_pointer, num_layers - layer - 1); } double *probs; probs = (double * ) malloc(layer_size[num_layers - 1]* sizeof(double)); int probs_pointer = data_ind(num_layers - 1, layer_size); cudaMemcpy(probs, d_data + probs_pointer, layer_size[num_layers - 1] * sizeof(double), cudaMemcpyDeviceToHost); /* Softmax */ sum = 0.0; for ( i = 0; i < layer_size[num_layers - 1]; i++) { sum += exp(probs[i]); } for ( i = 0; i < layer_size[num_layers - 1]; i++) { probs[i] = exp(probs[i]) / sum; } /* Loss computation */ /* Compute the cross-entropy loss by adding the log probabilities assigned to the correct classes */ global_loss += -log(probs[label[batch_index * BATCH_SIZE + img]]); /* Gradient computation for the last (score) layer */ probs[label[batch_index * BATCH_SIZE + img]] -= 1.0; cudaMemcpy(d_data+probs_pointer, probs, layer_size[num_layers - 1]*sizeof(double), cudaMemcpyHostToDevice); /* Backpropagation */ for (layer = num_layers - 1; layer > 0; layer--) { input_pointer = data_ind(layer, layer_size); output_pointer = data_ind(layer - 1, layer_size); param_pointer = param_ind(layer, layer_size); dim3 dimBlock( 128, 1, 1 ); dim3 dimGrid( ceil((float)layer_size[layer]/128.0), 1, 1 ); backward_part1<<<dimGrid, dimBlock>>>(d_data + input_pointer, layer_size[layer], d_data + output_pointer, layer_size[layer - 1], d_param + param_pointer, d_grad_param + param_pointer, layer, lambda); if (layer > 1) { dim3 dimBlock( 128, 1, 1 ); dim3 dimGrid( ceil((float)layer_size[layer - 1]/128.0), 1, 1 ); backward_part2<<<dimGrid, dimBlock>>>(d_data + input_pointer, layer_size[layer], d_data + output_pointer, layer_size[layer - 1], d_param + param_pointer, d_grad_param + param_pointer, layer, lambda); } } } /* Update parameters using gradient averaged over batch)*/ dim3 dimBlock(512, 1, 1 ); dim3 dimGrid( ceil((float)num_param/512.0), 1 ,1 ); update_param<<<dimGrid, dimBlock>>>(d_param, d_grad_param, num_param, learning_rate, BATCH_SIZE); /* Compute global loss averaged over batch */ global_loss /= BATCH_SIZE; if (batch_index % 100 == 0) { printf("Ep: %zu/%d\t batch:%d \t train loss: %f\n", ep, epochs, batch_index, global_loss); } } } cudaFree(d_data); free(images); } /* Test the network */ void test(const char filename[], int* label, double *d_param, int *layer_size, int num_layers) { int it, layer, i; int input_pointer, output_pointer, param_pointer, probs_pointer, predicted_class; double sum, predicted_class_score, global_loss = 0.0; int image_size = layer_size[0]; int acc = 0; /* Allocate image data array */ double *images; images = (double *) malloc(image_size* sizeof(double)); /* Allocate data array */ double *d_data ; cudaMalloc( (void**)&d_data, data_size(layer_size,num_layers)*sizeof(double)); /* Loop over batches */ for (it = 0; it < TEST_SET_SIZE; it++) { /* Load image data */ read_images(filename, images, image_size, 1, it); /* Copy one image to the 1st layer of the data */ cudaMemcpy(d_data, images, image_size * sizeof(double),cudaMemcpyHostToDevice); /* Forward pass */ for (layer = 1; layer < num_layers; layer++) { input_pointer = data_ind(layer - 1, layer_size); output_pointer = data_ind(layer, layer_size); param_pointer = param_ind(layer, layer_size); dim3 dimBlock( 128, 1, 1 ); dim3 dimGrid( ceil((float)layer_size[layer]/128.0), 1, 1 ); forward<<<dimGrid, dimBlock>>>(d_data + input_pointer, layer_size[layer - 1], d_data + output_pointer, layer_size[layer], d_param + param_pointer, num_layers - layer - 1); } /* Softmax and Predict class */ double *probs; probs = (double * ) malloc(layer_size[num_layers - 1] * sizeof(double)); probs_pointer = data_ind(num_layers - 1, layer_size); cudaMemcpy(probs, d_data+probs_pointer, layer_size[num_layers - 1]*sizeof(double), cudaMemcpyDeviceToHost); sum = 0.0; for (i = 0; i < layer_size[num_layers - 1]; i++) { sum += exp(probs[i]); } predicted_class_score = -1.0; predicted_class = -1; for (i = 0; i < layer_size[num_layers - 1]; i++) { /* Compute normalized probability of the correct class */ probs[i] = exp(probs[i])/sum; /* Argmax */ if (probs[i] > predicted_class_score) { predicted_class_score = probs[i]; predicted_class = i; } } /* Compute accuracy */ if (predicted_class == label[it]) { acc++; } /* Compute the cross-entropy loss by adding the log probabilities assigned to the correct classes */ global_loss += -log(probs[label[it]]); if ((it+1) % 1000 == 0) { printf("Test set size: %d\tloss: %f\taccuracy:%f\n", it + 1, global_loss / (it + 1), (float) acc / (float) (it + 1)); } } cudaFree(d_data); free(images); } int main(int argc, char **argv) { /* local variables */ int epochs, W = 28, H = 28; const char train_images_filename[] = "./mnist_data/train_images.csv"; const char train_labels_filename[] = "./mnist_data/train_labels.csv"; const char test_images_filename[] = "./mnist_data/test_images.csv"; const char test_labels_filename[] = "./mnist_data/test_labels.csv"; /* Read number of epochs from command line */ if (argc < 2) { printf("Number of epochs is not given \n"); exit(1); } epochs = atoi(argv[1]); /* Set neural network parameters */ double lambda = 0.0005, learning_rate = 0.01; int num_layers = 4; int layer_size[4] = {W*H,300,100,10}; int p_size = param_size(layer_size, num_layers) * sizeof(double); double *h_param, *d_param; double *h_grad_param, *d_grad_param; int *train_label, *test_label; /* Allocate memory and read training labels */ train_label = (int *) malloc(TRAIN_SET_SIZE * sizeof(int)); read_labels(train_labels_filename, train_label, TRAIN_SET_SIZE); /* Allocate memory for parameters */ h_param = (double *) malloc(p_size); h_grad_param = (double *) malloc(p_size); cudaMalloc( (void**)&d_param, p_size); cudaMalloc( (void**)&d_grad_param, p_size); srandom(1); /* Initialize parameters using "xavier" initialization */ int layer, i, j; for (layer = 1; layer < num_layers; layer++) { double a = sqrt(3.0/(float)layer_size[layer-1]); // uniform interval limit int param_pointer = param_ind(layer, layer_size); for (i = 0; i < layer_size[layer]; i++) { h_param[param_pointer + i] = 0.0; // bias } for (j = 1; j < layer_size[layer - 1] +1; j++) { for (i = 0; i < layer_size[layer]; i++) { h_param[param_pointer + j * layer_size[layer] + i] = 2.0 * a * (double) random() / RAND_MAX - a; } } } /* Initialize gradients with 0 */ memset(h_grad_param, 0, p_size); cudaMemcpy( d_param, h_param, p_size, cudaMemcpyHostToDevice ); cudaMemcpy( d_grad_param, h_grad_param, p_size, cudaMemcpyHostToDevice ); /* Train network */ train(train_images_filename, train_label, d_param, d_grad_param, layer_size, num_layers, epochs, lambda, learning_rate); /* Deallocate arrays */ free(h_grad_param); cudaFree(d_grad_param); free(train_label); /* Allocate memory and read testing labels */ test_label = (int *) malloc(TEST_SET_SIZE * sizeof(int)); read_labels(test_labels_filename, test_label, TEST_SET_SIZE); /* Test network */ test(test_images_filename, test_label, d_param, layer_size, num_layers); /* Deallocate arrays */ free(h_param); cudaFree(d_param); free(test_label); return 0; }
/** * CUDA implementation of a fully-connected feed forward neural network * * Authors: Jevgenija Aksjonova (jevaks@kth.se) * Beatrice Ionascu (bionascu@kth.se) * * Last changed: 04/30/2017 */ #include <hip/hip_runtime.h> #include <stdio.h> #include <assert.h> #include <stdlib.h> #include <string.h> #include <math.h> #define TOL 0.00000001 #define BATCH_SIZE 100 #define TRAIN_SET_SIZE 60000 #define TEST_SET_SIZE 10000 #define FUNCTION_TYPE 1 // 0 - tanh, 1 - relu /* Read labels data */ void read_labels(const char filename[], int *array, int data_size) { memset(array, 0, sizeof(int) * data_size); FILE *file = fopen(filename, "r"); if (file) { size_t i; /* Read each line from the file. */ for (i = 0; i < data_size && !feof(file); i++) { fscanf(file, "%d", array + i); } assert(i == data_size); fclose(file); } else { perror(filename); } } /* Read image data */ void read_images(const char *filename, double *array, int data_size, int batch_size, int batch_ind) { FILE *file = fopen(filename, "r"); if (file) { size_t i, j; char buffer[data_size * 4], *ptr; /* Read each line from the file. */ for (i = 0; fgets(buffer, sizeof buffer, file); ++i) { /* Only parse data from desired batch. */ if (i >= batch_ind * batch_size && i < (batch_ind + 1) * batch_size ) { /* Parse the comma-separated values from each line into 'array'. */ for (j = 0, ptr = buffer; j < data_size; j++, ptr++) { array[(i - batch_ind * batch_size) * data_size + j] = strtol(ptr, &ptr, 10) / 256.0; } } } fclose(file); } else { perror(filename); } } /* Compute local number of parameters (includes weights and biases) */ int param_size(int *layer_size, int num_layers) { int size = 0; size_t i; for (i = 1; i < num_layers; i++) { size = size + layer_size[i] * (layer_size[i-1] +1); } return size; } /* Compute total amount of data (includes the input data features and the layer outputs) */ int data_size(int *layer_size, int num_layers) { int size = 0; size_t i; for (i = 0; i < num_layers; i++) { size += layer_size[i]; } return size; } /* Compute local starting index of the parameters corresponding to layer */ int param_ind(int layer, int *layer_size) { int ind = 0; size_t i; for (i = 1; i < layer; i++) { ind += layer_size[i] * (layer_size[i - 1] + 1); } return ind; } /* Compute local starting index of the data corresponding to layer */ int data_ind(int layer, int *layer_size) { int ind = 0; size_t i; for (i = 0; i < layer; i++) { ind += layer_size[i]; } return ind; } /* Forward pass */ __global__ void forward(double *input, int inputSize, double *output, int outputSize, double *param, int fun) { size_t i, j; double l_output; i = blockIdx.x * blockDim.x + threadIdx.x; if (i < outputSize ) { /* Initialize output with the bias term */ l_output = param[i]; for (j = 0; j < inputSize; j++) { // Add weighted inputs l_output += param[(j+1)*outputSize + i] * input[j]; } /* Activation */ if (fun > 0) { if (FUNCTION_TYPE == 0) { l_output = tanh(l_output); } else if (FUNCTION_TYPE == 1) { // relu function if (l_output <= 0) { l_output = 0.0; } } } output[i] = l_output; } } /* Backward pass (backpropagate gradients from current_layer to prev_layer) */ __global__ void backward_part1(double *current_layer, int layer_size, double *prev_layer, int prev_layer_size, double *param, double *grad_param, int layer,double lambda) { size_t i, j; double l_current_layer; i = blockIdx.x * blockDim.x + threadIdx.x; if (i < layer_size) { l_current_layer = current_layer[i]; // Update parameters grad_param[i] += l_current_layer; for (j = 0; j < prev_layer_size; j++) { grad_param[(j+1) * layer_size + i ] += l_current_layer * prev_layer[j] + 2.0 * lambda * param[(j+1) * layer_size + i]; // regularization } } } __global__ void backward_part2(double *current_layer, int layer_size, double *prev_layer, int prev_layer_size, double *param, double *grad_param, int layer,double lambda) { size_t i, k; double l_prev_layer; i = blockIdx.x * blockDim.x + threadIdx.x; if (i < prev_layer_size ) { l_prev_layer = prev_layer[i]; // backprapogate partial derivatives if (FUNCTION_TYPE == 0) { double localGrad = (1-pow(l_prev_layer,2)); // tanh derivative l_prev_layer = 0.0; for (k = 0; k < layer_size; k++) { l_prev_layer += current_layer[k] * param[(i+1) * layer_size + k]; } l_prev_layer *= localGrad; } else if (FUNCTION_TYPE == 1) { if (l_prev_layer >0.0) { l_prev_layer = 0.0; for (k = 0; k < layer_size; k++) { l_prev_layer += current_layer[k] * param[(i+1) * layer_size + k]; } } else { l_prev_layer = 0.0; } } prev_layer[i] = l_prev_layer; } } __global__ void update_param(double *param, double *grad_param, int num_param, double learning_rate, int batch_size) { int i = blockIdx.x * blockDim.x + threadIdx.x; if (i < num_param) { param[i] = param[i] - learning_rate * grad_param[i] / (float)batch_size; } } /* Train the network */ void train(const char filename[], int* label, double *d_param, double *d_grad_param, int *layer_size, int num_layers, int epochs, double lambda, double learning_rate) { size_t img, ep = 0; int i, layer, batch_index; int input_pointer, output_pointer, param_pointer; double sum, global_loss = 1; int image_size = layer_size[0]; int num_param = param_size(layer_size, num_layers); /* Allocate image data array */ double *images; images = (double *) malloc(BATCH_SIZE * image_size * sizeof(double)); /* Allocate data array */ double *d_data ; hipMalloc( (void**)&d_data, data_size(layer_size, num_layers) * sizeof(double)); /* Loop over epochs (one epoch = loop over all images) */ while (ep++ < epochs && global_loss > TOL) { /* Loop over batches */ for (batch_index = 0; batch_index < TRAIN_SET_SIZE / BATCH_SIZE; batch_index++) { /* Initialize loss */ global_loss = 0.0; /* Reset gradients to 0 */ hipMemset(d_grad_param, 0.0, num_param * sizeof(double)); /* Load images in batch */ read_images(filename, images, image_size, BATCH_SIZE, batch_index); /* Loop over images in batch */ for (img = 0; img < BATCH_SIZE; img++) { /* Copy one image to the 1st layer of the data */ hipMemcpy(d_data, images + img * image_size, image_size * sizeof(double), hipMemcpyHostToDevice ); /* Forward pass */ for (layer = 1; layer < num_layers; layer++) { input_pointer = data_ind(layer - 1, layer_size); output_pointer = data_ind(layer, layer_size); param_pointer = param_ind(layer, layer_size); dim3 dimBlock( 128,1,1 ); dim3 dimGrid(ceil((float)layer_size[layer]/128.0),1,1); forward<<<dimGrid, dimBlock>>>(d_data + input_pointer, layer_size[layer - 1], d_data + output_pointer, layer_size[layer], d_param + param_pointer, num_layers - layer - 1); } double *probs; probs = (double * ) malloc(layer_size[num_layers - 1]* sizeof(double)); int probs_pointer = data_ind(num_layers - 1, layer_size); hipMemcpy(probs, d_data + probs_pointer, layer_size[num_layers - 1] * sizeof(double), hipMemcpyDeviceToHost); /* Softmax */ sum = 0.0; for ( i = 0; i < layer_size[num_layers - 1]; i++) { sum += exp(probs[i]); } for ( i = 0; i < layer_size[num_layers - 1]; i++) { probs[i] = exp(probs[i]) / sum; } /* Loss computation */ /* Compute the cross-entropy loss by adding the log probabilities assigned to the correct classes */ global_loss += -log(probs[label[batch_index * BATCH_SIZE + img]]); /* Gradient computation for the last (score) layer */ probs[label[batch_index * BATCH_SIZE + img]] -= 1.0; hipMemcpy(d_data+probs_pointer, probs, layer_size[num_layers - 1]*sizeof(double), hipMemcpyHostToDevice); /* Backpropagation */ for (layer = num_layers - 1; layer > 0; layer--) { input_pointer = data_ind(layer, layer_size); output_pointer = data_ind(layer - 1, layer_size); param_pointer = param_ind(layer, layer_size); dim3 dimBlock( 128, 1, 1 ); dim3 dimGrid( ceil((float)layer_size[layer]/128.0), 1, 1 ); backward_part1<<<dimGrid, dimBlock>>>(d_data + input_pointer, layer_size[layer], d_data + output_pointer, layer_size[layer - 1], d_param + param_pointer, d_grad_param + param_pointer, layer, lambda); if (layer > 1) { dim3 dimBlock( 128, 1, 1 ); dim3 dimGrid( ceil((float)layer_size[layer - 1]/128.0), 1, 1 ); backward_part2<<<dimGrid, dimBlock>>>(d_data + input_pointer, layer_size[layer], d_data + output_pointer, layer_size[layer - 1], d_param + param_pointer, d_grad_param + param_pointer, layer, lambda); } } } /* Update parameters using gradient averaged over batch)*/ dim3 dimBlock(512, 1, 1 ); dim3 dimGrid( ceil((float)num_param/512.0), 1 ,1 ); update_param<<<dimGrid, dimBlock>>>(d_param, d_grad_param, num_param, learning_rate, BATCH_SIZE); /* Compute global loss averaged over batch */ global_loss /= BATCH_SIZE; if (batch_index % 100 == 0) { printf("Ep: %zu/%d\t batch:%d \t train loss: %f\n", ep, epochs, batch_index, global_loss); } } } hipFree(d_data); free(images); } /* Test the network */ void test(const char filename[], int* label, double *d_param, int *layer_size, int num_layers) { int it, layer, i; int input_pointer, output_pointer, param_pointer, probs_pointer, predicted_class; double sum, predicted_class_score, global_loss = 0.0; int image_size = layer_size[0]; int acc = 0; /* Allocate image data array */ double *images; images = (double *) malloc(image_size* sizeof(double)); /* Allocate data array */ double *d_data ; hipMalloc( (void**)&d_data, data_size(layer_size,num_layers)*sizeof(double)); /* Loop over batches */ for (it = 0; it < TEST_SET_SIZE; it++) { /* Load image data */ read_images(filename, images, image_size, 1, it); /* Copy one image to the 1st layer of the data */ hipMemcpy(d_data, images, image_size * sizeof(double),hipMemcpyHostToDevice); /* Forward pass */ for (layer = 1; layer < num_layers; layer++) { input_pointer = data_ind(layer - 1, layer_size); output_pointer = data_ind(layer, layer_size); param_pointer = param_ind(layer, layer_size); dim3 dimBlock( 128, 1, 1 ); dim3 dimGrid( ceil((float)layer_size[layer]/128.0), 1, 1 ); forward<<<dimGrid, dimBlock>>>(d_data + input_pointer, layer_size[layer - 1], d_data + output_pointer, layer_size[layer], d_param + param_pointer, num_layers - layer - 1); } /* Softmax and Predict class */ double *probs; probs = (double * ) malloc(layer_size[num_layers - 1] * sizeof(double)); probs_pointer = data_ind(num_layers - 1, layer_size); hipMemcpy(probs, d_data+probs_pointer, layer_size[num_layers - 1]*sizeof(double), hipMemcpyDeviceToHost); sum = 0.0; for (i = 0; i < layer_size[num_layers - 1]; i++) { sum += exp(probs[i]); } predicted_class_score = -1.0; predicted_class = -1; for (i = 0; i < layer_size[num_layers - 1]; i++) { /* Compute normalized probability of the correct class */ probs[i] = exp(probs[i])/sum; /* Argmax */ if (probs[i] > predicted_class_score) { predicted_class_score = probs[i]; predicted_class = i; } } /* Compute accuracy */ if (predicted_class == label[it]) { acc++; } /* Compute the cross-entropy loss by adding the log probabilities assigned to the correct classes */ global_loss += -log(probs[label[it]]); if ((it+1) % 1000 == 0) { printf("Test set size: %d\tloss: %f\taccuracy:%f\n", it + 1, global_loss / (it + 1), (float) acc / (float) (it + 1)); } } hipFree(d_data); free(images); } int main(int argc, char **argv) { /* local variables */ int epochs, W = 28, H = 28; const char train_images_filename[] = "./mnist_data/train_images.csv"; const char train_labels_filename[] = "./mnist_data/train_labels.csv"; const char test_images_filename[] = "./mnist_data/test_images.csv"; const char test_labels_filename[] = "./mnist_data/test_labels.csv"; /* Read number of epochs from command line */ if (argc < 2) { printf("Number of epochs is not given \n"); exit(1); } epochs = atoi(argv[1]); /* Set neural network parameters */ double lambda = 0.0005, learning_rate = 0.01; int num_layers = 4; int layer_size[4] = {W*H,300,100,10}; int p_size = param_size(layer_size, num_layers) * sizeof(double); double *h_param, *d_param; double *h_grad_param, *d_grad_param; int *train_label, *test_label; /* Allocate memory and read training labels */ train_label = (int *) malloc(TRAIN_SET_SIZE * sizeof(int)); read_labels(train_labels_filename, train_label, TRAIN_SET_SIZE); /* Allocate memory for parameters */ h_param = (double *) malloc(p_size); h_grad_param = (double *) malloc(p_size); hipMalloc( (void**)&d_param, p_size); hipMalloc( (void**)&d_grad_param, p_size); srandom(1); /* Initialize parameters using "xavier" initialization */ int layer, i, j; for (layer = 1; layer < num_layers; layer++) { double a = sqrt(3.0/(float)layer_size[layer-1]); // uniform interval limit int param_pointer = param_ind(layer, layer_size); for (i = 0; i < layer_size[layer]; i++) { h_param[param_pointer + i] = 0.0; // bias } for (j = 1; j < layer_size[layer - 1] +1; j++) { for (i = 0; i < layer_size[layer]; i++) { h_param[param_pointer + j * layer_size[layer] + i] = 2.0 * a * (double) random() / RAND_MAX - a; } } } /* Initialize gradients with 0 */ memset(h_grad_param, 0, p_size); hipMemcpy( d_param, h_param, p_size, hipMemcpyHostToDevice ); hipMemcpy( d_grad_param, h_grad_param, p_size, hipMemcpyHostToDevice ); /* Train network */ train(train_images_filename, train_label, d_param, d_grad_param, layer_size, num_layers, epochs, lambda, learning_rate); /* Deallocate arrays */ free(h_grad_param); hipFree(d_grad_param); free(train_label); /* Allocate memory and read testing labels */ test_label = (int *) malloc(TEST_SET_SIZE * sizeof(int)); read_labels(test_labels_filename, test_label, TEST_SET_SIZE); /* Test network */ test(test_images_filename, test_label, d_param, layer_size, num_layers); /* Deallocate arrays */ free(h_param); hipFree(d_param); free(test_label); return 0; }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
/** * CUDA implementation of a fully-connected feed forward neural network * * Authors: Jevgenija Aksjonova (jevaks@kth.se) * Beatrice Ionascu (bionascu@kth.se) * * Last changed: 04/30/2017 */ #include <hip/hip_runtime.h> #include <stdio.h> #include <assert.h> #include <stdlib.h> #include <string.h> #include <math.h> #define TOL 0.00000001 #define BATCH_SIZE 100 #define TRAIN_SET_SIZE 60000 #define TEST_SET_SIZE 10000 #define FUNCTION_TYPE 1 // 0 - tanh, 1 - relu /* Read labels data */ void read_labels(const char filename[], int *array, int data_size) { memset(array, 0, sizeof(int) * data_size); FILE *file = fopen(filename, "r"); if (file) { size_t i; /* Read each line from the file. */ for (i = 0; i < data_size && !feof(file); i++) { fscanf(file, "%d", array + i); } assert(i == data_size); fclose(file); } else { perror(filename); } } /* Read image data */ void read_images(const char *filename, double *array, int data_size, int batch_size, int batch_ind) { FILE *file = fopen(filename, "r"); if (file) { size_t i, j; char buffer[data_size * 4], *ptr; /* Read each line from the file. */ for (i = 0; fgets(buffer, sizeof buffer, file); ++i) { /* Only parse data from desired batch. */ if (i >= batch_ind * batch_size && i < (batch_ind + 1) * batch_size ) { /* Parse the comma-separated values from each line into 'array'. */ for (j = 0, ptr = buffer; j < data_size; j++, ptr++) { array[(i - batch_ind * batch_size) * data_size + j] = strtol(ptr, &ptr, 10) / 256.0; } } } fclose(file); } else { perror(filename); } } /* Compute local number of parameters (includes weights and biases) */ int param_size(int *layer_size, int num_layers) { int size = 0; size_t i; for (i = 1; i < num_layers; i++) { size = size + layer_size[i] * (layer_size[i-1] +1); } return size; } /* Compute total amount of data (includes the input data features and the layer outputs) */ int data_size(int *layer_size, int num_layers) { int size = 0; size_t i; for (i = 0; i < num_layers; i++) { size += layer_size[i]; } return size; } /* Compute local starting index of the parameters corresponding to layer */ int param_ind(int layer, int *layer_size) { int ind = 0; size_t i; for (i = 1; i < layer; i++) { ind += layer_size[i] * (layer_size[i - 1] + 1); } return ind; } /* Compute local starting index of the data corresponding to layer */ int data_ind(int layer, int *layer_size) { int ind = 0; size_t i; for (i = 0; i < layer; i++) { ind += layer_size[i]; } return ind; } /* Forward pass */ __global__ void forward(double *input, int inputSize, double *output, int outputSize, double *param, int fun) { size_t i, j; double l_output; i = blockIdx.x * blockDim.x + threadIdx.x; if (i < outputSize ) { /* Initialize output with the bias term */ l_output = param[i]; for (j = 0; j < inputSize; j++) { // Add weighted inputs l_output += param[(j+1)*outputSize + i] * input[j]; } /* Activation */ if (fun > 0) { if (FUNCTION_TYPE == 0) { l_output = tanh(l_output); } else if (FUNCTION_TYPE == 1) { // relu function if (l_output <= 0) { l_output = 0.0; } } } output[i] = l_output; } } /* Backward pass (backpropagate gradients from current_layer to prev_layer) */ __global__ void backward_part1(double *current_layer, int layer_size, double *prev_layer, int prev_layer_size, double *param, double *grad_param, int layer,double lambda) { size_t i, j; double l_current_layer; i = blockIdx.x * blockDim.x + threadIdx.x; if (i < layer_size) { l_current_layer = current_layer[i]; // Update parameters grad_param[i] += l_current_layer; for (j = 0; j < prev_layer_size; j++) { grad_param[(j+1) * layer_size + i ] += l_current_layer * prev_layer[j] + 2.0 * lambda * param[(j+1) * layer_size + i]; // regularization } } } __global__ void backward_part2(double *current_layer, int layer_size, double *prev_layer, int prev_layer_size, double *param, double *grad_param, int layer,double lambda) { size_t i, k; double l_prev_layer; i = blockIdx.x * blockDim.x + threadIdx.x; if (i < prev_layer_size ) { l_prev_layer = prev_layer[i]; // backprapogate partial derivatives if (FUNCTION_TYPE == 0) { double localGrad = (1-pow(l_prev_layer,2)); // tanh derivative l_prev_layer = 0.0; for (k = 0; k < layer_size; k++) { l_prev_layer += current_layer[k] * param[(i+1) * layer_size + k]; } l_prev_layer *= localGrad; } else if (FUNCTION_TYPE == 1) { if (l_prev_layer >0.0) { l_prev_layer = 0.0; for (k = 0; k < layer_size; k++) { l_prev_layer += current_layer[k] * param[(i+1) * layer_size + k]; } } else { l_prev_layer = 0.0; } } prev_layer[i] = l_prev_layer; } } __global__ void update_param(double *param, double *grad_param, int num_param, double learning_rate, int batch_size) { int i = blockIdx.x * blockDim.x + threadIdx.x; if (i < num_param) { param[i] = param[i] - learning_rate * grad_param[i] / (float)batch_size; } } /* Train the network */ void train(const char filename[], int* label, double *d_param, double *d_grad_param, int *layer_size, int num_layers, int epochs, double lambda, double learning_rate) { size_t img, ep = 0; int i, layer, batch_index; int input_pointer, output_pointer, param_pointer; double sum, global_loss = 1; int image_size = layer_size[0]; int num_param = param_size(layer_size, num_layers); /* Allocate image data array */ double *images; images = (double *) malloc(BATCH_SIZE * image_size * sizeof(double)); /* Allocate data array */ double *d_data ; hipMalloc( (void**)&d_data, data_size(layer_size, num_layers) * sizeof(double)); /* Loop over epochs (one epoch = loop over all images) */ while (ep++ < epochs && global_loss > TOL) { /* Loop over batches */ for (batch_index = 0; batch_index < TRAIN_SET_SIZE / BATCH_SIZE; batch_index++) { /* Initialize loss */ global_loss = 0.0; /* Reset gradients to 0 */ hipMemset(d_grad_param, 0.0, num_param * sizeof(double)); /* Load images in batch */ read_images(filename, images, image_size, BATCH_SIZE, batch_index); /* Loop over images in batch */ for (img = 0; img < BATCH_SIZE; img++) { /* Copy one image to the 1st layer of the data */ hipMemcpy(d_data, images + img * image_size, image_size * sizeof(double), hipMemcpyHostToDevice ); /* Forward pass */ for (layer = 1; layer < num_layers; layer++) { input_pointer = data_ind(layer - 1, layer_size); output_pointer = data_ind(layer, layer_size); param_pointer = param_ind(layer, layer_size); dim3 dimBlock( 128,1,1 ); dim3 dimGrid(ceil((float)layer_size[layer]/128.0),1,1); forward<<<dimGrid, dimBlock>>>(d_data + input_pointer, layer_size[layer - 1], d_data + output_pointer, layer_size[layer], d_param + param_pointer, num_layers - layer - 1); } double *probs; probs = (double * ) malloc(layer_size[num_layers - 1]* sizeof(double)); int probs_pointer = data_ind(num_layers - 1, layer_size); hipMemcpy(probs, d_data + probs_pointer, layer_size[num_layers - 1] * sizeof(double), hipMemcpyDeviceToHost); /* Softmax */ sum = 0.0; for ( i = 0; i < layer_size[num_layers - 1]; i++) { sum += exp(probs[i]); } for ( i = 0; i < layer_size[num_layers - 1]; i++) { probs[i] = exp(probs[i]) / sum; } /* Loss computation */ /* Compute the cross-entropy loss by adding the log probabilities assigned to the correct classes */ global_loss += -log(probs[label[batch_index * BATCH_SIZE + img]]); /* Gradient computation for the last (score) layer */ probs[label[batch_index * BATCH_SIZE + img]] -= 1.0; hipMemcpy(d_data+probs_pointer, probs, layer_size[num_layers - 1]*sizeof(double), hipMemcpyHostToDevice); /* Backpropagation */ for (layer = num_layers - 1; layer > 0; layer--) { input_pointer = data_ind(layer, layer_size); output_pointer = data_ind(layer - 1, layer_size); param_pointer = param_ind(layer, layer_size); dim3 dimBlock( 128, 1, 1 ); dim3 dimGrid( ceil((float)layer_size[layer]/128.0), 1, 1 ); backward_part1<<<dimGrid, dimBlock>>>(d_data + input_pointer, layer_size[layer], d_data + output_pointer, layer_size[layer - 1], d_param + param_pointer, d_grad_param + param_pointer, layer, lambda); if (layer > 1) { dim3 dimBlock( 128, 1, 1 ); dim3 dimGrid( ceil((float)layer_size[layer - 1]/128.0), 1, 1 ); backward_part2<<<dimGrid, dimBlock>>>(d_data + input_pointer, layer_size[layer], d_data + output_pointer, layer_size[layer - 1], d_param + param_pointer, d_grad_param + param_pointer, layer, lambda); } } } /* Update parameters using gradient averaged over batch)*/ dim3 dimBlock(512, 1, 1 ); dim3 dimGrid( ceil((float)num_param/512.0), 1 ,1 ); update_param<<<dimGrid, dimBlock>>>(d_param, d_grad_param, num_param, learning_rate, BATCH_SIZE); /* Compute global loss averaged over batch */ global_loss /= BATCH_SIZE; if (batch_index % 100 == 0) { printf("Ep: %zu/%d\t batch:%d \t train loss: %f\n", ep, epochs, batch_index, global_loss); } } } hipFree(d_data); free(images); } /* Test the network */ void test(const char filename[], int* label, double *d_param, int *layer_size, int num_layers) { int it, layer, i; int input_pointer, output_pointer, param_pointer, probs_pointer, predicted_class; double sum, predicted_class_score, global_loss = 0.0; int image_size = layer_size[0]; int acc = 0; /* Allocate image data array */ double *images; images = (double *) malloc(image_size* sizeof(double)); /* Allocate data array */ double *d_data ; hipMalloc( (void**)&d_data, data_size(layer_size,num_layers)*sizeof(double)); /* Loop over batches */ for (it = 0; it < TEST_SET_SIZE; it++) { /* Load image data */ read_images(filename, images, image_size, 1, it); /* Copy one image to the 1st layer of the data */ hipMemcpy(d_data, images, image_size * sizeof(double),hipMemcpyHostToDevice); /* Forward pass */ for (layer = 1; layer < num_layers; layer++) { input_pointer = data_ind(layer - 1, layer_size); output_pointer = data_ind(layer, layer_size); param_pointer = param_ind(layer, layer_size); dim3 dimBlock( 128, 1, 1 ); dim3 dimGrid( ceil((float)layer_size[layer]/128.0), 1, 1 ); forward<<<dimGrid, dimBlock>>>(d_data + input_pointer, layer_size[layer - 1], d_data + output_pointer, layer_size[layer], d_param + param_pointer, num_layers - layer - 1); } /* Softmax and Predict class */ double *probs; probs = (double * ) malloc(layer_size[num_layers - 1] * sizeof(double)); probs_pointer = data_ind(num_layers - 1, layer_size); hipMemcpy(probs, d_data+probs_pointer, layer_size[num_layers - 1]*sizeof(double), hipMemcpyDeviceToHost); sum = 0.0; for (i = 0; i < layer_size[num_layers - 1]; i++) { sum += exp(probs[i]); } predicted_class_score = -1.0; predicted_class = -1; for (i = 0; i < layer_size[num_layers - 1]; i++) { /* Compute normalized probability of the correct class */ probs[i] = exp(probs[i])/sum; /* Argmax */ if (probs[i] > predicted_class_score) { predicted_class_score = probs[i]; predicted_class = i; } } /* Compute accuracy */ if (predicted_class == label[it]) { acc++; } /* Compute the cross-entropy loss by adding the log probabilities assigned to the correct classes */ global_loss += -log(probs[label[it]]); if ((it+1) % 1000 == 0) { printf("Test set size: %d\tloss: %f\taccuracy:%f\n", it + 1, global_loss / (it + 1), (float) acc / (float) (it + 1)); } } hipFree(d_data); free(images); } int main(int argc, char **argv) { /* local variables */ int epochs, W = 28, H = 28; const char train_images_filename[] = "./mnist_data/train_images.csv"; const char train_labels_filename[] = "./mnist_data/train_labels.csv"; const char test_images_filename[] = "./mnist_data/test_images.csv"; const char test_labels_filename[] = "./mnist_data/test_labels.csv"; /* Read number of epochs from command line */ if (argc < 2) { printf("Number of epochs is not given \n"); exit(1); } epochs = atoi(argv[1]); /* Set neural network parameters */ double lambda = 0.0005, learning_rate = 0.01; int num_layers = 4; int layer_size[4] = {W*H,300,100,10}; int p_size = param_size(layer_size, num_layers) * sizeof(double); double *h_param, *d_param; double *h_grad_param, *d_grad_param; int *train_label, *test_label; /* Allocate memory and read training labels */ train_label = (int *) malloc(TRAIN_SET_SIZE * sizeof(int)); read_labels(train_labels_filename, train_label, TRAIN_SET_SIZE); /* Allocate memory for parameters */ h_param = (double *) malloc(p_size); h_grad_param = (double *) malloc(p_size); hipMalloc( (void**)&d_param, p_size); hipMalloc( (void**)&d_grad_param, p_size); srandom(1); /* Initialize parameters using "xavier" initialization */ int layer, i, j; for (layer = 1; layer < num_layers; layer++) { double a = sqrt(3.0/(float)layer_size[layer-1]); // uniform interval limit int param_pointer = param_ind(layer, layer_size); for (i = 0; i < layer_size[layer]; i++) { h_param[param_pointer + i] = 0.0; // bias } for (j = 1; j < layer_size[layer - 1] +1; j++) { for (i = 0; i < layer_size[layer]; i++) { h_param[param_pointer + j * layer_size[layer] + i] = 2.0 * a * (double) random() / RAND_MAX - a; } } } /* Initialize gradients with 0 */ memset(h_grad_param, 0, p_size); hipMemcpy( d_param, h_param, p_size, hipMemcpyHostToDevice ); hipMemcpy( d_grad_param, h_grad_param, p_size, hipMemcpyHostToDevice ); /* Train network */ train(train_images_filename, train_label, d_param, d_grad_param, layer_size, num_layers, epochs, lambda, learning_rate); /* Deallocate arrays */ free(h_grad_param); hipFree(d_grad_param); free(train_label); /* Allocate memory and read testing labels */ test_label = (int *) malloc(TEST_SET_SIZE * sizeof(int)); read_labels(test_labels_filename, test_label, TEST_SET_SIZE); /* Test network */ test(test_images_filename, test_label, d_param, layer_size, num_layers); /* Deallocate arrays */ free(h_param); hipFree(d_param); free(test_label); return 0; }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z7forwardPdiS_iS_i .globl _Z7forwardPdiS_iS_i .p2align 8 .type _Z7forwardPdiS_iS_i,@function _Z7forwardPdiS_iS_i: s_clause 0x1 s_load_b32 s2, s[0:1], 0x3c s_load_b32 s6, s[0:1], 0x18 s_waitcnt lgkmcnt(0) s_and_b32 s2, s2, 0xffff s_ashr_i32 s7, s6, 31 v_mad_u64_u32 v[1:2], null, s15, s2, v[0:1] v_mov_b32_e32 v2, 0 s_mov_b32 s2, exec_lo s_delay_alu instid0(VALU_DEP_1) v_cmpx_gt_u64_e64 s[6:7], v[1:2] s_cbranch_execz .LBB0_5 s_clause 0x1 s_load_b64 s[8:9], s[0:1], 0x20 s_load_b32 s2, s[0:1], 0x8 v_lshlrev_b64 v[5:6], 3, v[1:2] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v3, vcc_lo, s8, v5 v_add_co_ci_u32_e32 v4, vcc_lo, s9, v6, vcc_lo s_cmp_eq_u32 s2, 0 global_load_b64 v[3:4], v[3:4], off s_cbranch_scc1 .LBB0_4 s_load_b64 s[4:5], s[0:1], 0x0 s_lshl_b64 s[6:7], s[6:7], 3 s_ashr_i32 s3, s2, 31 s_add_u32 s8, s8, s6 s_addc_u32 s9, s9, s7 v_add_co_u32 v5, vcc_lo, s8, v5 v_add_co_ci_u32_e32 v6, vcc_lo, s9, v6, vcc_lo .LBB0_3: global_load_b64 v[7:8], v[5:6], off s_waitcnt lgkmcnt(0) s_load_b64 s[8:9], s[4:5], 0x0 v_add_co_u32 v5, vcc_lo, v5, s6 s_add_u32 s2, s2, -1 v_add_co_ci_u32_e32 v6, vcc_lo, s7, v6, vcc_lo s_addc_u32 s3, s3, -1 s_add_u32 s4, s4, 8 s_addc_u32 s5, s5, 0 s_cmp_lg_u64 s[2:3], 0 s_waitcnt vmcnt(0) lgkmcnt(0) v_fma_f64 v[3:4], v[7:8], s[8:9], v[3:4] s_cbranch_scc1 .LBB0_3 .LBB0_4: s_waitcnt vmcnt(0) s_delay_alu instid0(VALU_DEP_1) v_cmp_ge_f64_e32 vcc_lo, 0, v[3:4] s_clause 0x1 s_load_b32 s2, s[0:1], 0x28 s_load_b64 s[0:1], s[0:1], 0x10 v_lshlrev_b64 v[0:1], 3, v[1:2] s_waitcnt lgkmcnt(0) s_cmp_gt_i32 s2, 0 s_cselect_b32 s2, -1, 0 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1) s_and_b32 s2, s2, vcc_lo v_add_co_u32 v0, vcc_lo, s0, v0 v_cndmask_b32_e64 v4, v4, 0, s2 v_cndmask_b32_e64 v3, v3, 0, s2 v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo global_store_b64 v[0:1], v[3:4], off .LBB0_5: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z7forwardPdiS_iS_i .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 304 .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 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 _Z7forwardPdiS_iS_i, .Lfunc_end0-_Z7forwardPdiS_iS_i .section .AMDGPU.csdata,"",@progbits .text .protected _Z14backward_part1PdiS_iS_S_id .globl _Z14backward_part1PdiS_iS_S_id .p2align 8 .type _Z14backward_part1PdiS_iS_S_id,@function _Z14backward_part1PdiS_iS_S_id: s_clause 0x1 s_load_b32 s2, s[0:1], 0x4c s_load_b32 s6, s[0:1], 0x8 s_waitcnt lgkmcnt(0) s_and_b32 s2, s2, 0xffff s_ashr_i32 s7, s6, 31 v_mad_u64_u32 v[4:5], null, s15, s2, v[0:1] v_mov_b32_e32 v5, 0 s_mov_b32 s2, exec_lo s_delay_alu instid0(VALU_DEP_1) v_cmpx_gt_u64_e64 s[6:7], v[4:5] s_cbranch_execz .LBB1_4 s_clause 0x1 s_load_b64 s[8:9], s[0:1], 0x0 s_load_b64 s[2:3], s[0:1], 0x28 v_lshlrev_b64 v[0:1], 3, v[4:5] s_load_b32 s4, s[0:1], 0x18 s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v2, vcc_lo, s8, v0 v_add_co_ci_u32_e32 v3, vcc_lo, s9, v1, vcc_lo v_add_co_u32 v6, vcc_lo, s2, v0 v_add_co_ci_u32_e32 v7, vcc_lo, s3, v1, vcc_lo s_cmp_eq_u32 s4, 0 global_load_b64 v[0:1], v[2:3], off global_load_b64 v[2:3], v[6:7], off s_waitcnt vmcnt(0) v_add_f64 v[2:3], v[0:1], v[2:3] global_store_b64 v[6:7], v[2:3], off s_cbranch_scc1 .LBB1_4 s_load_b64 s[8:9], s[0:1], 0x38 v_add_co_u32 v4, vcc_lo, s6, v4 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v5, vcc_lo v_mov_b32_e32 v6, 0 s_ashr_i32 s5, s4, 31 s_lshl_b64 s[6:7], s[6:7], 3 s_delay_alu instid0(VALU_DEP_2) v_lshlrev_b64 v[4:5], 3, v[4:5] s_waitcnt lgkmcnt(0) v_add_f64 v[2:3], s[8:9], s[8:9] s_clause 0x1 s_load_b64 s[8:9], s[0:1], 0x10 s_load_b64 s[0:1], s[0:1], 0x20 .p2align 6 .LBB1_3: s_waitcnt lgkmcnt(0) v_add_co_u32 v7, vcc_lo, s0, v4 v_add_co_ci_u32_e32 v8, vcc_lo, s1, v5, vcc_lo v_add_co_u32 v11, vcc_lo, s2, v4 v_add_co_ci_u32_e32 v12, vcc_lo, s3, v5, vcc_lo global_load_b64 v[7:8], v[7:8], off global_load_b64 v[9:10], v6, s[8:9] v_add_co_u32 v4, vcc_lo, v4, s6 global_load_b64 v[13:14], v[11:12], off s_add_u32 s4, s4, -1 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v5, vcc_lo s_addc_u32 s5, s5, -1 s_add_u32 s8, s8, 8 s_addc_u32 s9, s9, 0 s_cmp_lg_u64 s[4:5], 0 s_waitcnt vmcnt(2) v_mul_f64 v[7:8], v[2:3], v[7:8] s_waitcnt vmcnt(1) s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_fma_f64 v[7:8], v[0:1], v[9:10], v[7:8] s_waitcnt vmcnt(0) v_add_f64 v[7:8], v[13:14], v[7:8] global_store_b64 v[11:12], v[7:8], off s_cbranch_scc1 .LBB1_3 .LBB1_4: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z14backward_part1PdiS_iS_S_id .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 320 .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_end1: .size _Z14backward_part1PdiS_iS_S_id, .Lfunc_end1-_Z14backward_part1PdiS_iS_S_id .section .AMDGPU.csdata,"",@progbits .text .protected _Z14backward_part2PdiS_iS_S_id .globl _Z14backward_part2PdiS_iS_S_id .p2align 8 .type _Z14backward_part2PdiS_iS_S_id,@function _Z14backward_part2PdiS_iS_S_id: s_clause 0x1 s_load_b32 s3, s[0:1], 0x4c s_load_b32 s2, s[0:1], 0x18 s_waitcnt lgkmcnt(0) s_and_b32 s3, s3, 0xffff s_delay_alu instid0(SALU_CYCLE_1) v_mad_u64_u32 v[4:5], null, s15, s3, v[0:1] v_mov_b32_e32 v5, 0 s_ashr_i32 s3, s2, 31 s_delay_alu instid0(VALU_DEP_1) | instid1(SALU_CYCLE_1) v_cmp_gt_u64_e32 vcc_lo, s[2:3], v[4:5] s_and_saveexec_b32 s2, vcc_lo s_cbranch_execz .LBB2_7 s_load_b64 s[2:3], s[0:1], 0x10 v_lshlrev_b64 v[0:1], 3, v[4:5] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v0, vcc_lo, s2, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s3, v1, vcc_lo global_load_b64 v[2:3], v[0:1], off s_waitcnt vmcnt(0) v_cmp_lt_f64_e32 vcc_lo, 0, v[2:3] v_mov_b32_e32 v2, 0 v_mov_b32_e32 v3, 0 s_and_saveexec_b32 s4, vcc_lo s_cbranch_execz .LBB2_6 s_load_b32 s2, s[0:1], 0x8 s_waitcnt lgkmcnt(0) s_cmp_eq_u32 s2, 0 s_cbranch_scc1 .LBB2_5 s_ashr_i32 s3, s2, 31 s_load_b64 s[6:7], s[0:1], 0x20 v_mad_u64_u32 v[2:3], null, s2, v4, s[2:3] s_load_b64 s[0:1], s[0:1], 0x0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mad_u64_u32 v[5:6], null, s3, v4, v[3:4] v_mov_b32_e32 v3, v5 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_3) | instid1(VALU_DEP_3) v_lshlrev_b64 v[4:5], 3, v[2:3] v_mov_b32_e32 v2, 0 v_mov_b32_e32 v3, 0 s_waitcnt lgkmcnt(0) v_add_co_u32 v4, vcc_lo, s6, v4 s_delay_alu instid0(VALU_DEP_4) v_add_co_ci_u32_e32 v5, vcc_lo, s7, v5, vcc_lo .LBB2_4: global_load_b64 v[6:7], v[4:5], off s_load_b64 s[6:7], s[0:1], 0x0 v_add_co_u32 v4, vcc_lo, v4, 8 s_add_u32 s2, s2, -1 v_add_co_ci_u32_e32 v5, vcc_lo, 0, v5, vcc_lo s_addc_u32 s3, s3, -1 s_add_u32 s0, s0, 8 s_addc_u32 s1, s1, 0 s_cmp_lg_u64 s[2:3], 0 s_waitcnt vmcnt(0) lgkmcnt(0) v_fma_f64 v[2:3], s[6:7], v[6:7], v[2:3] s_cbranch_scc1 .LBB2_4 s_branch .LBB2_6 .LBB2_5: v_mov_b32_e32 v2, 0 v_mov_b32_e32 v3, 0 .LBB2_6: s_or_b32 exec_lo, exec_lo, s4 global_store_b64 v[0:1], v[2:3], off .LBB2_7: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z14backward_part2PdiS_iS_S_id .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 320 .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 8 .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_end2: .size _Z14backward_part2PdiS_iS_S_id, .Lfunc_end2-_Z14backward_part2PdiS_iS_S_id .section .AMDGPU.csdata,"",@progbits .text .protected _Z12update_paramPdS_idi .globl _Z12update_paramPdS_idi .p2align 8 .type _Z12update_paramPdS_idi,@function _Z12update_paramPdS_idi: s_clause 0x1 s_load_b32 s2, s[0:1], 0x34 s_load_b32 s3, s[0:1], 0x10 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_e64 s3, v1 s_cbranch_execz .LBB3_2 s_clause 0x1 s_load_b128 s[4:7], s[0:1], 0x0 s_load_b64 s[2:3], s[0:1], 0x18 v_ashrrev_i32_e32 v2, 31, v1 s_load_b32 s0, s[0:1], 0x20 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[0:1], 3, v[1:2] s_waitcnt lgkmcnt(0) v_add_co_u32 v2, vcc_lo, s6, v0 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v3, vcc_lo, s7, v1, vcc_lo v_cvt_f32_i32_e32 v4, s0 v_add_co_u32 v0, vcc_lo, s4, v0 global_load_b64 v[2:3], v[2:3], off v_add_co_ci_u32_e32 v1, vcc_lo, s5, v1, vcc_lo v_cvt_f64_f32_e32 v[4:5], v4 global_load_b64 v[10:11], v[0:1], off s_waitcnt vmcnt(1) v_mul_f64 v[2:3], v[2:3], s[2:3] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_div_scale_f64 v[6:7], null, v[4:5], v[4:5], v[2:3] v_rcp_f64_e32 v[8:9], v[6:7] s_waitcnt_depctr 0xfff v_fma_f64 v[12:13], -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[8:9], v[12:13], v[8:9] v_fma_f64 v[12:13], -v[6:7], v[8:9], 1.0 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_fma_f64 v[8:9], v[8:9], v[12:13], v[8:9] v_div_scale_f64 v[12:13], vcc_lo, v[2:3], v[4:5], v[2:3] v_mul_f64 v[14:15], v[12:13], v[8:9] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f64 v[6:7], -v[6:7], v[14:15], v[12:13] v_div_fmas_f64 v[6:7], v[6:7], v[8:9], v[14:15] s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_div_fixup_f64 v[2:3], v[6:7], v[4:5], v[2:3] s_waitcnt vmcnt(0) v_add_f64 v[2:3], v[10:11], -v[2:3] global_store_b64 v[0:1], v[2:3], off .LBB3_2: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z12update_paramPdS_idi .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 296 .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 16 .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_end3: .size _Z12update_paramPdS_idi, .Lfunc_end3-_Z12update_paramPdS_idi .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 - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .offset: 24 .size: 4 .value_kind: by_value - .address_space: global .offset: 32 .size: 8 .value_kind: global_buffer - .offset: 40 .size: 4 .value_kind: by_value - .offset: 48 .size: 4 .value_kind: hidden_block_count_x - .offset: 52 .size: 4 .value_kind: hidden_block_count_y - .offset: 56 .size: 4 .value_kind: hidden_block_count_z - .offset: 60 .size: 2 .value_kind: hidden_group_size_x - .offset: 62 .size: 2 .value_kind: hidden_group_size_y - .offset: 64 .size: 2 .value_kind: hidden_group_size_z - .offset: 66 .size: 2 .value_kind: hidden_remainder_x - .offset: 68 .size: 2 .value_kind: hidden_remainder_y - .offset: 70 .size: 2 .value_kind: hidden_remainder_z - .offset: 88 .size: 8 .value_kind: hidden_global_offset_x - .offset: 96 .size: 8 .value_kind: hidden_global_offset_y - .offset: 104 .size: 8 .value_kind: hidden_global_offset_z - .offset: 112 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 304 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z7forwardPdiS_iS_i .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z7forwardPdiS_iS_i.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 9 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .offset: 8 .size: 4 .value_kind: by_value - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .offset: 24 .size: 4 .value_kind: by_value - .address_space: global .offset: 32 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 40 .size: 8 .value_kind: global_buffer - .offset: 48 .size: 4 .value_kind: by_value - .offset: 56 .size: 8 .value_kind: by_value - .offset: 64 .size: 4 .value_kind: hidden_block_count_x - .offset: 68 .size: 4 .value_kind: hidden_block_count_y - .offset: 72 .size: 4 .value_kind: hidden_block_count_z - .offset: 76 .size: 2 .value_kind: hidden_group_size_x - .offset: 78 .size: 2 .value_kind: hidden_group_size_y - .offset: 80 .size: 2 .value_kind: hidden_group_size_z - .offset: 82 .size: 2 .value_kind: hidden_remainder_x - .offset: 84 .size: 2 .value_kind: hidden_remainder_y - .offset: 86 .size: 2 .value_kind: hidden_remainder_z - .offset: 104 .size: 8 .value_kind: hidden_global_offset_x - .offset: 112 .size: 8 .value_kind: hidden_global_offset_y - .offset: 120 .size: 8 .value_kind: hidden_global_offset_z - .offset: 128 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 320 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z14backward_part1PdiS_iS_S_id .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z14backward_part1PdiS_iS_S_id.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 15 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .args: - .address_space: global .offset: 0 .size: 8 .value_kind: global_buffer - .offset: 8 .size: 4 .value_kind: by_value - .address_space: global .offset: 16 .size: 8 .value_kind: global_buffer - .offset: 24 .size: 4 .value_kind: by_value - .address_space: global .offset: 32 .size: 8 .value_kind: global_buffer - .address_space: global .offset: 40 .size: 8 .value_kind: global_buffer - .offset: 48 .size: 4 .value_kind: by_value - .offset: 56 .size: 8 .value_kind: by_value - .offset: 64 .size: 4 .value_kind: hidden_block_count_x - .offset: 68 .size: 4 .value_kind: hidden_block_count_y - .offset: 72 .size: 4 .value_kind: hidden_block_count_z - .offset: 76 .size: 2 .value_kind: hidden_group_size_x - .offset: 78 .size: 2 .value_kind: hidden_group_size_y - .offset: 80 .size: 2 .value_kind: hidden_group_size_z - .offset: 82 .size: 2 .value_kind: hidden_remainder_x - .offset: 84 .size: 2 .value_kind: hidden_remainder_y - .offset: 86 .size: 2 .value_kind: hidden_remainder_z - .offset: 104 .size: 8 .value_kind: hidden_global_offset_x - .offset: 112 .size: 8 .value_kind: hidden_global_offset_y - .offset: 120 .size: 8 .value_kind: hidden_global_offset_z - .offset: 128 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 320 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z14backward_part2PdiS_iS_S_id .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z14backward_part2PdiS_iS_S_id.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 8 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .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: 24 .size: 8 .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: _Z12update_paramPdS_idi .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z12update_paramPdS_idi.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 16 .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 <stdlib.h> #include <stdio.h> extern int HI; // the total number of point patterns extern int HN; // the total number of points extern int HK_star; // the total number of covariates extern int Hd; // the dimensionality of the problem extern int HV; // the total number of elements in the grid void read_files(int ** foci , float ** Z , int * foci_counts , int * brain , int* author) { int i , j; // those are for the for loops // read the file with the points FILE *file; file = fopen("./inputs/foci.txt","r"); for (i=0 ; i<HN ; i++) { for (j=0 ; j<(Hd+1) ; j++) { if(!fscanf(file,"%d",&foci[i][j])) break; } } fclose(file); // read the design matrix file = fopen("./inputs/Z.txt","r"); for (i=0 ; i<HI ; i++) { for (j=0 ; j<HK_star ; j++) { if(!fscanf(file,"%f",&Z[i][j])) break; } } fclose(file); // read the total number of points per pattern file = fopen("./inputs/counts.txt","r"); for (i=0 ; i<HI ; i++) { if(!fscanf(file,"%d",&foci_counts[i])) break; } fclose(file); /* Binary vector indicating which voxels in the grid are within the brain mask */ file = fopen("mask.txt","r"); for (i=0 ; i<HV ; i++) { if(!fscanf(file,"%d",&brain[i])) break; } fclose(file); /* Read the publication identifier. Studies from the same paper must appear consecutive */ file = fopen("./inputs/paper.txt","r"); for (i=0 ; i<HI ; i++) { if(!fscanf(file,"%d",&author[i])) break; } fclose(file); }
code for sm_80
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
#include <stdlib.h> #include <stdio.h> extern int HI; // the total number of point patterns extern int HN; // the total number of points extern int HK_star; // the total number of covariates extern int Hd; // the dimensionality of the problem extern int HV; // the total number of elements in the grid void read_files(int ** foci , float ** Z , int * foci_counts , int * brain , int* author) { int i , j; // those are for the for loops // read the file with the points FILE *file; file = fopen("./inputs/foci.txt","r"); for (i=0 ; i<HN ; i++) { for (j=0 ; j<(Hd+1) ; j++) { if(!fscanf(file,"%d",&foci[i][j])) break; } } fclose(file); // read the design matrix file = fopen("./inputs/Z.txt","r"); for (i=0 ; i<HI ; i++) { for (j=0 ; j<HK_star ; j++) { if(!fscanf(file,"%f",&Z[i][j])) break; } } fclose(file); // read the total number of points per pattern file = fopen("./inputs/counts.txt","r"); for (i=0 ; i<HI ; i++) { if(!fscanf(file,"%d",&foci_counts[i])) break; } fclose(file); /* Binary vector indicating which voxels in the grid are within the brain mask */ file = fopen("mask.txt","r"); for (i=0 ; i<HV ; i++) { if(!fscanf(file,"%d",&brain[i])) break; } fclose(file); /* Read the publication identifier. Studies from the same paper must appear consecutive */ file = fopen("./inputs/paper.txt","r"); for (i=0 ; i<HI ; i++) { if(!fscanf(file,"%d",&author[i])) break; } fclose(file); }
.file "tmpxft_000daf33_00000000-6_read_files.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.str1.1,"aMS",@progbits,1 .LC0: .string "r" .LC1: .string "./inputs/foci.txt" .LC2: .string "%d" .LC3: .string "./inputs/Z.txt" .LC4: .string "%f" .LC5: .string "./inputs/counts.txt" .LC6: .string "mask.txt" .LC7: .string "./inputs/paper.txt" .text .globl _Z10read_filesPPiPPfS_S_S_ .type _Z10read_filesPPiPPfS_S_S_, @function _Z10read_filesPPiPPfS_S_S_: .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 $40, %rsp .cfi_def_cfa_offset 96 movq %rdi, %rbp movq %rsi, 24(%rsp) movq %rdx, %r13 movq %rcx, 8(%rsp) movq %r8, 16(%rsp) leaq .LC0(%rip), %rsi leaq .LC1(%rip), %rdi call fopen@PLT movq %rax, %r12 cmpl $0, HN(%rip) jle .L4 movl $0, %r15d leaq .LC2(%rip), %r14 jmp .L5 .L6: addl $1, %r15d addq $8, %rbp cmpl %r15d, HN(%rip) jle .L4 .L5: movl $0, %ebx cmpl $0, Hd(%rip) js .L6 .L7: movq 0(%rbp), %rax leaq (%rax,%rbx,4), %rdx movq %r14, %rsi movq %r12, %rdi movl $0, %eax call __isoc23_fscanf@PLT testl %eax, %eax je .L6 addq $1, %rbx cmpl %ebx, Hd(%rip) jge .L7 jmp .L6 .L4: movq %r12, %rdi call fclose@PLT leaq .LC0(%rip), %rsi leaq .LC3(%rip), %rdi call fopen@PLT movq %rax, %r12 cmpl $0, HI(%rip) jle .L8 movq 24(%rsp), %rbp movl $0, %r15d leaq .LC4(%rip), %r14 jmp .L9 .L10: addl $1, %r15d addq $8, %rbp cmpl %r15d, HI(%rip) jle .L8 .L9: movl $0, %ebx cmpl $0, HK_star(%rip) jle .L10 .L11: movq 0(%rbp), %rax leaq (%rax,%rbx,4), %rdx movq %r14, %rsi movq %r12, %rdi movl $0, %eax call __isoc23_fscanf@PLT testl %eax, %eax je .L10 addq $1, %rbx cmpl %ebx, HK_star(%rip) jg .L11 jmp .L10 .L8: movq %r12, %rdi call fclose@PLT leaq .LC0(%rip), %rsi leaq .LC5(%rip), %rdi call fopen@PLT movq %rax, %rbp cmpl $0, HI(%rip) jle .L12 movl $0, %ebx leaq .LC2(%rip), %r12 .L13: movq %r13, %rdx movq %r12, %rsi movq %rbp, %rdi movl $0, %eax call __isoc23_fscanf@PLT testl %eax, %eax je .L12 addl $1, %ebx addq $4, %r13 cmpl %ebx, HI(%rip) jg .L13 .L12: movq %rbp, %rdi call fclose@PLT leaq .LC0(%rip), %rsi leaq .LC6(%rip), %rdi call fopen@PLT movq %rax, %r12 cmpl $0, HV(%rip) jle .L14 movq 8(%rsp), %rbx movl $0, %ebp leaq .LC2(%rip), %r13 .L15: movq %rbx, %rdx movq %r13, %rsi movq %r12, %rdi movl $0, %eax call __isoc23_fscanf@PLT testl %eax, %eax je .L14 addl $1, %ebp addq $4, %rbx cmpl %ebp, HV(%rip) jg .L15 .L14: movq %r12, %rdi call fclose@PLT leaq .LC0(%rip), %rsi leaq .LC7(%rip), %rdi call fopen@PLT movq %rax, %r12 cmpl $0, HI(%rip) jle .L16 movq 16(%rsp), %rbx movl $0, %ebp leaq .LC2(%rip), %r13 .L17: movq %rbx, %rdx movq %r13, %rsi movq %r12, %rdi movl $0, %eax call __isoc23_fscanf@PLT testl %eax, %eax je .L16 addl $1, %ebp addq $4, %rbx cmpl %ebp, HI(%rip) jg .L17 .L16: movq %r12, %rdi call fclose@PLT 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 .LFE2057: .size _Z10read_filesPPiPPfS_S_S_, .-_Z10read_filesPPiPPfS_S_S_ .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 <stdlib.h> #include <stdio.h> extern int HI; // the total number of point patterns extern int HN; // the total number of points extern int HK_star; // the total number of covariates extern int Hd; // the dimensionality of the problem extern int HV; // the total number of elements in the grid void read_files(int ** foci , float ** Z , int * foci_counts , int * brain , int* author) { int i , j; // those are for the for loops // read the file with the points FILE *file; file = fopen("./inputs/foci.txt","r"); for (i=0 ; i<HN ; i++) { for (j=0 ; j<(Hd+1) ; j++) { if(!fscanf(file,"%d",&foci[i][j])) break; } } fclose(file); // read the design matrix file = fopen("./inputs/Z.txt","r"); for (i=0 ; i<HI ; i++) { for (j=0 ; j<HK_star ; j++) { if(!fscanf(file,"%f",&Z[i][j])) break; } } fclose(file); // read the total number of points per pattern file = fopen("./inputs/counts.txt","r"); for (i=0 ; i<HI ; i++) { if(!fscanf(file,"%d",&foci_counts[i])) break; } fclose(file); /* Binary vector indicating which voxels in the grid are within the brain mask */ file = fopen("mask.txt","r"); for (i=0 ; i<HV ; i++) { if(!fscanf(file,"%d",&brain[i])) break; } fclose(file); /* Read the publication identifier. Studies from the same paper must appear consecutive */ file = fopen("./inputs/paper.txt","r"); for (i=0 ; i<HI ; i++) { if(!fscanf(file,"%d",&author[i])) break; } fclose(file); }
#include <hip/hip_runtime.h> #include <stdlib.h> #include <stdio.h> extern int HI; // the total number of point patterns extern int HN; // the total number of points extern int HK_star; // the total number of covariates extern int Hd; // the dimensionality of the problem extern int HV; // the total number of elements in the grid void read_files(int ** foci , float ** Z , int * foci_counts , int * brain , int* author) { int i , j; // those are for the for loops // read the file with the points FILE *file; file = fopen("./inputs/foci.txt","r"); for (i=0 ; i<HN ; i++) { for (j=0 ; j<(Hd+1) ; j++) { if(!fscanf(file,"%d",&foci[i][j])) break; } } fclose(file); // read the design matrix file = fopen("./inputs/Z.txt","r"); for (i=0 ; i<HI ; i++) { for (j=0 ; j<HK_star ; j++) { if(!fscanf(file,"%f",&Z[i][j])) break; } } fclose(file); // read the total number of points per pattern file = fopen("./inputs/counts.txt","r"); for (i=0 ; i<HI ; i++) { if(!fscanf(file,"%d",&foci_counts[i])) break; } fclose(file); /* Binary vector indicating which voxels in the grid are within the brain mask */ file = fopen("mask.txt","r"); for (i=0 ; i<HV ; i++) { if(!fscanf(file,"%d",&brain[i])) break; } fclose(file); /* Read the publication identifier. Studies from the same paper must appear consecutive */ file = fopen("./inputs/paper.txt","r"); for (i=0 ; i<HI ; i++) { if(!fscanf(file,"%d",&author[i])) break; } fclose(file); }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <hip/hip_runtime.h> #include <stdlib.h> #include <stdio.h> extern int HI; // the total number of point patterns extern int HN; // the total number of points extern int HK_star; // the total number of covariates extern int Hd; // the dimensionality of the problem extern int HV; // the total number of elements in the grid void read_files(int ** foci , float ** Z , int * foci_counts , int * brain , int* author) { int i , j; // those are for the for loops // read the file with the points FILE *file; file = fopen("./inputs/foci.txt","r"); for (i=0 ; i<HN ; i++) { for (j=0 ; j<(Hd+1) ; j++) { if(!fscanf(file,"%d",&foci[i][j])) break; } } fclose(file); // read the design matrix file = fopen("./inputs/Z.txt","r"); for (i=0 ; i<HI ; i++) { for (j=0 ; j<HK_star ; j++) { if(!fscanf(file,"%f",&Z[i][j])) break; } } fclose(file); // read the total number of points per pattern file = fopen("./inputs/counts.txt","r"); for (i=0 ; i<HI ; i++) { if(!fscanf(file,"%d",&foci_counts[i])) break; } fclose(file); /* Binary vector indicating which voxels in the grid are within the brain mask */ file = fopen("mask.txt","r"); for (i=0 ; i<HV ; i++) { if(!fscanf(file,"%d",&brain[i])) break; } fclose(file); /* Read the publication identifier. Studies from the same paper must appear consecutive */ file = fopen("./inputs/paper.txt","r"); for (i=0 ; i<HI ; i++) { if(!fscanf(file,"%d",&author[i])) break; } fclose(file); }
.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 <stdlib.h> #include <stdio.h> extern int HI; // the total number of point patterns extern int HN; // the total number of points extern int HK_star; // the total number of covariates extern int Hd; // the dimensionality of the problem extern int HV; // the total number of elements in the grid void read_files(int ** foci , float ** Z , int * foci_counts , int * brain , int* author) { int i , j; // those are for the for loops // read the file with the points FILE *file; file = fopen("./inputs/foci.txt","r"); for (i=0 ; i<HN ; i++) { for (j=0 ; j<(Hd+1) ; j++) { if(!fscanf(file,"%d",&foci[i][j])) break; } } fclose(file); // read the design matrix file = fopen("./inputs/Z.txt","r"); for (i=0 ; i<HI ; i++) { for (j=0 ; j<HK_star ; j++) { if(!fscanf(file,"%f",&Z[i][j])) break; } } fclose(file); // read the total number of points per pattern file = fopen("./inputs/counts.txt","r"); for (i=0 ; i<HI ; i++) { if(!fscanf(file,"%d",&foci_counts[i])) break; } fclose(file); /* Binary vector indicating which voxels in the grid are within the brain mask */ file = fopen("mask.txt","r"); for (i=0 ; i<HV ; i++) { if(!fscanf(file,"%d",&brain[i])) break; } fclose(file); /* Read the publication identifier. Studies from the same paper must appear consecutive */ file = fopen("./inputs/paper.txt","r"); for (i=0 ; i<HI ; i++) { if(!fscanf(file,"%d",&author[i])) break; } fclose(file); }
.text .file "read_files.hip" .globl _Z10read_filesPPiPPfS_S_S_ # -- Begin function _Z10read_filesPPiPPfS_S_S_ .p2align 4, 0x90 .type _Z10read_filesPPiPPfS_S_S_,@function _Z10read_filesPPiPPfS_S_S_: # @_Z10read_filesPPiPPfS_S_S_ .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 %r8, 16(%rsp) # 8-byte Spill movq %rcx, 8(%rsp) # 8-byte Spill movq %rdx, (%rsp) # 8-byte Spill movq %rsi, %r12 movq %rdi, %r13 movl $.L.str, %edi movl $.L.str.1, %esi callq fopen movq %rax, %rbp cmpl $0, HN(%rip) jle .LBB0_7 # %bb.1: # %.preheader42.preheader xorl %ebx, %ebx jmp .LBB0_2 .p2align 4, 0x90 .LBB0_6: # %._crit_edge # in Loop: Header=BB0_2 Depth=1 incq %rbx movslq HN(%rip), %rax cmpq %rax, %rbx jge .LBB0_7 .LBB0_2: # %.preheader42 # =>This Loop Header: Depth=1 # Child Loop BB0_4 Depth 2 cmpl $0, Hd(%rip) js .LBB0_6 # %bb.3: # %.lr.ph.preheader # in Loop: Header=BB0_2 Depth=1 xorl %r14d, %r14d xorl %r15d, %r15d .p2align 4, 0x90 .LBB0_4: # %.lr.ph # Parent Loop BB0_2 Depth=1 # => This Inner Loop Header: Depth=2 movq (%r13,%rbx,8), %rdx addq %r14, %rdx movl $.L.str.2, %esi movq %rbp, %rdi xorl %eax, %eax callq __isoc23_fscanf testl %eax, %eax je .LBB0_6 # %bb.5: # %.lr.ph # in Loop: Header=BB0_4 Depth=2 leaq 1(%r15), %rax addq $4, %r14 movslq Hd(%rip), %rcx cmpq %rcx, %r15 movq %rax, %r15 jl .LBB0_4 jmp .LBB0_6 .LBB0_7: # %._crit_edge47 movq %rbp, %rdi callq fclose movl $.L.str.3, %edi movl $.L.str.1, %esi callq fopen movq %rax, %r13 cmpl $0, HI(%rip) jle .LBB0_14 # %bb.8: # %.preheader.preheader xorl %ebp, %ebp jmp .LBB0_9 .p2align 4, 0x90 .LBB0_13: # %._crit_edge50 # in Loop: Header=BB0_9 Depth=1 incq %rbp movslq HI(%rip), %rax cmpq %rax, %rbp jge .LBB0_14 .LBB0_9: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB0_11 Depth 2 cmpl $0, HK_star(%rip) jle .LBB0_13 # %bb.10: # %.lr.ph49.preheader # in Loop: Header=BB0_9 Depth=1 movl $1, %ebx xorl %r14d, %r14d .p2align 4, 0x90 .LBB0_11: # %.lr.ph49 # Parent Loop BB0_9 Depth=1 # => This Inner Loop Header: Depth=2 movq (%r12,%rbp,8), %rdx addq %r14, %rdx movl $.L.str.4, %esi movq %r13, %rdi xorl %eax, %eax callq __isoc23_fscanf testl %eax, %eax je .LBB0_13 # %bb.12: # %.lr.ph49 # in Loop: Header=BB0_11 Depth=2 addq $4, %r14 leaq 1(%rbx), %rax movslq HK_star(%rip), %rcx cmpq %rcx, %rbx movq %rax, %rbx jl .LBB0_11 jmp .LBB0_13 .LBB0_14: # %._crit_edge53 movq %r13, %rdi callq fclose movl $.L.str.5, %edi movl $.L.str.1, %esi callq fopen movq %rax, %r12 cmpl $0, HI(%rip) movq (%rsp), %r14 # 8-byte Reload jle .LBB0_18 # %bb.15: # %.lr.ph56.preheader xorl %ebx, %ebx .p2align 4, 0x90 .LBB0_17: # %.lr.ph56 # =>This Inner Loop Header: Depth=1 movl $.L.str.2, %esi movq %r12, %rdi movq %r14, %rdx xorl %eax, %eax callq __isoc23_fscanf testl %eax, %eax je .LBB0_18 # %bb.16: # in Loop: Header=BB0_17 Depth=1 incq %rbx movslq HI(%rip), %rax addq $4, %r14 cmpq %rax, %rbx jl .LBB0_17 .LBB0_18: # %._crit_edge57 movq %r12, %rdi callq fclose movl $.L.str.6, %edi movl $.L.str.1, %esi callq fopen movq %rax, %r15 cmpl $0, HV(%rip) movq 8(%rsp), %r14 # 8-byte Reload jle .LBB0_22 # %bb.19: # %.lr.ph61.preheader xorl %ebx, %ebx .p2align 4, 0x90 .LBB0_21: # %.lr.ph61 # =>This Inner Loop Header: Depth=1 movl $.L.str.2, %esi movq %r15, %rdi movq %r14, %rdx xorl %eax, %eax callq __isoc23_fscanf testl %eax, %eax je .LBB0_22 # %bb.20: # in Loop: Header=BB0_21 Depth=1 incq %rbx movslq HV(%rip), %rax addq $4, %r14 cmpq %rax, %rbx jl .LBB0_21 .LBB0_22: # %._crit_edge62 movq %r15, %rdi callq fclose movl $.L.str.7, %edi movl $.L.str.1, %esi callq fopen movq %rax, %r14 cmpl $0, HI(%rip) movq 16(%rsp), %r15 # 8-byte Reload jle .LBB0_26 # %bb.23: # %.lr.ph66.preheader xorl %ebx, %ebx .p2align 4, 0x90 .LBB0_25: # %.lr.ph66 # =>This Inner Loop Header: Depth=1 movl $.L.str.2, %esi movq %r14, %rdi movq %r15, %rdx xorl %eax, %eax callq __isoc23_fscanf testl %eax, %eax je .LBB0_26 # %bb.24: # in Loop: Header=BB0_25 Depth=1 incq %rbx movslq HI(%rip), %rax addq $4, %r15 cmpq %rax, %rbx jl .LBB0_25 .LBB0_26: # %._crit_edge67 movq %r14, %rdi 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 jmp fclose # TAILCALL .Lfunc_end0: .size _Z10read_filesPPiPPfS_S_S_, .Lfunc_end0-_Z10read_filesPPiPPfS_S_S_ .cfi_endproc # -- End function .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "./inputs/foci.txt" .size .L.str, 18 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "r" .size .L.str.1, 2 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "%d" .size .L.str.2, 3 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "./inputs/Z.txt" .size .L.str.3, 15 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "%f" .size .L.str.4, 3 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "./inputs/counts.txt" .size .L.str.5, 20 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz "mask.txt" .size .L.str.6, 9 .type .L.str.7,@object # @.str.7 .L.str.7: .asciz "./inputs/paper.txt" .size .L.str.7, 19 .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_000daf33_00000000-6_read_files.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.str1.1,"aMS",@progbits,1 .LC0: .string "r" .LC1: .string "./inputs/foci.txt" .LC2: .string "%d" .LC3: .string "./inputs/Z.txt" .LC4: .string "%f" .LC5: .string "./inputs/counts.txt" .LC6: .string "mask.txt" .LC7: .string "./inputs/paper.txt" .text .globl _Z10read_filesPPiPPfS_S_S_ .type _Z10read_filesPPiPPfS_S_S_, @function _Z10read_filesPPiPPfS_S_S_: .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 $40, %rsp .cfi_def_cfa_offset 96 movq %rdi, %rbp movq %rsi, 24(%rsp) movq %rdx, %r13 movq %rcx, 8(%rsp) movq %r8, 16(%rsp) leaq .LC0(%rip), %rsi leaq .LC1(%rip), %rdi call fopen@PLT movq %rax, %r12 cmpl $0, HN(%rip) jle .L4 movl $0, %r15d leaq .LC2(%rip), %r14 jmp .L5 .L6: addl $1, %r15d addq $8, %rbp cmpl %r15d, HN(%rip) jle .L4 .L5: movl $0, %ebx cmpl $0, Hd(%rip) js .L6 .L7: movq 0(%rbp), %rax leaq (%rax,%rbx,4), %rdx movq %r14, %rsi movq %r12, %rdi movl $0, %eax call __isoc23_fscanf@PLT testl %eax, %eax je .L6 addq $1, %rbx cmpl %ebx, Hd(%rip) jge .L7 jmp .L6 .L4: movq %r12, %rdi call fclose@PLT leaq .LC0(%rip), %rsi leaq .LC3(%rip), %rdi call fopen@PLT movq %rax, %r12 cmpl $0, HI(%rip) jle .L8 movq 24(%rsp), %rbp movl $0, %r15d leaq .LC4(%rip), %r14 jmp .L9 .L10: addl $1, %r15d addq $8, %rbp cmpl %r15d, HI(%rip) jle .L8 .L9: movl $0, %ebx cmpl $0, HK_star(%rip) jle .L10 .L11: movq 0(%rbp), %rax leaq (%rax,%rbx,4), %rdx movq %r14, %rsi movq %r12, %rdi movl $0, %eax call __isoc23_fscanf@PLT testl %eax, %eax je .L10 addq $1, %rbx cmpl %ebx, HK_star(%rip) jg .L11 jmp .L10 .L8: movq %r12, %rdi call fclose@PLT leaq .LC0(%rip), %rsi leaq .LC5(%rip), %rdi call fopen@PLT movq %rax, %rbp cmpl $0, HI(%rip) jle .L12 movl $0, %ebx leaq .LC2(%rip), %r12 .L13: movq %r13, %rdx movq %r12, %rsi movq %rbp, %rdi movl $0, %eax call __isoc23_fscanf@PLT testl %eax, %eax je .L12 addl $1, %ebx addq $4, %r13 cmpl %ebx, HI(%rip) jg .L13 .L12: movq %rbp, %rdi call fclose@PLT leaq .LC0(%rip), %rsi leaq .LC6(%rip), %rdi call fopen@PLT movq %rax, %r12 cmpl $0, HV(%rip) jle .L14 movq 8(%rsp), %rbx movl $0, %ebp leaq .LC2(%rip), %r13 .L15: movq %rbx, %rdx movq %r13, %rsi movq %r12, %rdi movl $0, %eax call __isoc23_fscanf@PLT testl %eax, %eax je .L14 addl $1, %ebp addq $4, %rbx cmpl %ebp, HV(%rip) jg .L15 .L14: movq %r12, %rdi call fclose@PLT leaq .LC0(%rip), %rsi leaq .LC7(%rip), %rdi call fopen@PLT movq %rax, %r12 cmpl $0, HI(%rip) jle .L16 movq 16(%rsp), %rbx movl $0, %ebp leaq .LC2(%rip), %r13 .L17: movq %rbx, %rdx movq %r13, %rsi movq %r12, %rdi movl $0, %eax call __isoc23_fscanf@PLT testl %eax, %eax je .L16 addl $1, %ebp addq $4, %rbx cmpl %ebp, HI(%rip) jg .L17 .L16: movq %r12, %rdi call fclose@PLT 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 .LFE2057: .size _Z10read_filesPPiPPfS_S_S_, .-_Z10read_filesPPiPPfS_S_S_ .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 "read_files.hip" .globl _Z10read_filesPPiPPfS_S_S_ # -- Begin function _Z10read_filesPPiPPfS_S_S_ .p2align 4, 0x90 .type _Z10read_filesPPiPPfS_S_S_,@function _Z10read_filesPPiPPfS_S_S_: # @_Z10read_filesPPiPPfS_S_S_ .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 %r8, 16(%rsp) # 8-byte Spill movq %rcx, 8(%rsp) # 8-byte Spill movq %rdx, (%rsp) # 8-byte Spill movq %rsi, %r12 movq %rdi, %r13 movl $.L.str, %edi movl $.L.str.1, %esi callq fopen movq %rax, %rbp cmpl $0, HN(%rip) jle .LBB0_7 # %bb.1: # %.preheader42.preheader xorl %ebx, %ebx jmp .LBB0_2 .p2align 4, 0x90 .LBB0_6: # %._crit_edge # in Loop: Header=BB0_2 Depth=1 incq %rbx movslq HN(%rip), %rax cmpq %rax, %rbx jge .LBB0_7 .LBB0_2: # %.preheader42 # =>This Loop Header: Depth=1 # Child Loop BB0_4 Depth 2 cmpl $0, Hd(%rip) js .LBB0_6 # %bb.3: # %.lr.ph.preheader # in Loop: Header=BB0_2 Depth=1 xorl %r14d, %r14d xorl %r15d, %r15d .p2align 4, 0x90 .LBB0_4: # %.lr.ph # Parent Loop BB0_2 Depth=1 # => This Inner Loop Header: Depth=2 movq (%r13,%rbx,8), %rdx addq %r14, %rdx movl $.L.str.2, %esi movq %rbp, %rdi xorl %eax, %eax callq __isoc23_fscanf testl %eax, %eax je .LBB0_6 # %bb.5: # %.lr.ph # in Loop: Header=BB0_4 Depth=2 leaq 1(%r15), %rax addq $4, %r14 movslq Hd(%rip), %rcx cmpq %rcx, %r15 movq %rax, %r15 jl .LBB0_4 jmp .LBB0_6 .LBB0_7: # %._crit_edge47 movq %rbp, %rdi callq fclose movl $.L.str.3, %edi movl $.L.str.1, %esi callq fopen movq %rax, %r13 cmpl $0, HI(%rip) jle .LBB0_14 # %bb.8: # %.preheader.preheader xorl %ebp, %ebp jmp .LBB0_9 .p2align 4, 0x90 .LBB0_13: # %._crit_edge50 # in Loop: Header=BB0_9 Depth=1 incq %rbp movslq HI(%rip), %rax cmpq %rax, %rbp jge .LBB0_14 .LBB0_9: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB0_11 Depth 2 cmpl $0, HK_star(%rip) jle .LBB0_13 # %bb.10: # %.lr.ph49.preheader # in Loop: Header=BB0_9 Depth=1 movl $1, %ebx xorl %r14d, %r14d .p2align 4, 0x90 .LBB0_11: # %.lr.ph49 # Parent Loop BB0_9 Depth=1 # => This Inner Loop Header: Depth=2 movq (%r12,%rbp,8), %rdx addq %r14, %rdx movl $.L.str.4, %esi movq %r13, %rdi xorl %eax, %eax callq __isoc23_fscanf testl %eax, %eax je .LBB0_13 # %bb.12: # %.lr.ph49 # in Loop: Header=BB0_11 Depth=2 addq $4, %r14 leaq 1(%rbx), %rax movslq HK_star(%rip), %rcx cmpq %rcx, %rbx movq %rax, %rbx jl .LBB0_11 jmp .LBB0_13 .LBB0_14: # %._crit_edge53 movq %r13, %rdi callq fclose movl $.L.str.5, %edi movl $.L.str.1, %esi callq fopen movq %rax, %r12 cmpl $0, HI(%rip) movq (%rsp), %r14 # 8-byte Reload jle .LBB0_18 # %bb.15: # %.lr.ph56.preheader xorl %ebx, %ebx .p2align 4, 0x90 .LBB0_17: # %.lr.ph56 # =>This Inner Loop Header: Depth=1 movl $.L.str.2, %esi movq %r12, %rdi movq %r14, %rdx xorl %eax, %eax callq __isoc23_fscanf testl %eax, %eax je .LBB0_18 # %bb.16: # in Loop: Header=BB0_17 Depth=1 incq %rbx movslq HI(%rip), %rax addq $4, %r14 cmpq %rax, %rbx jl .LBB0_17 .LBB0_18: # %._crit_edge57 movq %r12, %rdi callq fclose movl $.L.str.6, %edi movl $.L.str.1, %esi callq fopen movq %rax, %r15 cmpl $0, HV(%rip) movq 8(%rsp), %r14 # 8-byte Reload jle .LBB0_22 # %bb.19: # %.lr.ph61.preheader xorl %ebx, %ebx .p2align 4, 0x90 .LBB0_21: # %.lr.ph61 # =>This Inner Loop Header: Depth=1 movl $.L.str.2, %esi movq %r15, %rdi movq %r14, %rdx xorl %eax, %eax callq __isoc23_fscanf testl %eax, %eax je .LBB0_22 # %bb.20: # in Loop: Header=BB0_21 Depth=1 incq %rbx movslq HV(%rip), %rax addq $4, %r14 cmpq %rax, %rbx jl .LBB0_21 .LBB0_22: # %._crit_edge62 movq %r15, %rdi callq fclose movl $.L.str.7, %edi movl $.L.str.1, %esi callq fopen movq %rax, %r14 cmpl $0, HI(%rip) movq 16(%rsp), %r15 # 8-byte Reload jle .LBB0_26 # %bb.23: # %.lr.ph66.preheader xorl %ebx, %ebx .p2align 4, 0x90 .LBB0_25: # %.lr.ph66 # =>This Inner Loop Header: Depth=1 movl $.L.str.2, %esi movq %r14, %rdi movq %r15, %rdx xorl %eax, %eax callq __isoc23_fscanf testl %eax, %eax je .LBB0_26 # %bb.24: # in Loop: Header=BB0_25 Depth=1 incq %rbx movslq HI(%rip), %rax addq $4, %r15 cmpq %rax, %rbx jl .LBB0_25 .LBB0_26: # %._crit_edge67 movq %r14, %rdi 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 jmp fclose # TAILCALL .Lfunc_end0: .size _Z10read_filesPPiPPfS_S_S_, .Lfunc_end0-_Z10read_filesPPiPPfS_S_S_ .cfi_endproc # -- End function .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "./inputs/foci.txt" .size .L.str, 18 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "r" .size .L.str.1, 2 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "%d" .size .L.str.2, 3 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "./inputs/Z.txt" .size .L.str.3, 15 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "%f" .size .L.str.4, 3 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "./inputs/counts.txt" .size .L.str.5, 20 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz "mask.txt" .size .L.str.6, 9 .type .L.str.7,@object # @.str.7 .L.str.7: .asciz "./inputs/paper.txt" .size .L.str.7, 19 .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 CUDA device assembly.
#include <stdio.h> #include <stdlib.h> #include <sys/time.h> #define MAX_SIZE 1024 #define GET_TIME(now){ \ struct timeval t; \ gettimeofday(&t, NULL); \ now = t.tv_sec * 1000 + t.tv_usec / 1000.0; \ } int arr[MAX_SIZE * MAX_SIZE]; int brr[MAX_SIZE * MAX_SIZE]; int crr[MAX_SIZE * MAX_SIZE]; __global__ void mp(int *a, int *b, int *c) { int i, sum; int id_0, id_1, id_2; int col = blockIdx.x * blockDim.x + threadIdx.x; int row = blockIdx.y * blockDim.y + threadIdx.y; sum = 0; for(i = 0; i < MAX_SIZE; i++){ id_0 = row * MAX_SIZE + i; id_1 = i * MAX_SIZE + col; sum += a[id_0] * b[id_1]; } id_2 = row * MAX_SIZE + col; c[id_2] = sum; } int main(void) { int *d_a, *d_b, *d_c; int block_size; int size = sizeof(int) * MAX_SIZE * MAX_SIZE; int i, j; double start_time, end_time; FILE *fp = fopen("matrix.txt","r"); printf("Input the Block Size: "); scanf("%d",&block_size); for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fscanf(fp,"%d",&arr[i * MAX_SIZE + j]); } } for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fscanf(fp,"%d",&brr[i * MAX_SIZE + j]); } } fclose(fp); cudaMalloc((void **)&d_a, size); cudaMalloc((void **)&d_b, size); cudaMalloc((void **)&d_c, size); cudaMemcpy(d_a, arr, size, cudaMemcpyHostToDevice); cudaMemcpy(d_b, brr, size, cudaMemcpyHostToDevice); GET_TIME(start_time); dim3 dimBlock (block_size, block_size); dim3 dimGrid(MAX_SIZE / dimBlock.x, MAX_SIZE/dimBlock.y); mp<<<dimGrid, dimBlock>>>(d_a, d_b, d_c); GET_TIME(end_time); cudaMemcpy(crr, d_c, size, cudaMemcpyDeviceToHost); fp = fopen("matrix_cu.txt","w"); for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fprintf(fp,"%d ",crr[i * MAX_SIZE + j]); } fprintf(fp,"\n"); } fclose(fp); printf("Matrix Mulitplication Cuda VER: Elapsed time is %e (msec)\n",end_time - start_time); cudaFree(d_a); cudaFree(d_b); cudaFree(d_c); return 0; }
code for sm_80 Function : _Z2mpPiS_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 R9, SR_CTAID.Y ; /* 0x0000000000097919 */ /* 0x000e220000002600 */ /*0020*/ HFMA2.MMA R12, -RZ, RZ, 0, 0 ; /* 0x00000000ff0c7435 */ /* 0x000fe200000001ff */ /*0030*/ MOV R4, c[0x0][0x160] ; /* 0x0000580000047a02 */ /* 0x000fe20000000f00 */ /*0040*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*0050*/ S2R R2, SR_TID.Y ; /* 0x0000000000027919 */ /* 0x000e220000002200 */ /*0060*/ MOV R5, c[0x0][0x164] ; /* 0x0000590000057a02 */ /* 0x000fe20000000f00 */ /*0070*/ ULDC.64 UR6, c[0x0][0x168] ; /* 0x00005a0000067ab9 */ /* 0x000fe20000000a00 */ /*0080*/ MOV R8, RZ ; /* 0x000000ff00087202 */ /* 0x000fe20000000f00 */ /*0090*/ S2R R0, SR_CTAID.X ; /* 0x0000000000007919 */ /* 0x000e680000002500 */ /*00a0*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e620000002100 */ /*00b0*/ IMAD R9, R9, c[0x0][0x4], R2 ; /* 0x0000010009097a24 */ /* 0x001fca00078e0202 */ /*00c0*/ SHF.L.U32 R9, R9, 0xa, RZ ; /* 0x0000000a09097819 */ /* 0x000fe200000006ff */ /*00d0*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x002fc600078e0203 */ /*00e0*/ IADD3 R11, R9, 0x1, RZ ; /* 0x00000001090b7810 */ /* 0x000fe40007ffe0ff */ /*00f0*/ MOV R6, UR6 ; /* 0x0000000600067c02 */ /* 0x000fe20008000f00 */ /*0100*/ IMAD.WIDE R14, R9, 0x4, R4 ; /* 0x00000004090e7825 */ /* 0x000fe200078e0204 */ /*0110*/ MOV R7, UR7 ; /* 0x0000000700077c02 */ /* 0x000fc60008000f00 */ /*0120*/ IMAD.WIDE R2, R11, 0x4, R4 ; /* 0x000000040b027825 */ /* 0x000fe400078e0204 */ /*0130*/ LDG.E R14, [R14.64] ; /* 0x000000040e0e7981 */ /* 0x0000a4000c1e1900 */ /*0140*/ IMAD.WIDE R6, R0, 0x4, R6 ; /* 0x0000000400067825 */ /* 0x000fe400078e0206 */ /*0150*/ LDG.E R24, [R2.64] ; /* 0x0000000402187981 */ /* 0x000ee8000c1e1900 */ /*0160*/ LDG.E R17, [R6.64] ; /* 0x0000000406117981 */ /* 0x000ea8000c1e1900 */ /*0170*/ LDG.E R27, [R6.64+0x1000] ; /* 0x00100004061b7981 */ /* 0x000ee8000c1e1900 */ /*0180*/ LDG.E R16, [R2.64+0x4] ; /* 0x0000040402107981 */ /* 0x000f28000c1e1900 */ /*0190*/ LDG.E R19, [R6.64+0x2000] ; /* 0x0020000406137981 */ /* 0x000f28000c1e1900 */ /*01a0*/ LDG.E R23, [R2.64+0x8] ; /* 0x0000080402177981 */ /* 0x000f68000c1e1900 */ /*01b0*/ LDG.E R20, [R6.64+0x3000] ; /* 0x0030000406147981 */ /* 0x000f68000c1e1900 */ /*01c0*/ LDG.E R22, [R2.64+0xc] ; /* 0x00000c0402167981 */ /* 0x000f68000c1e1900 */ /*01d0*/ LDG.E R25, [R6.64+0x4000] ; /* 0x0040000406197981 */ /* 0x000f68000c1e1900 */ /*01e0*/ LDG.E R18, [R2.64+0x10] ; /* 0x0000100402127981 */ /* 0x000f68000c1e1900 */ /*01f0*/ LDG.E R21, [R6.64+0x5000] ; /* 0x0050000406157981 */ /* 0x000f68000c1e1900 */ /*0200*/ LDG.E R10, [R2.64+0x14] ; /* 0x00001404020a7981 */ /* 0x000f68000c1e1900 */ /*0210*/ LDG.E R13, [R6.64+0x6000] ; /* 0x00600004060d7981 */ /* 0x000f68000c1e1900 */ /*0220*/ LDG.E R15, [R6.64+0x7000] ; /* 0x00700004060f7981 */ /* 0x001f68000c1e1900 */ /*0230*/ LDG.E R26, [R6.64+0xf000] ; /* 0x00f00004061a7981 */ /* 0x000f68000c1e1900 */ /*0240*/ LDG.E R29, [R2.64+0x38] ; /* 0x00003804021d7981 */ /* 0x000f62000c1e1900 */ /*0250*/ IMAD R14, R17, R14, R12 ; /* 0x0000000e110e7224 */ /* 0x004fc600078e020c */ /*0260*/ LDG.E R12, [R2.64+0x18] ; /* 0x00001804020c7981 */ /* 0x000ea2000c1e1900 */ /*0270*/ IMAD R24, R27, R24, R14 ; /* 0x000000181b187224 */ /* 0x008fc600078e020e */ /*0280*/ LDG.E R14, [R2.64+0x1c] ; /* 0x00001c04020e7981 */ /* 0x000ee8000c1e1900 */ /*0290*/ LDG.E R17, [R6.64+0x8000] ; /* 0x0080000406117981 */ /* 0x000ee2000c1e1900 */ /*02a0*/ IMAD R24, R19, R16, R24 ; /* 0x0000001013187224 */ /* 0x010fc600078e0218 */ /*02b0*/ LDG.E R16, [R2.64+0x20] ; /* 0x0000200402107981 */ /* 0x000f28000c1e1900 */ /*02c0*/ LDG.E R19, [R6.64+0x9000] ; /* 0x0090000406137981 */ /* 0x000f22000c1e1900 */ /*02d0*/ IMAD R24, R20, R23, R24 ; /* 0x0000001714187224 */ /* 0x020fc600078e0218 */ /*02e0*/ LDG.E R23, [R2.64+0x24] ; /* 0x0000240402177981 */ /* 0x000f68000c1e1900 */ /*02f0*/ LDG.E R20, [R6.64+0xa000] ; /* 0x00a0000406147981 */ /* 0x000f62000c1e1900 */ /*0300*/ IMAD R24, R25, R22, R24 ; /* 0x0000001619187224 */ /* 0x000fc600078e0218 */ /*0310*/ LDG.E R25, [R2.64+0x28] ; /* 0x0000280402197981 */ /* 0x000f68000c1e1900 */ /*0320*/ LDG.E R22, [R6.64+0xb000] ; /* 0x00b0000406167981 */ /* 0x000f62000c1e1900 */ /*0330*/ IMAD R24, R21, R18, R24 ; /* 0x0000001215187224 */ /* 0x000fc600078e0218 */ /*0340*/ LDG.E R21, [R2.64+0x2c] ; /* 0x00002c0402157981 */ /* 0x000f68000c1e1900 */ /*0350*/ LDG.E R18, [R6.64+0xc000] ; /* 0x00c0000406127981 */ /* 0x000f62000c1e1900 */ /*0360*/ IMAD R28, R13, R10, R24 ; /* 0x0000000a0d1c7224 */ /* 0x000fc600078e0218 */ /*0370*/ LDG.E R24, [R2.64+0x30] ; /* 0x0000300402187981 */ /* 0x000f68000c1e1900 */ /*0380*/ LDG.E R13, [R6.64+0xd000] ; /* 0x00d00004060d7981 */ /* 0x000f68000c1e1900 */ /*0390*/ LDG.E R10, [R6.64+0xe000] ; /* 0x00e00004060a7981 */ /* 0x000f68000c1e1900 */ /*03a0*/ LDG.E R27, [R2.64+0x34] ; /* 0x00003404021b7981 */ /* 0x000f62000c1e1900 */ /*03b0*/ IADD3 R8, R8, 0x10, RZ ; /* 0x0000001008087810 */ /* 0x000fc80007ffe0ff */ /*03c0*/ ISETP.NE.AND P0, PT, R8, 0x400, PT ; /* 0x000004000800780c */ /* 0x000fe20003f05270 */ /*03d0*/ UIADD3 UR6, UP0, UR6, 0x10000, URZ ; /* 0x0001000006067890 */ /* 0x000fe2000ff1e03f */ /*03e0*/ IADD3 R4, P1, R4, 0x40, RZ ; /* 0x0000004004047810 */ /* 0x000fc60007f3e0ff */ /*03f0*/ UIADD3.X UR7, URZ, UR7, URZ, UP0, !UPT ; /* 0x000000073f077290 */ /* 0x000fe200087fe43f */ /*0400*/ IADD3.X R5, RZ, R5, RZ, P1, !PT ; /* 0x00000005ff057210 */ /* 0x000fe20000ffe4ff */ /*0410*/ IMAD R12, R15, R12, R28 ; /* 0x0000000c0f0c7224 */ /* 0x004fc800078e021c */ /*0420*/ IMAD R12, R17, R14, R12 ; /* 0x0000000e110c7224 */ /* 0x008fc800078e020c */ /*0430*/ IMAD R12, R19, R16, R12 ; /* 0x00000010130c7224 */ /* 0x010fc800078e020c */ /*0440*/ IMAD R12, R20, R23, R12 ; /* 0x00000017140c7224 */ /* 0x020fc800078e020c */ /*0450*/ IMAD R12, R22, R25, R12 ; /* 0x00000019160c7224 */ /* 0x000fc800078e020c */ /*0460*/ IMAD R12, R18, R21, R12 ; /* 0x00000015120c7224 */ /* 0x000fc800078e020c */ /*0470*/ IMAD R12, R13, R24, R12 ; /* 0x000000180d0c7224 */ /* 0x000fc800078e020c */ /*0480*/ IMAD R12, R10, R27, R12 ; /* 0x0000001b0a0c7224 */ /* 0x000fc800078e020c */ /*0490*/ IMAD R12, R26, R29, R12 ; /* 0x0000001d1a0c7224 */ /* 0x000fe200078e020c */ /*04a0*/ @P0 BRA 0xf0 ; /* 0xfffffc4000000947 */ /* 0x000fea000383ffff */ /*04b0*/ HFMA2.MMA R3, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff037435 */ /* 0x000fe200000001ff */ /*04c0*/ IADD3 R2, R0, R9, RZ ; /* 0x0000000900027210 */ /* 0x000fd20007ffe0ff */ /*04d0*/ IMAD.WIDE R2, R2, R3, c[0x0][0x170] ; /* 0x00005c0002027625 */ /* 0x000fca00078e0203 */ /*04e0*/ STG.E [R2.64], R12 ; /* 0x0000000c02007986 */ /* 0x000fe2000c101904 */ /*04f0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0500*/ BRA 0x500; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0510*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0520*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0530*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0540*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0550*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*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> #define MAX_SIZE 1024 #define GET_TIME(now){ \ struct timeval t; \ gettimeofday(&t, NULL); \ now = t.tv_sec * 1000 + t.tv_usec / 1000.0; \ } int arr[MAX_SIZE * MAX_SIZE]; int brr[MAX_SIZE * MAX_SIZE]; int crr[MAX_SIZE * MAX_SIZE]; __global__ void mp(int *a, int *b, int *c) { int i, sum; int id_0, id_1, id_2; int col = blockIdx.x * blockDim.x + threadIdx.x; int row = blockIdx.y * blockDim.y + threadIdx.y; sum = 0; for(i = 0; i < MAX_SIZE; i++){ id_0 = row * MAX_SIZE + i; id_1 = i * MAX_SIZE + col; sum += a[id_0] * b[id_1]; } id_2 = row * MAX_SIZE + col; c[id_2] = sum; } int main(void) { int *d_a, *d_b, *d_c; int block_size; int size = sizeof(int) * MAX_SIZE * MAX_SIZE; int i, j; double start_time, end_time; FILE *fp = fopen("matrix.txt","r"); printf("Input the Block Size: "); scanf("%d",&block_size); for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fscanf(fp,"%d",&arr[i * MAX_SIZE + j]); } } for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fscanf(fp,"%d",&brr[i * MAX_SIZE + j]); } } fclose(fp); cudaMalloc((void **)&d_a, size); cudaMalloc((void **)&d_b, size); cudaMalloc((void **)&d_c, size); cudaMemcpy(d_a, arr, size, cudaMemcpyHostToDevice); cudaMemcpy(d_b, brr, size, cudaMemcpyHostToDevice); GET_TIME(start_time); dim3 dimBlock (block_size, block_size); dim3 dimGrid(MAX_SIZE / dimBlock.x, MAX_SIZE/dimBlock.y); mp<<<dimGrid, dimBlock>>>(d_a, d_b, d_c); GET_TIME(end_time); cudaMemcpy(crr, d_c, size, cudaMemcpyDeviceToHost); fp = fopen("matrix_cu.txt","w"); for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fprintf(fp,"%d ",crr[i * MAX_SIZE + j]); } fprintf(fp,"\n"); } fclose(fp); printf("Matrix Mulitplication Cuda VER: Elapsed time is %e (msec)\n",end_time - start_time); cudaFree(d_a); cudaFree(d_b); cudaFree(d_c); return 0; }
.file "tmpxft_00125d87_00000000-6_mp_cu.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 _Z25__device_stub__Z2mpPiS_S_PiS_S_ .type _Z25__device_stub__Z2mpPiS_S_PiS_S_, @function _Z25__device_stub__Z2mpPiS_S_PiS_S_: .LFB2082: .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 _Z2mpPiS_S_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 144 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _Z25__device_stub__Z2mpPiS_S_PiS_S_, .-_Z25__device_stub__Z2mpPiS_S_PiS_S_ .globl _Z2mpPiS_S_ .type _Z2mpPiS_S_, @function _Z2mpPiS_S_: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z25__device_stub__Z2mpPiS_S_PiS_S_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _Z2mpPiS_S_, .-_Z2mpPiS_S_ .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "r" .LC1: .string "matrix.txt" .LC2: .string "Input the Block Size: " .LC3: .string "%d" .LC5: .string "w" .LC6: .string "matrix_cu.txt" .LC7: .string "%d " .LC8: .string "\n" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC9: .string "Matrix Mulitplication Cuda VER: Elapsed time is %e (msec)\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 $120, %rsp .cfi_def_cfa_offset 176 movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax leaq .LC0(%rip), %rsi leaq .LC1(%rip), %rdi call fopen@PLT movq %rax, %rbp leaq .LC2(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT leaq 28(%rsp), %rsi leaq .LC3(%rip), %rdi movl $0, %eax call __isoc23_scanf@PLT leaq 4096+arr(%rip), %r12 leaq 4194304(%r12), %r14 leaq .LC3(%rip), %r13 .L12: leaq -4096(%r12), %rbx .L13: movq %rbx, %rdx movq %r13, %rsi movq %rbp, %rdi movl $0, %eax call __isoc23_fscanf@PLT addq $4, %rbx cmpq %r12, %rbx jne .L13 addq $4096, %r12 cmpq %r14, %r12 jne .L12 leaq 4096+brr(%rip), %r12 leaq 4194304(%r12), %r14 leaq .LC3(%rip), %r13 .L14: leaq -4096(%r12), %rbx .L15: movq %rbx, %rdx movq %r13, %rsi movq %rbp, %rdi movl $0, %eax call __isoc23_fscanf@PLT addq $4, %rbx cmpq %r12, %rbx jne .L15 addq $4096, %r12 cmpq %r14, %r12 jne .L14 movq %rbp, %rdi call fclose@PLT leaq 32(%rsp), %rdi movl $4194304, %esi call cudaMalloc@PLT leaq 40(%rsp), %rdi movl $4194304, %esi call cudaMalloc@PLT leaq 48(%rsp), %rdi movl $4194304, %esi call cudaMalloc@PLT movl $1, %ecx movl $4194304, %edx leaq arr(%rip), %rsi movq 32(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movl $4194304, %edx leaq brr(%rip), %rsi movq 40(%rsp), %rdi call cudaMemcpy@PLT leaq 80(%rsp), %rdi movl $0, %esi call gettimeofday@PLT imulq $1000, 80(%rsp), %rax pxor %xmm1, %xmm1 cvtsi2sdq %rax, %xmm1 pxor %xmm0, %xmm0 cvtsi2sdq 88(%rsp), %xmm0 divsd .LC4(%rip), %xmm0 addsd %xmm0, %xmm1 movsd %xmm1, (%rsp) movl 28(%rsp), %ecx movl $1024, %eax movl $0, %edx divl %ecx movl %eax, 68(%rsp) movl %eax, 72(%rsp) movl %ecx, 56(%rsp) movl %ecx, 60(%rsp) movl $0, %r9d movl $0, %r8d movq 56(%rsp), %rdx movl $1, %ecx movq 68(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L27 .L17: leaq 80(%rsp), %rdi movl $0, %esi call gettimeofday@PLT imulq $1000, 80(%rsp), %rax pxor %xmm1, %xmm1 cvtsi2sdq %rax, %xmm1 pxor %xmm0, %xmm0 cvtsi2sdq 88(%rsp), %xmm0 divsd .LC4(%rip), %xmm0 addsd %xmm0, %xmm1 movsd %xmm1, 8(%rsp) movl $2, %ecx movl $4194304, %edx movq 48(%rsp), %rsi leaq crr(%rip), %rdi call cudaMemcpy@PLT leaq .LC5(%rip), %rsi leaq .LC6(%rip), %rdi call fopen@PLT movq %rax, %r12 leaq 4096+crr(%rip), %rbp leaq 4194304(%rbp), %r15 leaq .LC7(%rip), %r13 leaq .LC8(%rip), %r14 .L18: leaq -4096(%rbp), %rbx .L19: movl (%rbx), %ecx movq %r13, %rdx movl $2, %esi movq %r12, %rdi movl $0, %eax call __fprintf_chk@PLT addq $4, %rbx cmpq %rbp, %rbx jne .L19 movq %r14, %rdx movl $2, %esi movq %r12, %rdi movl $0, %eax call __fprintf_chk@PLT addq $4096, %rbp cmpq %r15, %rbp jne .L18 movq %r12, %rdi call fclose@PLT movsd 8(%rsp), %xmm0 subsd (%rsp), %xmm0 leaq .LC9(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movq 32(%rsp), %rdi call cudaFree@PLT movq 40(%rsp), %rdi call cudaFree@PLT movq 48(%rsp), %rdi call cudaFree@PLT movq 104(%rsp), %rax subq %fs:40, %rax jne .L28 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 .L27: .cfi_restore_state movq 48(%rsp), %rdx movq 40(%rsp), %rsi movq 32(%rsp), %rdi call _Z25__device_stub__Z2mpPiS_S_PiS_S_ jmp .L17 .L28: call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size main, .-main .section .rodata.str1.1 .LC10: .string "_Z2mpPiS_S_" .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 .LC10(%rip), %rdx movq %rdx, %rcx leaq _Z2mpPiS_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 .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 .globl crr .bss .align 32 .type crr, @object .size crr, 4194304 crr: .zero 4194304 .globl brr .align 32 .type brr, @object .size brr, 4194304 brr: .zero 4194304 .globl arr .align 32 .type arr, @object .size arr, 4194304 arr: .zero 4194304 .section .rodata.cst8,"aM",@progbits,8 .align 8 .LC4: .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> #define MAX_SIZE 1024 #define GET_TIME(now){ \ struct timeval t; \ gettimeofday(&t, NULL); \ now = t.tv_sec * 1000 + t.tv_usec / 1000.0; \ } int arr[MAX_SIZE * MAX_SIZE]; int brr[MAX_SIZE * MAX_SIZE]; int crr[MAX_SIZE * MAX_SIZE]; __global__ void mp(int *a, int *b, int *c) { int i, sum; int id_0, id_1, id_2; int col = blockIdx.x * blockDim.x + threadIdx.x; int row = blockIdx.y * blockDim.y + threadIdx.y; sum = 0; for(i = 0; i < MAX_SIZE; i++){ id_0 = row * MAX_SIZE + i; id_1 = i * MAX_SIZE + col; sum += a[id_0] * b[id_1]; } id_2 = row * MAX_SIZE + col; c[id_2] = sum; } int main(void) { int *d_a, *d_b, *d_c; int block_size; int size = sizeof(int) * MAX_SIZE * MAX_SIZE; int i, j; double start_time, end_time; FILE *fp = fopen("matrix.txt","r"); printf("Input the Block Size: "); scanf("%d",&block_size); for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fscanf(fp,"%d",&arr[i * MAX_SIZE + j]); } } for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fscanf(fp,"%d",&brr[i * MAX_SIZE + j]); } } fclose(fp); cudaMalloc((void **)&d_a, size); cudaMalloc((void **)&d_b, size); cudaMalloc((void **)&d_c, size); cudaMemcpy(d_a, arr, size, cudaMemcpyHostToDevice); cudaMemcpy(d_b, brr, size, cudaMemcpyHostToDevice); GET_TIME(start_time); dim3 dimBlock (block_size, block_size); dim3 dimGrid(MAX_SIZE / dimBlock.x, MAX_SIZE/dimBlock.y); mp<<<dimGrid, dimBlock>>>(d_a, d_b, d_c); GET_TIME(end_time); cudaMemcpy(crr, d_c, size, cudaMemcpyDeviceToHost); fp = fopen("matrix_cu.txt","w"); for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fprintf(fp,"%d ",crr[i * MAX_SIZE + j]); } fprintf(fp,"\n"); } fclose(fp); printf("Matrix Mulitplication Cuda VER: Elapsed time is %e (msec)\n",end_time - start_time); cudaFree(d_a); cudaFree(d_b); cudaFree(d_c); return 0; }
#include <hip/hip_runtime.h> #include <stdio.h> #include <stdlib.h> #include <sys/time.h> #define MAX_SIZE 1024 #define GET_TIME(now){ \ struct timeval t; \ gettimeofday(&t, NULL); \ now = t.tv_sec * 1000 + t.tv_usec / 1000.0; \ } int arr[MAX_SIZE * MAX_SIZE]; int brr[MAX_SIZE * MAX_SIZE]; int crr[MAX_SIZE * MAX_SIZE]; __global__ void mp(int *a, int *b, int *c) { int i, sum; int id_0, id_1, id_2; int col = blockIdx.x * blockDim.x + threadIdx.x; int row = blockIdx.y * blockDim.y + threadIdx.y; sum = 0; for(i = 0; i < MAX_SIZE; i++){ id_0 = row * MAX_SIZE + i; id_1 = i * MAX_SIZE + col; sum += a[id_0] * b[id_1]; } id_2 = row * MAX_SIZE + col; c[id_2] = sum; } int main(void) { int *d_a, *d_b, *d_c; int block_size; int size = sizeof(int) * MAX_SIZE * MAX_SIZE; int i, j; double start_time, end_time; FILE *fp = fopen("matrix.txt","r"); printf("Input the Block Size: "); scanf("%d",&block_size); for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fscanf(fp,"%d",&arr[i * MAX_SIZE + j]); } } for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fscanf(fp,"%d",&brr[i * MAX_SIZE + j]); } } fclose(fp); hipMalloc((void **)&d_a, size); hipMalloc((void **)&d_b, size); hipMalloc((void **)&d_c, size); hipMemcpy(d_a, arr, size, hipMemcpyHostToDevice); hipMemcpy(d_b, brr, size, hipMemcpyHostToDevice); GET_TIME(start_time); dim3 dimBlock (block_size, block_size); dim3 dimGrid(MAX_SIZE / dimBlock.x, MAX_SIZE/dimBlock.y); mp<<<dimGrid, dimBlock>>>(d_a, d_b, d_c); GET_TIME(end_time); hipMemcpy(crr, d_c, size, hipMemcpyDeviceToHost); fp = fopen("matrix_cu.txt","w"); for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fprintf(fp,"%d ",crr[i * MAX_SIZE + j]); } fprintf(fp,"\n"); } fclose(fp); printf("Matrix Mulitplication Cuda VER: Elapsed time is %e (msec)\n",end_time - start_time); hipFree(d_a); hipFree(d_b); hipFree(d_c); 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> #include <sys/time.h> #define MAX_SIZE 1024 #define GET_TIME(now){ \ struct timeval t; \ gettimeofday(&t, NULL); \ now = t.tv_sec * 1000 + t.tv_usec / 1000.0; \ } int arr[MAX_SIZE * MAX_SIZE]; int brr[MAX_SIZE * MAX_SIZE]; int crr[MAX_SIZE * MAX_SIZE]; __global__ void mp(int *a, int *b, int *c) { int i, sum; int id_0, id_1, id_2; int col = blockIdx.x * blockDim.x + threadIdx.x; int row = blockIdx.y * blockDim.y + threadIdx.y; sum = 0; for(i = 0; i < MAX_SIZE; i++){ id_0 = row * MAX_SIZE + i; id_1 = i * MAX_SIZE + col; sum += a[id_0] * b[id_1]; } id_2 = row * MAX_SIZE + col; c[id_2] = sum; } int main(void) { int *d_a, *d_b, *d_c; int block_size; int size = sizeof(int) * MAX_SIZE * MAX_SIZE; int i, j; double start_time, end_time; FILE *fp = fopen("matrix.txt","r"); printf("Input the Block Size: "); scanf("%d",&block_size); for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fscanf(fp,"%d",&arr[i * MAX_SIZE + j]); } } for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fscanf(fp,"%d",&brr[i * MAX_SIZE + j]); } } fclose(fp); hipMalloc((void **)&d_a, size); hipMalloc((void **)&d_b, size); hipMalloc((void **)&d_c, size); hipMemcpy(d_a, arr, size, hipMemcpyHostToDevice); hipMemcpy(d_b, brr, size, hipMemcpyHostToDevice); GET_TIME(start_time); dim3 dimBlock (block_size, block_size); dim3 dimGrid(MAX_SIZE / dimBlock.x, MAX_SIZE/dimBlock.y); mp<<<dimGrid, dimBlock>>>(d_a, d_b, d_c); GET_TIME(end_time); hipMemcpy(crr, d_c, size, hipMemcpyDeviceToHost); fp = fopen("matrix_cu.txt","w"); for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fprintf(fp,"%d ",crr[i * MAX_SIZE + j]); } fprintf(fp,"\n"); } fclose(fp); printf("Matrix Mulitplication Cuda VER: Elapsed time is %e (msec)\n",end_time - start_time); hipFree(d_a); hipFree(d_b); hipFree(d_c); return 0; }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z2mpPiS_S_ .globl _Z2mpPiS_S_ .p2align 8 .type _Z2mpPiS_S_,@function _Z2mpPiS_S_: s_clause 0x1 s_load_b32 s2, s[0:1], 0x24 s_load_b128 s[4:7], s[0:1], 0x0 v_bfe_u32 v3, v0, 10, 10 v_and_b32_e32 v0, 0x3ff, v0 s_waitcnt lgkmcnt(0) s_lshr_b32 s3, s2, 16 s_and_b32 s2, s2, 0xffff v_mad_u64_u32 v[1:2], null, s15, s3, v[3:4] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b32_e32 v2, 10, v1 v_ashrrev_i32_e32 v3, 31, v2 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_3) | instid1(VALU_DEP_3) v_lshlrev_b64 v[4:5], 2, v[2:3] v_mad_u64_u32 v[2:3], null, s14, s2, v[0:1] v_mov_b32_e32 v0, 0 s_mov_b64 s[2:3], 0 v_add_co_u32 v6, vcc_lo, s4, v4 s_delay_alu instid0(VALU_DEP_4) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_ci_u32_e32 v7, vcc_lo, s5, v5, vcc_lo v_mov_b32_e32 v3, v2 .p2align 6 .LBB0_1: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_4) v_ashrrev_i32_e32 v4, 31, v3 v_add_co_u32 v8, vcc_lo, v6, s2 s_delay_alu instid0(VALU_DEP_4) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_co_ci_u32_e32 v9, vcc_lo, s3, v7, vcc_lo v_lshlrev_b64 v[4:5], 2, v[3:4] s_add_u32 s2, s2, 4 s_addc_u32 s3, s3, 0 s_cmpk_lg_i32 s2, 0x1000 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v4, vcc_lo, s6, v4 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v5, vcc_lo global_load_b32 v8, v[8:9], off global_load_b32 v9, v[4:5], off s_waitcnt vmcnt(0) v_mad_u64_u32 v[4:5], null, v9, v8, v[0:1] s_delay_alu instid0(VALU_DEP_1) v_dual_mov_b32 v0, v4 :: v_dual_add_nc_u32 v3, 0x400, v3 s_cbranch_scc1 .LBB0_1 s_load_b64 s[0:1], s[0:1], 0x10 v_lshl_add_u32 v0, v1, 10, v2 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v1, 31, v0 v_lshlrev_b64 v[0:1], 2, v[0:1] 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], v4, off s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z2mpPiS_S_ .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 280 .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 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 _Z2mpPiS_S_, .Lfunc_end0-_Z2mpPiS_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: - .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: hidden_block_count_x - .offset: 28 .size: 4 .value_kind: hidden_block_count_y - .offset: 32 .size: 4 .value_kind: hidden_block_count_z - .offset: 36 .size: 2 .value_kind: hidden_group_size_x - .offset: 38 .size: 2 .value_kind: hidden_group_size_y - .offset: 40 .size: 2 .value_kind: hidden_group_size_z - .offset: 42 .size: 2 .value_kind: hidden_remainder_x - .offset: 44 .size: 2 .value_kind: hidden_remainder_y - .offset: 46 .size: 2 .value_kind: hidden_remainder_z - .offset: 64 .size: 8 .value_kind: hidden_global_offset_x - .offset: 72 .size: 8 .value_kind: hidden_global_offset_y - .offset: 80 .size: 8 .value_kind: hidden_global_offset_z - .offset: 88 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 280 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z2mpPiS_S_ .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z2mpPiS_S_.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> #include <sys/time.h> #define MAX_SIZE 1024 #define GET_TIME(now){ \ struct timeval t; \ gettimeofday(&t, NULL); \ now = t.tv_sec * 1000 + t.tv_usec / 1000.0; \ } int arr[MAX_SIZE * MAX_SIZE]; int brr[MAX_SIZE * MAX_SIZE]; int crr[MAX_SIZE * MAX_SIZE]; __global__ void mp(int *a, int *b, int *c) { int i, sum; int id_0, id_1, id_2; int col = blockIdx.x * blockDim.x + threadIdx.x; int row = blockIdx.y * blockDim.y + threadIdx.y; sum = 0; for(i = 0; i < MAX_SIZE; i++){ id_0 = row * MAX_SIZE + i; id_1 = i * MAX_SIZE + col; sum += a[id_0] * b[id_1]; } id_2 = row * MAX_SIZE + col; c[id_2] = sum; } int main(void) { int *d_a, *d_b, *d_c; int block_size; int size = sizeof(int) * MAX_SIZE * MAX_SIZE; int i, j; double start_time, end_time; FILE *fp = fopen("matrix.txt","r"); printf("Input the Block Size: "); scanf("%d",&block_size); for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fscanf(fp,"%d",&arr[i * MAX_SIZE + j]); } } for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fscanf(fp,"%d",&brr[i * MAX_SIZE + j]); } } fclose(fp); hipMalloc((void **)&d_a, size); hipMalloc((void **)&d_b, size); hipMalloc((void **)&d_c, size); hipMemcpy(d_a, arr, size, hipMemcpyHostToDevice); hipMemcpy(d_b, brr, size, hipMemcpyHostToDevice); GET_TIME(start_time); dim3 dimBlock (block_size, block_size); dim3 dimGrid(MAX_SIZE / dimBlock.x, MAX_SIZE/dimBlock.y); mp<<<dimGrid, dimBlock>>>(d_a, d_b, d_c); GET_TIME(end_time); hipMemcpy(crr, d_c, size, hipMemcpyDeviceToHost); fp = fopen("matrix_cu.txt","w"); for(i = 0; i < MAX_SIZE; i++){ for(j = 0; j < MAX_SIZE; j++){ fprintf(fp,"%d ",crr[i * MAX_SIZE + j]); } fprintf(fp,"\n"); } fclose(fp); printf("Matrix Mulitplication Cuda VER: Elapsed time is %e (msec)\n",end_time - start_time); hipFree(d_a); hipFree(d_b); hipFree(d_c); return 0; }
.text .file "mp_cu.hip" .globl _Z17__device_stub__mpPiS_S_ # -- Begin function _Z17__device_stub__mpPiS_S_ .p2align 4, 0x90 .type _Z17__device_stub__mpPiS_S_,@function _Z17__device_stub__mpPiS_S_: # @_Z17__device_stub__mpPiS_S_ .cfi_startproc # %bb.0: subq $104, %rsp .cfi_def_cfa_offset 112 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movq %rdx, 56(%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 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 $_Z2mpPiS_S_, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $120, %rsp .cfi_adjust_cfa_offset -120 retq .Lfunc_end0: .size _Z17__device_stub__mpPiS_S_, .Lfunc_end0-_Z17__device_stub__mpPiS_S_ .cfi_endproc # -- End function .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 # -- Begin function main .LCPI1_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 $152, %rsp .cfi_def_cfa_offset 208 .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 $.L.str, %edi movl $.L.str.1, %esi callq fopen movq %rax, %rbx movl $.L.str.2, %edi xorl %eax, %eax callq printf leaq 44(%rsp), %rsi movl $.L.str.3, %edi xorl %eax, %eax callq __isoc23_scanf movl $arr, %r14d xorl %r15d, %r15d .p2align 4, 0x90 .LBB1_1: # %.preheader44 # =>This Loop Header: Depth=1 # Child Loop BB1_2 Depth 2 xorl %r12d, %r12d .p2align 4, 0x90 .LBB1_2: # Parent Loop BB1_1 Depth=1 # => This Inner Loop Header: Depth=2 leaq (%r14,%r12), %rdx movl $.L.str.3, %esi movq %rbx, %rdi xorl %eax, %eax callq __isoc23_fscanf addq $4, %r12 cmpq $4096, %r12 # imm = 0x1000 jne .LBB1_2 # %bb.3: # in Loop: Header=BB1_1 Depth=1 incq %r15 addq $4096, %r14 # imm = 0x1000 cmpq $1024, %r15 # imm = 0x400 jne .LBB1_1 # %bb.4: # %.preheader42.preheader movl $brr, %r14d xorl %r15d, %r15d .p2align 4, 0x90 .LBB1_5: # %.preheader42 # =>This Loop Header: Depth=1 # Child Loop BB1_6 Depth 2 xorl %r12d, %r12d .p2align 4, 0x90 .LBB1_6: # Parent Loop BB1_5 Depth=1 # => This Inner Loop Header: Depth=2 leaq (%r14,%r12), %rdx movl $.L.str.3, %esi movq %rbx, %rdi xorl %eax, %eax callq __isoc23_fscanf addq $4, %r12 cmpq $4096, %r12 # imm = 0x1000 jne .LBB1_6 # %bb.7: # in Loop: Header=BB1_5 Depth=1 incq %r15 addq $4096, %r14 # imm = 0x1000 cmpq $1024, %r15 # imm = 0x400 jne .LBB1_5 # %bb.8: movq %rbx, %rdi callq fclose leaq 32(%rsp), %rdi movl $4194304, %esi # imm = 0x400000 callq hipMalloc leaq 24(%rsp), %rdi movl $4194304, %esi # imm = 0x400000 callq hipMalloc leaq 16(%rsp), %rdi movl $4194304, %esi # imm = 0x400000 callq hipMalloc movq 32(%rsp), %rdi movl $arr, %esi movl $4194304, %edx # imm = 0x400000 movl $1, %ecx callq hipMemcpy movq 24(%rsp), %rdi movl $brr, %esi movl $4194304, %edx # imm = 0x400000 movl $1, %ecx callq hipMemcpy leaq 48(%rsp), %rdi xorl %esi, %esi callq gettimeofday movq 48(%rsp), %rax movq %rax, 72(%rsp) # 8-byte Spill movq 56(%rsp), %rax movq %rax, 8(%rsp) # 8-byte Spill movl 44(%rsp), %esi movq %rsi, %rcx shlq $32, %rcx orq %rsi, %rcx movl $1024, %eax # imm = 0x400 xorl %edx, %edx divl %esi # kill: def $eax killed $eax def $rax movq %rax, %rdi shlq $32, %rdi orq %rax, %rdi movl $1, %esi movq %rcx, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_10 # %bb.9: movq 32(%rsp), %rax movq 24(%rsp), %rcx movq 16(%rsp), %rdx movq %rax, 144(%rsp) movq %rcx, 136(%rsp) movq %rdx, 128(%rsp) leaq 144(%rsp), %rax movq %rax, 48(%rsp) leaq 136(%rsp), %rax movq %rax, 56(%rsp) leaq 128(%rsp), %rax movq %rax, 64(%rsp) leaq 112(%rsp), %rdi leaq 96(%rsp), %rsi leaq 88(%rsp), %rdx leaq 80(%rsp), %rcx callq __hipPopCallConfiguration movq 112(%rsp), %rsi movl 120(%rsp), %edx movq 96(%rsp), %rcx movl 104(%rsp), %r8d leaq 48(%rsp), %r9 movl $_Z2mpPiS_S_, %edi pushq 80(%rsp) .cfi_adjust_cfa_offset 8 pushq 96(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_10: leaq 48(%rsp), %rdi xorl %esi, %esi callq gettimeofday movq 48(%rsp), %r13 movq 56(%rsp), %r12 movq 16(%rsp), %rsi movl $crr, %ebp movl $crr, %edi movl $4194304, %edx # imm = 0x400000 movl $2, %ecx callq hipMemcpy movl $.L.str.4, %edi movl $.L.str.5, %esi callq fopen movq %rax, %rbx xorl %r14d, %r14d .p2align 4, 0x90 .LBB1_11: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB1_12 Depth 2 xorl %r15d, %r15d .p2align 4, 0x90 .LBB1_12: # Parent Loop BB1_11 Depth=1 # => This Inner Loop Header: Depth=2 movl (%rbp,%r15,4), %edx movl $.L.str.6, %esi movq %rbx, %rdi xorl %eax, %eax callq fprintf incq %r15 cmpq $1024, %r15 # imm = 0x400 jne .LBB1_12 # %bb.13: # in Loop: Header=BB1_11 Depth=1 movl $10, %edi movq %rbx, %rsi callq fputc@PLT incq %r14 addq $4096, %rbp # imm = 0x1000 cmpq $1024, %r14 # imm = 0x400 jne .LBB1_11 # %bb.14: imulq $1000, %r13, %rax # imm = 0x3E8 cvtsi2sd %rax, %xmm0 cvtsi2sd %r12, %xmm4 movsd .LCPI1_0(%rip), %xmm1 # xmm1 = mem[0],zero divsd %xmm1, %xmm4 imulq $1000, 72(%rsp), %rax # 8-byte Folded Reload # imm = 0x3E8 cvtsi2sd %rax, %xmm2 cvtsi2sdq 8(%rsp), %xmm3 # 8-byte Folded Reload addsd %xmm0, %xmm4 divsd %xmm1, %xmm3 addsd %xmm2, %xmm3 subsd %xmm3, %xmm4 movsd %xmm4, 8(%rsp) # 8-byte Spill movq %rbx, %rdi callq fclose movl $.L.str.8, %edi movsd 8(%rsp), %xmm0 # 8-byte Reload # xmm0 = mem[0],zero movb $1, %al callq printf movq 32(%rsp), %rdi callq hipFree movq 24(%rsp), %rdi callq hipFree movq 16(%rsp), %rdi callq hipFree xorl %eax, %eax addq $152, %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 $_Z2mpPiS_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 arr,@object # @arr .bss .globl arr .p2align 4, 0x0 arr: .zero 4194304 .size arr, 4194304 .type brr,@object # @brr .globl brr .p2align 4, 0x0 brr: .zero 4194304 .size brr, 4194304 .type crr,@object # @crr .globl crr .p2align 4, 0x0 crr: .zero 4194304 .size crr, 4194304 .type _Z2mpPiS_S_,@object # @_Z2mpPiS_S_ .section .rodata,"a",@progbits .globl _Z2mpPiS_S_ .p2align 3, 0x0 _Z2mpPiS_S_: .quad _Z17__device_stub__mpPiS_S_ .size _Z2mpPiS_S_, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "matrix.txt" .size .L.str, 11 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "r" .size .L.str.1, 2 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "Input the Block Size: " .size .L.str.2, 23 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "%d" .size .L.str.3, 3 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "matrix_cu.txt" .size .L.str.4, 14 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "w" .size .L.str.5, 2 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz "%d " .size .L.str.6, 4 .type .L.str.8,@object # @.str.8 .L.str.8: .asciz "Matrix Mulitplication Cuda VER: Elapsed time is %e (msec)\n" .size .L.str.8, 59 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z2mpPiS_S_" .size .L__unnamed_1, 12 .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 _Z17__device_stub__mpPiS_S_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym arr .addrsig_sym brr .addrsig_sym crr .addrsig_sym _Z2mpPiS_S_ .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 : _Z2mpPiS_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 R9, SR_CTAID.Y ; /* 0x0000000000097919 */ /* 0x000e220000002600 */ /*0020*/ HFMA2.MMA R12, -RZ, RZ, 0, 0 ; /* 0x00000000ff0c7435 */ /* 0x000fe200000001ff */ /*0030*/ MOV R4, c[0x0][0x160] ; /* 0x0000580000047a02 */ /* 0x000fe20000000f00 */ /*0040*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe20000000a00 */ /*0050*/ S2R R2, SR_TID.Y ; /* 0x0000000000027919 */ /* 0x000e220000002200 */ /*0060*/ MOV R5, c[0x0][0x164] ; /* 0x0000590000057a02 */ /* 0x000fe20000000f00 */ /*0070*/ ULDC.64 UR6, c[0x0][0x168] ; /* 0x00005a0000067ab9 */ /* 0x000fe20000000a00 */ /*0080*/ MOV R8, RZ ; /* 0x000000ff00087202 */ /* 0x000fe20000000f00 */ /*0090*/ S2R R0, SR_CTAID.X ; /* 0x0000000000007919 */ /* 0x000e680000002500 */ /*00a0*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e620000002100 */ /*00b0*/ IMAD R9, R9, c[0x0][0x4], R2 ; /* 0x0000010009097a24 */ /* 0x001fca00078e0202 */ /*00c0*/ SHF.L.U32 R9, R9, 0xa, RZ ; /* 0x0000000a09097819 */ /* 0x000fe200000006ff */ /*00d0*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x002fc600078e0203 */ /*00e0*/ IADD3 R11, R9, 0x1, RZ ; /* 0x00000001090b7810 */ /* 0x000fe40007ffe0ff */ /*00f0*/ MOV R6, UR6 ; /* 0x0000000600067c02 */ /* 0x000fe20008000f00 */ /*0100*/ IMAD.WIDE R14, R9, 0x4, R4 ; /* 0x00000004090e7825 */ /* 0x000fe200078e0204 */ /*0110*/ MOV R7, UR7 ; /* 0x0000000700077c02 */ /* 0x000fc60008000f00 */ /*0120*/ IMAD.WIDE R2, R11, 0x4, R4 ; /* 0x000000040b027825 */ /* 0x000fe400078e0204 */ /*0130*/ LDG.E R14, [R14.64] ; /* 0x000000040e0e7981 */ /* 0x0000a4000c1e1900 */ /*0140*/ IMAD.WIDE R6, R0, 0x4, R6 ; /* 0x0000000400067825 */ /* 0x000fe400078e0206 */ /*0150*/ LDG.E R24, [R2.64] ; /* 0x0000000402187981 */ /* 0x000ee8000c1e1900 */ /*0160*/ LDG.E R17, [R6.64] ; /* 0x0000000406117981 */ /* 0x000ea8000c1e1900 */ /*0170*/ LDG.E R27, [R6.64+0x1000] ; /* 0x00100004061b7981 */ /* 0x000ee8000c1e1900 */ /*0180*/ LDG.E R16, [R2.64+0x4] ; /* 0x0000040402107981 */ /* 0x000f28000c1e1900 */ /*0190*/ LDG.E R19, [R6.64+0x2000] ; /* 0x0020000406137981 */ /* 0x000f28000c1e1900 */ /*01a0*/ LDG.E R23, [R2.64+0x8] ; /* 0x0000080402177981 */ /* 0x000f68000c1e1900 */ /*01b0*/ LDG.E R20, [R6.64+0x3000] ; /* 0x0030000406147981 */ /* 0x000f68000c1e1900 */ /*01c0*/ LDG.E R22, [R2.64+0xc] ; /* 0x00000c0402167981 */ /* 0x000f68000c1e1900 */ /*01d0*/ LDG.E R25, [R6.64+0x4000] ; /* 0x0040000406197981 */ /* 0x000f68000c1e1900 */ /*01e0*/ LDG.E R18, [R2.64+0x10] ; /* 0x0000100402127981 */ /* 0x000f68000c1e1900 */ /*01f0*/ LDG.E R21, [R6.64+0x5000] ; /* 0x0050000406157981 */ /* 0x000f68000c1e1900 */ /*0200*/ LDG.E R10, [R2.64+0x14] ; /* 0x00001404020a7981 */ /* 0x000f68000c1e1900 */ /*0210*/ LDG.E R13, [R6.64+0x6000] ; /* 0x00600004060d7981 */ /* 0x000f68000c1e1900 */ /*0220*/ LDG.E R15, [R6.64+0x7000] ; /* 0x00700004060f7981 */ /* 0x001f68000c1e1900 */ /*0230*/ LDG.E R26, [R6.64+0xf000] ; /* 0x00f00004061a7981 */ /* 0x000f68000c1e1900 */ /*0240*/ LDG.E R29, [R2.64+0x38] ; /* 0x00003804021d7981 */ /* 0x000f62000c1e1900 */ /*0250*/ IMAD R14, R17, R14, R12 ; /* 0x0000000e110e7224 */ /* 0x004fc600078e020c */ /*0260*/ LDG.E R12, [R2.64+0x18] ; /* 0x00001804020c7981 */ /* 0x000ea2000c1e1900 */ /*0270*/ IMAD R24, R27, R24, R14 ; /* 0x000000181b187224 */ /* 0x008fc600078e020e */ /*0280*/ LDG.E R14, [R2.64+0x1c] ; /* 0x00001c04020e7981 */ /* 0x000ee8000c1e1900 */ /*0290*/ LDG.E R17, [R6.64+0x8000] ; /* 0x0080000406117981 */ /* 0x000ee2000c1e1900 */ /*02a0*/ IMAD R24, R19, R16, R24 ; /* 0x0000001013187224 */ /* 0x010fc600078e0218 */ /*02b0*/ LDG.E R16, [R2.64+0x20] ; /* 0x0000200402107981 */ /* 0x000f28000c1e1900 */ /*02c0*/ LDG.E R19, [R6.64+0x9000] ; /* 0x0090000406137981 */ /* 0x000f22000c1e1900 */ /*02d0*/ IMAD R24, R20, R23, R24 ; /* 0x0000001714187224 */ /* 0x020fc600078e0218 */ /*02e0*/ LDG.E R23, [R2.64+0x24] ; /* 0x0000240402177981 */ /* 0x000f68000c1e1900 */ /*02f0*/ LDG.E R20, [R6.64+0xa000] ; /* 0x00a0000406147981 */ /* 0x000f62000c1e1900 */ /*0300*/ IMAD R24, R25, R22, R24 ; /* 0x0000001619187224 */ /* 0x000fc600078e0218 */ /*0310*/ LDG.E R25, [R2.64+0x28] ; /* 0x0000280402197981 */ /* 0x000f68000c1e1900 */ /*0320*/ LDG.E R22, [R6.64+0xb000] ; /* 0x00b0000406167981 */ /* 0x000f62000c1e1900 */ /*0330*/ IMAD R24, R21, R18, R24 ; /* 0x0000001215187224 */ /* 0x000fc600078e0218 */ /*0340*/ LDG.E R21, [R2.64+0x2c] ; /* 0x00002c0402157981 */ /* 0x000f68000c1e1900 */ /*0350*/ LDG.E R18, [R6.64+0xc000] ; /* 0x00c0000406127981 */ /* 0x000f62000c1e1900 */ /*0360*/ IMAD R28, R13, R10, R24 ; /* 0x0000000a0d1c7224 */ /* 0x000fc600078e0218 */ /*0370*/ LDG.E R24, [R2.64+0x30] ; /* 0x0000300402187981 */ /* 0x000f68000c1e1900 */ /*0380*/ LDG.E R13, [R6.64+0xd000] ; /* 0x00d00004060d7981 */ /* 0x000f68000c1e1900 */ /*0390*/ LDG.E R10, [R6.64+0xe000] ; /* 0x00e00004060a7981 */ /* 0x000f68000c1e1900 */ /*03a0*/ LDG.E R27, [R2.64+0x34] ; /* 0x00003404021b7981 */ /* 0x000f62000c1e1900 */ /*03b0*/ IADD3 R8, R8, 0x10, RZ ; /* 0x0000001008087810 */ /* 0x000fc80007ffe0ff */ /*03c0*/ ISETP.NE.AND P0, PT, R8, 0x400, PT ; /* 0x000004000800780c */ /* 0x000fe20003f05270 */ /*03d0*/ UIADD3 UR6, UP0, UR6, 0x10000, URZ ; /* 0x0001000006067890 */ /* 0x000fe2000ff1e03f */ /*03e0*/ IADD3 R4, P1, R4, 0x40, RZ ; /* 0x0000004004047810 */ /* 0x000fc60007f3e0ff */ /*03f0*/ UIADD3.X UR7, URZ, UR7, URZ, UP0, !UPT ; /* 0x000000073f077290 */ /* 0x000fe200087fe43f */ /*0400*/ IADD3.X R5, RZ, R5, RZ, P1, !PT ; /* 0x00000005ff057210 */ /* 0x000fe20000ffe4ff */ /*0410*/ IMAD R12, R15, R12, R28 ; /* 0x0000000c0f0c7224 */ /* 0x004fc800078e021c */ /*0420*/ IMAD R12, R17, R14, R12 ; /* 0x0000000e110c7224 */ /* 0x008fc800078e020c */ /*0430*/ IMAD R12, R19, R16, R12 ; /* 0x00000010130c7224 */ /* 0x010fc800078e020c */ /*0440*/ IMAD R12, R20, R23, R12 ; /* 0x00000017140c7224 */ /* 0x020fc800078e020c */ /*0450*/ IMAD R12, R22, R25, R12 ; /* 0x00000019160c7224 */ /* 0x000fc800078e020c */ /*0460*/ IMAD R12, R18, R21, R12 ; /* 0x00000015120c7224 */ /* 0x000fc800078e020c */ /*0470*/ IMAD R12, R13, R24, R12 ; /* 0x000000180d0c7224 */ /* 0x000fc800078e020c */ /*0480*/ IMAD R12, R10, R27, R12 ; /* 0x0000001b0a0c7224 */ /* 0x000fc800078e020c */ /*0490*/ IMAD R12, R26, R29, R12 ; /* 0x0000001d1a0c7224 */ /* 0x000fe200078e020c */ /*04a0*/ @P0 BRA 0xf0 ; /* 0xfffffc4000000947 */ /* 0x000fea000383ffff */ /*04b0*/ HFMA2.MMA R3, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff037435 */ /* 0x000fe200000001ff */ /*04c0*/ IADD3 R2, R0, R9, RZ ; /* 0x0000000900027210 */ /* 0x000fd20007ffe0ff */ /*04d0*/ IMAD.WIDE R2, R2, R3, c[0x0][0x170] ; /* 0x00005c0002027625 */ /* 0x000fca00078e0203 */ /*04e0*/ STG.E [R2.64], R12 ; /* 0x0000000c02007986 */ /* 0x000fe2000c101904 */ /*04f0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0500*/ BRA 0x500; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0510*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0520*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0530*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0540*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0550*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*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 _Z2mpPiS_S_ .globl _Z2mpPiS_S_ .p2align 8 .type _Z2mpPiS_S_,@function _Z2mpPiS_S_: s_clause 0x1 s_load_b32 s2, s[0:1], 0x24 s_load_b128 s[4:7], s[0:1], 0x0 v_bfe_u32 v3, v0, 10, 10 v_and_b32_e32 v0, 0x3ff, v0 s_waitcnt lgkmcnt(0) s_lshr_b32 s3, s2, 16 s_and_b32 s2, s2, 0xffff v_mad_u64_u32 v[1:2], null, s15, s3, v[3:4] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b32_e32 v2, 10, v1 v_ashrrev_i32_e32 v3, 31, v2 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_3) | instid1(VALU_DEP_3) v_lshlrev_b64 v[4:5], 2, v[2:3] v_mad_u64_u32 v[2:3], null, s14, s2, v[0:1] v_mov_b32_e32 v0, 0 s_mov_b64 s[2:3], 0 v_add_co_u32 v6, vcc_lo, s4, v4 s_delay_alu instid0(VALU_DEP_4) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_ci_u32_e32 v7, vcc_lo, s5, v5, vcc_lo v_mov_b32_e32 v3, v2 .p2align 6 .LBB0_1: s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_4) v_ashrrev_i32_e32 v4, 31, v3 v_add_co_u32 v8, vcc_lo, v6, s2 s_delay_alu instid0(VALU_DEP_4) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_co_ci_u32_e32 v9, vcc_lo, s3, v7, vcc_lo v_lshlrev_b64 v[4:5], 2, v[3:4] s_add_u32 s2, s2, 4 s_addc_u32 s3, s3, 0 s_cmpk_lg_i32 s2, 0x1000 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v4, vcc_lo, s6, v4 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v5, vcc_lo global_load_b32 v8, v[8:9], off global_load_b32 v9, v[4:5], off s_waitcnt vmcnt(0) v_mad_u64_u32 v[4:5], null, v9, v8, v[0:1] s_delay_alu instid0(VALU_DEP_1) v_dual_mov_b32 v0, v4 :: v_dual_add_nc_u32 v3, 0x400, v3 s_cbranch_scc1 .LBB0_1 s_load_b64 s[0:1], s[0:1], 0x10 v_lshl_add_u32 v0, v1, 10, v2 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v1, 31, v0 v_lshlrev_b64 v[0:1], 2, v[0:1] 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], v4, off s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z2mpPiS_S_ .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 280 .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 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 _Z2mpPiS_S_, .Lfunc_end0-_Z2mpPiS_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: - .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: hidden_block_count_x - .offset: 28 .size: 4 .value_kind: hidden_block_count_y - .offset: 32 .size: 4 .value_kind: hidden_block_count_z - .offset: 36 .size: 2 .value_kind: hidden_group_size_x - .offset: 38 .size: 2 .value_kind: hidden_group_size_y - .offset: 40 .size: 2 .value_kind: hidden_group_size_z - .offset: 42 .size: 2 .value_kind: hidden_remainder_x - .offset: 44 .size: 2 .value_kind: hidden_remainder_y - .offset: 46 .size: 2 .value_kind: hidden_remainder_z - .offset: 64 .size: 8 .value_kind: hidden_global_offset_x - .offset: 72 .size: 8 .value_kind: hidden_global_offset_y - .offset: 80 .size: 8 .value_kind: hidden_global_offset_z - .offset: 88 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 280 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z2mpPiS_S_ .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z2mpPiS_S_.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_00125d87_00000000-6_mp_cu.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 _Z25__device_stub__Z2mpPiS_S_PiS_S_ .type _Z25__device_stub__Z2mpPiS_S_PiS_S_, @function _Z25__device_stub__Z2mpPiS_S_PiS_S_: .LFB2082: .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 _Z2mpPiS_S_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 144 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _Z25__device_stub__Z2mpPiS_S_PiS_S_, .-_Z25__device_stub__Z2mpPiS_S_PiS_S_ .globl _Z2mpPiS_S_ .type _Z2mpPiS_S_, @function _Z2mpPiS_S_: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z25__device_stub__Z2mpPiS_S_PiS_S_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _Z2mpPiS_S_, .-_Z2mpPiS_S_ .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "r" .LC1: .string "matrix.txt" .LC2: .string "Input the Block Size: " .LC3: .string "%d" .LC5: .string "w" .LC6: .string "matrix_cu.txt" .LC7: .string "%d " .LC8: .string "\n" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC9: .string "Matrix Mulitplication Cuda VER: Elapsed time is %e (msec)\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 $120, %rsp .cfi_def_cfa_offset 176 movq %fs:40, %rax movq %rax, 104(%rsp) xorl %eax, %eax leaq .LC0(%rip), %rsi leaq .LC1(%rip), %rdi call fopen@PLT movq %rax, %rbp leaq .LC2(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT leaq 28(%rsp), %rsi leaq .LC3(%rip), %rdi movl $0, %eax call __isoc23_scanf@PLT leaq 4096+arr(%rip), %r12 leaq 4194304(%r12), %r14 leaq .LC3(%rip), %r13 .L12: leaq -4096(%r12), %rbx .L13: movq %rbx, %rdx movq %r13, %rsi movq %rbp, %rdi movl $0, %eax call __isoc23_fscanf@PLT addq $4, %rbx cmpq %r12, %rbx jne .L13 addq $4096, %r12 cmpq %r14, %r12 jne .L12 leaq 4096+brr(%rip), %r12 leaq 4194304(%r12), %r14 leaq .LC3(%rip), %r13 .L14: leaq -4096(%r12), %rbx .L15: movq %rbx, %rdx movq %r13, %rsi movq %rbp, %rdi movl $0, %eax call __isoc23_fscanf@PLT addq $4, %rbx cmpq %r12, %rbx jne .L15 addq $4096, %r12 cmpq %r14, %r12 jne .L14 movq %rbp, %rdi call fclose@PLT leaq 32(%rsp), %rdi movl $4194304, %esi call cudaMalloc@PLT leaq 40(%rsp), %rdi movl $4194304, %esi call cudaMalloc@PLT leaq 48(%rsp), %rdi movl $4194304, %esi call cudaMalloc@PLT movl $1, %ecx movl $4194304, %edx leaq arr(%rip), %rsi movq 32(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movl $4194304, %edx leaq brr(%rip), %rsi movq 40(%rsp), %rdi call cudaMemcpy@PLT leaq 80(%rsp), %rdi movl $0, %esi call gettimeofday@PLT imulq $1000, 80(%rsp), %rax pxor %xmm1, %xmm1 cvtsi2sdq %rax, %xmm1 pxor %xmm0, %xmm0 cvtsi2sdq 88(%rsp), %xmm0 divsd .LC4(%rip), %xmm0 addsd %xmm0, %xmm1 movsd %xmm1, (%rsp) movl 28(%rsp), %ecx movl $1024, %eax movl $0, %edx divl %ecx movl %eax, 68(%rsp) movl %eax, 72(%rsp) movl %ecx, 56(%rsp) movl %ecx, 60(%rsp) movl $0, %r9d movl $0, %r8d movq 56(%rsp), %rdx movl $1, %ecx movq 68(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax je .L27 .L17: leaq 80(%rsp), %rdi movl $0, %esi call gettimeofday@PLT imulq $1000, 80(%rsp), %rax pxor %xmm1, %xmm1 cvtsi2sdq %rax, %xmm1 pxor %xmm0, %xmm0 cvtsi2sdq 88(%rsp), %xmm0 divsd .LC4(%rip), %xmm0 addsd %xmm0, %xmm1 movsd %xmm1, 8(%rsp) movl $2, %ecx movl $4194304, %edx movq 48(%rsp), %rsi leaq crr(%rip), %rdi call cudaMemcpy@PLT leaq .LC5(%rip), %rsi leaq .LC6(%rip), %rdi call fopen@PLT movq %rax, %r12 leaq 4096+crr(%rip), %rbp leaq 4194304(%rbp), %r15 leaq .LC7(%rip), %r13 leaq .LC8(%rip), %r14 .L18: leaq -4096(%rbp), %rbx .L19: movl (%rbx), %ecx movq %r13, %rdx movl $2, %esi movq %r12, %rdi movl $0, %eax call __fprintf_chk@PLT addq $4, %rbx cmpq %rbp, %rbx jne .L19 movq %r14, %rdx movl $2, %esi movq %r12, %rdi movl $0, %eax call __fprintf_chk@PLT addq $4096, %rbp cmpq %r15, %rbp jne .L18 movq %r12, %rdi call fclose@PLT movsd 8(%rsp), %xmm0 subsd (%rsp), %xmm0 leaq .LC9(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movq 32(%rsp), %rdi call cudaFree@PLT movq 40(%rsp), %rdi call cudaFree@PLT movq 48(%rsp), %rdi call cudaFree@PLT movq 104(%rsp), %rax subq %fs:40, %rax jne .L28 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 .L27: .cfi_restore_state movq 48(%rsp), %rdx movq 40(%rsp), %rsi movq 32(%rsp), %rdi call _Z25__device_stub__Z2mpPiS_S_PiS_S_ jmp .L17 .L28: call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size main, .-main .section .rodata.str1.1 .LC10: .string "_Z2mpPiS_S_" .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 .LC10(%rip), %rdx movq %rdx, %rcx leaq _Z2mpPiS_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 .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 .globl crr .bss .align 32 .type crr, @object .size crr, 4194304 crr: .zero 4194304 .globl brr .align 32 .type brr, @object .size brr, 4194304 brr: .zero 4194304 .globl arr .align 32 .type arr, @object .size arr, 4194304 arr: .zero 4194304 .section .rodata.cst8,"aM",@progbits,8 .align 8 .LC4: .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 "mp_cu.hip" .globl _Z17__device_stub__mpPiS_S_ # -- Begin function _Z17__device_stub__mpPiS_S_ .p2align 4, 0x90 .type _Z17__device_stub__mpPiS_S_,@function _Z17__device_stub__mpPiS_S_: # @_Z17__device_stub__mpPiS_S_ .cfi_startproc # %bb.0: subq $104, %rsp .cfi_def_cfa_offset 112 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movq %rdx, 56(%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 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 $_Z2mpPiS_S_, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $120, %rsp .cfi_adjust_cfa_offset -120 retq .Lfunc_end0: .size _Z17__device_stub__mpPiS_S_, .Lfunc_end0-_Z17__device_stub__mpPiS_S_ .cfi_endproc # -- End function .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 # -- Begin function main .LCPI1_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 $152, %rsp .cfi_def_cfa_offset 208 .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 $.L.str, %edi movl $.L.str.1, %esi callq fopen movq %rax, %rbx movl $.L.str.2, %edi xorl %eax, %eax callq printf leaq 44(%rsp), %rsi movl $.L.str.3, %edi xorl %eax, %eax callq __isoc23_scanf movl $arr, %r14d xorl %r15d, %r15d .p2align 4, 0x90 .LBB1_1: # %.preheader44 # =>This Loop Header: Depth=1 # Child Loop BB1_2 Depth 2 xorl %r12d, %r12d .p2align 4, 0x90 .LBB1_2: # Parent Loop BB1_1 Depth=1 # => This Inner Loop Header: Depth=2 leaq (%r14,%r12), %rdx movl $.L.str.3, %esi movq %rbx, %rdi xorl %eax, %eax callq __isoc23_fscanf addq $4, %r12 cmpq $4096, %r12 # imm = 0x1000 jne .LBB1_2 # %bb.3: # in Loop: Header=BB1_1 Depth=1 incq %r15 addq $4096, %r14 # imm = 0x1000 cmpq $1024, %r15 # imm = 0x400 jne .LBB1_1 # %bb.4: # %.preheader42.preheader movl $brr, %r14d xorl %r15d, %r15d .p2align 4, 0x90 .LBB1_5: # %.preheader42 # =>This Loop Header: Depth=1 # Child Loop BB1_6 Depth 2 xorl %r12d, %r12d .p2align 4, 0x90 .LBB1_6: # Parent Loop BB1_5 Depth=1 # => This Inner Loop Header: Depth=2 leaq (%r14,%r12), %rdx movl $.L.str.3, %esi movq %rbx, %rdi xorl %eax, %eax callq __isoc23_fscanf addq $4, %r12 cmpq $4096, %r12 # imm = 0x1000 jne .LBB1_6 # %bb.7: # in Loop: Header=BB1_5 Depth=1 incq %r15 addq $4096, %r14 # imm = 0x1000 cmpq $1024, %r15 # imm = 0x400 jne .LBB1_5 # %bb.8: movq %rbx, %rdi callq fclose leaq 32(%rsp), %rdi movl $4194304, %esi # imm = 0x400000 callq hipMalloc leaq 24(%rsp), %rdi movl $4194304, %esi # imm = 0x400000 callq hipMalloc leaq 16(%rsp), %rdi movl $4194304, %esi # imm = 0x400000 callq hipMalloc movq 32(%rsp), %rdi movl $arr, %esi movl $4194304, %edx # imm = 0x400000 movl $1, %ecx callq hipMemcpy movq 24(%rsp), %rdi movl $brr, %esi movl $4194304, %edx # imm = 0x400000 movl $1, %ecx callq hipMemcpy leaq 48(%rsp), %rdi xorl %esi, %esi callq gettimeofday movq 48(%rsp), %rax movq %rax, 72(%rsp) # 8-byte Spill movq 56(%rsp), %rax movq %rax, 8(%rsp) # 8-byte Spill movl 44(%rsp), %esi movq %rsi, %rcx shlq $32, %rcx orq %rsi, %rcx movl $1024, %eax # imm = 0x400 xorl %edx, %edx divl %esi # kill: def $eax killed $eax def $rax movq %rax, %rdi shlq $32, %rdi orq %rax, %rdi movl $1, %esi movq %rcx, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_10 # %bb.9: movq 32(%rsp), %rax movq 24(%rsp), %rcx movq 16(%rsp), %rdx movq %rax, 144(%rsp) movq %rcx, 136(%rsp) movq %rdx, 128(%rsp) leaq 144(%rsp), %rax movq %rax, 48(%rsp) leaq 136(%rsp), %rax movq %rax, 56(%rsp) leaq 128(%rsp), %rax movq %rax, 64(%rsp) leaq 112(%rsp), %rdi leaq 96(%rsp), %rsi leaq 88(%rsp), %rdx leaq 80(%rsp), %rcx callq __hipPopCallConfiguration movq 112(%rsp), %rsi movl 120(%rsp), %edx movq 96(%rsp), %rcx movl 104(%rsp), %r8d leaq 48(%rsp), %r9 movl $_Z2mpPiS_S_, %edi pushq 80(%rsp) .cfi_adjust_cfa_offset 8 pushq 96(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_10: leaq 48(%rsp), %rdi xorl %esi, %esi callq gettimeofday movq 48(%rsp), %r13 movq 56(%rsp), %r12 movq 16(%rsp), %rsi movl $crr, %ebp movl $crr, %edi movl $4194304, %edx # imm = 0x400000 movl $2, %ecx callq hipMemcpy movl $.L.str.4, %edi movl $.L.str.5, %esi callq fopen movq %rax, %rbx xorl %r14d, %r14d .p2align 4, 0x90 .LBB1_11: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB1_12 Depth 2 xorl %r15d, %r15d .p2align 4, 0x90 .LBB1_12: # Parent Loop BB1_11 Depth=1 # => This Inner Loop Header: Depth=2 movl (%rbp,%r15,4), %edx movl $.L.str.6, %esi movq %rbx, %rdi xorl %eax, %eax callq fprintf incq %r15 cmpq $1024, %r15 # imm = 0x400 jne .LBB1_12 # %bb.13: # in Loop: Header=BB1_11 Depth=1 movl $10, %edi movq %rbx, %rsi callq fputc@PLT incq %r14 addq $4096, %rbp # imm = 0x1000 cmpq $1024, %r14 # imm = 0x400 jne .LBB1_11 # %bb.14: imulq $1000, %r13, %rax # imm = 0x3E8 cvtsi2sd %rax, %xmm0 cvtsi2sd %r12, %xmm4 movsd .LCPI1_0(%rip), %xmm1 # xmm1 = mem[0],zero divsd %xmm1, %xmm4 imulq $1000, 72(%rsp), %rax # 8-byte Folded Reload # imm = 0x3E8 cvtsi2sd %rax, %xmm2 cvtsi2sdq 8(%rsp), %xmm3 # 8-byte Folded Reload addsd %xmm0, %xmm4 divsd %xmm1, %xmm3 addsd %xmm2, %xmm3 subsd %xmm3, %xmm4 movsd %xmm4, 8(%rsp) # 8-byte Spill movq %rbx, %rdi callq fclose movl $.L.str.8, %edi movsd 8(%rsp), %xmm0 # 8-byte Reload # xmm0 = mem[0],zero movb $1, %al callq printf movq 32(%rsp), %rdi callq hipFree movq 24(%rsp), %rdi callq hipFree movq 16(%rsp), %rdi callq hipFree xorl %eax, %eax addq $152, %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 $_Z2mpPiS_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 arr,@object # @arr .bss .globl arr .p2align 4, 0x0 arr: .zero 4194304 .size arr, 4194304 .type brr,@object # @brr .globl brr .p2align 4, 0x0 brr: .zero 4194304 .size brr, 4194304 .type crr,@object # @crr .globl crr .p2align 4, 0x0 crr: .zero 4194304 .size crr, 4194304 .type _Z2mpPiS_S_,@object # @_Z2mpPiS_S_ .section .rodata,"a",@progbits .globl _Z2mpPiS_S_ .p2align 3, 0x0 _Z2mpPiS_S_: .quad _Z17__device_stub__mpPiS_S_ .size _Z2mpPiS_S_, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "matrix.txt" .size .L.str, 11 .type .L.str.1,@object # @.str.1 .L.str.1: .asciz "r" .size .L.str.1, 2 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "Input the Block Size: " .size .L.str.2, 23 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "%d" .size .L.str.3, 3 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "matrix_cu.txt" .size .L.str.4, 14 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "w" .size .L.str.5, 2 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz "%d " .size .L.str.6, 4 .type .L.str.8,@object # @.str.8 .L.str.8: .asciz "Matrix Mulitplication Cuda VER: Elapsed time is %e (msec)\n" .size .L.str.8, 59 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z2mpPiS_S_" .size .L__unnamed_1, 12 .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 _Z17__device_stub__mpPiS_S_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym arr .addrsig_sym brr .addrsig_sym crr .addrsig_sym _Z2mpPiS_S_ .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 <stdio.h> #include <stdlib.h> __global__ void calc_meanshift2(float* y_new, float* y_old, float* meanshift) { int i = blockDim.x*blockIdx.x + threadIdx.x; float tempY_new = y_new[i]; float tempY_old = y_old[i]; meanshift[i] = (tempY_new-tempY_old)*(tempY_new-tempY_old); } __device__ float kernel_fun(float x, float sigma2) { if( x > sigma2) return 0; else return exp(-x/2/sigma2); } __global__ void calc_Kernel_Matrix(int N, int D, float *x, float *y, float *K, int sigma2) { int i = blockDim.x*blockIdx.x + threadIdx.x; int j = blockDim.y*blockIdx.y + threadIdx.y; // It is also true that ( k == (N*i + j) ) // Calc Dist... float dist = 0; for(int d=0; d<D; d++) dist+= (y[i*D+d] - x[j*D+d])*(y[i*D+d] - x[j*D+d]); K[i*N+j] = kernel_fun(dist, sigma2); } __global__ void kernel_sum_div(int D, float* y_new, float* K_sum) { int i = blockDim.x*blockIdx.x + threadIdx.x; for(int d=0; d<D; d++) y_new[i*D+d] = y_new[i*D+d]/K_sum[i]; } __global__ void kernel_Dvec_mult(int N, int D, float* K, float* x, float* Kx, int d) { int i = blockDim.x*blockIdx.x + threadIdx.x; int j = blockDim.y*blockIdx.y + threadIdx.y; Kx[i*N+j] = K[i*N+j]*x[j*D+d]; } __global__ void copy_to_y(int D, float* d_y_new, float* kernelXsum, int d) { int i = blockDim.x*blockIdx.x + threadIdx.x; d_y_new[i*D+d] = kernelXsum[i]; } __global__ void calc_reduce_meanshift(int N, float* y_new, float* y_old, float* reducted_vec) { extern __shared__ float reduction_cache[] ; //thread ID on each row of blocks int tid = blockDim.x * blockIdx.x + threadIdx.x; int cache_i = threadIdx.x; float temp=0; float tempY_new; // This are usuful to ensure that only on GB access is gonna happen for each vector float tempY_old; while (tid < N) { tempY_new = y_new[tid]; tempY_old= y_old[tid]; temp += (tempY_new-tempY_old)*(tempY_new-tempY_old); tid += blockDim.x * gridDim.x; } reduction_cache[cache_i] = temp; __syncthreads(); // Begin the reduction per shared-memory-block for(int i=blockDim.x/2; i>0; i>>=1) { if(cache_i < i) reduction_cache[cache_i] += reduction_cache[cache_i+i]; __syncthreads(); } // Final Sum is stored in global array. if(cache_i==0) reducted_vec[blockIdx.x] = reduction_cache[0]; } // __global__ // void kernelX_dot_product(int N, int D, int d, float* K, float* x, float* reducted_vec) // { // extern __shared__ float reduction_cache[] ; // //thread ID on each row of blocks // int tid = blockDim.x * blockIdx.x + threadIdx.x; // int cache_i = threadIdx.x; // /* This UNROLLS the elements of x, "outside" the grid's index range. // In the case of N=600, threadsPerBlock=256 and 2 blocks in total, // we have 600-256*2=88 additions done in parallel, before the reduction of the 512 threads. // incase the index-range > N, the reduction scheme will simply add some zeros to the vector. // This allows as to oversubscribe in terms of threads and blocks. // */ // int offset = N*blockIdx.y; // float temp=0; // while (tid < N) // { // temp += K[tid+offset]*x[tid*D+d]; // tid += blockDim.x * gridDim.x; // } // /* Load x-data into local shared memory. // As mentioned before, some entries are small sums of // x's outside the grid's range */ // reduction_cache[cache_i] = temp; // __syncthreads(); // // Begin the reduction per shared-memory-block // for(int i=blockDim.x/2; i>0; i>>=1) // { // if(cache_i < i) // reduction_cache[cache_i] += reduction_cache[cache_i+i]; // __syncthreads(); // } // // Final Sum is stored in global array, with stride d, to match the NxD dimensionality of the input dataset. // if(cache_i==0) // reducted_vec[blockIdx.y*gridDim.x + blockIdx.x + d] = reduction_cache[cache_i]; // } // void WR_kernelX_dow_product(int N, float* d_K, float* d_x, /*out*/ ReductionCache* rc ) // { // dim3 blockDim2(4, 1, 1); // dim3 gridDim2(N/4,N,1); // size_t cache_size = 4*N*sizeof(float); // kernelX_dot_product<<<gridDim2, blockDim2, cache_size>>>(N,D,0, d_KernelMatrix, d_x, d_y_new); // kernelX_dot_product<<<gridDim2, blockDim2, cache_size>>>(N,D,1, d_KernelMatrix, d_x, d_y_new); // //reduction_sum<<<L/256, 256, 256*sizeof(float) >>>(N/4, d_y_new, d_y_new); // if(rc->blocksNum == 1) // { // kernelX_dot_product<<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N,D,0, d_K,d_x, rc->d_sum); // kernelX_dot_product<<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N,D,1, d_K,d_x, rc->d_sum); // } // else // { // // We need multiple reduction calls! // reduction_sum <<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N, d_A, rc->d_reduced_vec); // /* Reduct the final reduction vector! */ // // Ideally we would like threads_num==length(reduced_vec)/numRow. // However threads_num2 must be a power of 2. Thus: // int threads_num2 = exp2f(floor(log2f(rc->reduced_vec_length/rc->rowNum))); // if(threads_num2>512) // threads_num2=512; // //printf("THREADS: %d RED_VEC %d\n", threads_num2, rc->reduced_vec_length/rc->rowNum ); // dim3 gridDim2(1,rc->rowNum,1); // dim3 blockDim2(threads_num2,1,1); // reduction_sum<<<gridDim2, blockDim2, threads_num2*sizeof(float)>>>\ // (rc->gridDim.x, rc->d_reduced_vec, rc->d_sum); // // // WARNING: launching with original thread_num might be too much. // // SOLUTION: Find power-of-2 nearest to block_num // } // }
.file "tmpxft_000c76d3_00000000-6_cuda_meanshift.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 _Z10kernel_funff .type _Z10kernel_funff, @function _Z10kernel_funff: .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 _Z10kernel_funff, .-_Z10kernel_funff .globl _Z39__device_stub__Z15calc_meanshift2PfS_S_PfS_S_ .type _Z39__device_stub__Z15calc_meanshift2PfS_S_PfS_S_, @function _Z39__device_stub__Z15calc_meanshift2PfS_S_PfS_S_: .LFB2082: .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 .L9 .L5: movq 120(%rsp), %rax subq %fs:40, %rax jne .L10 addq $136, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L9: .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 _Z15calc_meanshift2PfS_S_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 144 jmp .L5 .L10: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _Z39__device_stub__Z15calc_meanshift2PfS_S_PfS_S_, .-_Z39__device_stub__Z15calc_meanshift2PfS_S_PfS_S_ .globl _Z15calc_meanshift2PfS_S_ .type _Z15calc_meanshift2PfS_S_, @function _Z15calc_meanshift2PfS_S_: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z39__device_stub__Z15calc_meanshift2PfS_S_PfS_S_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _Z15calc_meanshift2PfS_S_, .-_Z15calc_meanshift2PfS_S_ .globl _Z45__device_stub__Z18calc_Kernel_MatrixiiPfS_S_iiiPfS_S_i .type _Z45__device_stub__Z18calc_Kernel_MatrixiiPfS_S_iiiPfS_S_i, @function _Z45__device_stub__Z18calc_Kernel_MatrixiiPfS_S_iiiPfS_S_i: .LFB2084: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movl %edi, 44(%rsp) movl %esi, 40(%rsp) movq %rdx, 32(%rsp) movq %rcx, 24(%rsp) movq %r8, 16(%rsp) movl %r9d, 12(%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 32(%rsp), %rax movq %rax, 128(%rsp) leaq 24(%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 .L17 .L13: movq 168(%rsp), %rax subq %fs:40, %rax jne .L18 addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L17: .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 _Z18calc_Kernel_MatrixiiPfS_S_i(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 192 jmp .L13 .L18: call __stack_chk_fail@PLT .cfi_endproc .LFE2084: .size _Z45__device_stub__Z18calc_Kernel_MatrixiiPfS_S_iiiPfS_S_i, .-_Z45__device_stub__Z18calc_Kernel_MatrixiiPfS_S_iiiPfS_S_i .globl _Z18calc_Kernel_MatrixiiPfS_S_i .type _Z18calc_Kernel_MatrixiiPfS_S_i, @function _Z18calc_Kernel_MatrixiiPfS_S_i: .LFB2085: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z45__device_stub__Z18calc_Kernel_MatrixiiPfS_S_iiiPfS_S_i addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2085: .size _Z18calc_Kernel_MatrixiiPfS_S_i, .-_Z18calc_Kernel_MatrixiiPfS_S_i .globl _Z37__device_stub__Z14kernel_sum_diviPfS_iPfS_ .type _Z37__device_stub__Z14kernel_sum_diviPfS_iPfS_, @function _Z37__device_stub__Z14kernel_sum_diviPfS_iPfS_: .LFB2086: .cfi_startproc endbr64 subq $136, %rsp .cfi_def_cfa_offset 144 movl %edi, 28(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movq %fs:40, %rax movq %rax, 120(%rsp) xorl %eax, %eax leaq 28(%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 .L25 .L21: movq 120(%rsp), %rax subq %fs:40, %rax jne .L26 addq $136, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L25: .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 _Z14kernel_sum_diviPfS_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 144 jmp .L21 .L26: call __stack_chk_fail@PLT .cfi_endproc .LFE2086: .size _Z37__device_stub__Z14kernel_sum_diviPfS_iPfS_, .-_Z37__device_stub__Z14kernel_sum_diviPfS_iPfS_ .globl _Z14kernel_sum_diviPfS_ .type _Z14kernel_sum_diviPfS_, @function _Z14kernel_sum_diviPfS_: .LFB2087: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z37__device_stub__Z14kernel_sum_diviPfS_iPfS_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2087: .size _Z14kernel_sum_diviPfS_, .-_Z14kernel_sum_diviPfS_ .globl _Z43__device_stub__Z16kernel_Dvec_multiiPfS_S_iiiPfS_S_i .type _Z43__device_stub__Z16kernel_Dvec_multiiPfS_S_iiiPfS_S_i, @function _Z43__device_stub__Z16kernel_Dvec_multiiPfS_S_iiiPfS_S_i: .LFB2088: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movl %edi, 44(%rsp) movl %esi, 40(%rsp) movq %rdx, 32(%rsp) movq %rcx, 24(%rsp) movq %r8, 16(%rsp) movl %r9d, 12(%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 32(%rsp), %rax movq %rax, 128(%rsp) leaq 24(%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 .L33 .L29: movq 168(%rsp), %rax subq %fs:40, %rax jne .L34 addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L33: .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 _Z16kernel_Dvec_multiiPfS_S_i(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 192 jmp .L29 .L34: call __stack_chk_fail@PLT .cfi_endproc .LFE2088: .size _Z43__device_stub__Z16kernel_Dvec_multiiPfS_S_iiiPfS_S_i, .-_Z43__device_stub__Z16kernel_Dvec_multiiPfS_S_iiiPfS_S_i .globl _Z16kernel_Dvec_multiiPfS_S_i .type _Z16kernel_Dvec_multiiPfS_S_i, @function _Z16kernel_Dvec_multiiPfS_S_i: .LFB2089: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z43__device_stub__Z16kernel_Dvec_multiiPfS_S_iiiPfS_S_i addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2089: .size _Z16kernel_Dvec_multiiPfS_S_i, .-_Z16kernel_Dvec_multiiPfS_S_i .globl _Z32__device_stub__Z9copy_to_yiPfS_iiPfS_i .type _Z32__device_stub__Z9copy_to_yiPfS_iiPfS_i, @function _Z32__device_stub__Z9copy_to_yiPfS_iiPfS_i: .LFB2090: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movl %edi, 28(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movl %ecx, 24(%rsp) movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax leaq 28(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) leaq 24(%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 .L41 .L37: movq 136(%rsp), %rax subq %fs:40, %rax jne .L42 addq $152, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L41: .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 _Z9copy_to_yiPfS_i(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L37 .L42: call __stack_chk_fail@PLT .cfi_endproc .LFE2090: .size _Z32__device_stub__Z9copy_to_yiPfS_iiPfS_i, .-_Z32__device_stub__Z9copy_to_yiPfS_iiPfS_i .globl _Z9copy_to_yiPfS_i .type _Z9copy_to_yiPfS_i, @function _Z9copy_to_yiPfS_i: .LFB2091: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z32__device_stub__Z9copy_to_yiPfS_iiPfS_i addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2091: .size _Z9copy_to_yiPfS_i, .-_Z9copy_to_yiPfS_i .globl _Z46__device_stub__Z21calc_reduce_meanshiftiPfS_S_iPfS_S_ .type _Z46__device_stub__Z21calc_reduce_meanshiftiPfS_S_iPfS_S_, @function _Z46__device_stub__Z21calc_reduce_meanshiftiPfS_S_iPfS_S_: .LFB2092: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movl %edi, 28(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movq %rcx, (%rsp) movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax leaq 28(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) movq %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 .L49 .L45: movq 136(%rsp), %rax subq %fs:40, %rax jne .L50 addq $152, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L49: .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 _Z21calc_reduce_meanshiftiPfS_S_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L45 .L50: call __stack_chk_fail@PLT .cfi_endproc .LFE2092: .size _Z46__device_stub__Z21calc_reduce_meanshiftiPfS_S_iPfS_S_, .-_Z46__device_stub__Z21calc_reduce_meanshiftiPfS_S_iPfS_S_ .globl _Z21calc_reduce_meanshiftiPfS_S_ .type _Z21calc_reduce_meanshiftiPfS_S_, @function _Z21calc_reduce_meanshiftiPfS_S_: .LFB2093: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z46__device_stub__Z21calc_reduce_meanshiftiPfS_S_iPfS_S_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2093: .size _Z21calc_reduce_meanshiftiPfS_S_, .-_Z21calc_reduce_meanshiftiPfS_S_ .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "_Z21calc_reduce_meanshiftiPfS_S_" .section .rodata.str1.1,"aMS",@progbits,1 .LC1: .string "_Z9copy_to_yiPfS_i" .LC2: .string "_Z16kernel_Dvec_multiiPfS_S_i" .LC3: .string "_Z14kernel_sum_diviPfS_" .section .rodata.str1.8 .align 8 .LC4: .string "_Z18calc_Kernel_MatrixiiPfS_S_i" .section .rodata.str1.1 .LC5: .string "_Z15calc_meanshift2PfS_S_" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2095: .cfi_startproc endbr64 pushq %rbx .cfi_def_cfa_offset 16 .cfi_offset 3, -16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rbx 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 _Z21calc_reduce_meanshiftiPfS_S_(%rip), %rsi movq %rax, %rdi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 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 _Z9copy_to_yiPfS_i(%rip), %rsi movq %rbx, %rdi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 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 .LC2(%rip), %rdx movq %rdx, %rcx leaq _Z16kernel_Dvec_multiiPfS_S_i(%rip), %rsi movq %rbx, %rdi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 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 .LC3(%rip), %rdx movq %rdx, %rcx leaq _Z14kernel_sum_diviPfS_(%rip), %rsi movq %rbx, %rdi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 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 _Z18calc_Kernel_MatrixiiPfS_S_i(%rip), %rsi movq %rbx, %rdi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 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 _Z15calc_meanshift2PfS_S_(%rip), %rsi movq %rbx, %rdi 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 popq %rbx .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2095: .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> __global__ void calc_meanshift2(float* y_new, float* y_old, float* meanshift) { int i = blockDim.x*blockIdx.x + threadIdx.x; float tempY_new = y_new[i]; float tempY_old = y_old[i]; meanshift[i] = (tempY_new-tempY_old)*(tempY_new-tempY_old); } __device__ float kernel_fun(float x, float sigma2) { if( x > sigma2) return 0; else return exp(-x/2/sigma2); } __global__ void calc_Kernel_Matrix(int N, int D, float *x, float *y, float *K, int sigma2) { int i = blockDim.x*blockIdx.x + threadIdx.x; int j = blockDim.y*blockIdx.y + threadIdx.y; // It is also true that ( k == (N*i + j) ) // Calc Dist... float dist = 0; for(int d=0; d<D; d++) dist+= (y[i*D+d] - x[j*D+d])*(y[i*D+d] - x[j*D+d]); K[i*N+j] = kernel_fun(dist, sigma2); } __global__ void kernel_sum_div(int D, float* y_new, float* K_sum) { int i = blockDim.x*blockIdx.x + threadIdx.x; for(int d=0; d<D; d++) y_new[i*D+d] = y_new[i*D+d]/K_sum[i]; } __global__ void kernel_Dvec_mult(int N, int D, float* K, float* x, float* Kx, int d) { int i = blockDim.x*blockIdx.x + threadIdx.x; int j = blockDim.y*blockIdx.y + threadIdx.y; Kx[i*N+j] = K[i*N+j]*x[j*D+d]; } __global__ void copy_to_y(int D, float* d_y_new, float* kernelXsum, int d) { int i = blockDim.x*blockIdx.x + threadIdx.x; d_y_new[i*D+d] = kernelXsum[i]; } __global__ void calc_reduce_meanshift(int N, float* y_new, float* y_old, float* reducted_vec) { extern __shared__ float reduction_cache[] ; //thread ID on each row of blocks int tid = blockDim.x * blockIdx.x + threadIdx.x; int cache_i = threadIdx.x; float temp=0; float tempY_new; // This are usuful to ensure that only on GB access is gonna happen for each vector float tempY_old; while (tid < N) { tempY_new = y_new[tid]; tempY_old= y_old[tid]; temp += (tempY_new-tempY_old)*(tempY_new-tempY_old); tid += blockDim.x * gridDim.x; } reduction_cache[cache_i] = temp; __syncthreads(); // Begin the reduction per shared-memory-block for(int i=blockDim.x/2; i>0; i>>=1) { if(cache_i < i) reduction_cache[cache_i] += reduction_cache[cache_i+i]; __syncthreads(); } // Final Sum is stored in global array. if(cache_i==0) reducted_vec[blockIdx.x] = reduction_cache[0]; } // __global__ // void kernelX_dot_product(int N, int D, int d, float* K, float* x, float* reducted_vec) // { // extern __shared__ float reduction_cache[] ; // //thread ID on each row of blocks // int tid = blockDim.x * blockIdx.x + threadIdx.x; // int cache_i = threadIdx.x; // /* This UNROLLS the elements of x, "outside" the grid's index range. // In the case of N=600, threadsPerBlock=256 and 2 blocks in total, // we have 600-256*2=88 additions done in parallel, before the reduction of the 512 threads. // incase the index-range > N, the reduction scheme will simply add some zeros to the vector. // This allows as to oversubscribe in terms of threads and blocks. // */ // int offset = N*blockIdx.y; // float temp=0; // while (tid < N) // { // temp += K[tid+offset]*x[tid*D+d]; // tid += blockDim.x * gridDim.x; // } // /* Load x-data into local shared memory. // As mentioned before, some entries are small sums of // x's outside the grid's range */ // reduction_cache[cache_i] = temp; // __syncthreads(); // // Begin the reduction per shared-memory-block // for(int i=blockDim.x/2; i>0; i>>=1) // { // if(cache_i < i) // reduction_cache[cache_i] += reduction_cache[cache_i+i]; // __syncthreads(); // } // // Final Sum is stored in global array, with stride d, to match the NxD dimensionality of the input dataset. // if(cache_i==0) // reducted_vec[blockIdx.y*gridDim.x + blockIdx.x + d] = reduction_cache[cache_i]; // } // void WR_kernelX_dow_product(int N, float* d_K, float* d_x, /*out*/ ReductionCache* rc ) // { // dim3 blockDim2(4, 1, 1); // dim3 gridDim2(N/4,N,1); // size_t cache_size = 4*N*sizeof(float); // kernelX_dot_product<<<gridDim2, blockDim2, cache_size>>>(N,D,0, d_KernelMatrix, d_x, d_y_new); // kernelX_dot_product<<<gridDim2, blockDim2, cache_size>>>(N,D,1, d_KernelMatrix, d_x, d_y_new); // //reduction_sum<<<L/256, 256, 256*sizeof(float) >>>(N/4, d_y_new, d_y_new); // if(rc->blocksNum == 1) // { // kernelX_dot_product<<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N,D,0, d_K,d_x, rc->d_sum); // kernelX_dot_product<<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N,D,1, d_K,d_x, rc->d_sum); // } // else // { // // We need multiple reduction calls! // reduction_sum <<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N, d_A, rc->d_reduced_vec); // /* Reduct the final reduction vector! */ // // Ideally we would like threads_num==length(reduced_vec)/numRow. // However threads_num2 must be a power of 2. Thus: // int threads_num2 = exp2f(floor(log2f(rc->reduced_vec_length/rc->rowNum))); // if(threads_num2>512) // threads_num2=512; // //printf("THREADS: %d RED_VEC %d\n", threads_num2, rc->reduced_vec_length/rc->rowNum ); // dim3 gridDim2(1,rc->rowNum,1); // dim3 blockDim2(threads_num2,1,1); // reduction_sum<<<gridDim2, blockDim2, threads_num2*sizeof(float)>>>\ // (rc->gridDim.x, rc->d_reduced_vec, rc->d_sum); // // // WARNING: launching with original thread_num might be too much. // // SOLUTION: Find power-of-2 nearest to block_num // } // }
#include <hip/hip_runtime.h> #include <stdio.h> #include <stdlib.h> __global__ void calc_meanshift2(float* y_new, float* y_old, float* meanshift) { int i = blockDim.x*blockIdx.x + threadIdx.x; float tempY_new = y_new[i]; float tempY_old = y_old[i]; meanshift[i] = (tempY_new-tempY_old)*(tempY_new-tempY_old); } __device__ float kernel_fun(float x, float sigma2) { if( x > sigma2) return 0; else return exp(-x/2/sigma2); } __global__ void calc_Kernel_Matrix(int N, int D, float *x, float *y, float *K, int sigma2) { int i = blockDim.x*blockIdx.x + threadIdx.x; int j = blockDim.y*blockIdx.y + threadIdx.y; // It is also true that ( k == (N*i + j) ) // Calc Dist... float dist = 0; for(int d=0; d<D; d++) dist+= (y[i*D+d] - x[j*D+d])*(y[i*D+d] - x[j*D+d]); K[i*N+j] = kernel_fun(dist, sigma2); } __global__ void kernel_sum_div(int D, float* y_new, float* K_sum) { int i = blockDim.x*blockIdx.x + threadIdx.x; for(int d=0; d<D; d++) y_new[i*D+d] = y_new[i*D+d]/K_sum[i]; } __global__ void kernel_Dvec_mult(int N, int D, float* K, float* x, float* Kx, int d) { int i = blockDim.x*blockIdx.x + threadIdx.x; int j = blockDim.y*blockIdx.y + threadIdx.y; Kx[i*N+j] = K[i*N+j]*x[j*D+d]; } __global__ void copy_to_y(int D, float* d_y_new, float* kernelXsum, int d) { int i = blockDim.x*blockIdx.x + threadIdx.x; d_y_new[i*D+d] = kernelXsum[i]; } __global__ void calc_reduce_meanshift(int N, float* y_new, float* y_old, float* reducted_vec) { extern __shared__ float reduction_cache[] ; //thread ID on each row of blocks int tid = blockDim.x * blockIdx.x + threadIdx.x; int cache_i = threadIdx.x; float temp=0; float tempY_new; // This are usuful to ensure that only on GB access is gonna happen for each vector float tempY_old; while (tid < N) { tempY_new = y_new[tid]; tempY_old= y_old[tid]; temp += (tempY_new-tempY_old)*(tempY_new-tempY_old); tid += blockDim.x * gridDim.x; } reduction_cache[cache_i] = temp; __syncthreads(); // Begin the reduction per shared-memory-block for(int i=blockDim.x/2; i>0; i>>=1) { if(cache_i < i) reduction_cache[cache_i] += reduction_cache[cache_i+i]; __syncthreads(); } // Final Sum is stored in global array. if(cache_i==0) reducted_vec[blockIdx.x] = reduction_cache[0]; } // __global__ // void kernelX_dot_product(int N, int D, int d, float* K, float* x, float* reducted_vec) // { // extern __shared__ float reduction_cache[] ; // //thread ID on each row of blocks // int tid = blockDim.x * blockIdx.x + threadIdx.x; // int cache_i = threadIdx.x; // /* This UNROLLS the elements of x, "outside" the grid's index range. // In the case of N=600, threadsPerBlock=256 and 2 blocks in total, // we have 600-256*2=88 additions done in parallel, before the reduction of the 512 threads. // incase the index-range > N, the reduction scheme will simply add some zeros to the vector. // This allows as to oversubscribe in terms of threads and blocks. // */ // int offset = N*blockIdx.y; // float temp=0; // while (tid < N) // { // temp += K[tid+offset]*x[tid*D+d]; // tid += blockDim.x * gridDim.x; // } // /* Load x-data into local shared memory. // As mentioned before, some entries are small sums of // x's outside the grid's range */ // reduction_cache[cache_i] = temp; // __syncthreads(); // // Begin the reduction per shared-memory-block // for(int i=blockDim.x/2; i>0; i>>=1) // { // if(cache_i < i) // reduction_cache[cache_i] += reduction_cache[cache_i+i]; // __syncthreads(); // } // // Final Sum is stored in global array, with stride d, to match the NxD dimensionality of the input dataset. // if(cache_i==0) // reducted_vec[blockIdx.y*gridDim.x + blockIdx.x + d] = reduction_cache[cache_i]; // } // void WR_kernelX_dow_product(int N, float* d_K, float* d_x, /*out*/ ReductionCache* rc ) // { // dim3 blockDim2(4, 1, 1); // dim3 gridDim2(N/4,N,1); // size_t cache_size = 4*N*sizeof(float); // kernelX_dot_product<<<gridDim2, blockDim2, cache_size>>>(N,D,0, d_KernelMatrix, d_x, d_y_new); // kernelX_dot_product<<<gridDim2, blockDim2, cache_size>>>(N,D,1, d_KernelMatrix, d_x, d_y_new); // //reduction_sum<<<L/256, 256, 256*sizeof(float) >>>(N/4, d_y_new, d_y_new); // if(rc->blocksNum == 1) // { // kernelX_dot_product<<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N,D,0, d_K,d_x, rc->d_sum); // kernelX_dot_product<<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N,D,1, d_K,d_x, rc->d_sum); // } // else // { // // We need multiple reduction calls! // reduction_sum <<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N, d_A, rc->d_reduced_vec); // /* Reduct the final reduction vector! */ // // Ideally we would like threads_num==length(reduced_vec)/numRow. // However threads_num2 must be a power of 2. Thus: // int threads_num2 = exp2f(floor(log2f(rc->reduced_vec_length/rc->rowNum))); // if(threads_num2>512) // threads_num2=512; // //printf("THREADS: %d RED_VEC %d\n", threads_num2, rc->reduced_vec_length/rc->rowNum ); // dim3 gridDim2(1,rc->rowNum,1); // dim3 blockDim2(threads_num2,1,1); // reduction_sum<<<gridDim2, blockDim2, threads_num2*sizeof(float)>>>\ // (rc->gridDim.x, rc->d_reduced_vec, rc->d_sum); // // // WARNING: launching with original thread_num might be too much. // // SOLUTION: Find power-of-2 nearest to block_num // } // }
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> __global__ void calc_meanshift2(float* y_new, float* y_old, float* meanshift) { int i = blockDim.x*blockIdx.x + threadIdx.x; float tempY_new = y_new[i]; float tempY_old = y_old[i]; meanshift[i] = (tempY_new-tempY_old)*(tempY_new-tempY_old); } __device__ float kernel_fun(float x, float sigma2) { if( x > sigma2) return 0; else return exp(-x/2/sigma2); } __global__ void calc_Kernel_Matrix(int N, int D, float *x, float *y, float *K, int sigma2) { int i = blockDim.x*blockIdx.x + threadIdx.x; int j = blockDim.y*blockIdx.y + threadIdx.y; // It is also true that ( k == (N*i + j) ) // Calc Dist... float dist = 0; for(int d=0; d<D; d++) dist+= (y[i*D+d] - x[j*D+d])*(y[i*D+d] - x[j*D+d]); K[i*N+j] = kernel_fun(dist, sigma2); } __global__ void kernel_sum_div(int D, float* y_new, float* K_sum) { int i = blockDim.x*blockIdx.x + threadIdx.x; for(int d=0; d<D; d++) y_new[i*D+d] = y_new[i*D+d]/K_sum[i]; } __global__ void kernel_Dvec_mult(int N, int D, float* K, float* x, float* Kx, int d) { int i = blockDim.x*blockIdx.x + threadIdx.x; int j = blockDim.y*blockIdx.y + threadIdx.y; Kx[i*N+j] = K[i*N+j]*x[j*D+d]; } __global__ void copy_to_y(int D, float* d_y_new, float* kernelXsum, int d) { int i = blockDim.x*blockIdx.x + threadIdx.x; d_y_new[i*D+d] = kernelXsum[i]; } __global__ void calc_reduce_meanshift(int N, float* y_new, float* y_old, float* reducted_vec) { extern __shared__ float reduction_cache[] ; //thread ID on each row of blocks int tid = blockDim.x * blockIdx.x + threadIdx.x; int cache_i = threadIdx.x; float temp=0; float tempY_new; // This are usuful to ensure that only on GB access is gonna happen for each vector float tempY_old; while (tid < N) { tempY_new = y_new[tid]; tempY_old= y_old[tid]; temp += (tempY_new-tempY_old)*(tempY_new-tempY_old); tid += blockDim.x * gridDim.x; } reduction_cache[cache_i] = temp; __syncthreads(); // Begin the reduction per shared-memory-block for(int i=blockDim.x/2; i>0; i>>=1) { if(cache_i < i) reduction_cache[cache_i] += reduction_cache[cache_i+i]; __syncthreads(); } // Final Sum is stored in global array. if(cache_i==0) reducted_vec[blockIdx.x] = reduction_cache[0]; } // __global__ // void kernelX_dot_product(int N, int D, int d, float* K, float* x, float* reducted_vec) // { // extern __shared__ float reduction_cache[] ; // //thread ID on each row of blocks // int tid = blockDim.x * blockIdx.x + threadIdx.x; // int cache_i = threadIdx.x; // /* This UNROLLS the elements of x, "outside" the grid's index range. // In the case of N=600, threadsPerBlock=256 and 2 blocks in total, // we have 600-256*2=88 additions done in parallel, before the reduction of the 512 threads. // incase the index-range > N, the reduction scheme will simply add some zeros to the vector. // This allows as to oversubscribe in terms of threads and blocks. // */ // int offset = N*blockIdx.y; // float temp=0; // while (tid < N) // { // temp += K[tid+offset]*x[tid*D+d]; // tid += blockDim.x * gridDim.x; // } // /* Load x-data into local shared memory. // As mentioned before, some entries are small sums of // x's outside the grid's range */ // reduction_cache[cache_i] = temp; // __syncthreads(); // // Begin the reduction per shared-memory-block // for(int i=blockDim.x/2; i>0; i>>=1) // { // if(cache_i < i) // reduction_cache[cache_i] += reduction_cache[cache_i+i]; // __syncthreads(); // } // // Final Sum is stored in global array, with stride d, to match the NxD dimensionality of the input dataset. // if(cache_i==0) // reducted_vec[blockIdx.y*gridDim.x + blockIdx.x + d] = reduction_cache[cache_i]; // } // void WR_kernelX_dow_product(int N, float* d_K, float* d_x, /*out*/ ReductionCache* rc ) // { // dim3 blockDim2(4, 1, 1); // dim3 gridDim2(N/4,N,1); // size_t cache_size = 4*N*sizeof(float); // kernelX_dot_product<<<gridDim2, blockDim2, cache_size>>>(N,D,0, d_KernelMatrix, d_x, d_y_new); // kernelX_dot_product<<<gridDim2, blockDim2, cache_size>>>(N,D,1, d_KernelMatrix, d_x, d_y_new); // //reduction_sum<<<L/256, 256, 256*sizeof(float) >>>(N/4, d_y_new, d_y_new); // if(rc->blocksNum == 1) // { // kernelX_dot_product<<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N,D,0, d_K,d_x, rc->d_sum); // kernelX_dot_product<<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N,D,1, d_K,d_x, rc->d_sum); // } // else // { // // We need multiple reduction calls! // reduction_sum <<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N, d_A, rc->d_reduced_vec); // /* Reduct the final reduction vector! */ // // Ideally we would like threads_num==length(reduced_vec)/numRow. // However threads_num2 must be a power of 2. Thus: // int threads_num2 = exp2f(floor(log2f(rc->reduced_vec_length/rc->rowNum))); // if(threads_num2>512) // threads_num2=512; // //printf("THREADS: %d RED_VEC %d\n", threads_num2, rc->reduced_vec_length/rc->rowNum ); // dim3 gridDim2(1,rc->rowNum,1); // dim3 blockDim2(threads_num2,1,1); // reduction_sum<<<gridDim2, blockDim2, threads_num2*sizeof(float)>>>\ // (rc->gridDim.x, rc->d_reduced_vec, rc->d_sum); // // // WARNING: launching with original thread_num might be too much. // // SOLUTION: Find power-of-2 nearest to block_num // } // }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z15calc_meanshift2PfS_S_ .globl _Z15calc_meanshift2PfS_S_ .p2align 8 .type _Z15calc_meanshift2PfS_S_,@function _Z15calc_meanshift2PfS_S_: s_clause 0x2 s_load_b32 s2, s[0:1], 0x24 s_load_b128 s[4:7], s[0:1], 0x0 s_load_b64 s[0:1], s[0:1], 0x10 s_waitcnt lgkmcnt(0) s_and_b32 s2, s2, 0xffff s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mad_u64_u32 v[1:2], null, s15, s2, v[0:1] v_ashrrev_i32_e32 v2, 31, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[0:1], 2, v[1:2] v_add_co_u32 v2, vcc_lo, s4, v0 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v3, vcc_lo, s5, v1, vcc_lo v_add_co_u32 v4, vcc_lo, s6, v0 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v1, vcc_lo v_add_co_u32 v0, vcc_lo, s0, v0 global_load_b32 v2, v[2:3], off global_load_b32 v3, v[4:5], off v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo s_waitcnt vmcnt(0) v_sub_f32_e32 v2, v2, v3 s_delay_alu instid0(VALU_DEP_1) v_mul_f32_e32 v2, v2, v2 global_store_b32 v[0:1], v2, off s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z15calc_meanshift2PfS_S_ .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 280 .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 6 .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 _Z15calc_meanshift2PfS_S_, .Lfunc_end0-_Z15calc_meanshift2PfS_S_ .section .AMDGPU.csdata,"",@progbits .text .protected _Z18calc_Kernel_MatrixiiPfS_S_i .globl _Z18calc_Kernel_MatrixiiPfS_S_i .p2align 8 .type _Z18calc_Kernel_MatrixiiPfS_S_i,@function _Z18calc_Kernel_MatrixiiPfS_S_i: s_clause 0x1 s_load_b32 s3, s[0:1], 0x34 s_load_b32 s2, s[0:1], 0x4 v_and_b32_e32 v2, 0x3ff, v0 v_bfe_u32 v3, v0, 10, 10 s_waitcnt lgkmcnt(0) s_and_b32 s4, s3, 0xffff s_lshr_b32 s3, s3, 16 s_delay_alu instid0(VALU_DEP_1) v_mad_u64_u32 v[0:1], null, s14, s4, v[2:3] v_mad_u64_u32 v[1:2], null, s15, s3, v[3:4] s_cmp_lt_i32 s2, 1 s_cbranch_scc1 .LBB1_3 s_load_b128 s[4:7], s[0:1], 0x8 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_3) v_mul_lo_u32 v2, v1, s2 v_mul_lo_u32 v4, v0, s2 v_mov_b32_e32 v6, 0 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_ashrrev_i32_e32 v3, 31, v2 v_ashrrev_i32_e32 v5, 31, v4 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_lshlrev_b64 v[2:3], 2, v[2:3] v_lshlrev_b64 v[4:5], 2, v[4:5] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_co_u32 v2, vcc_lo, s4, v2 v_add_co_ci_u32_e32 v3, vcc_lo, s5, v3, vcc_lo s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_u32 v4, vcc_lo, s6, v4 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v5, vcc_lo .LBB1_2: global_load_b32 v7, v[4:5], off global_load_b32 v8, v[2:3], off v_add_co_u32 v2, vcc_lo, v2, 4 v_add_co_ci_u32_e32 v3, vcc_lo, 0, v3, vcc_lo v_add_co_u32 v4, vcc_lo, v4, 4 v_add_co_ci_u32_e32 v5, vcc_lo, 0, v5, vcc_lo s_add_i32 s2, s2, -1 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_2) | instid1(VALU_DEP_1) s_cmp_eq_u32 s2, 0 s_waitcnt vmcnt(0) v_sub_f32_e32 v7, v7, v8 v_fmac_f32_e32 v6, v7, v7 s_cbranch_scc0 .LBB1_2 s_branch .LBB1_4 .LBB1_3: v_mov_b32_e32 v6, 0 .LBB1_4: s_load_b32 s2, s[0:1], 0x20 v_mov_b32_e32 v3, 0 s_waitcnt lgkmcnt(0) v_cvt_f32_i32_e32 v2, s2 s_mov_b32 s2, exec_lo s_delay_alu instid0(VALU_DEP_1) v_cmpx_ngt_f32_e32 v6, v2 s_cbranch_execz .LBB1_6 v_mul_f32_e32 v3, -0.5, v6 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_div_scale_f32 v4, null, v2, v2, v3 v_div_scale_f32 v7, vcc_lo, v3, v2, v3 v_rcp_f32_e32 v5, v4 s_waitcnt_depctr 0xfff v_fma_f32 v6, -v4, v5, 1.0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmac_f32_e32 v5, v6, v5 v_mul_f32_e32 v6, v7, v5 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f32 v8, -v4, v6, v7 v_fmac_f32_e32 v6, v8, v5 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fma_f32 v4, -v4, v6, v7 v_div_fmas_f32 v4, v4, v5, v6 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_div_fixup_f32 v2, v4, v2, v3 v_mul_f32_e32 v3, 0x3fb8aa3b, v2 v_cmp_ngt_f32_e32 vcc_lo, 0xc2ce8ed0, v2 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_fma_f32 v4, v2, 0x3fb8aa3b, -v3 v_rndne_f32_e32 v5, v3 v_dual_fmamk_f32 v4, v2, 0x32a5705f, v4 :: v_dual_sub_f32 v3, v3, v5 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_add_f32_e32 v3, v3, v4 v_cvt_i32_f32_e32 v4, v5 v_exp_f32_e32 v3, v3 s_waitcnt_depctr 0xfff v_ldexp_f32 v3, v3, v4 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_cndmask_b32_e32 v3, 0, v3, vcc_lo v_cmp_nlt_f32_e32 vcc_lo, 0x42b17218, v2 v_cndmask_b32_e32 v3, 0x7f800000, v3, vcc_lo .LBB1_6: s_or_b32 exec_lo, exec_lo, s2 s_clause 0x1 s_load_b32 s2, s[0:1], 0x0 s_load_b64 s[0:1], s[0:1], 0x18 s_waitcnt lgkmcnt(0) v_mad_u64_u32 v[4:5], null, v0, s2, v[1:2] s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_ashrrev_i32_e32 v5, 31, v4 v_lshlrev_b64 v[0:1], 2, v[4:5] 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 s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z18calc_Kernel_MatrixiiPfS_S_i .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_end1: .size _Z18calc_Kernel_MatrixiiPfS_S_i, .Lfunc_end1-_Z18calc_Kernel_MatrixiiPfS_S_i .section .AMDGPU.csdata,"",@progbits .text .protected _Z14kernel_sum_diviPfS_ .globl _Z14kernel_sum_diviPfS_ .p2align 8 .type _Z14kernel_sum_diviPfS_,@function _Z14kernel_sum_diviPfS_: s_load_b32 s2, s[0:1], 0x0 s_waitcnt lgkmcnt(0) s_cmp_lt_i32 s2, 1 s_cbranch_scc1 .LBB2_3 s_clause 0x1 s_load_b32 s3, s[0:1], 0x24 s_load_b128 s[4:7], s[0:1], 0x8 s_waitcnt lgkmcnt(0) s_and_b32 s3, s3, 0xffff s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mad_u64_u32 v[1:2], null, s15, s3, v[0:1] v_mul_lo_u32 v3, v1, s2 v_ashrrev_i32_e32 v2, 31, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_3) v_lshlrev_b64 v[0:1], 2, v[1:2] v_ashrrev_i32_e32 v4, 31, v3 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_u32 v0, vcc_lo, s6, v0 v_lshlrev_b64 v[2:3], 2, v[3:4] s_delay_alu instid0(VALU_DEP_4) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_co_ci_u32_e32 v1, vcc_lo, s7, v1, vcc_lo v_add_co_u32 v2, vcc_lo, s4, v2 s_delay_alu instid0(VALU_DEP_3) v_add_co_ci_u32_e32 v3, vcc_lo, s5, v3, vcc_lo .p2align 6 .LBB2_2: global_load_b32 v4, v[2:3], off global_load_b32 v5, v[0:1], off s_add_i32 s2, s2, -1 s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_3) | instid1(VALU_DEP_2) s_cmp_eq_u32 s2, 0 s_waitcnt vmcnt(0) v_div_scale_f32 v6, null, v5, v5, v4 v_div_scale_f32 v9, vcc_lo, v4, v5, v4 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) v_div_fixup_f32 v4, v6, v5, v4 global_store_b32 v[2:3], v4, off v_add_co_u32 v2, vcc_lo, v2, 4 v_add_co_ci_u32_e32 v3, vcc_lo, 0, v3, vcc_lo s_cbranch_scc0 .LBB2_2 .LBB2_3: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z14kernel_sum_diviPfS_ .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 280 .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_end2: .size _Z14kernel_sum_diviPfS_, .Lfunc_end2-_Z14kernel_sum_diviPfS_ .section .AMDGPU.csdata,"",@progbits .text .protected _Z16kernel_Dvec_multiiPfS_S_i .globl _Z16kernel_Dvec_multiiPfS_S_i .p2align 8 .type _Z16kernel_Dvec_multiiPfS_S_i,@function _Z16kernel_Dvec_multiiPfS_S_i: s_clause 0x2 s_load_b32 s2, s[0:1], 0x34 s_load_b256 s[4:11], s[0:1], 0x0 s_load_b32 s0, s[0:1], 0x20 v_and_b32_e32 v1, 0x3ff, v0 v_bfe_u32 v0, v0, 10, 10 s_waitcnt lgkmcnt(0) s_and_b32 s1, s2, 0xffff s_lshr_b32 s2, s2, 16 v_mad_u64_u32 v[2:3], null, s14, s1, v[1:2] v_mad_u64_u32 v[3:4], null, s15, s2, v[0:1] s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_2) v_mad_u64_u32 v[0:1], null, v2, s4, v[3:4] v_mad_u64_u32 v[4:5], null, v3, s5, s[0:1] v_ashrrev_i32_e32 v1, 31, v0 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_ashrrev_i32_e32 v5, 31, v4 v_lshlrev_b64 v[0:1], 2, v[0:1] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_lshlrev_b64 v[2:3], 2, v[4:5] v_add_co_u32 v4, vcc_lo, s6, v0 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_co_ci_u32_e32 v5, vcc_lo, s7, v1, vcc_lo v_add_co_u32 v2, vcc_lo, s8, v2 s_delay_alu instid0(VALU_DEP_4) v_add_co_ci_u32_e32 v3, vcc_lo, s9, v3, vcc_lo global_load_b32 v4, v[4:5], off global_load_b32 v2, v[2:3], off v_add_co_u32 v0, vcc_lo, s10, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s11, v1, vcc_lo s_waitcnt vmcnt(0) v_mul_f32_e32 v2, v4, v2 global_store_b32 v[0:1], v2, off s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z16kernel_Dvec_multiiPfS_S_i .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 6 .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_end3: .size _Z16kernel_Dvec_multiiPfS_S_i, .Lfunc_end3-_Z16kernel_Dvec_multiiPfS_S_i .section .AMDGPU.csdata,"",@progbits .text .protected _Z9copy_to_yiPfS_i .globl _Z9copy_to_yiPfS_i .p2align 8 .type _Z9copy_to_yiPfS_i,@function _Z9copy_to_yiPfS_i: s_clause 0x1 s_load_b32 s2, s[0:1], 0x2c s_load_b128 s[4:7], s[0:1], 0x8 s_waitcnt lgkmcnt(0) s_and_b32 s2, s2, 0xffff s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_3) | instid1(VALU_DEP_1) v_mad_u64_u32 v[1:2], null, s15, s2, v[0:1] s_clause 0x1 s_load_b32 s2, s[0:1], 0x18 s_load_b32 s0, s[0:1], 0x0 v_ashrrev_i32_e32 v2, 31, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[2:3], 2, v[1:2] v_add_co_u32 v2, vcc_lo, s6, v2 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_3) | instid1(VALU_DEP_1) v_add_co_ci_u32_e32 v3, vcc_lo, s7, v3, vcc_lo global_load_b32 v4, v[2:3], off s_waitcnt lgkmcnt(0) v_mad_u64_u32 v[2:3], null, v1, s0, s[2:3] v_ashrrev_i32_e32 v3, 31, v2 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[0:1], 2, v[2:3] v_add_co_u32 v0, vcc_lo, s4, v0 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v1, vcc_lo, s5, v1, vcc_lo s_waitcnt vmcnt(0) global_store_b32 v[0:1], v4, off s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z9copy_to_yiPfS_i .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 5 .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_end4: .size _Z9copy_to_yiPfS_i, .Lfunc_end4-_Z9copy_to_yiPfS_i .section .AMDGPU.csdata,"",@progbits .text .protected _Z21calc_reduce_meanshiftiPfS_S_ .globl _Z21calc_reduce_meanshiftiPfS_S_ .p2align 8 .type _Z21calc_reduce_meanshiftiPfS_S_,@function _Z21calc_reduce_meanshiftiPfS_S_: s_clause 0x1 s_load_b32 s3, s[0:1], 0x2c s_load_b32 s8, s[0:1], 0x0 s_add_u32 s4, s0, 32 s_mov_b32 s2, s15 s_addc_u32 s5, s1, 0 v_mov_b32_e32 v3, 0 s_mov_b32 s9, exec_lo s_waitcnt lgkmcnt(0) s_and_b32 s3, s3, 0xffff s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_mad_u64_u32 v[1:2], null, s2, s3, v[0:1] v_cmpx_gt_i32_e64 s8, v1 s_cbranch_execz .LBB5_4 s_load_b32 s11, s[4:5], 0x0 s_load_b128 s[4:7], s[0:1], 0x8 v_mov_b32_e32 v3, 0 s_mov_b32 s10, 0 s_waitcnt lgkmcnt(0) s_mul_i32 s11, s11, s3 .p2align 6 .LBB5_2: v_ashrrev_i32_e32 v2, 31, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[4:5], 2, v[1:2] v_add_co_u32 v6, vcc_lo, s4, v4 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v7, vcc_lo, s5, v5, vcc_lo v_add_co_u32 v4, vcc_lo, s6, v4 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v5, vcc_lo global_load_b32 v2, v[6:7], off global_load_b32 v4, v[4:5], off s_waitcnt vmcnt(0) v_dual_sub_f32 v2, v2, v4 :: v_dual_add_nc_u32 v1, s11, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_2) v_cmp_le_i32_e32 vcc_lo, s8, v1 v_fmac_f32_e32 v3, v2, v2 s_or_b32 s10, vcc_lo, s10 s_delay_alu instid0(SALU_CYCLE_1) s_and_not1_b32 exec_lo, exec_lo, s10 s_cbranch_execnz .LBB5_2 s_or_b32 exec_lo, exec_lo, s10 .LBB5_4: s_delay_alu instid0(SALU_CYCLE_1) s_or_b32 exec_lo, exec_lo, s9 v_lshl_add_u32 v1, v0, 2, 0 s_cmp_lt_u32 s3, 2 ds_store_b32 v1, v3 s_waitcnt lgkmcnt(0) s_barrier s_branch .LBB5_6 .p2align 6 .LBB5_5: s_or_b32 exec_lo, exec_lo, s5 s_waitcnt lgkmcnt(0) s_barrier s_cmp_lt_u32 s3, 4 s_mov_b32 s3, s4 .LBB5_6: buffer_gl0_inv s_cbranch_scc1 .LBB5_9 s_lshr_b32 s4, s3, 1 s_mov_b32 s5, exec_lo v_cmpx_gt_u32_e64 s4, v0 s_cbranch_execz .LBB5_5 v_add_nc_u32_e32 v2, s4, v0 s_delay_alu instid0(VALU_DEP_1) v_lshl_add_u32 v2, v2, 2, 0 ds_load_b32 v2, v2 ds_load_b32 v3, v1 s_waitcnt lgkmcnt(0) v_add_f32_e32 v2, v2, v3 ds_store_b32 v1, v2 s_branch .LBB5_5 .LBB5_9: s_mov_b32 s3, exec_lo v_cmpx_eq_u32_e32 0, v0 s_cbranch_execz .LBB5_11 v_dual_mov_b32 v0, 0 :: v_dual_mov_b32 v1, 0 s_load_b64 s[0:1], s[0:1], 0x18 s_mov_b32 s3, 0 s_delay_alu instid0(SALU_CYCLE_1) s_lshl_b64 s[2:3], s[2:3], 2 ds_load_b32 v0, v0 s_waitcnt lgkmcnt(0) s_add_u32 s0, s0, s2 s_addc_u32 s1, s1, s3 global_store_b32 v1, v0, s[0:1] .LBB5_11: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z21calc_reduce_meanshiftiPfS_S_ .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 8 .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_end5: .size _Z21calc_reduce_meanshiftiPfS_S_, .Lfunc_end5-_Z21calc_reduce_meanshiftiPfS_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: - .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: hidden_block_count_x - .offset: 28 .size: 4 .value_kind: hidden_block_count_y - .offset: 32 .size: 4 .value_kind: hidden_block_count_z - .offset: 36 .size: 2 .value_kind: hidden_group_size_x - .offset: 38 .size: 2 .value_kind: hidden_group_size_y - .offset: 40 .size: 2 .value_kind: hidden_group_size_z - .offset: 42 .size: 2 .value_kind: hidden_remainder_x - .offset: 44 .size: 2 .value_kind: hidden_remainder_y - .offset: 46 .size: 2 .value_kind: hidden_remainder_z - .offset: 64 .size: 8 .value_kind: hidden_global_offset_x - .offset: 72 .size: 8 .value_kind: hidden_global_offset_y - .offset: 80 .size: 8 .value_kind: hidden_global_offset_z - .offset: 88 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 280 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z15calc_meanshift2PfS_S_ .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z15calc_meanshift2PfS_S_.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 6 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .args: - .offset: 0 .size: 4 .value_kind: by_value - .offset: 4 .size: 4 .value_kind: by_value - .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 - .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: _Z18calc_Kernel_MatrixiiPfS_S_i .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z18calc_Kernel_MatrixiiPfS_S_i.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 9 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .args: - .offset: 0 .size: 4 .value_kind: by_value - .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: hidden_block_count_x - .offset: 28 .size: 4 .value_kind: hidden_block_count_y - .offset: 32 .size: 4 .value_kind: hidden_block_count_z - .offset: 36 .size: 2 .value_kind: hidden_group_size_x - .offset: 38 .size: 2 .value_kind: hidden_group_size_y - .offset: 40 .size: 2 .value_kind: hidden_group_size_z - .offset: 42 .size: 2 .value_kind: hidden_remainder_x - .offset: 44 .size: 2 .value_kind: hidden_remainder_y - .offset: 46 .size: 2 .value_kind: hidden_remainder_z - .offset: 64 .size: 8 .value_kind: hidden_global_offset_x - .offset: 72 .size: 8 .value_kind: hidden_global_offset_y - .offset: 80 .size: 8 .value_kind: hidden_global_offset_z - .offset: 88 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 280 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z14kernel_sum_diviPfS_ .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z14kernel_sum_diviPfS_.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 11 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .args: - .offset: 0 .size: 4 .value_kind: by_value - .offset: 4 .size: 4 .value_kind: by_value - .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 - .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: _Z16kernel_Dvec_multiiPfS_S_i .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z16kernel_Dvec_multiiPfS_S_i.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 6 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .args: - .offset: 0 .size: 4 .value_kind: by_value - .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: _Z9copy_to_yiPfS_i .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z9copy_to_yiPfS_i.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 5 .vgpr_spill_count: 0 .wavefront_size: 32 .workgroup_processor_mode: 1 - .args: - .offset: 0 .size: 4 .value_kind: by_value - .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 - .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: _Z21calc_reduce_meanshiftiPfS_S_ .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z21calc_reduce_meanshiftiPfS_S_.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 8 .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> __global__ void calc_meanshift2(float* y_new, float* y_old, float* meanshift) { int i = blockDim.x*blockIdx.x + threadIdx.x; float tempY_new = y_new[i]; float tempY_old = y_old[i]; meanshift[i] = (tempY_new-tempY_old)*(tempY_new-tempY_old); } __device__ float kernel_fun(float x, float sigma2) { if( x > sigma2) return 0; else return exp(-x/2/sigma2); } __global__ void calc_Kernel_Matrix(int N, int D, float *x, float *y, float *K, int sigma2) { int i = blockDim.x*blockIdx.x + threadIdx.x; int j = blockDim.y*blockIdx.y + threadIdx.y; // It is also true that ( k == (N*i + j) ) // Calc Dist... float dist = 0; for(int d=0; d<D; d++) dist+= (y[i*D+d] - x[j*D+d])*(y[i*D+d] - x[j*D+d]); K[i*N+j] = kernel_fun(dist, sigma2); } __global__ void kernel_sum_div(int D, float* y_new, float* K_sum) { int i = blockDim.x*blockIdx.x + threadIdx.x; for(int d=0; d<D; d++) y_new[i*D+d] = y_new[i*D+d]/K_sum[i]; } __global__ void kernel_Dvec_mult(int N, int D, float* K, float* x, float* Kx, int d) { int i = blockDim.x*blockIdx.x + threadIdx.x; int j = blockDim.y*blockIdx.y + threadIdx.y; Kx[i*N+j] = K[i*N+j]*x[j*D+d]; } __global__ void copy_to_y(int D, float* d_y_new, float* kernelXsum, int d) { int i = blockDim.x*blockIdx.x + threadIdx.x; d_y_new[i*D+d] = kernelXsum[i]; } __global__ void calc_reduce_meanshift(int N, float* y_new, float* y_old, float* reducted_vec) { extern __shared__ float reduction_cache[] ; //thread ID on each row of blocks int tid = blockDim.x * blockIdx.x + threadIdx.x; int cache_i = threadIdx.x; float temp=0; float tempY_new; // This are usuful to ensure that only on GB access is gonna happen for each vector float tempY_old; while (tid < N) { tempY_new = y_new[tid]; tempY_old= y_old[tid]; temp += (tempY_new-tempY_old)*(tempY_new-tempY_old); tid += blockDim.x * gridDim.x; } reduction_cache[cache_i] = temp; __syncthreads(); // Begin the reduction per shared-memory-block for(int i=blockDim.x/2; i>0; i>>=1) { if(cache_i < i) reduction_cache[cache_i] += reduction_cache[cache_i+i]; __syncthreads(); } // Final Sum is stored in global array. if(cache_i==0) reducted_vec[blockIdx.x] = reduction_cache[0]; } // __global__ // void kernelX_dot_product(int N, int D, int d, float* K, float* x, float* reducted_vec) // { // extern __shared__ float reduction_cache[] ; // //thread ID on each row of blocks // int tid = blockDim.x * blockIdx.x + threadIdx.x; // int cache_i = threadIdx.x; // /* This UNROLLS the elements of x, "outside" the grid's index range. // In the case of N=600, threadsPerBlock=256 and 2 blocks in total, // we have 600-256*2=88 additions done in parallel, before the reduction of the 512 threads. // incase the index-range > N, the reduction scheme will simply add some zeros to the vector. // This allows as to oversubscribe in terms of threads and blocks. // */ // int offset = N*blockIdx.y; // float temp=0; // while (tid < N) // { // temp += K[tid+offset]*x[tid*D+d]; // tid += blockDim.x * gridDim.x; // } // /* Load x-data into local shared memory. // As mentioned before, some entries are small sums of // x's outside the grid's range */ // reduction_cache[cache_i] = temp; // __syncthreads(); // // Begin the reduction per shared-memory-block // for(int i=blockDim.x/2; i>0; i>>=1) // { // if(cache_i < i) // reduction_cache[cache_i] += reduction_cache[cache_i+i]; // __syncthreads(); // } // // Final Sum is stored in global array, with stride d, to match the NxD dimensionality of the input dataset. // if(cache_i==0) // reducted_vec[blockIdx.y*gridDim.x + blockIdx.x + d] = reduction_cache[cache_i]; // } // void WR_kernelX_dow_product(int N, float* d_K, float* d_x, /*out*/ ReductionCache* rc ) // { // dim3 blockDim2(4, 1, 1); // dim3 gridDim2(N/4,N,1); // size_t cache_size = 4*N*sizeof(float); // kernelX_dot_product<<<gridDim2, blockDim2, cache_size>>>(N,D,0, d_KernelMatrix, d_x, d_y_new); // kernelX_dot_product<<<gridDim2, blockDim2, cache_size>>>(N,D,1, d_KernelMatrix, d_x, d_y_new); // //reduction_sum<<<L/256, 256, 256*sizeof(float) >>>(N/4, d_y_new, d_y_new); // if(rc->blocksNum == 1) // { // kernelX_dot_product<<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N,D,0, d_K,d_x, rc->d_sum); // kernelX_dot_product<<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N,D,1, d_K,d_x, rc->d_sum); // } // else // { // // We need multiple reduction calls! // reduction_sum <<<rc->gridDim, rc->blockDim, rc->cache_size>>>(N, d_A, rc->d_reduced_vec); // /* Reduct the final reduction vector! */ // // Ideally we would like threads_num==length(reduced_vec)/numRow. // However threads_num2 must be a power of 2. Thus: // int threads_num2 = exp2f(floor(log2f(rc->reduced_vec_length/rc->rowNum))); // if(threads_num2>512) // threads_num2=512; // //printf("THREADS: %d RED_VEC %d\n", threads_num2, rc->reduced_vec_length/rc->rowNum ); // dim3 gridDim2(1,rc->rowNum,1); // dim3 blockDim2(threads_num2,1,1); // reduction_sum<<<gridDim2, blockDim2, threads_num2*sizeof(float)>>>\ // (rc->gridDim.x, rc->d_reduced_vec, rc->d_sum); // // // WARNING: launching with original thread_num might be too much. // // SOLUTION: Find power-of-2 nearest to block_num // } // }
.text .file "cuda_meanshift.hip" .globl _Z30__device_stub__calc_meanshift2PfS_S_ # -- Begin function _Z30__device_stub__calc_meanshift2PfS_S_ .p2align 4, 0x90 .type _Z30__device_stub__calc_meanshift2PfS_S_,@function _Z30__device_stub__calc_meanshift2PfS_S_: # @_Z30__device_stub__calc_meanshift2PfS_S_ .cfi_startproc # %bb.0: subq $104, %rsp .cfi_def_cfa_offset 112 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movq %rdx, 56(%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 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 $_Z15calc_meanshift2PfS_S_, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $120, %rsp .cfi_adjust_cfa_offset -120 retq .Lfunc_end0: .size _Z30__device_stub__calc_meanshift2PfS_S_, .Lfunc_end0-_Z30__device_stub__calc_meanshift2PfS_S_ .cfi_endproc # -- End function .globl _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i # -- Begin function _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i .p2align 4, 0x90 .type _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i,@function _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i: # @_Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movl %edi, 20(%rsp) movl %esi, 16(%rsp) movq %rdx, 88(%rsp) movq %rcx, 80(%rsp) movq %r8, 72(%rsp) movl %r9d, 12(%rsp) leaq 20(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 88(%rsp), %rax movq %rax, 112(%rsp) leaq 80(%rsp), %rax movq %rax, 120(%rsp) leaq 72(%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 $_Z18calc_Kernel_MatrixiiPfS_S_i, %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_end1: .size _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i, .Lfunc_end1-_Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i .cfi_endproc # -- End function .globl _Z29__device_stub__kernel_sum_diviPfS_ # -- Begin function _Z29__device_stub__kernel_sum_diviPfS_ .p2align 4, 0x90 .type _Z29__device_stub__kernel_sum_diviPfS_,@function _Z29__device_stub__kernel_sum_diviPfS_: # @_Z29__device_stub__kernel_sum_diviPfS_ .cfi_startproc # %bb.0: subq $104, %rsp .cfi_def_cfa_offset 112 movl %edi, 12(%rsp) movq %rsi, 72(%rsp) movq %rdx, 64(%rsp) leaq 12(%rsp), %rax movq %rax, 80(%rsp) leaq 72(%rsp), %rax movq %rax, 88(%rsp) leaq 64(%rsp), %rax movq %rax, 96(%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 $_Z14kernel_sum_diviPfS_, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $120, %rsp .cfi_adjust_cfa_offset -120 retq .Lfunc_end2: .size _Z29__device_stub__kernel_sum_diviPfS_, .Lfunc_end2-_Z29__device_stub__kernel_sum_diviPfS_ .cfi_endproc # -- End function .globl _Z31__device_stub__kernel_Dvec_multiiPfS_S_i # -- Begin function _Z31__device_stub__kernel_Dvec_multiiPfS_S_i .p2align 4, 0x90 .type _Z31__device_stub__kernel_Dvec_multiiPfS_S_i,@function _Z31__device_stub__kernel_Dvec_multiiPfS_S_i: # @_Z31__device_stub__kernel_Dvec_multiiPfS_S_i .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movl %edi, 20(%rsp) movl %esi, 16(%rsp) movq %rdx, 88(%rsp) movq %rcx, 80(%rsp) movq %r8, 72(%rsp) movl %r9d, 12(%rsp) leaq 20(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 88(%rsp), %rax movq %rax, 112(%rsp) leaq 80(%rsp), %rax movq %rax, 120(%rsp) leaq 72(%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 $_Z16kernel_Dvec_multiiPfS_S_i, %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_end3: .size _Z31__device_stub__kernel_Dvec_multiiPfS_S_i, .Lfunc_end3-_Z31__device_stub__kernel_Dvec_multiiPfS_S_i .cfi_endproc # -- End function .globl _Z24__device_stub__copy_to_yiPfS_i # -- Begin function _Z24__device_stub__copy_to_yiPfS_i .p2align 4, 0x90 .type _Z24__device_stub__copy_to_yiPfS_i,@function _Z24__device_stub__copy_to_yiPfS_i: # @_Z24__device_stub__copy_to_yiPfS_i .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movl %edi, 12(%rsp) movq %rsi, 72(%rsp) movq %rdx, 64(%rsp) movl %ecx, 8(%rsp) leaq 12(%rsp), %rax movq %rax, 80(%rsp) leaq 72(%rsp), %rax movq %rax, 88(%rsp) leaq 64(%rsp), %rax movq %rax, 96(%rsp) leaq 8(%rsp), %rax movq %rax, 104(%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 $_Z9copy_to_yiPfS_i, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $136, %rsp .cfi_adjust_cfa_offset -136 retq .Lfunc_end4: .size _Z24__device_stub__copy_to_yiPfS_i, .Lfunc_end4-_Z24__device_stub__copy_to_yiPfS_i .cfi_endproc # -- End function .globl _Z36__device_stub__calc_reduce_meanshiftiPfS_S_ # -- Begin function _Z36__device_stub__calc_reduce_meanshiftiPfS_S_ .p2align 4, 0x90 .type _Z36__device_stub__calc_reduce_meanshiftiPfS_S_,@function _Z36__device_stub__calc_reduce_meanshiftiPfS_S_: # @_Z36__device_stub__calc_reduce_meanshiftiPfS_S_ .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movl %edi, 4(%rsp) movq %rsi, 72(%rsp) movq %rdx, 64(%rsp) movq %rcx, 56(%rsp) leaq 4(%rsp), %rax movq %rax, 80(%rsp) leaq 72(%rsp), %rax movq %rax, 88(%rsp) leaq 64(%rsp), %rax movq %rax, 96(%rsp) leaq 56(%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 $_Z21calc_reduce_meanshiftiPfS_S_, %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_end5: .size _Z36__device_stub__calc_reduce_meanshiftiPfS_S_, .Lfunc_end5-_Z36__device_stub__calc_reduce_meanshiftiPfS_S_ .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: pushq %rbx .cfi_def_cfa_offset 16 subq $32, %rsp .cfi_def_cfa_offset 48 .cfi_offset %rbx, -16 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB6_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB6_2: movq __hip_gpubin_handle(%rip), %rbx xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z15calc_meanshift2PfS_S_, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z18calc_Kernel_MatrixiiPfS_S_i, %esi movl $.L__unnamed_2, %edx movl $.L__unnamed_2, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z14kernel_sum_diviPfS_, %esi movl $.L__unnamed_3, %edx movl $.L__unnamed_3, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z16kernel_Dvec_multiiPfS_S_i, %esi movl $.L__unnamed_4, %edx movl $.L__unnamed_4, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z9copy_to_yiPfS_i, %esi movl $.L__unnamed_5, %edx movl $.L__unnamed_5, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z21calc_reduce_meanshiftiPfS_S_, %esi movl $.L__unnamed_6, %edx movl $.L__unnamed_6, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $32, %rsp .cfi_def_cfa_offset 16 popq %rbx .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end6: .size __hip_module_ctor, .Lfunc_end6-__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 .LBB7_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 .LBB7_2: retq .Lfunc_end7: .size __hip_module_dtor, .Lfunc_end7-__hip_module_dtor .cfi_endproc # -- End function .type _Z15calc_meanshift2PfS_S_,@object # @_Z15calc_meanshift2PfS_S_ .section .rodata,"a",@progbits .globl _Z15calc_meanshift2PfS_S_ .p2align 3, 0x0 _Z15calc_meanshift2PfS_S_: .quad _Z30__device_stub__calc_meanshift2PfS_S_ .size _Z15calc_meanshift2PfS_S_, 8 .type _Z18calc_Kernel_MatrixiiPfS_S_i,@object # @_Z18calc_Kernel_MatrixiiPfS_S_i .globl _Z18calc_Kernel_MatrixiiPfS_S_i .p2align 3, 0x0 _Z18calc_Kernel_MatrixiiPfS_S_i: .quad _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i .size _Z18calc_Kernel_MatrixiiPfS_S_i, 8 .type _Z14kernel_sum_diviPfS_,@object # @_Z14kernel_sum_diviPfS_ .globl _Z14kernel_sum_diviPfS_ .p2align 3, 0x0 _Z14kernel_sum_diviPfS_: .quad _Z29__device_stub__kernel_sum_diviPfS_ .size _Z14kernel_sum_diviPfS_, 8 .type _Z16kernel_Dvec_multiiPfS_S_i,@object # @_Z16kernel_Dvec_multiiPfS_S_i .globl _Z16kernel_Dvec_multiiPfS_S_i .p2align 3, 0x0 _Z16kernel_Dvec_multiiPfS_S_i: .quad _Z31__device_stub__kernel_Dvec_multiiPfS_S_i .size _Z16kernel_Dvec_multiiPfS_S_i, 8 .type _Z9copy_to_yiPfS_i,@object # @_Z9copy_to_yiPfS_i .globl _Z9copy_to_yiPfS_i .p2align 3, 0x0 _Z9copy_to_yiPfS_i: .quad _Z24__device_stub__copy_to_yiPfS_i .size _Z9copy_to_yiPfS_i, 8 .type _Z21calc_reduce_meanshiftiPfS_S_,@object # @_Z21calc_reduce_meanshiftiPfS_S_ .globl _Z21calc_reduce_meanshiftiPfS_S_ .p2align 3, 0x0 _Z21calc_reduce_meanshiftiPfS_S_: .quad _Z36__device_stub__calc_reduce_meanshiftiPfS_S_ .size _Z21calc_reduce_meanshiftiPfS_S_, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z15calc_meanshift2PfS_S_" .size .L__unnamed_1, 26 .type .L__unnamed_2,@object # @1 .L__unnamed_2: .asciz "_Z18calc_Kernel_MatrixiiPfS_S_i" .size .L__unnamed_2, 32 .type .L__unnamed_3,@object # @2 .L__unnamed_3: .asciz "_Z14kernel_sum_diviPfS_" .size .L__unnamed_3, 24 .type .L__unnamed_4,@object # @3 .L__unnamed_4: .asciz "_Z16kernel_Dvec_multiiPfS_S_i" .size .L__unnamed_4, 30 .type .L__unnamed_5,@object # @4 .L__unnamed_5: .asciz "_Z9copy_to_yiPfS_i" .size .L__unnamed_5, 19 .type .L__unnamed_6,@object # @5 .L__unnamed_6: .asciz "_Z21calc_reduce_meanshiftiPfS_S_" .size .L__unnamed_6, 33 .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__calc_meanshift2PfS_S_ .addrsig_sym _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i .addrsig_sym _Z29__device_stub__kernel_sum_diviPfS_ .addrsig_sym _Z31__device_stub__kernel_Dvec_multiiPfS_S_i .addrsig_sym _Z24__device_stub__copy_to_yiPfS_i .addrsig_sym _Z36__device_stub__calc_reduce_meanshiftiPfS_S_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z15calc_meanshift2PfS_S_ .addrsig_sym _Z18calc_Kernel_MatrixiiPfS_S_i .addrsig_sym _Z14kernel_sum_diviPfS_ .addrsig_sym _Z16kernel_Dvec_multiiPfS_S_i .addrsig_sym _Z9copy_to_yiPfS_i .addrsig_sym _Z21calc_reduce_meanshiftiPfS_S_ .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_000c76d3_00000000-6_cuda_meanshift.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 _Z10kernel_funff .type _Z10kernel_funff, @function _Z10kernel_funff: .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 _Z10kernel_funff, .-_Z10kernel_funff .globl _Z39__device_stub__Z15calc_meanshift2PfS_S_PfS_S_ .type _Z39__device_stub__Z15calc_meanshift2PfS_S_PfS_S_, @function _Z39__device_stub__Z15calc_meanshift2PfS_S_PfS_S_: .LFB2082: .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 .L9 .L5: movq 120(%rsp), %rax subq %fs:40, %rax jne .L10 addq $136, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L9: .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 _Z15calc_meanshift2PfS_S_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 144 jmp .L5 .L10: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _Z39__device_stub__Z15calc_meanshift2PfS_S_PfS_S_, .-_Z39__device_stub__Z15calc_meanshift2PfS_S_PfS_S_ .globl _Z15calc_meanshift2PfS_S_ .type _Z15calc_meanshift2PfS_S_, @function _Z15calc_meanshift2PfS_S_: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z39__device_stub__Z15calc_meanshift2PfS_S_PfS_S_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _Z15calc_meanshift2PfS_S_, .-_Z15calc_meanshift2PfS_S_ .globl _Z45__device_stub__Z18calc_Kernel_MatrixiiPfS_S_iiiPfS_S_i .type _Z45__device_stub__Z18calc_Kernel_MatrixiiPfS_S_iiiPfS_S_i, @function _Z45__device_stub__Z18calc_Kernel_MatrixiiPfS_S_iiiPfS_S_i: .LFB2084: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movl %edi, 44(%rsp) movl %esi, 40(%rsp) movq %rdx, 32(%rsp) movq %rcx, 24(%rsp) movq %r8, 16(%rsp) movl %r9d, 12(%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 32(%rsp), %rax movq %rax, 128(%rsp) leaq 24(%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 .L17 .L13: movq 168(%rsp), %rax subq %fs:40, %rax jne .L18 addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L17: .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 _Z18calc_Kernel_MatrixiiPfS_S_i(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 192 jmp .L13 .L18: call __stack_chk_fail@PLT .cfi_endproc .LFE2084: .size _Z45__device_stub__Z18calc_Kernel_MatrixiiPfS_S_iiiPfS_S_i, .-_Z45__device_stub__Z18calc_Kernel_MatrixiiPfS_S_iiiPfS_S_i .globl _Z18calc_Kernel_MatrixiiPfS_S_i .type _Z18calc_Kernel_MatrixiiPfS_S_i, @function _Z18calc_Kernel_MatrixiiPfS_S_i: .LFB2085: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z45__device_stub__Z18calc_Kernel_MatrixiiPfS_S_iiiPfS_S_i addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2085: .size _Z18calc_Kernel_MatrixiiPfS_S_i, .-_Z18calc_Kernel_MatrixiiPfS_S_i .globl _Z37__device_stub__Z14kernel_sum_diviPfS_iPfS_ .type _Z37__device_stub__Z14kernel_sum_diviPfS_iPfS_, @function _Z37__device_stub__Z14kernel_sum_diviPfS_iPfS_: .LFB2086: .cfi_startproc endbr64 subq $136, %rsp .cfi_def_cfa_offset 144 movl %edi, 28(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movq %fs:40, %rax movq %rax, 120(%rsp) xorl %eax, %eax leaq 28(%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 .L25 .L21: movq 120(%rsp), %rax subq %fs:40, %rax jne .L26 addq $136, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L25: .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 _Z14kernel_sum_diviPfS_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 144 jmp .L21 .L26: call __stack_chk_fail@PLT .cfi_endproc .LFE2086: .size _Z37__device_stub__Z14kernel_sum_diviPfS_iPfS_, .-_Z37__device_stub__Z14kernel_sum_diviPfS_iPfS_ .globl _Z14kernel_sum_diviPfS_ .type _Z14kernel_sum_diviPfS_, @function _Z14kernel_sum_diviPfS_: .LFB2087: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z37__device_stub__Z14kernel_sum_diviPfS_iPfS_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2087: .size _Z14kernel_sum_diviPfS_, .-_Z14kernel_sum_diviPfS_ .globl _Z43__device_stub__Z16kernel_Dvec_multiiPfS_S_iiiPfS_S_i .type _Z43__device_stub__Z16kernel_Dvec_multiiPfS_S_iiiPfS_S_i, @function _Z43__device_stub__Z16kernel_Dvec_multiiPfS_S_iiiPfS_S_i: .LFB2088: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movl %edi, 44(%rsp) movl %esi, 40(%rsp) movq %rdx, 32(%rsp) movq %rcx, 24(%rsp) movq %r8, 16(%rsp) movl %r9d, 12(%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 32(%rsp), %rax movq %rax, 128(%rsp) leaq 24(%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 .L33 .L29: movq 168(%rsp), %rax subq %fs:40, %rax jne .L34 addq $184, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L33: .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 _Z16kernel_Dvec_multiiPfS_S_i(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 192 jmp .L29 .L34: call __stack_chk_fail@PLT .cfi_endproc .LFE2088: .size _Z43__device_stub__Z16kernel_Dvec_multiiPfS_S_iiiPfS_S_i, .-_Z43__device_stub__Z16kernel_Dvec_multiiPfS_S_iiiPfS_S_i .globl _Z16kernel_Dvec_multiiPfS_S_i .type _Z16kernel_Dvec_multiiPfS_S_i, @function _Z16kernel_Dvec_multiiPfS_S_i: .LFB2089: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z43__device_stub__Z16kernel_Dvec_multiiPfS_S_iiiPfS_S_i addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2089: .size _Z16kernel_Dvec_multiiPfS_S_i, .-_Z16kernel_Dvec_multiiPfS_S_i .globl _Z32__device_stub__Z9copy_to_yiPfS_iiPfS_i .type _Z32__device_stub__Z9copy_to_yiPfS_iiPfS_i, @function _Z32__device_stub__Z9copy_to_yiPfS_iiPfS_i: .LFB2090: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movl %edi, 28(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movl %ecx, 24(%rsp) movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax leaq 28(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) leaq 24(%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 .L41 .L37: movq 136(%rsp), %rax subq %fs:40, %rax jne .L42 addq $152, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L41: .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 _Z9copy_to_yiPfS_i(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L37 .L42: call __stack_chk_fail@PLT .cfi_endproc .LFE2090: .size _Z32__device_stub__Z9copy_to_yiPfS_iiPfS_i, .-_Z32__device_stub__Z9copy_to_yiPfS_iiPfS_i .globl _Z9copy_to_yiPfS_i .type _Z9copy_to_yiPfS_i, @function _Z9copy_to_yiPfS_i: .LFB2091: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z32__device_stub__Z9copy_to_yiPfS_iiPfS_i addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2091: .size _Z9copy_to_yiPfS_i, .-_Z9copy_to_yiPfS_i .globl _Z46__device_stub__Z21calc_reduce_meanshiftiPfS_S_iPfS_S_ .type _Z46__device_stub__Z21calc_reduce_meanshiftiPfS_S_iPfS_S_, @function _Z46__device_stub__Z21calc_reduce_meanshiftiPfS_S_iPfS_S_: .LFB2092: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movl %edi, 28(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movq %rcx, (%rsp) movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax leaq 28(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 8(%rsp), %rax movq %rax, 112(%rsp) movq %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 .L49 .L45: movq 136(%rsp), %rax subq %fs:40, %rax jne .L50 addq $152, %rsp .cfi_remember_state .cfi_def_cfa_offset 8 ret .L49: .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 _Z21calc_reduce_meanshiftiPfS_S_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L45 .L50: call __stack_chk_fail@PLT .cfi_endproc .LFE2092: .size _Z46__device_stub__Z21calc_reduce_meanshiftiPfS_S_iPfS_S_, .-_Z46__device_stub__Z21calc_reduce_meanshiftiPfS_S_iPfS_S_ .globl _Z21calc_reduce_meanshiftiPfS_S_ .type _Z21calc_reduce_meanshiftiPfS_S_, @function _Z21calc_reduce_meanshiftiPfS_S_: .LFB2093: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z46__device_stub__Z21calc_reduce_meanshiftiPfS_S_iPfS_S_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2093: .size _Z21calc_reduce_meanshiftiPfS_S_, .-_Z21calc_reduce_meanshiftiPfS_S_ .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "_Z21calc_reduce_meanshiftiPfS_S_" .section .rodata.str1.1,"aMS",@progbits,1 .LC1: .string "_Z9copy_to_yiPfS_i" .LC2: .string "_Z16kernel_Dvec_multiiPfS_S_i" .LC3: .string "_Z14kernel_sum_diviPfS_" .section .rodata.str1.8 .align 8 .LC4: .string "_Z18calc_Kernel_MatrixiiPfS_S_i" .section .rodata.str1.1 .LC5: .string "_Z15calc_meanshift2PfS_S_" .text .type _ZL24__sti____cudaRegisterAllv, @function _ZL24__sti____cudaRegisterAllv: .LFB2095: .cfi_startproc endbr64 pushq %rbx .cfi_def_cfa_offset 16 .cfi_offset 3, -16 leaq _ZL15__fatDeviceText(%rip), %rdi call __cudaRegisterFatBinary@PLT movq %rax, %rbx 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 _Z21calc_reduce_meanshiftiPfS_S_(%rip), %rsi movq %rax, %rdi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 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 _Z9copy_to_yiPfS_i(%rip), %rsi movq %rbx, %rdi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 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 .LC2(%rip), %rdx movq %rdx, %rcx leaq _Z16kernel_Dvec_multiiPfS_S_i(%rip), %rsi movq %rbx, %rdi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 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 .LC3(%rip), %rdx movq %rdx, %rcx leaq _Z14kernel_sum_diviPfS_(%rip), %rsi movq %rbx, %rdi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 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 _Z18calc_Kernel_MatrixiiPfS_S_i(%rip), %rsi movq %rbx, %rdi call __cudaRegisterFunction@PLT addq $32, %rsp .cfi_def_cfa_offset 16 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 _Z15calc_meanshift2PfS_S_(%rip), %rsi movq %rbx, %rdi 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 popq %rbx .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2095: .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 "cuda_meanshift.hip" .globl _Z30__device_stub__calc_meanshift2PfS_S_ # -- Begin function _Z30__device_stub__calc_meanshift2PfS_S_ .p2align 4, 0x90 .type _Z30__device_stub__calc_meanshift2PfS_S_,@function _Z30__device_stub__calc_meanshift2PfS_S_: # @_Z30__device_stub__calc_meanshift2PfS_S_ .cfi_startproc # %bb.0: subq $104, %rsp .cfi_def_cfa_offset 112 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movq %rdx, 56(%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 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 $_Z15calc_meanshift2PfS_S_, %edi pushq 8(%rsp) .cfi_adjust_cfa_offset 8 pushq 24(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $120, %rsp .cfi_adjust_cfa_offset -120 retq .Lfunc_end0: .size _Z30__device_stub__calc_meanshift2PfS_S_, .Lfunc_end0-_Z30__device_stub__calc_meanshift2PfS_S_ .cfi_endproc # -- End function .globl _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i # -- Begin function _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i .p2align 4, 0x90 .type _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i,@function _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i: # @_Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movl %edi, 20(%rsp) movl %esi, 16(%rsp) movq %rdx, 88(%rsp) movq %rcx, 80(%rsp) movq %r8, 72(%rsp) movl %r9d, 12(%rsp) leaq 20(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 88(%rsp), %rax movq %rax, 112(%rsp) leaq 80(%rsp), %rax movq %rax, 120(%rsp) leaq 72(%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 $_Z18calc_Kernel_MatrixiiPfS_S_i, %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_end1: .size _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i, .Lfunc_end1-_Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i .cfi_endproc # -- End function .globl _Z29__device_stub__kernel_sum_diviPfS_ # -- Begin function _Z29__device_stub__kernel_sum_diviPfS_ .p2align 4, 0x90 .type _Z29__device_stub__kernel_sum_diviPfS_,@function _Z29__device_stub__kernel_sum_diviPfS_: # @_Z29__device_stub__kernel_sum_diviPfS_ .cfi_startproc # %bb.0: subq $104, %rsp .cfi_def_cfa_offset 112 movl %edi, 12(%rsp) movq %rsi, 72(%rsp) movq %rdx, 64(%rsp) leaq 12(%rsp), %rax movq %rax, 80(%rsp) leaq 72(%rsp), %rax movq %rax, 88(%rsp) leaq 64(%rsp), %rax movq %rax, 96(%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 $_Z14kernel_sum_diviPfS_, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $120, %rsp .cfi_adjust_cfa_offset -120 retq .Lfunc_end2: .size _Z29__device_stub__kernel_sum_diviPfS_, .Lfunc_end2-_Z29__device_stub__kernel_sum_diviPfS_ .cfi_endproc # -- End function .globl _Z31__device_stub__kernel_Dvec_multiiPfS_S_i # -- Begin function _Z31__device_stub__kernel_Dvec_multiiPfS_S_i .p2align 4, 0x90 .type _Z31__device_stub__kernel_Dvec_multiiPfS_S_i,@function _Z31__device_stub__kernel_Dvec_multiiPfS_S_i: # @_Z31__device_stub__kernel_Dvec_multiiPfS_S_i .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movl %edi, 20(%rsp) movl %esi, 16(%rsp) movq %rdx, 88(%rsp) movq %rcx, 80(%rsp) movq %r8, 72(%rsp) movl %r9d, 12(%rsp) leaq 20(%rsp), %rax movq %rax, 96(%rsp) leaq 16(%rsp), %rax movq %rax, 104(%rsp) leaq 88(%rsp), %rax movq %rax, 112(%rsp) leaq 80(%rsp), %rax movq %rax, 120(%rsp) leaq 72(%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 $_Z16kernel_Dvec_multiiPfS_S_i, %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_end3: .size _Z31__device_stub__kernel_Dvec_multiiPfS_S_i, .Lfunc_end3-_Z31__device_stub__kernel_Dvec_multiiPfS_S_i .cfi_endproc # -- End function .globl _Z24__device_stub__copy_to_yiPfS_i # -- Begin function _Z24__device_stub__copy_to_yiPfS_i .p2align 4, 0x90 .type _Z24__device_stub__copy_to_yiPfS_i,@function _Z24__device_stub__copy_to_yiPfS_i: # @_Z24__device_stub__copy_to_yiPfS_i .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movl %edi, 12(%rsp) movq %rsi, 72(%rsp) movq %rdx, 64(%rsp) movl %ecx, 8(%rsp) leaq 12(%rsp), %rax movq %rax, 80(%rsp) leaq 72(%rsp), %rax movq %rax, 88(%rsp) leaq 64(%rsp), %rax movq %rax, 96(%rsp) leaq 8(%rsp), %rax movq %rax, 104(%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 $_Z9copy_to_yiPfS_i, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $136, %rsp .cfi_adjust_cfa_offset -136 retq .Lfunc_end4: .size _Z24__device_stub__copy_to_yiPfS_i, .Lfunc_end4-_Z24__device_stub__copy_to_yiPfS_i .cfi_endproc # -- End function .globl _Z36__device_stub__calc_reduce_meanshiftiPfS_S_ # -- Begin function _Z36__device_stub__calc_reduce_meanshiftiPfS_S_ .p2align 4, 0x90 .type _Z36__device_stub__calc_reduce_meanshiftiPfS_S_,@function _Z36__device_stub__calc_reduce_meanshiftiPfS_S_: # @_Z36__device_stub__calc_reduce_meanshiftiPfS_S_ .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movl %edi, 4(%rsp) movq %rsi, 72(%rsp) movq %rdx, 64(%rsp) movq %rcx, 56(%rsp) leaq 4(%rsp), %rax movq %rax, 80(%rsp) leaq 72(%rsp), %rax movq %rax, 88(%rsp) leaq 64(%rsp), %rax movq %rax, 96(%rsp) leaq 56(%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 $_Z21calc_reduce_meanshiftiPfS_S_, %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_end5: .size _Z36__device_stub__calc_reduce_meanshiftiPfS_S_, .Lfunc_end5-_Z36__device_stub__calc_reduce_meanshiftiPfS_S_ .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: pushq %rbx .cfi_def_cfa_offset 16 subq $32, %rsp .cfi_def_cfa_offset 48 .cfi_offset %rbx, -16 cmpq $0, __hip_gpubin_handle(%rip) jne .LBB6_2 # %bb.1: movl $__hip_fatbin_wrapper, %edi callq __hipRegisterFatBinary movq %rax, __hip_gpubin_handle(%rip) .LBB6_2: movq __hip_gpubin_handle(%rip), %rbx xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z15calc_meanshift2PfS_S_, %esi movl $.L__unnamed_1, %edx movl $.L__unnamed_1, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z18calc_Kernel_MatrixiiPfS_S_i, %esi movl $.L__unnamed_2, %edx movl $.L__unnamed_2, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z14kernel_sum_diviPfS_, %esi movl $.L__unnamed_3, %edx movl $.L__unnamed_3, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z16kernel_Dvec_multiiPfS_S_i, %esi movl $.L__unnamed_4, %edx movl $.L__unnamed_4, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z9copy_to_yiPfS_i, %esi movl $.L__unnamed_5, %edx movl $.L__unnamed_5, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction xorps %xmm0, %xmm0 movups %xmm0, 16(%rsp) movups %xmm0, (%rsp) movl $_Z21calc_reduce_meanshiftiPfS_S_, %esi movl $.L__unnamed_6, %edx movl $.L__unnamed_6, %ecx movq %rbx, %rdi movl $-1, %r8d xorl %r9d, %r9d callq __hipRegisterFunction movl $__hip_module_dtor, %edi addq $32, %rsp .cfi_def_cfa_offset 16 popq %rbx .cfi_def_cfa_offset 8 jmp atexit # TAILCALL .Lfunc_end6: .size __hip_module_ctor, .Lfunc_end6-__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 .LBB7_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 .LBB7_2: retq .Lfunc_end7: .size __hip_module_dtor, .Lfunc_end7-__hip_module_dtor .cfi_endproc # -- End function .type _Z15calc_meanshift2PfS_S_,@object # @_Z15calc_meanshift2PfS_S_ .section .rodata,"a",@progbits .globl _Z15calc_meanshift2PfS_S_ .p2align 3, 0x0 _Z15calc_meanshift2PfS_S_: .quad _Z30__device_stub__calc_meanshift2PfS_S_ .size _Z15calc_meanshift2PfS_S_, 8 .type _Z18calc_Kernel_MatrixiiPfS_S_i,@object # @_Z18calc_Kernel_MatrixiiPfS_S_i .globl _Z18calc_Kernel_MatrixiiPfS_S_i .p2align 3, 0x0 _Z18calc_Kernel_MatrixiiPfS_S_i: .quad _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i .size _Z18calc_Kernel_MatrixiiPfS_S_i, 8 .type _Z14kernel_sum_diviPfS_,@object # @_Z14kernel_sum_diviPfS_ .globl _Z14kernel_sum_diviPfS_ .p2align 3, 0x0 _Z14kernel_sum_diviPfS_: .quad _Z29__device_stub__kernel_sum_diviPfS_ .size _Z14kernel_sum_diviPfS_, 8 .type _Z16kernel_Dvec_multiiPfS_S_i,@object # @_Z16kernel_Dvec_multiiPfS_S_i .globl _Z16kernel_Dvec_multiiPfS_S_i .p2align 3, 0x0 _Z16kernel_Dvec_multiiPfS_S_i: .quad _Z31__device_stub__kernel_Dvec_multiiPfS_S_i .size _Z16kernel_Dvec_multiiPfS_S_i, 8 .type _Z9copy_to_yiPfS_i,@object # @_Z9copy_to_yiPfS_i .globl _Z9copy_to_yiPfS_i .p2align 3, 0x0 _Z9copy_to_yiPfS_i: .quad _Z24__device_stub__copy_to_yiPfS_i .size _Z9copy_to_yiPfS_i, 8 .type _Z21calc_reduce_meanshiftiPfS_S_,@object # @_Z21calc_reduce_meanshiftiPfS_S_ .globl _Z21calc_reduce_meanshiftiPfS_S_ .p2align 3, 0x0 _Z21calc_reduce_meanshiftiPfS_S_: .quad _Z36__device_stub__calc_reduce_meanshiftiPfS_S_ .size _Z21calc_reduce_meanshiftiPfS_S_, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z15calc_meanshift2PfS_S_" .size .L__unnamed_1, 26 .type .L__unnamed_2,@object # @1 .L__unnamed_2: .asciz "_Z18calc_Kernel_MatrixiiPfS_S_i" .size .L__unnamed_2, 32 .type .L__unnamed_3,@object # @2 .L__unnamed_3: .asciz "_Z14kernel_sum_diviPfS_" .size .L__unnamed_3, 24 .type .L__unnamed_4,@object # @3 .L__unnamed_4: .asciz "_Z16kernel_Dvec_multiiPfS_S_i" .size .L__unnamed_4, 30 .type .L__unnamed_5,@object # @4 .L__unnamed_5: .asciz "_Z9copy_to_yiPfS_i" .size .L__unnamed_5, 19 .type .L__unnamed_6,@object # @5 .L__unnamed_6: .asciz "_Z21calc_reduce_meanshiftiPfS_S_" .size .L__unnamed_6, 33 .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__calc_meanshift2PfS_S_ .addrsig_sym _Z33__device_stub__calc_Kernel_MatrixiiPfS_S_i .addrsig_sym _Z29__device_stub__kernel_sum_diviPfS_ .addrsig_sym _Z31__device_stub__kernel_Dvec_multiiPfS_S_i .addrsig_sym _Z24__device_stub__copy_to_yiPfS_i .addrsig_sym _Z36__device_stub__calc_reduce_meanshiftiPfS_S_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z15calc_meanshift2PfS_S_ .addrsig_sym _Z18calc_Kernel_MatrixiiPfS_S_i .addrsig_sym _Z14kernel_sum_diviPfS_ .addrsig_sym _Z16kernel_Dvec_multiiPfS_S_i .addrsig_sym _Z9copy_to_yiPfS_i .addrsig_sym _Z21calc_reduce_meanshiftiPfS_S_ .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.
// includes #include <stdio.h> #include <stdlib.h> //matriz de vecinos void liberaVecinos(int *vec) { free(vec); return; }
code for sm_80
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
// includes #include <stdio.h> #include <stdlib.h> //matriz de vecinos void liberaVecinos(int *vec) { free(vec); return; }
.file "tmpxft_0005107a_00000000-6_libera_vecinos.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 _Z13liberaVecinosPi .type _Z13liberaVecinosPi, @function _Z13liberaVecinosPi: .LFB2057: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call free@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2057: .size _Z13liberaVecinosPi, .-_Z13liberaVecinosPi .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.
// includes #include <stdio.h> #include <stdlib.h> //matriz de vecinos void liberaVecinos(int *vec) { free(vec); return; }
// includes #include <hip/hip_runtime.h> #include <stdio.h> #include <stdlib.h> //matriz de vecinos void liberaVecinos(int *vec) { free(vec); return; }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
// includes #include <hip/hip_runtime.h> #include <stdio.h> #include <stdlib.h> //matriz de vecinos void liberaVecinos(int *vec) { free(vec); return; }
.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.
// includes #include <hip/hip_runtime.h> #include <stdio.h> #include <stdlib.h> //matriz de vecinos void liberaVecinos(int *vec) { free(vec); return; }
.text .file "libera_vecinos.hip" .globl _Z13liberaVecinosPi # -- Begin function _Z13liberaVecinosPi .p2align 4, 0x90 .type _Z13liberaVecinosPi,@function _Z13liberaVecinosPi: # @_Z13liberaVecinosPi .cfi_startproc # %bb.0: jmp free # TAILCALL .Lfunc_end0: .size _Z13liberaVecinosPi, .Lfunc_end0-_Z13liberaVecinosPi .cfi_endproc # -- End function .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_0005107a_00000000-6_libera_vecinos.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 _Z13liberaVecinosPi .type _Z13liberaVecinosPi, @function _Z13liberaVecinosPi: .LFB2057: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call free@PLT addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2057: .size _Z13liberaVecinosPi, .-_Z13liberaVecinosPi .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 "libera_vecinos.hip" .globl _Z13liberaVecinosPi # -- Begin function _Z13liberaVecinosPi .p2align 4, 0x90 .type _Z13liberaVecinosPi,@function _Z13liberaVecinosPi: # @_Z13liberaVecinosPi .cfi_startproc # %bb.0: jmp free # TAILCALL .Lfunc_end0: .size _Z13liberaVecinosPi, .Lfunc_end0-_Z13liberaVecinosPi .cfi_endproc # -- End function .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 CUDA device assembly.
/*************************************************************************** *cr *cr (C) Copyright 2010 The Board of Trustees of the *cr University of Illinois *cr All Rights Reserved *cr ***************************************************************************/ #include "common.h" __global__ void block2D_hybrid_coarsen_x(float c0,float c1,float *A0,float *Anext, int nx, int ny, int nz) { //thread coarsening along x direction const int i = blockIdx.x*blockDim.x*2+threadIdx.x; const int i2= blockIdx.x*blockDim.x*2+threadIdx.x+blockDim.x; const int j = blockIdx.y*blockDim.y+threadIdx.y; const int sh_id=threadIdx.x + threadIdx.y*blockDim.x*2; const int sh_id2=threadIdx.x +blockDim.x+ threadIdx.y*blockDim.x*2; //shared memeory extern __shared__ float sh_A0[]; sh_A0[sh_id]=0.0f; sh_A0[sh_id2]=0.0f; __syncthreads(); //get available region for load and store const bool w_region = i>0 && j>0 &&(i<nx-1) &&(j<ny-1) ; const bool w_region2 = j>0 &&(i2<nx-1) &&(j<ny-1) ; const bool x_l_bound = (threadIdx.x==0); const bool x_h_bound = ((threadIdx.x+blockDim.x)==(blockDim.x*2-1)); const bool y_l_bound = (threadIdx.y==0); const bool y_h_bound = (threadIdx.y==(blockDim.y-1)); //register for bottom and top planes //because of thread coarsening, we need to doulbe registers float bottom=0.0f,bottom2=0.0f,top=0.0f,top2=0.0f; //load data for bottom and current if((i<nx) &&(j<ny)) { bottom=A0[Index3D (nx, ny, i, j, 0)]; sh_A0[sh_id]=A0[Index3D (nx, ny, i, j, 1)]; } if((i2<nx) &&(j<ny)) { bottom2=A0[Index3D (nx, ny, i2, j, 0)]; sh_A0[sh_id2]=A0[Index3D (nx, ny, i2, j, 1)]; } __syncthreads(); for(int k=1;k<nz-1;k++) { float a_left_right,a_top,a_down; //load required data on xy planes //if it on shared memory, load from shared memory //if not, load from global memory if((i<nx) &&(j<ny)) top=A0[Index3D (nx, ny, i, j, k+1)]; if(w_region) { a_left_right=x_l_bound?A0[Index3D (nx, ny, i-1, j, k )]:sh_A0[sh_id-1]; a_down=y_l_bound?A0[Index3D (nx, ny, i, j-1, k )]:sh_A0[sh_id-2*blockDim.x]; a_top=y_h_bound?A0[Index3D (nx, ny, i, j+1, k )]:sh_A0[sh_id+2*blockDim.x]; Anext[Index3D (nx, ny, i, j, k)] = (a_left_right + sh_A0[sh_id+1] + a_top + a_down + top + bottom)*c1 - sh_A0[sh_id]*c0 ; } //load another block if((i2<nx) &&(j<ny)) top2=A0[Index3D (nx, ny, i2, j, k+1)]; if(w_region2) { a_left_right=x_h_bound?A0[Index3D (nx, ny, i2+1, j, k )]:sh_A0[sh_id2+1]; a_down=y_l_bound?A0[Index3D (nx, ny, i2, j-1, k )]:sh_A0[sh_id2-2*blockDim.x]; a_top=y_h_bound?A0[Index3D (nx, ny, i2, j+1, k )]:sh_A0[sh_id2+2*blockDim.x]; Anext[Index3D (nx, ny, i2, j, k)] = (a_left_right + sh_A0[sh_id2-1] + a_top + a_down + top2 +bottom2)*c1 - sh_A0[sh_id2]*c0; } //swap data __syncthreads(); bottom=sh_A0[sh_id]; sh_A0[sh_id]=top; bottom2=sh_A0[sh_id2]; sh_A0[sh_id2]=top2; __syncthreads(); } }
code for sm_80 Function : _Z24block2D_hybrid_coarsen_xffPfS_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 R0, SR_TID.Y ; /* 0x0000000000007919 */ /* 0x000e220000002200 */ /*0020*/ MOV R13, c[0x0][0x0] ; /* 0x00000000000d7a02 */ /* 0x000fe20000000f00 */ /*0030*/ ULDC.64 UR8, c[0x0][0x118] ; /* 0x0000460000087ab9 */ /* 0x000fe40000000a00 */ /*0040*/ S2R R21, SR_CTAID.Y ; /* 0x0000000000157919 */ /* 0x000e220000002600 */ /*0050*/ SHF.L.U32 R15, R13, 0x1, RZ ; /* 0x000000010d0f7819 */ /* 0x000fc600000006ff */ /*0060*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e680000002100 */ /*0070*/ S2R R10, SR_CTAID.X ; /* 0x00000000000a7919 */ /* 0x000ea20000002500 */ /*0080*/ IMAD R12, R21, c[0x0][0x4], R0 ; /* 0x00000100150c7a24 */ /* 0x001fe200078e0200 */ /*0090*/ IADD3 R14, R3, c[0x0][0x0], RZ ; /* 0x00000000030e7a10 */ /* 0x002fe20007ffe0ff */ /*00a0*/ IMAD R2, R15, R0, R3 ; /* 0x000000000f027224 */ /* 0x000fc600078e0203 */ /*00b0*/ ISETP.GE.AND P0, PT, R12, c[0x0][0x17c], PT ; /* 0x00005f000c007a0c */ /* 0x000fe20003f06270 */ /*00c0*/ IMAD R11, R15, R10, R3 ; /* 0x0000000a0f0b7224 */ /* 0x004fe200078e0203 */ /*00d0*/ STS [R2.X4], RZ ; /* 0x000000ff02007388 */ /* 0x0001e80000004800 */ /*00e0*/ ISETP.LT.AND P3, PT, R11.reuse, c[0x0][0x178], !P0 ; /* 0x00005e000b007a0c */ /* 0x040fe40004761270 */ /*00f0*/ IADD3 R23, R11, c[0x0][0x0], RZ ; /* 0x000000000b177a10 */ /* 0x000fc80007ffe0ff */ /*0100*/ ISETP.LT.AND P0, PT, R23, c[0x0][0x178], !P0 ; /* 0x00005e0017007a0c */ /* 0x000fce0004701270 */ /*0110*/ @P3 IADD3 R4, R12, c[0x0][0x17c], RZ ; /* 0x00005f000c043a10 */ /* 0x000fe40007ffe0ff */ /*0120*/ @P3 MOV R7, 0x4 ; /* 0x0000000400073802 */ /* 0x000fc60000000f00 */ /*0130*/ @P3 IMAD R18, R4, c[0x0][0x178], R11 ; /* 0x00005e0004123a24 */ /* 0x000fe200078e020b */ /*0140*/ @P0 IADD3 R16, R12, c[0x0][0x17c], RZ ; /* 0x00005f000c100a10 */ /* 0x000fe20007ffe0ff */ /*0150*/ IMAD R4, R15, R0, R14 ; /* 0x000000000f047224 */ /* 0x000fe200078e020e */ /*0160*/ @P0 MOV R9, 0x4 ; /* 0x0000000400090802 */ /* 0x000fe20000000f00 */ /*0170*/ @P3 IMAD.WIDE R18, R18, R7, c[0x0][0x168] ; /* 0x00005a0012123625 */ /* 0x000fc600078e0207 */ /*0180*/ STS [R4.X4], RZ ; /* 0x000000ff04007388 */ /* 0x0001e20000004800 */ /*0190*/ @P0 IMAD R16, R16, c[0x0][0x178], R23 ; /* 0x00005e0010100a24 */ /* 0x000fc600078e0217 */ /*01a0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fe20000010000 */ /*01b0*/ @P0 IMAD.WIDE R16, R16, R9, c[0x0][0x168] ; /* 0x00005a0010100625 */ /* 0x000fca00078e0209 */ /*01c0*/ @P3 LDG.E R19, [R18.64] ; /* 0x0000000812133981 */ /* 0x000ea8000c1e1900 */ /*01d0*/ @P0 LDG.E R17, [R16.64] ; /* 0x0000000810110981 */ /* 0x000ee2000c1e1900 */ /*01e0*/ HFMA2.MMA R26, -RZ, RZ, 0, 0 ; /* 0x00000000ff1a7435 */ /* 0x000fe200000001ff */ /*01f0*/ @P3 IMAD R6, R12.reuse, c[0x0][0x178], R11 ; /* 0x00005e000c063a24 */ /* 0x040fe200078e020b */ /*0200*/ MOV R5, RZ ; /* 0x000000ff00057202 */ /* 0x000fe20000000f00 */ /*0210*/ @P0 IMAD R8, R12, c[0x0][0x178], R23 ; /* 0x00005e000c080a24 */ /* 0x000fe400078e0217 */ /*0220*/ @P3 IMAD.WIDE R6, R6, R7, c[0x0][0x168] ; /* 0x00005a0006063625 */ /* 0x000fc800078e0207 */ /*0230*/ @P0 IMAD.WIDE R8, R8, R9, c[0x0][0x168] ; /* 0x00005a0008080625 */ /* 0x000fe200078e0209 */ /*0240*/ @P3 LDG.E R26, [R6.64] ; /* 0x00000008061a3981 */ /* 0x000168000c1e1900 */ /*0250*/ @P0 LDG.E R5, [R8.64] ; /* 0x0000000808050981 */ /* 0x000162000c1e1900 */ /*0260*/ MOV R22, c[0x0][0x180] ; /* 0x0000600000167a02 */ /* 0x000fc80000000f00 */ /*0270*/ ISETP.GE.AND P1, PT, R22, 0x3, PT ; /* 0x000000031600780c */ /* 0x000fe20003f26270 */ /*0280*/ @P3 STS [R2.X4], R19 ; /* 0x0000001302003388 */ /* 0x0041e80000004800 */ /*0290*/ @P0 STS [R4.X4], R17 ; /* 0x0000001104000388 */ /* 0x0081e80000004800 */ /*02a0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*02b0*/ @!P1 EXIT ; /* 0x000000000000994d */ /* 0x000fea0003800000 */ /*02c0*/ UMOV UR6, 0x1 ; /* 0x0000000100067882 */ /* 0x001fe20000000000 */ /*02d0*/ ISETP.GT.AND P1, PT, R12, RZ, PT ; /* 0x000000ff0c00720c */ /* 0x000fe20003f24270 */ /*02e0*/ ULDC.64 UR4, c[0x0][0x178] ; /* 0x00005e0000047ab9 */ /* 0x000fe20000000a00 */ /*02f0*/ IADD3 R7, R15, -0x1, RZ ; /* 0xffffffff0f077810 */ /* 0x000fe20007ffe0ff */ /*0300*/ UIADD3 UR4, -UR6, UR4, URZ ; /* 0x0000000406047290 */ /* 0x000fe2000fffe13f */ /*0310*/ IADD3 R6, R0, c[0x0][0x17c], RZ ; /* 0x00005f0000067a10 */ /* 0x000fe20007ffe0ff */ /*0320*/ UIADD3 UR5, -UR6, UR5, URZ ; /* 0x0000000506057290 */ /* 0x000fe2000fffe13f */ /*0330*/ MOV R9, c[0x0][0x17c] ; /* 0x00005f0000097a02 */ /* 0x000fc40000000f00 */ /*0340*/ ISETP.NE.AND P4, PT, R14, R7, PT ; /* 0x000000070e00720c */ /* 0x000fe20003f85270 */ /*0350*/ IMAD R14, R21, c[0x0][0x4], R6 ; /* 0x00000100150e7a24 */ /* 0x000fe200078e0206 */ /*0360*/ ISETP.LT.AND P2, PT, R23, UR4, P1 ; /* 0x0000000417007c0c */ /* 0x000fe40008f41270 */ /*0370*/ ISETP.GT.AND P1, PT, R11, RZ, P1 ; /* 0x000000ff0b00720c */ /* 0x000fe40000f24270 */ /*0380*/ SHF.L.U32 R8, R4, 0x2, RZ ; /* 0x0000000204087819 */ /* 0x000fe400000006ff */ /*0390*/ LEA R18, R9, R12, 0x1 ; /* 0x0000000c09127211 */ /* 0x000fe200078e08ff */ /*03a0*/ IMAD R9, R14, c[0x0][0x178], R11.reuse ; /* 0x00005e000e097a24 */ /* 0x100fe200078e020b */ /*03b0*/ ISETP.LT.AND P5, PT, R11, UR4, P1 ; /* 0x000000040b007c0c */ /* 0x000fe20008fa1270 */ /*03c0*/ ULDC UR4, c[0x0][0x4] ; /* 0x0000010000047ab9 */ /* 0x000fe20000000800 */ /*03d0*/ LEA R7, R13, R8, 0x3 ; /* 0x000000080d077211 */ /* 0x000fe200078e18ff */ /*03e0*/ IMAD R8, R18, c[0x0][0x178], R11 ; /* 0x00005e0012087a24 */ /* 0x000fe200078e020b */ /*03f0*/ ISETP.LT.AND P1, PT, R12, UR5, P2 ; /* 0x000000050c007c0c */ /* 0x000fe20009721270 */ /*0400*/ IMAD R17, R18, c[0x0][0x178], R3 ; /* 0x00005e0012117a24 */ /* 0x000fe200078e0203 */ /*0410*/ ISETP.LT.AND P2, PT, R12, UR5, P5 ; /* 0x000000050c007c0c */ /* 0x000fe2000af41270 */ /*0420*/ UIADD3 UR4, -UR6, UR4, URZ ; /* 0x0000000406047290 */ /* 0x000fe2000fffe13f */ /*0430*/ SHF.L.U32 R6, R2, 0x2, RZ ; /* 0x0000000202067819 */ /* 0x000fe200000006ff */ /*0440*/ UMOV UR5, 0x1 ; /* 0x0000000100057882 */ /* 0x000fe20000000000 */ /*0450*/ IADD3 R12, R14, -0x1, RZ ; /* 0xffffffff0e0c7810 */ /* 0x000fc40007ffe0ff */ /*0460*/ IADD3 R18, R14.reuse, 0x1, RZ ; /* 0x000000010e127810 */ /* 0x040fe40007ffe0ff */ /*0470*/ LEA R6, R13, R6, 0x3 ; /* 0x000000060d067211 */ /* 0x000fe200078e18ff */ /*0480*/ IMAD R13, R14, c[0x0][0x178], R3.reuse ; /* 0x00005e000e0d7a24 */ /* 0x100fe200078e0203 */ /*0490*/ LEA R16, R10, 0x1, 0x1 ; /* 0x000000010a107811 */ /* 0x000fe200078e08ff */ /*04a0*/ IMAD R19, R12, c[0x0][0x178], R3.reuse ; /* 0x00005e000c137a24 */ /* 0x100fe200078e0203 */ /*04b0*/ IADD3 R22, R22, -0x1, RZ ; /* 0xffffffff16167810 */ /* 0x000fe20007ffe0ff */ /*04c0*/ IMAD R21, R18, c[0x0][0x178], R3 ; /* 0x00005e0012157a24 */ /* 0x000fe200078e0203 */ /*04d0*/ IADD3 R23, -R15.reuse, R2, RZ ; /* 0x000000020f177210 */ /* 0x040fe20007ffe1ff */ /*04e0*/ IMAD R10, R12, c[0x0][0x178], R11.reuse ; /* 0x00005e000c0a7a24 */ /* 0x100fe200078e020b */ /*04f0*/ IADD3 R24, -R15, R4, RZ ; /* 0x000000040f187210 */ /* 0x000fe20007ffe1ff */ /*0500*/ IMAD R11, R18, c[0x0][0x178], R11 ; /* 0x00005e00120b7a24 */ /* 0x000fe200078e020b */ /*0510*/ MOV R25, RZ ; /* 0x000000ff00197202 */ /* 0x000fe20000000f00 */ /*0520*/ IMAD R12, R16.reuse, c[0x0][0x0], R17 ; /* 0x00000000100c7a24 */ /* 0x040fe200078e0211 */ /*0530*/ MOV R27, RZ ; /* 0x000000ff001b7202 */ /* 0x000fe20000000f00 */ /*0540*/ IMAD R13, R16, c[0x0][0x0], R13 ; /* 0x00000000100d7a24 */ /* 0x000fc400078e020d */ /*0550*/ IMAD R20, R16.reuse, c[0x0][0x0], R19 ; /* 0x0000000010147a24 */ /* 0x040fe400078e0213 */ /*0560*/ IMAD R21, R16, c[0x0][0x0], R21 ; /* 0x0000000010157a24 */ /* 0x000fe400078e0215 */ /*0570*/ @P3 MOV R15, 0x4 ; /* 0x00000004000f3802 */ /* 0x000fca0000000f00 */ /*0580*/ @P3 IMAD.WIDE R14, R8, R15, c[0x0][0x168] ; /* 0x00005a00080e3625 */ /* 0x000fca00078e020f */ /*0590*/ @P3 LDG.E R27, [R14.64] ; /* 0x000000080e1b3981 */ /* 0x020162000c1e1900 */ /*05a0*/ BSSY B0, 0x760 ; /* 0x000001b000007945 */ /* 0x000fe20003800000 */ /*05b0*/ @!P2 BRA 0x750 ; /* 0x000001900000a947 */ /* 0x006fea0003800000 */ /*05c0*/ ISETP.NE.AND P5, PT, R3, RZ, PT ; /* 0x000000ff0300720c */ /* 0x000fe20003fa5270 */ /*05d0*/ HFMA2.MMA R28, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff1c7435 */ /* 0x000fe200000001ff */ /*05e0*/ LDS R15, [R2.X4+0x4] ; /* 0x00000400020f7984 */ /* 0x001ff60000004800 */ /*05f0*/ @P5 LDS R14, [R2.X4+-0x4] ; /* 0xfffffc00020e5984 */ /* 0x000e220000004800 */ /*0600*/ @!P5 IADD3 R19, R9, -0x1, RZ ; /* 0xffffffff0913d810 */ /* 0x000fca0007ffe0ff */ /*0610*/ @!P5 IMAD.WIDE R18, R19, R28, c[0x0][0x168] ; /* 0x00005a001312d625 */ /* 0x000fca00078e021c */ /*0620*/ @!P5 LDG.E R14, [R18.64] ; /* 0x00000008120ed981 */ /* 0x000e22000c1e1900 */ /*0630*/ ISETP.NE.AND P5, PT, R0, UR4, PT ; /* 0x0000000400007c0c */ /* 0x000fda000bfa5270 */ /*0640*/ @P5 LDS R16, [R6] ; /* 0x0000000006105984 */ /* 0x000e620000000800 */ /*0650*/ FADD R17, R15, R14 ; /* 0x0000000e0f117221 */ /* 0x001fe40000000000 */ /*0660*/ @!P5 IMAD.WIDE R14, R11, R28, c[0x0][0x168] ; /* 0x00005a000b0ed625 */ /* 0x000fca00078e021c */ /*0670*/ @!P5 LDG.E R16, [R14.64] ; /* 0x000000080e10d981 */ /* 0x000062000c1e1900 */ /*0680*/ ISETP.NE.AND P5, PT, R0, RZ, PT ; /* 0x000000ff0000720c */ /* 0x000fc60003fa5270 */ /*0690*/ LDS R15, [R2.X4] ; /* 0x00000000020f7984 */ /* 0x001ff40000004800 */ /*06a0*/ @P5 LDS R29, [R23.X4] ; /* 0x00000000171d5984 */ /* 0x000e220000004800 */ /*06b0*/ FADD R14, R17, R16 ; /* 0x00000010110e7221 */ /* 0x002fc40000000000 */ /*06c0*/ @!P5 IMAD.WIDE R16, R10, R28, c[0x0][0x168] ; /* 0x00005a000a10d625 */ /* 0x000fca00078e021c */ /*06d0*/ @!P5 LDG.E R29, [R16.64] ; /* 0x00000008101dd981 */ /* 0x000e24000c1e1900 */ /*06e0*/ FADD R18, R14, R29 ; /* 0x0000001d0e127221 */ /* 0x001fc80000000000 */ /*06f0*/ FADD R19, R27, R18 ; /* 0x000000121b137221 */ /* 0x020fe40000000000 */ /*0700*/ FMUL R18, R15, c[0x0][0x160] ; /* 0x000058000f127a20 */ /* 0x000fe40000400000 */ /*0710*/ FADD R19, R19, R26 ; /* 0x0000001a13137221 */ /* 0x000fe40000000000 */ /*0720*/ IMAD.WIDE R14, R9, R28, c[0x0][0x170] ; /* 0x00005c00090e7625 */ /* 0x000fc800078e021c */ /*0730*/ FFMA R19, R19, c[0x0][0x164], -R18 ; /* 0x0000590013137a23 */ /* 0x000fca0000000812 */ /*0740*/ STG.E [R14.64], R19 ; /* 0x000000130e007986 */ /* 0x0001e4000c101908 */ /*0750*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0760*/ BSSY B0, 0x7c0 ; /* 0x0000005000007945 */ /* 0x000fe20003800000 */ /*0770*/ @!P0 BRA 0x7b0 ; /* 0x0000003000008947 */ /* 0x000fea0003800000 */ /*0780*/ MOV R15, 0x4 ; /* 0x00000004000f7802 */ /* 0x001fca0000000f00 */ /*0790*/ IMAD.WIDE R14, R12, R15, c[0x0][0x168] ; /* 0x00005a000c0e7625 */ /* 0x000fca00078e020f */ /*07a0*/ LDG.E R25, [R14.64] ; /* 0x000000080e197981 */ /* 0x000164000c1e1900 */ /*07b0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*07c0*/ BSSY B0, 0x970 ; /* 0x000001a000007945 */ /* 0x000fe20003800000 */ /*07d0*/ @!P1 BRA 0x960 ; /* 0x0000018000009947 */ /* 0x000fea0003800000 */ /*07e0*/ @P4 LDS R18, [R4.X4+0x4] ; /* 0x0000040004124984 */ /* 0x000fe20000004800 */ /*07f0*/ HFMA2.MMA R26, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff1a7435 */ /* 0x000fe200000001ff */ /*0800*/ @!P4 IADD3 R15, R13, 0x1, RZ ; /* 0x000000010d0fc810 */ /* 0x001fe40007ffe0ff */ /*0810*/ ISETP.NE.AND P6, PT, R0.reuse, UR4, PT ; /* 0x0000000400007c0c */ /* 0x040fe2000bfc5270 */ /*0820*/ LDS R19, [R4.X4+-0x4] ; /* 0xfffffc0004137984 */ /* 0x000e2c0000004800 */ /*0830*/ @!P4 IMAD.WIDE R14, R15, R26, c[0x0][0x168] ; /* 0x00005a000f0ec625 */ /* 0x000fe200078e021a */ /*0840*/ ISETP.NE.AND P5, PT, R0, RZ, PT ; /* 0x000000ff0000720c */ /* 0x000fca0003fa5270 */ /*0850*/ @P6 LDS R28, [R7] ; /* 0x00000000071c6984 */ /* 0x000e700000000800 */ /*0860*/ @P5 LDS R29, [R24.X4] ; /* 0x00000000181d5984 */ /* 0x000ea80000004800 */ /*0870*/ @!P4 LDG.E R18, [R14.64] ; /* 0x000000080e12c981 */ /* 0x000622000c1e1900 */ /*0880*/ @!P6 IMAD.WIDE R16, R21, R26, c[0x0][0x168] ; /* 0x00005a001510e625 */ /* 0x000fc600078e021a */ /*0890*/ LDS R14, [R4.X4] ; /* 0x00000000040e7984 */ /* 0x008ee80000004800 */ /*08a0*/ @!P6 LDG.E R28, [R16.64] ; /* 0x00000008101ce981 */ /* 0x000e62000c1e1900 */ /*08b0*/ FADD R15, R19, R18 ; /* 0x00000012130f7221 */ /* 0x001fe40000000000 */ /*08c0*/ @!P5 IMAD.WIDE R18, R20, R26, c[0x0][0x168] ; /* 0x00005a001412d625 */ /* 0x000fca00078e021a */ /*08d0*/ @!P5 LDG.E R29, [R18.64] ; /* 0x00000008121dd981 */ /* 0x000ea2000c1e1900 */ /*08e0*/ FADD R28, R15, R28 ; /* 0x0000001c0f1c7221 */ /* 0x002fc80000000000 */ /*08f0*/ FADD R28, R28, R29 ; /* 0x0000001d1c1c7221 */ /* 0x004fe40000000000 */ /*0900*/ FMUL R29, R14, c[0x0][0x160] ; /* 0x000058000e1d7a20 */ /* 0x008fe40000400000 */ /*0910*/ FADD R28, R25, R28 ; /* 0x0000001c191c7221 */ /* 0x020fe40000000000 */ /*0920*/ IMAD.WIDE R14, R13, R26, c[0x0][0x170] ; /* 0x00005c000d0e7625 */ /* 0x000fc800078e021a */ /*0930*/ FADD R5, R28, R5 ; /* 0x000000051c057221 */ /* 0x000fc80000000000 */ /*0940*/ FFMA R5, R5, c[0x0][0x164], -R29 ; /* 0x0000590005057a23 */ /* 0x000fca000000081d */ /*0950*/ STG.E [R14.64], R5 ; /* 0x000000050e007986 */ /* 0x0001e4000c101908 */ /*0960*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0970*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fe20000010000 */ /*0980*/ UIADD3 UR5, UR5, 0x1, URZ ; /* 0x0000000105057890 */ /* 0x000fe2000fffe03f */ /*0990*/ MOV R14, c[0x0][0x17c] ; /* 0x00005f00000e7a02 */ /* 0x001fca0000000f00 */ /*09a0*/ ISETP.LE.AND P5, PT, R22, UR5, PT ; /* 0x0000000516007c0c */ /* 0x000fe2000bfa3270 */ /*09b0*/ IMAD R13, R14.reuse, c[0x0][0x178], R13 ; /* 0x00005e000e0d7a24 */ /* 0x040fe400078e020d */ /*09c0*/ IMAD R9, R14.reuse, c[0x0][0x178], R9 ; /* 0x00005e000e097a24 */ /* 0x040fe400078e0209 */ /*09d0*/ IMAD R8, R14.reuse, c[0x0][0x178], R8 ; /* 0x00005e000e087a24 */ /* 0x040fe400078e0208 */ /*09e0*/ IMAD R12, R14.reuse, c[0x0][0x178], R12 ; /* 0x00005e000e0c7a24 */ /* 0x040fe400078e020c */ /*09f0*/ IMAD R10, R14.reuse, c[0x0][0x178], R10 ; /* 0x00005e000e0a7a24 */ /* 0x040fe400078e020a */ /*0a00*/ IMAD R11, R14, c[0x0][0x178], R11 ; /* 0x00005e000e0b7a24 */ /* 0x000fc400078e020b */ /*0a10*/ IMAD R20, R14.reuse, c[0x0][0x178], R20 ; /* 0x00005e000e147a24 */ /* 0x040fe400078e0214 */ /*0a20*/ IMAD R21, R14, c[0x0][0x178], R21 ; /* 0x00005e000e157a24 */ /* 0x000fe200078e0215 */ /*0a30*/ LDS R26, [R2.X4] ; /* 0x00000000021a7984 */ /* 0x0000680000004800 */ /*0a40*/ STS [R2.X4], R27 ; /* 0x0000001b02007388 */ /* 0x0201e80000004800 */ /*0a50*/ LDS R5, [R4.X4] ; /* 0x0000000004057984 */ /* 0x0000a80000004800 */ /*0a60*/ STS [R4.X4], R25 ; /* 0x0000001904007388 */ /* 0x0001e80000004800 */ /*0a70*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0a80*/ @!P5 BRA 0x570 ; /* 0xfffffae00000d947 */ /* 0x001fea000383ffff */ /*0a90*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0aa0*/ BRA 0xaa0; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0ab0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ac0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ad0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ae0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0af0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b00*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b10*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b20*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b30*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b40*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b50*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b60*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b70*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
You are a helpful coding assistant specialized in converting CUDA source code to CUDA host assembly.
/*************************************************************************** *cr *cr (C) Copyright 2010 The Board of Trustees of the *cr University of Illinois *cr All Rights Reserved *cr ***************************************************************************/ #include "common.h" __global__ void block2D_hybrid_coarsen_x(float c0,float c1,float *A0,float *Anext, int nx, int ny, int nz) { //thread coarsening along x direction const int i = blockIdx.x*blockDim.x*2+threadIdx.x; const int i2= blockIdx.x*blockDim.x*2+threadIdx.x+blockDim.x; const int j = blockIdx.y*blockDim.y+threadIdx.y; const int sh_id=threadIdx.x + threadIdx.y*blockDim.x*2; const int sh_id2=threadIdx.x +blockDim.x+ threadIdx.y*blockDim.x*2; //shared memeory extern __shared__ float sh_A0[]; sh_A0[sh_id]=0.0f; sh_A0[sh_id2]=0.0f; __syncthreads(); //get available region for load and store const bool w_region = i>0 && j>0 &&(i<nx-1) &&(j<ny-1) ; const bool w_region2 = j>0 &&(i2<nx-1) &&(j<ny-1) ; const bool x_l_bound = (threadIdx.x==0); const bool x_h_bound = ((threadIdx.x+blockDim.x)==(blockDim.x*2-1)); const bool y_l_bound = (threadIdx.y==0); const bool y_h_bound = (threadIdx.y==(blockDim.y-1)); //register for bottom and top planes //because of thread coarsening, we need to doulbe registers float bottom=0.0f,bottom2=0.0f,top=0.0f,top2=0.0f; //load data for bottom and current if((i<nx) &&(j<ny)) { bottom=A0[Index3D (nx, ny, i, j, 0)]; sh_A0[sh_id]=A0[Index3D (nx, ny, i, j, 1)]; } if((i2<nx) &&(j<ny)) { bottom2=A0[Index3D (nx, ny, i2, j, 0)]; sh_A0[sh_id2]=A0[Index3D (nx, ny, i2, j, 1)]; } __syncthreads(); for(int k=1;k<nz-1;k++) { float a_left_right,a_top,a_down; //load required data on xy planes //if it on shared memory, load from shared memory //if not, load from global memory if((i<nx) &&(j<ny)) top=A0[Index3D (nx, ny, i, j, k+1)]; if(w_region) { a_left_right=x_l_bound?A0[Index3D (nx, ny, i-1, j, k )]:sh_A0[sh_id-1]; a_down=y_l_bound?A0[Index3D (nx, ny, i, j-1, k )]:sh_A0[sh_id-2*blockDim.x]; a_top=y_h_bound?A0[Index3D (nx, ny, i, j+1, k )]:sh_A0[sh_id+2*blockDim.x]; Anext[Index3D (nx, ny, i, j, k)] = (a_left_right + sh_A0[sh_id+1] + a_top + a_down + top + bottom)*c1 - sh_A0[sh_id]*c0 ; } //load another block if((i2<nx) &&(j<ny)) top2=A0[Index3D (nx, ny, i2, j, k+1)]; if(w_region2) { a_left_right=x_h_bound?A0[Index3D (nx, ny, i2+1, j, k )]:sh_A0[sh_id2+1]; a_down=y_l_bound?A0[Index3D (nx, ny, i2, j-1, k )]:sh_A0[sh_id2-2*blockDim.x]; a_top=y_h_bound?A0[Index3D (nx, ny, i2, j+1, k )]:sh_A0[sh_id2+2*blockDim.x]; Anext[Index3D (nx, ny, i2, j, k)] = (a_left_right + sh_A0[sh_id2-1] + a_top + a_down + top2 +bottom2)*c1 - sh_A0[sh_id2]*c0; } //swap data __syncthreads(); bottom=sh_A0[sh_id]; sh_A0[sh_id]=top; bottom2=sh_A0[sh_id2]; sh_A0[sh_id2]=top2; __syncthreads(); } }
.file "tmpxft_000601f4_00000000-6_kernels.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 _Z51__device_stub__Z24block2D_hybrid_coarsen_xffPfS_iiiffPfS_iii .type _Z51__device_stub__Z24block2D_hybrid_coarsen_xffPfS_iiiffPfS_iii, @function _Z51__device_stub__Z24block2D_hybrid_coarsen_xffPfS_iiiffPfS_iii: .LFB2051: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movss %xmm0, 44(%rsp) movss %xmm1, 40(%rsp) movq %rdi, 32(%rsp) movq %rsi, 24(%rsp) movl %edx, 20(%rsp) movl %ecx, 16(%rsp) movl %r8d, 12(%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 32(%rsp), %rax movq %rax, 128(%rsp) leaq 24(%rsp), %rax movq %rax, 136(%rsp) leaq 20(%rsp), %rax movq %rax, 144(%rsp) leaq 16(%rsp), %rax movq %rax, 152(%rsp) leaq 12(%rsp), %rax movq %rax, 160(%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 _Z24block2D_hybrid_coarsen_xffPfS_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 _Z51__device_stub__Z24block2D_hybrid_coarsen_xffPfS_iiiffPfS_iii, .-_Z51__device_stub__Z24block2D_hybrid_coarsen_xffPfS_iiiffPfS_iii .globl _Z24block2D_hybrid_coarsen_xffPfS_iii .type _Z24block2D_hybrid_coarsen_xffPfS_iii, @function _Z24block2D_hybrid_coarsen_xffPfS_iii: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z51__device_stub__Z24block2D_hybrid_coarsen_xffPfS_iiiffPfS_iii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z24block2D_hybrid_coarsen_xffPfS_iii, .-_Z24block2D_hybrid_coarsen_xffPfS_iii .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "_Z24block2D_hybrid_coarsen_xffPfS_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 _Z24block2D_hybrid_coarsen_xffPfS_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.
/*************************************************************************** *cr *cr (C) Copyright 2010 The Board of Trustees of the *cr University of Illinois *cr All Rights Reserved *cr ***************************************************************************/ #include "common.h" __global__ void block2D_hybrid_coarsen_x(float c0,float c1,float *A0,float *Anext, int nx, int ny, int nz) { //thread coarsening along x direction const int i = blockIdx.x*blockDim.x*2+threadIdx.x; const int i2= blockIdx.x*blockDim.x*2+threadIdx.x+blockDim.x; const int j = blockIdx.y*blockDim.y+threadIdx.y; const int sh_id=threadIdx.x + threadIdx.y*blockDim.x*2; const int sh_id2=threadIdx.x +blockDim.x+ threadIdx.y*blockDim.x*2; //shared memeory extern __shared__ float sh_A0[]; sh_A0[sh_id]=0.0f; sh_A0[sh_id2]=0.0f; __syncthreads(); //get available region for load and store const bool w_region = i>0 && j>0 &&(i<nx-1) &&(j<ny-1) ; const bool w_region2 = j>0 &&(i2<nx-1) &&(j<ny-1) ; const bool x_l_bound = (threadIdx.x==0); const bool x_h_bound = ((threadIdx.x+blockDim.x)==(blockDim.x*2-1)); const bool y_l_bound = (threadIdx.y==0); const bool y_h_bound = (threadIdx.y==(blockDim.y-1)); //register for bottom and top planes //because of thread coarsening, we need to doulbe registers float bottom=0.0f,bottom2=0.0f,top=0.0f,top2=0.0f; //load data for bottom and current if((i<nx) &&(j<ny)) { bottom=A0[Index3D (nx, ny, i, j, 0)]; sh_A0[sh_id]=A0[Index3D (nx, ny, i, j, 1)]; } if((i2<nx) &&(j<ny)) { bottom2=A0[Index3D (nx, ny, i2, j, 0)]; sh_A0[sh_id2]=A0[Index3D (nx, ny, i2, j, 1)]; } __syncthreads(); for(int k=1;k<nz-1;k++) { float a_left_right,a_top,a_down; //load required data on xy planes //if it on shared memory, load from shared memory //if not, load from global memory if((i<nx) &&(j<ny)) top=A0[Index3D (nx, ny, i, j, k+1)]; if(w_region) { a_left_right=x_l_bound?A0[Index3D (nx, ny, i-1, j, k )]:sh_A0[sh_id-1]; a_down=y_l_bound?A0[Index3D (nx, ny, i, j-1, k )]:sh_A0[sh_id-2*blockDim.x]; a_top=y_h_bound?A0[Index3D (nx, ny, i, j+1, k )]:sh_A0[sh_id+2*blockDim.x]; Anext[Index3D (nx, ny, i, j, k)] = (a_left_right + sh_A0[sh_id+1] + a_top + a_down + top + bottom)*c1 - sh_A0[sh_id]*c0 ; } //load another block if((i2<nx) &&(j<ny)) top2=A0[Index3D (nx, ny, i2, j, k+1)]; if(w_region2) { a_left_right=x_h_bound?A0[Index3D (nx, ny, i2+1, j, k )]:sh_A0[sh_id2+1]; a_down=y_l_bound?A0[Index3D (nx, ny, i2, j-1, k )]:sh_A0[sh_id2-2*blockDim.x]; a_top=y_h_bound?A0[Index3D (nx, ny, i2, j+1, k )]:sh_A0[sh_id2+2*blockDim.x]; Anext[Index3D (nx, ny, i2, j, k)] = (a_left_right + sh_A0[sh_id2-1] + a_top + a_down + top2 +bottom2)*c1 - sh_A0[sh_id2]*c0; } //swap data __syncthreads(); bottom=sh_A0[sh_id]; sh_A0[sh_id]=top; bottom2=sh_A0[sh_id2]; sh_A0[sh_id2]=top2; __syncthreads(); } }
/*************************************************************************** *cr *cr (C) Copyright 2010 The Board of Trustees of the *cr University of Illinois *cr All Rights Reserved *cr ***************************************************************************/ #include <hip/hip_runtime.h> #include "common.h" __global__ void block2D_hybrid_coarsen_x(float c0,float c1,float *A0,float *Anext, int nx, int ny, int nz) { //thread coarsening along x direction const int i = blockIdx.x*blockDim.x*2+threadIdx.x; const int i2= blockIdx.x*blockDim.x*2+threadIdx.x+blockDim.x; const int j = blockIdx.y*blockDim.y+threadIdx.y; const int sh_id=threadIdx.x + threadIdx.y*blockDim.x*2; const int sh_id2=threadIdx.x +blockDim.x+ threadIdx.y*blockDim.x*2; //shared memeory extern __shared__ float sh_A0[]; sh_A0[sh_id]=0.0f; sh_A0[sh_id2]=0.0f; __syncthreads(); //get available region for load and store const bool w_region = i>0 && j>0 &&(i<nx-1) &&(j<ny-1) ; const bool w_region2 = j>0 &&(i2<nx-1) &&(j<ny-1) ; const bool x_l_bound = (threadIdx.x==0); const bool x_h_bound = ((threadIdx.x+blockDim.x)==(blockDim.x*2-1)); const bool y_l_bound = (threadIdx.y==0); const bool y_h_bound = (threadIdx.y==(blockDim.y-1)); //register for bottom and top planes //because of thread coarsening, we need to doulbe registers float bottom=0.0f,bottom2=0.0f,top=0.0f,top2=0.0f; //load data for bottom and current if((i<nx) &&(j<ny)) { bottom=A0[Index3D (nx, ny, i, j, 0)]; sh_A0[sh_id]=A0[Index3D (nx, ny, i, j, 1)]; } if((i2<nx) &&(j<ny)) { bottom2=A0[Index3D (nx, ny, i2, j, 0)]; sh_A0[sh_id2]=A0[Index3D (nx, ny, i2, j, 1)]; } __syncthreads(); for(int k=1;k<nz-1;k++) { float a_left_right,a_top,a_down; //load required data on xy planes //if it on shared memory, load from shared memory //if not, load from global memory if((i<nx) &&(j<ny)) top=A0[Index3D (nx, ny, i, j, k+1)]; if(w_region) { a_left_right=x_l_bound?A0[Index3D (nx, ny, i-1, j, k )]:sh_A0[sh_id-1]; a_down=y_l_bound?A0[Index3D (nx, ny, i, j-1, k )]:sh_A0[sh_id-2*blockDim.x]; a_top=y_h_bound?A0[Index3D (nx, ny, i, j+1, k )]:sh_A0[sh_id+2*blockDim.x]; Anext[Index3D (nx, ny, i, j, k)] = (a_left_right + sh_A0[sh_id+1] + a_top + a_down + top + bottom)*c1 - sh_A0[sh_id]*c0 ; } //load another block if((i2<nx) &&(j<ny)) top2=A0[Index3D (nx, ny, i2, j, k+1)]; if(w_region2) { a_left_right=x_h_bound?A0[Index3D (nx, ny, i2+1, j, k )]:sh_A0[sh_id2+1]; a_down=y_l_bound?A0[Index3D (nx, ny, i2, j-1, k )]:sh_A0[sh_id2-2*blockDim.x]; a_top=y_h_bound?A0[Index3D (nx, ny, i2, j+1, k )]:sh_A0[sh_id2+2*blockDim.x]; Anext[Index3D (nx, ny, i2, j, k)] = (a_left_right + sh_A0[sh_id2-1] + a_top + a_down + top2 +bottom2)*c1 - sh_A0[sh_id2]*c0; } //swap data __syncthreads(); bottom=sh_A0[sh_id]; sh_A0[sh_id]=top; bottom2=sh_A0[sh_id2]; sh_A0[sh_id2]=top2; __syncthreads(); } }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
/*************************************************************************** *cr *cr (C) Copyright 2010 The Board of Trustees of the *cr University of Illinois *cr All Rights Reserved *cr ***************************************************************************/ #include <hip/hip_runtime.h> #include "common.h" __global__ void block2D_hybrid_coarsen_x(float c0,float c1,float *A0,float *Anext, int nx, int ny, int nz) { //thread coarsening along x direction const int i = blockIdx.x*blockDim.x*2+threadIdx.x; const int i2= blockIdx.x*blockDim.x*2+threadIdx.x+blockDim.x; const int j = blockIdx.y*blockDim.y+threadIdx.y; const int sh_id=threadIdx.x + threadIdx.y*blockDim.x*2; const int sh_id2=threadIdx.x +blockDim.x+ threadIdx.y*blockDim.x*2; //shared memeory extern __shared__ float sh_A0[]; sh_A0[sh_id]=0.0f; sh_A0[sh_id2]=0.0f; __syncthreads(); //get available region for load and store const bool w_region = i>0 && j>0 &&(i<nx-1) &&(j<ny-1) ; const bool w_region2 = j>0 &&(i2<nx-1) &&(j<ny-1) ; const bool x_l_bound = (threadIdx.x==0); const bool x_h_bound = ((threadIdx.x+blockDim.x)==(blockDim.x*2-1)); const bool y_l_bound = (threadIdx.y==0); const bool y_h_bound = (threadIdx.y==(blockDim.y-1)); //register for bottom and top planes //because of thread coarsening, we need to doulbe registers float bottom=0.0f,bottom2=0.0f,top=0.0f,top2=0.0f; //load data for bottom and current if((i<nx) &&(j<ny)) { bottom=A0[Index3D (nx, ny, i, j, 0)]; sh_A0[sh_id]=A0[Index3D (nx, ny, i, j, 1)]; } if((i2<nx) &&(j<ny)) { bottom2=A0[Index3D (nx, ny, i2, j, 0)]; sh_A0[sh_id2]=A0[Index3D (nx, ny, i2, j, 1)]; } __syncthreads(); for(int k=1;k<nz-1;k++) { float a_left_right,a_top,a_down; //load required data on xy planes //if it on shared memory, load from shared memory //if not, load from global memory if((i<nx) &&(j<ny)) top=A0[Index3D (nx, ny, i, j, k+1)]; if(w_region) { a_left_right=x_l_bound?A0[Index3D (nx, ny, i-1, j, k )]:sh_A0[sh_id-1]; a_down=y_l_bound?A0[Index3D (nx, ny, i, j-1, k )]:sh_A0[sh_id-2*blockDim.x]; a_top=y_h_bound?A0[Index3D (nx, ny, i, j+1, k )]:sh_A0[sh_id+2*blockDim.x]; Anext[Index3D (nx, ny, i, j, k)] = (a_left_right + sh_A0[sh_id+1] + a_top + a_down + top + bottom)*c1 - sh_A0[sh_id]*c0 ; } //load another block if((i2<nx) &&(j<ny)) top2=A0[Index3D (nx, ny, i2, j, k+1)]; if(w_region2) { a_left_right=x_h_bound?A0[Index3D (nx, ny, i2+1, j, k )]:sh_A0[sh_id2+1]; a_down=y_l_bound?A0[Index3D (nx, ny, i2, j-1, k )]:sh_A0[sh_id2-2*blockDim.x]; a_top=y_h_bound?A0[Index3D (nx, ny, i2, j+1, k )]:sh_A0[sh_id2+2*blockDim.x]; Anext[Index3D (nx, ny, i2, j, k)] = (a_left_right + sh_A0[sh_id2-1] + a_top + a_down + top2 +bottom2)*c1 - sh_A0[sh_id2]*c0; } //swap data __syncthreads(); bottom=sh_A0[sh_id]; sh_A0[sh_id]=top; bottom2=sh_A0[sh_id2]; sh_A0[sh_id2]=top2; __syncthreads(); } }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z24block2D_hybrid_coarsen_xffPfS_iii .globl _Z24block2D_hybrid_coarsen_xffPfS_iii .p2align 8 .type _Z24block2D_hybrid_coarsen_xffPfS_iii,@function _Z24block2D_hybrid_coarsen_xffPfS_iii: s_clause 0x1 s_load_b32 s2, s[0:1], 0x34 s_load_b64 s[12:13], s[0:1], 0x18 v_bfe_u32 v10, v0, 10, 10 v_dual_mov_b32 v1, 0 :: v_dual_and_b32 v14, 0x3ff, v0 s_lshl_b32 s19, s14, 1 s_mov_b32 s5, 0 s_delay_alu instid0(VALU_DEP_2) v_lshlrev_b32_e32 v5, 1, v10 s_mov_b32 s14, 0 s_waitcnt lgkmcnt(0) s_and_b32 s18, s2, 0xffff s_lshr_b32 s4, s2, 16 s_mul_i32 s17, s19, s18 s_mul_i32 s20, s15, s4 v_add_nc_u32_e32 v3, s18, v14 v_add_nc_u32_e32 v0, s17, v14 v_add_nc_u32_e32 v4, s20, v10 v_mad_u32_u24 v2, v5, s18, v14 s_add_i32 s6, s12, -1 v_mad_u32_u24 v7, v5, s18, v3 v_cmp_lt_i32_e64 s2, 0, v0 v_cmp_lt_i32_e32 vcc_lo, 0, v4 v_cmp_gt_i32_e64 s3, s6, v0 v_lshl_add_u32 v15, v2, 2, 0 v_lshl_add_u32 v16, v7, 2, 0 s_and_b32 s2, s2, vcc_lo s_delay_alu instid0(VALU_DEP_3) | instid1(SALU_CYCLE_1) s_and_b32 s2, s2, s3 ds_store_b32 v15, v1 ds_store_b32 v16, v1 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_and_saveexec_b32 s3, s2 s_add_i32 s2, s13, -1 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cmp_gt_i32_e64 s2, s2, v4 s_and_b32 s14, s2, exec_lo s_or_b32 exec_lo, exec_lo, s3 v_add_nc_u32_e32 v1, s18, v0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cmp_gt_i32_e64 s2, s6, v1 s_and_b32 s3, vcc_lo, s2 s_delay_alu instid0(SALU_CYCLE_1) s_and_saveexec_b32 s2, s3 s_add_i32 s3, s13, -1 s_delay_alu instid0(SALU_CYCLE_1) v_cmp_gt_i32_e32 vcc_lo, s3, v4 s_and_b32 s5, vcc_lo, exec_lo s_or_b32 exec_lo, exec_lo, s2 s_load_b64 s[6:7], s[0:1], 0x8 v_cmp_gt_i32_e64 s2, s12, v0 v_cmp_gt_i32_e32 vcc_lo, s13, v4 v_mov_b32_e32 v17, 0 v_add_nc_u32_e32 v5, s13, v4 v_mov_b32_e32 v19, 0 s_and_b32 s15, s2, vcc_lo s_delay_alu instid0(SALU_CYCLE_1) s_and_saveexec_b32 s3, s15 s_cbranch_execz .LBB0_6 v_mad_u64_u32 v[8:9], null, v5, s12, v[0:1] v_mad_u64_u32 v[11:12], null, v4, s12, v[0:1] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_ashrrev_i32_e32 v9, 31, v8 v_ashrrev_i32_e32 v12, 31, v11 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_lshlrev_b64 v[8:9], 2, v[8:9] v_lshlrev_b64 v[11:12], 2, v[11:12] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v8, s2, s6, v8 v_add_co_ci_u32_e64 v9, s2, s7, v9, s2 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v11, s2, s6, v11 v_add_co_ci_u32_e64 v12, s2, s7, v12, s2 s_clause 0x1 global_load_b32 v0, v[8:9], off global_load_b32 v19, v[11:12], off s_waitcnt vmcnt(1) ds_store_b32 v15, v0 .LBB0_6: s_or_b32 exec_lo, exec_lo, s3 v_cmp_gt_i32_e64 s2, s12, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_and_b32 s16, s2, vcc_lo s_and_saveexec_b32 s2, s16 s_cbranch_execz .LBB0_8 v_mad_u64_u32 v[8:9], null, v5, s12, v[1:2] v_mad_u64_u32 v[5:6], null, v4, s12, v[1:2] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_ashrrev_i32_e32 v9, 31, v8 v_ashrrev_i32_e32 v6, 31, v5 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_lshlrev_b64 v[0:1], 2, v[8:9] v_lshlrev_b64 v[4:5], 2, v[5:6] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_co_u32 v0, vcc_lo, s6, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s7, v1, vcc_lo s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_u32 v4, vcc_lo, s6, v4 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v5, vcc_lo s_clause 0x1 global_load_b32 v0, v[0:1], off global_load_b32 v17, v[4:5], off s_waitcnt vmcnt(1) ds_store_b32 v16, v0 .LBB0_8: s_or_b32 exec_lo, exec_lo, s2 s_load_b32 s21, s[0:1], 0x20 s_waitcnt vmcnt(0) lgkmcnt(0) s_barrier buffer_gl0_inv s_cmp_lt_i32 s21, 3 s_cbranch_scc1 .LBB0_31 v_lshl_add_u32 v6, v2, 2, 0 s_lshl_b32 s24, s18, 1 s_clause 0x1 s_load_b64 s[8:9], s[0:1], 0x0 s_load_b64 s[10:11], s[0:1], 0x10 v_subrev_nc_u32_e32 v1, s24, v2 s_add_i32 s2, s24, -1 v_dual_mov_b32 v29, 0 :: v_dual_add_nc_u32 v0, -4, v6 v_cmp_eq_u32_e64 s2, s2, v3 s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_4) v_lshl_add_u32 v3, v1, 2, 0 v_lshl_add_u32 v12, v7, 2, 0 v_cmp_ne_u32_e64 s0, -1, v0 s_mov_b64 s[22:23], src_shared_base v_add_nc_u32_e32 v2, s24, v2 v_subrev_nc_u32_e32 v5, s24, v7 v_add_nc_u32_e32 v8, 4, v12 v_cndmask_b32_e64 v0, 0, v0, s0 v_cndmask_b32_e64 v1, 0, s23, s0 v_cmp_ne_u32_e64 s0, -1, v3 v_lshl_add_u32 v4, v2, 2, 0 v_add3_u32 v13, s20, s13, v10 v_add_nc_u32_e32 v18, 4, v6 v_dual_mov_b32 v30, 0 :: v_dual_add_nc_u32 v11, s24, v7 v_cndmask_b32_e64 v2, 0, v3, s0 v_cndmask_b32_e64 v3, 0, s23, s0 v_cmp_ne_u32_e64 s0, -1, v8 v_cmp_ne_u32_e64 s1, -1, v4 v_lshl_add_u32 v9, v5, 2, 0 v_mul_lo_u32 v25, s12, v13 v_add_nc_u32_e32 v13, -1, v13 v_cndmask_b32_e64 v6, 0, v8, s0 v_cndmask_b32_e64 v7, 0, s23, s0 s_lshl_b32 s0, s13, 1 v_lshl_add_u32 v11, v11, 2, 0 v_add3_u32 v20, s0, s20, v10 v_cndmask_b32_e64 v4, 0, v4, s1 v_cndmask_b32_e64 v5, 0, s23, s1 v_cmp_ne_u32_e64 s1, -1, v9 v_mul_lo_u32 v13, s12, v13 v_mul_lo_u32 v28, s12, v20 v_cmp_ne_u32_e64 s0, -1, v11 v_add_nc_u32_e32 v26, s12, v25 v_cndmask_b32_e64 v8, 0, v9, s1 v_cndmask_b32_e64 v9, 0, s23, s1 s_add_i32 s1, s19, 1 s_add_i32 s4, s4, -1 s_mul_i32 s1, s1, s18 v_cmp_eq_u32_e32 vcc_lo, 0, v14 v_cmp_eq_u32_e64 s3, 0, v10 v_cmp_eq_u32_e64 s4, s4, v10 v_cndmask_b32_e64 v10, 0, v11, s0 v_cndmask_b32_e64 v11, 0, s23, s0 v_add_nc_u32_e32 v20, -4, v12 v_add_nc_u32_e32 v21, s1, v25 v_add_nc_u32_e32 v22, s1, v26 v_add_nc_u32_e32 v23, s1, v13 v_add_nc_u32_e32 v24, s1, v28 v_add_nc_u32_e32 v25, s17, v25 v_add_nc_u32_e32 v26, s17, v26 v_add_nc_u32_e32 v27, s17, v13 v_add_nc_u32_e32 v28, s17, v28 s_add_i32 s1, s21, -2 s_mul_i32 s12, s13, s12 s_branch .LBB0_12 .LBB0_10: s_or_b32 exec_lo, exec_lo, s17 flat_load_b32 v12, v[12:13] ds_load_b32 v13, v20 ds_load_b32 v32, v16 s_waitcnt vmcnt(2) lgkmcnt(0) v_add_f32_e32 v13, v19, v13 s_waitcnt vmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_f32_e32 v12, v12, v13 v_add_f32_e32 v13, v31, v12 v_dual_mul_f32 v31, s8, v32 :: v_dual_add_nc_u32 v12, v21, v14 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f32_e32 v19, v29, v13 v_ashrrev_i32_e32 v13, 31, v12 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f32_e32 v17, v17, v19 v_lshlrev_b64 v[12:13], 2, v[12:13] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f32 v17, v17, s9, -v31 v_add_co_u32 v12, s0, s10, v12 s_delay_alu instid0(VALU_DEP_1) v_add_co_ci_u32_e64 v13, s0, s11, v13, s0 global_store_b32 v[12:13], v17, off .LBB0_11: s_or_b32 exec_lo, exec_lo, s13 s_waitcnt vmcnt(0) lgkmcnt(0) s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv ds_load_b32 v19, v15 ds_store_b32 v15, v30 ds_load_b32 v17, v16 v_add_nc_u32_e32 v14, s12, v14 s_add_i32 s1, s1, -1 ds_store_b32 v16, v29 s_cmp_eq_u32 s1, 0 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_cbranch_scc1 .LBB0_31 .LBB0_12: s_and_saveexec_b32 s13, s15 s_cbranch_execz .LBB0_14 v_add_nc_u32_e32 v12, v28, v14 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 global_load_b32 v30, v[12:13], off .LBB0_14: s_or_b32 exec_lo, exec_lo, s13 s_and_saveexec_b32 s13, s14 s_cbranch_execz .LBB0_22 v_dual_mov_b32 v13, v1 :: v_dual_mov_b32 v12, v0 s_and_saveexec_b32 s17, vcc_lo v_add3_u32 v12, v25, v14, -1 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 s_or_b32 exec_lo, exec_lo, s17 flat_load_b32 v31, v[12:13] v_dual_mov_b32 v13, v3 :: v_dual_mov_b32 v12, v2 s_and_saveexec_b32 s17, s3 v_add_nc_u32_e32 v12, v27, v14 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 s_or_b32 exec_lo, exec_lo, s17 flat_load_b32 v32, v[12:13] v_dual_mov_b32 v13, v5 :: v_dual_mov_b32 v12, v4 s_and_saveexec_b32 s17, s4 v_add_nc_u32_e32 v12, v26, v14 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 s_or_b32 exec_lo, exec_lo, s17 flat_load_b32 v12, v[12:13] ds_load_b32 v13, v18 ds_load_b32 v33, v15 s_waitcnt vmcnt(2) lgkmcnt(0) v_add_f32_e32 v13, v31, v13 s_waitcnt vmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_f32_e32 v12, v12, v13 v_dual_add_f32 v13, v32, v12 :: v_dual_mul_f32 v32, s8, v33 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_dual_add_f32 v31, v30, v13 :: v_dual_add_nc_u32 v12, v25, v14 v_ashrrev_i32_e32 v13, 31, v12 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f32_e32 v19, v19, v31 v_lshlrev_b64 v[12:13], 2, v[12:13] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f32 v19, v19, s9, -v32 v_add_co_u32 v12, s0, s10, v12 s_delay_alu instid0(VALU_DEP_1) v_add_co_ci_u32_e64 v13, s0, s11, v13, s0 global_store_b32 v[12:13], v19, off .LBB0_22: s_or_b32 exec_lo, exec_lo, s13 s_and_saveexec_b32 s13, s16 s_cbranch_execz .LBB0_24 v_add_nc_u32_e32 v12, v24, v14 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 global_load_b32 v29, v[12:13], off .LBB0_24: s_or_b32 exec_lo, exec_lo, s13 s_and_saveexec_b32 s13, s5 s_cbranch_execz .LBB0_11 v_dual_mov_b32 v13, v7 :: v_dual_mov_b32 v12, v6 s_and_saveexec_b32 s17, s2 v_add3_u32 v12, v21, v14, 1 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 s_or_b32 exec_lo, exec_lo, s17 flat_load_b32 v19, v[12:13] v_dual_mov_b32 v13, v9 :: v_dual_mov_b32 v12, v8 s_and_saveexec_b32 s17, s3 v_add_nc_u32_e32 v12, v23, v14 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 s_or_b32 exec_lo, exec_lo, s17 flat_load_b32 v31, v[12:13] v_dual_mov_b32 v13, v11 :: v_dual_mov_b32 v12, v10 s_and_saveexec_b32 s17, s4 s_cbranch_execz .LBB0_10 v_add_nc_u32_e32 v12, v22, v14 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 s_branch .LBB0_10 .LBB0_31: s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z24block2D_hybrid_coarsen_xffPfS_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 34 .amdhsa_next_free_sgpr 25 .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 _Z24block2D_hybrid_coarsen_xffPfS_iii, .Lfunc_end0-_Z24block2D_hybrid_coarsen_xffPfS_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: - .offset: 0 .size: 4 .value_kind: by_value - .offset: 4 .size: 4 .value_kind: by_value - .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 - .offset: 160 .size: 4 .value_kind: hidden_dynamic_lds_size .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: _Z24block2D_hybrid_coarsen_xffPfS_iii .private_segment_fixed_size: 0 .sgpr_count: 27 .sgpr_spill_count: 0 .symbol: _Z24block2D_hybrid_coarsen_xffPfS_iii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 34 .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.
/*************************************************************************** *cr *cr (C) Copyright 2010 The Board of Trustees of the *cr University of Illinois *cr All Rights Reserved *cr ***************************************************************************/ #include <hip/hip_runtime.h> #include "common.h" __global__ void block2D_hybrid_coarsen_x(float c0,float c1,float *A0,float *Anext, int nx, int ny, int nz) { //thread coarsening along x direction const int i = blockIdx.x*blockDim.x*2+threadIdx.x; const int i2= blockIdx.x*blockDim.x*2+threadIdx.x+blockDim.x; const int j = blockIdx.y*blockDim.y+threadIdx.y; const int sh_id=threadIdx.x + threadIdx.y*blockDim.x*2; const int sh_id2=threadIdx.x +blockDim.x+ threadIdx.y*blockDim.x*2; //shared memeory extern __shared__ float sh_A0[]; sh_A0[sh_id]=0.0f; sh_A0[sh_id2]=0.0f; __syncthreads(); //get available region for load and store const bool w_region = i>0 && j>0 &&(i<nx-1) &&(j<ny-1) ; const bool w_region2 = j>0 &&(i2<nx-1) &&(j<ny-1) ; const bool x_l_bound = (threadIdx.x==0); const bool x_h_bound = ((threadIdx.x+blockDim.x)==(blockDim.x*2-1)); const bool y_l_bound = (threadIdx.y==0); const bool y_h_bound = (threadIdx.y==(blockDim.y-1)); //register for bottom and top planes //because of thread coarsening, we need to doulbe registers float bottom=0.0f,bottom2=0.0f,top=0.0f,top2=0.0f; //load data for bottom and current if((i<nx) &&(j<ny)) { bottom=A0[Index3D (nx, ny, i, j, 0)]; sh_A0[sh_id]=A0[Index3D (nx, ny, i, j, 1)]; } if((i2<nx) &&(j<ny)) { bottom2=A0[Index3D (nx, ny, i2, j, 0)]; sh_A0[sh_id2]=A0[Index3D (nx, ny, i2, j, 1)]; } __syncthreads(); for(int k=1;k<nz-1;k++) { float a_left_right,a_top,a_down; //load required data on xy planes //if it on shared memory, load from shared memory //if not, load from global memory if((i<nx) &&(j<ny)) top=A0[Index3D (nx, ny, i, j, k+1)]; if(w_region) { a_left_right=x_l_bound?A0[Index3D (nx, ny, i-1, j, k )]:sh_A0[sh_id-1]; a_down=y_l_bound?A0[Index3D (nx, ny, i, j-1, k )]:sh_A0[sh_id-2*blockDim.x]; a_top=y_h_bound?A0[Index3D (nx, ny, i, j+1, k )]:sh_A0[sh_id+2*blockDim.x]; Anext[Index3D (nx, ny, i, j, k)] = (a_left_right + sh_A0[sh_id+1] + a_top + a_down + top + bottom)*c1 - sh_A0[sh_id]*c0 ; } //load another block if((i2<nx) &&(j<ny)) top2=A0[Index3D (nx, ny, i2, j, k+1)]; if(w_region2) { a_left_right=x_h_bound?A0[Index3D (nx, ny, i2+1, j, k )]:sh_A0[sh_id2+1]; a_down=y_l_bound?A0[Index3D (nx, ny, i2, j-1, k )]:sh_A0[sh_id2-2*blockDim.x]; a_top=y_h_bound?A0[Index3D (nx, ny, i2, j+1, k )]:sh_A0[sh_id2+2*blockDim.x]; Anext[Index3D (nx, ny, i2, j, k)] = (a_left_right + sh_A0[sh_id2-1] + a_top + a_down + top2 +bottom2)*c1 - sh_A0[sh_id2]*c0; } //swap data __syncthreads(); bottom=sh_A0[sh_id]; sh_A0[sh_id]=top; bottom2=sh_A0[sh_id2]; sh_A0[sh_id2]=top2; __syncthreads(); } }
.text .file "kernels.hip" .globl _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii # -- Begin function _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii .p2align 4, 0x90 .type _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii,@function _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii: # @_Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movss %xmm0, 28(%rsp) movss %xmm1, 24(%rsp) movq %rdi, 88(%rsp) movq %rsi, 80(%rsp) movl %edx, 20(%rsp) movl %ecx, 16(%rsp) movl %r8d, 12(%rsp) leaq 28(%rsp), %rax movq %rax, 96(%rsp) leaq 24(%rsp), %rax movq %rax, 104(%rsp) leaq 88(%rsp), %rax movq %rax, 112(%rsp) leaq 80(%rsp), %rax movq %rax, 120(%rsp) leaq 20(%rsp), %rax movq %rax, 128(%rsp) leaq 16(%rsp), %rax movq %rax, 136(%rsp) leaq 12(%rsp), %rax movq %rax, 144(%rsp) leaq 64(%rsp), %rdi leaq 48(%rsp), %rsi leaq 40(%rsp), %rdx leaq 32(%rsp), %rcx callq __hipPopCallConfiguration movq 64(%rsp), %rsi movl 72(%rsp), %edx movq 48(%rsp), %rcx movl 56(%rsp), %r8d leaq 96(%rsp), %r9 movl $_Z24block2D_hybrid_coarsen_xffPfS_iii, %edi pushq 32(%rsp) .cfi_adjust_cfa_offset 8 pushq 48(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $168, %rsp .cfi_adjust_cfa_offset -168 retq .Lfunc_end0: .size _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii, .Lfunc_end0-_Z39__device_stub__block2D_hybrid_coarsen_xffPfS_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 $_Z24block2D_hybrid_coarsen_xffPfS_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 _Z24block2D_hybrid_coarsen_xffPfS_iii,@object # @_Z24block2D_hybrid_coarsen_xffPfS_iii .section .rodata,"a",@progbits .globl _Z24block2D_hybrid_coarsen_xffPfS_iii .p2align 3, 0x0 _Z24block2D_hybrid_coarsen_xffPfS_iii: .quad _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii .size _Z24block2D_hybrid_coarsen_xffPfS_iii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z24block2D_hybrid_coarsen_xffPfS_iii" .size .L__unnamed_1, 38 .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 _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z24block2D_hybrid_coarsen_xffPfS_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 : _Z24block2D_hybrid_coarsen_xffPfS_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 R0, SR_TID.Y ; /* 0x0000000000007919 */ /* 0x000e220000002200 */ /*0020*/ MOV R13, c[0x0][0x0] ; /* 0x00000000000d7a02 */ /* 0x000fe20000000f00 */ /*0030*/ ULDC.64 UR8, c[0x0][0x118] ; /* 0x0000460000087ab9 */ /* 0x000fe40000000a00 */ /*0040*/ S2R R21, SR_CTAID.Y ; /* 0x0000000000157919 */ /* 0x000e220000002600 */ /*0050*/ SHF.L.U32 R15, R13, 0x1, RZ ; /* 0x000000010d0f7819 */ /* 0x000fc600000006ff */ /*0060*/ S2R R3, SR_TID.X ; /* 0x0000000000037919 */ /* 0x000e680000002100 */ /*0070*/ S2R R10, SR_CTAID.X ; /* 0x00000000000a7919 */ /* 0x000ea20000002500 */ /*0080*/ IMAD R12, R21, c[0x0][0x4], R0 ; /* 0x00000100150c7a24 */ /* 0x001fe200078e0200 */ /*0090*/ IADD3 R14, R3, c[0x0][0x0], RZ ; /* 0x00000000030e7a10 */ /* 0x002fe20007ffe0ff */ /*00a0*/ IMAD R2, R15, R0, R3 ; /* 0x000000000f027224 */ /* 0x000fc600078e0203 */ /*00b0*/ ISETP.GE.AND P0, PT, R12, c[0x0][0x17c], PT ; /* 0x00005f000c007a0c */ /* 0x000fe20003f06270 */ /*00c0*/ IMAD R11, R15, R10, R3 ; /* 0x0000000a0f0b7224 */ /* 0x004fe200078e0203 */ /*00d0*/ STS [R2.X4], RZ ; /* 0x000000ff02007388 */ /* 0x0001e80000004800 */ /*00e0*/ ISETP.LT.AND P3, PT, R11.reuse, c[0x0][0x178], !P0 ; /* 0x00005e000b007a0c */ /* 0x040fe40004761270 */ /*00f0*/ IADD3 R23, R11, c[0x0][0x0], RZ ; /* 0x000000000b177a10 */ /* 0x000fc80007ffe0ff */ /*0100*/ ISETP.LT.AND P0, PT, R23, c[0x0][0x178], !P0 ; /* 0x00005e0017007a0c */ /* 0x000fce0004701270 */ /*0110*/ @P3 IADD3 R4, R12, c[0x0][0x17c], RZ ; /* 0x00005f000c043a10 */ /* 0x000fe40007ffe0ff */ /*0120*/ @P3 MOV R7, 0x4 ; /* 0x0000000400073802 */ /* 0x000fc60000000f00 */ /*0130*/ @P3 IMAD R18, R4, c[0x0][0x178], R11 ; /* 0x00005e0004123a24 */ /* 0x000fe200078e020b */ /*0140*/ @P0 IADD3 R16, R12, c[0x0][0x17c], RZ ; /* 0x00005f000c100a10 */ /* 0x000fe20007ffe0ff */ /*0150*/ IMAD R4, R15, R0, R14 ; /* 0x000000000f047224 */ /* 0x000fe200078e020e */ /*0160*/ @P0 MOV R9, 0x4 ; /* 0x0000000400090802 */ /* 0x000fe20000000f00 */ /*0170*/ @P3 IMAD.WIDE R18, R18, R7, c[0x0][0x168] ; /* 0x00005a0012123625 */ /* 0x000fc600078e0207 */ /*0180*/ STS [R4.X4], RZ ; /* 0x000000ff04007388 */ /* 0x0001e20000004800 */ /*0190*/ @P0 IMAD R16, R16, c[0x0][0x178], R23 ; /* 0x00005e0010100a24 */ /* 0x000fc600078e0217 */ /*01a0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fe20000010000 */ /*01b0*/ @P0 IMAD.WIDE R16, R16, R9, c[0x0][0x168] ; /* 0x00005a0010100625 */ /* 0x000fca00078e0209 */ /*01c0*/ @P3 LDG.E R19, [R18.64] ; /* 0x0000000812133981 */ /* 0x000ea8000c1e1900 */ /*01d0*/ @P0 LDG.E R17, [R16.64] ; /* 0x0000000810110981 */ /* 0x000ee2000c1e1900 */ /*01e0*/ HFMA2.MMA R26, -RZ, RZ, 0, 0 ; /* 0x00000000ff1a7435 */ /* 0x000fe200000001ff */ /*01f0*/ @P3 IMAD R6, R12.reuse, c[0x0][0x178], R11 ; /* 0x00005e000c063a24 */ /* 0x040fe200078e020b */ /*0200*/ MOV R5, RZ ; /* 0x000000ff00057202 */ /* 0x000fe20000000f00 */ /*0210*/ @P0 IMAD R8, R12, c[0x0][0x178], R23 ; /* 0x00005e000c080a24 */ /* 0x000fe400078e0217 */ /*0220*/ @P3 IMAD.WIDE R6, R6, R7, c[0x0][0x168] ; /* 0x00005a0006063625 */ /* 0x000fc800078e0207 */ /*0230*/ @P0 IMAD.WIDE R8, R8, R9, c[0x0][0x168] ; /* 0x00005a0008080625 */ /* 0x000fe200078e0209 */ /*0240*/ @P3 LDG.E R26, [R6.64] ; /* 0x00000008061a3981 */ /* 0x000168000c1e1900 */ /*0250*/ @P0 LDG.E R5, [R8.64] ; /* 0x0000000808050981 */ /* 0x000162000c1e1900 */ /*0260*/ MOV R22, c[0x0][0x180] ; /* 0x0000600000167a02 */ /* 0x000fc80000000f00 */ /*0270*/ ISETP.GE.AND P1, PT, R22, 0x3, PT ; /* 0x000000031600780c */ /* 0x000fe20003f26270 */ /*0280*/ @P3 STS [R2.X4], R19 ; /* 0x0000001302003388 */ /* 0x0041e80000004800 */ /*0290*/ @P0 STS [R4.X4], R17 ; /* 0x0000001104000388 */ /* 0x0081e80000004800 */ /*02a0*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*02b0*/ @!P1 EXIT ; /* 0x000000000000994d */ /* 0x000fea0003800000 */ /*02c0*/ UMOV UR6, 0x1 ; /* 0x0000000100067882 */ /* 0x001fe20000000000 */ /*02d0*/ ISETP.GT.AND P1, PT, R12, RZ, PT ; /* 0x000000ff0c00720c */ /* 0x000fe20003f24270 */ /*02e0*/ ULDC.64 UR4, c[0x0][0x178] ; /* 0x00005e0000047ab9 */ /* 0x000fe20000000a00 */ /*02f0*/ IADD3 R7, R15, -0x1, RZ ; /* 0xffffffff0f077810 */ /* 0x000fe20007ffe0ff */ /*0300*/ UIADD3 UR4, -UR6, UR4, URZ ; /* 0x0000000406047290 */ /* 0x000fe2000fffe13f */ /*0310*/ IADD3 R6, R0, c[0x0][0x17c], RZ ; /* 0x00005f0000067a10 */ /* 0x000fe20007ffe0ff */ /*0320*/ UIADD3 UR5, -UR6, UR5, URZ ; /* 0x0000000506057290 */ /* 0x000fe2000fffe13f */ /*0330*/ MOV R9, c[0x0][0x17c] ; /* 0x00005f0000097a02 */ /* 0x000fc40000000f00 */ /*0340*/ ISETP.NE.AND P4, PT, R14, R7, PT ; /* 0x000000070e00720c */ /* 0x000fe20003f85270 */ /*0350*/ IMAD R14, R21, c[0x0][0x4], R6 ; /* 0x00000100150e7a24 */ /* 0x000fe200078e0206 */ /*0360*/ ISETP.LT.AND P2, PT, R23, UR4, P1 ; /* 0x0000000417007c0c */ /* 0x000fe40008f41270 */ /*0370*/ ISETP.GT.AND P1, PT, R11, RZ, P1 ; /* 0x000000ff0b00720c */ /* 0x000fe40000f24270 */ /*0380*/ SHF.L.U32 R8, R4, 0x2, RZ ; /* 0x0000000204087819 */ /* 0x000fe400000006ff */ /*0390*/ LEA R18, R9, R12, 0x1 ; /* 0x0000000c09127211 */ /* 0x000fe200078e08ff */ /*03a0*/ IMAD R9, R14, c[0x0][0x178], R11.reuse ; /* 0x00005e000e097a24 */ /* 0x100fe200078e020b */ /*03b0*/ ISETP.LT.AND P5, PT, R11, UR4, P1 ; /* 0x000000040b007c0c */ /* 0x000fe20008fa1270 */ /*03c0*/ ULDC UR4, c[0x0][0x4] ; /* 0x0000010000047ab9 */ /* 0x000fe20000000800 */ /*03d0*/ LEA R7, R13, R8, 0x3 ; /* 0x000000080d077211 */ /* 0x000fe200078e18ff */ /*03e0*/ IMAD R8, R18, c[0x0][0x178], R11 ; /* 0x00005e0012087a24 */ /* 0x000fe200078e020b */ /*03f0*/ ISETP.LT.AND P1, PT, R12, UR5, P2 ; /* 0x000000050c007c0c */ /* 0x000fe20009721270 */ /*0400*/ IMAD R17, R18, c[0x0][0x178], R3 ; /* 0x00005e0012117a24 */ /* 0x000fe200078e0203 */ /*0410*/ ISETP.LT.AND P2, PT, R12, UR5, P5 ; /* 0x000000050c007c0c */ /* 0x000fe2000af41270 */ /*0420*/ UIADD3 UR4, -UR6, UR4, URZ ; /* 0x0000000406047290 */ /* 0x000fe2000fffe13f */ /*0430*/ SHF.L.U32 R6, R2, 0x2, RZ ; /* 0x0000000202067819 */ /* 0x000fe200000006ff */ /*0440*/ UMOV UR5, 0x1 ; /* 0x0000000100057882 */ /* 0x000fe20000000000 */ /*0450*/ IADD3 R12, R14, -0x1, RZ ; /* 0xffffffff0e0c7810 */ /* 0x000fc40007ffe0ff */ /*0460*/ IADD3 R18, R14.reuse, 0x1, RZ ; /* 0x000000010e127810 */ /* 0x040fe40007ffe0ff */ /*0470*/ LEA R6, R13, R6, 0x3 ; /* 0x000000060d067211 */ /* 0x000fe200078e18ff */ /*0480*/ IMAD R13, R14, c[0x0][0x178], R3.reuse ; /* 0x00005e000e0d7a24 */ /* 0x100fe200078e0203 */ /*0490*/ LEA R16, R10, 0x1, 0x1 ; /* 0x000000010a107811 */ /* 0x000fe200078e08ff */ /*04a0*/ IMAD R19, R12, c[0x0][0x178], R3.reuse ; /* 0x00005e000c137a24 */ /* 0x100fe200078e0203 */ /*04b0*/ IADD3 R22, R22, -0x1, RZ ; /* 0xffffffff16167810 */ /* 0x000fe20007ffe0ff */ /*04c0*/ IMAD R21, R18, c[0x0][0x178], R3 ; /* 0x00005e0012157a24 */ /* 0x000fe200078e0203 */ /*04d0*/ IADD3 R23, -R15.reuse, R2, RZ ; /* 0x000000020f177210 */ /* 0x040fe20007ffe1ff */ /*04e0*/ IMAD R10, R12, c[0x0][0x178], R11.reuse ; /* 0x00005e000c0a7a24 */ /* 0x100fe200078e020b */ /*04f0*/ IADD3 R24, -R15, R4, RZ ; /* 0x000000040f187210 */ /* 0x000fe20007ffe1ff */ /*0500*/ IMAD R11, R18, c[0x0][0x178], R11 ; /* 0x00005e00120b7a24 */ /* 0x000fe200078e020b */ /*0510*/ MOV R25, RZ ; /* 0x000000ff00197202 */ /* 0x000fe20000000f00 */ /*0520*/ IMAD R12, R16.reuse, c[0x0][0x0], R17 ; /* 0x00000000100c7a24 */ /* 0x040fe200078e0211 */ /*0530*/ MOV R27, RZ ; /* 0x000000ff001b7202 */ /* 0x000fe20000000f00 */ /*0540*/ IMAD R13, R16, c[0x0][0x0], R13 ; /* 0x00000000100d7a24 */ /* 0x000fc400078e020d */ /*0550*/ IMAD R20, R16.reuse, c[0x0][0x0], R19 ; /* 0x0000000010147a24 */ /* 0x040fe400078e0213 */ /*0560*/ IMAD R21, R16, c[0x0][0x0], R21 ; /* 0x0000000010157a24 */ /* 0x000fe400078e0215 */ /*0570*/ @P3 MOV R15, 0x4 ; /* 0x00000004000f3802 */ /* 0x000fca0000000f00 */ /*0580*/ @P3 IMAD.WIDE R14, R8, R15, c[0x0][0x168] ; /* 0x00005a00080e3625 */ /* 0x000fca00078e020f */ /*0590*/ @P3 LDG.E R27, [R14.64] ; /* 0x000000080e1b3981 */ /* 0x020162000c1e1900 */ /*05a0*/ BSSY B0, 0x760 ; /* 0x000001b000007945 */ /* 0x000fe20003800000 */ /*05b0*/ @!P2 BRA 0x750 ; /* 0x000001900000a947 */ /* 0x006fea0003800000 */ /*05c0*/ ISETP.NE.AND P5, PT, R3, RZ, PT ; /* 0x000000ff0300720c */ /* 0x000fe20003fa5270 */ /*05d0*/ HFMA2.MMA R28, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff1c7435 */ /* 0x000fe200000001ff */ /*05e0*/ LDS R15, [R2.X4+0x4] ; /* 0x00000400020f7984 */ /* 0x001ff60000004800 */ /*05f0*/ @P5 LDS R14, [R2.X4+-0x4] ; /* 0xfffffc00020e5984 */ /* 0x000e220000004800 */ /*0600*/ @!P5 IADD3 R19, R9, -0x1, RZ ; /* 0xffffffff0913d810 */ /* 0x000fca0007ffe0ff */ /*0610*/ @!P5 IMAD.WIDE R18, R19, R28, c[0x0][0x168] ; /* 0x00005a001312d625 */ /* 0x000fca00078e021c */ /*0620*/ @!P5 LDG.E R14, [R18.64] ; /* 0x00000008120ed981 */ /* 0x000e22000c1e1900 */ /*0630*/ ISETP.NE.AND P5, PT, R0, UR4, PT ; /* 0x0000000400007c0c */ /* 0x000fda000bfa5270 */ /*0640*/ @P5 LDS R16, [R6] ; /* 0x0000000006105984 */ /* 0x000e620000000800 */ /*0650*/ FADD R17, R15, R14 ; /* 0x0000000e0f117221 */ /* 0x001fe40000000000 */ /*0660*/ @!P5 IMAD.WIDE R14, R11, R28, c[0x0][0x168] ; /* 0x00005a000b0ed625 */ /* 0x000fca00078e021c */ /*0670*/ @!P5 LDG.E R16, [R14.64] ; /* 0x000000080e10d981 */ /* 0x000062000c1e1900 */ /*0680*/ ISETP.NE.AND P5, PT, R0, RZ, PT ; /* 0x000000ff0000720c */ /* 0x000fc60003fa5270 */ /*0690*/ LDS R15, [R2.X4] ; /* 0x00000000020f7984 */ /* 0x001ff40000004800 */ /*06a0*/ @P5 LDS R29, [R23.X4] ; /* 0x00000000171d5984 */ /* 0x000e220000004800 */ /*06b0*/ FADD R14, R17, R16 ; /* 0x00000010110e7221 */ /* 0x002fc40000000000 */ /*06c0*/ @!P5 IMAD.WIDE R16, R10, R28, c[0x0][0x168] ; /* 0x00005a000a10d625 */ /* 0x000fca00078e021c */ /*06d0*/ @!P5 LDG.E R29, [R16.64] ; /* 0x00000008101dd981 */ /* 0x000e24000c1e1900 */ /*06e0*/ FADD R18, R14, R29 ; /* 0x0000001d0e127221 */ /* 0x001fc80000000000 */ /*06f0*/ FADD R19, R27, R18 ; /* 0x000000121b137221 */ /* 0x020fe40000000000 */ /*0700*/ FMUL R18, R15, c[0x0][0x160] ; /* 0x000058000f127a20 */ /* 0x000fe40000400000 */ /*0710*/ FADD R19, R19, R26 ; /* 0x0000001a13137221 */ /* 0x000fe40000000000 */ /*0720*/ IMAD.WIDE R14, R9, R28, c[0x0][0x170] ; /* 0x00005c00090e7625 */ /* 0x000fc800078e021c */ /*0730*/ FFMA R19, R19, c[0x0][0x164], -R18 ; /* 0x0000590013137a23 */ /* 0x000fca0000000812 */ /*0740*/ STG.E [R14.64], R19 ; /* 0x000000130e007986 */ /* 0x0001e4000c101908 */ /*0750*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0760*/ BSSY B0, 0x7c0 ; /* 0x0000005000007945 */ /* 0x000fe20003800000 */ /*0770*/ @!P0 BRA 0x7b0 ; /* 0x0000003000008947 */ /* 0x000fea0003800000 */ /*0780*/ MOV R15, 0x4 ; /* 0x00000004000f7802 */ /* 0x001fca0000000f00 */ /*0790*/ IMAD.WIDE R14, R12, R15, c[0x0][0x168] ; /* 0x00005a000c0e7625 */ /* 0x000fca00078e020f */ /*07a0*/ LDG.E R25, [R14.64] ; /* 0x000000080e197981 */ /* 0x000164000c1e1900 */ /*07b0*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*07c0*/ BSSY B0, 0x970 ; /* 0x000001a000007945 */ /* 0x000fe20003800000 */ /*07d0*/ @!P1 BRA 0x960 ; /* 0x0000018000009947 */ /* 0x000fea0003800000 */ /*07e0*/ @P4 LDS R18, [R4.X4+0x4] ; /* 0x0000040004124984 */ /* 0x000fe20000004800 */ /*07f0*/ HFMA2.MMA R26, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff1a7435 */ /* 0x000fe200000001ff */ /*0800*/ @!P4 IADD3 R15, R13, 0x1, RZ ; /* 0x000000010d0fc810 */ /* 0x001fe40007ffe0ff */ /*0810*/ ISETP.NE.AND P6, PT, R0.reuse, UR4, PT ; /* 0x0000000400007c0c */ /* 0x040fe2000bfc5270 */ /*0820*/ LDS R19, [R4.X4+-0x4] ; /* 0xfffffc0004137984 */ /* 0x000e2c0000004800 */ /*0830*/ @!P4 IMAD.WIDE R14, R15, R26, c[0x0][0x168] ; /* 0x00005a000f0ec625 */ /* 0x000fe200078e021a */ /*0840*/ ISETP.NE.AND P5, PT, R0, RZ, PT ; /* 0x000000ff0000720c */ /* 0x000fca0003fa5270 */ /*0850*/ @P6 LDS R28, [R7] ; /* 0x00000000071c6984 */ /* 0x000e700000000800 */ /*0860*/ @P5 LDS R29, [R24.X4] ; /* 0x00000000181d5984 */ /* 0x000ea80000004800 */ /*0870*/ @!P4 LDG.E R18, [R14.64] ; /* 0x000000080e12c981 */ /* 0x000622000c1e1900 */ /*0880*/ @!P6 IMAD.WIDE R16, R21, R26, c[0x0][0x168] ; /* 0x00005a001510e625 */ /* 0x000fc600078e021a */ /*0890*/ LDS R14, [R4.X4] ; /* 0x00000000040e7984 */ /* 0x008ee80000004800 */ /*08a0*/ @!P6 LDG.E R28, [R16.64] ; /* 0x00000008101ce981 */ /* 0x000e62000c1e1900 */ /*08b0*/ FADD R15, R19, R18 ; /* 0x00000012130f7221 */ /* 0x001fe40000000000 */ /*08c0*/ @!P5 IMAD.WIDE R18, R20, R26, c[0x0][0x168] ; /* 0x00005a001412d625 */ /* 0x000fca00078e021a */ /*08d0*/ @!P5 LDG.E R29, [R18.64] ; /* 0x00000008121dd981 */ /* 0x000ea2000c1e1900 */ /*08e0*/ FADD R28, R15, R28 ; /* 0x0000001c0f1c7221 */ /* 0x002fc80000000000 */ /*08f0*/ FADD R28, R28, R29 ; /* 0x0000001d1c1c7221 */ /* 0x004fe40000000000 */ /*0900*/ FMUL R29, R14, c[0x0][0x160] ; /* 0x000058000e1d7a20 */ /* 0x008fe40000400000 */ /*0910*/ FADD R28, R25, R28 ; /* 0x0000001c191c7221 */ /* 0x020fe40000000000 */ /*0920*/ IMAD.WIDE R14, R13, R26, c[0x0][0x170] ; /* 0x00005c000d0e7625 */ /* 0x000fc800078e021a */ /*0930*/ FADD R5, R28, R5 ; /* 0x000000051c057221 */ /* 0x000fc80000000000 */ /*0940*/ FFMA R5, R5, c[0x0][0x164], -R29 ; /* 0x0000590005057a23 */ /* 0x000fca000000081d */ /*0950*/ STG.E [R14.64], R5 ; /* 0x000000050e007986 */ /* 0x0001e4000c101908 */ /*0960*/ BSYNC B0 ; /* 0x0000000000007941 */ /* 0x000fea0003800000 */ /*0970*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fe20000010000 */ /*0980*/ UIADD3 UR5, UR5, 0x1, URZ ; /* 0x0000000105057890 */ /* 0x000fe2000fffe03f */ /*0990*/ MOV R14, c[0x0][0x17c] ; /* 0x00005f00000e7a02 */ /* 0x001fca0000000f00 */ /*09a0*/ ISETP.LE.AND P5, PT, R22, UR5, PT ; /* 0x0000000516007c0c */ /* 0x000fe2000bfa3270 */ /*09b0*/ IMAD R13, R14.reuse, c[0x0][0x178], R13 ; /* 0x00005e000e0d7a24 */ /* 0x040fe400078e020d */ /*09c0*/ IMAD R9, R14.reuse, c[0x0][0x178], R9 ; /* 0x00005e000e097a24 */ /* 0x040fe400078e0209 */ /*09d0*/ IMAD R8, R14.reuse, c[0x0][0x178], R8 ; /* 0x00005e000e087a24 */ /* 0x040fe400078e0208 */ /*09e0*/ IMAD R12, R14.reuse, c[0x0][0x178], R12 ; /* 0x00005e000e0c7a24 */ /* 0x040fe400078e020c */ /*09f0*/ IMAD R10, R14.reuse, c[0x0][0x178], R10 ; /* 0x00005e000e0a7a24 */ /* 0x040fe400078e020a */ /*0a00*/ IMAD R11, R14, c[0x0][0x178], R11 ; /* 0x00005e000e0b7a24 */ /* 0x000fc400078e020b */ /*0a10*/ IMAD R20, R14.reuse, c[0x0][0x178], R20 ; /* 0x00005e000e147a24 */ /* 0x040fe400078e0214 */ /*0a20*/ IMAD R21, R14, c[0x0][0x178], R21 ; /* 0x00005e000e157a24 */ /* 0x000fe200078e0215 */ /*0a30*/ LDS R26, [R2.X4] ; /* 0x00000000021a7984 */ /* 0x0000680000004800 */ /*0a40*/ STS [R2.X4], R27 ; /* 0x0000001b02007388 */ /* 0x0201e80000004800 */ /*0a50*/ LDS R5, [R4.X4] ; /* 0x0000000004057984 */ /* 0x0000a80000004800 */ /*0a60*/ STS [R4.X4], R25 ; /* 0x0000001904007388 */ /* 0x0001e80000004800 */ /*0a70*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0a80*/ @!P5 BRA 0x570 ; /* 0xfffffae00000d947 */ /* 0x001fea000383ffff */ /*0a90*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0aa0*/ BRA 0xaa0; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*0ab0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ac0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ad0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0ae0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0af0*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b00*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b10*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b20*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b30*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b40*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b50*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b60*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ /*0b70*/ NOP; /* 0x0000000000007918 */ /* 0x000fc00000000000 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z24block2D_hybrid_coarsen_xffPfS_iii .globl _Z24block2D_hybrid_coarsen_xffPfS_iii .p2align 8 .type _Z24block2D_hybrid_coarsen_xffPfS_iii,@function _Z24block2D_hybrid_coarsen_xffPfS_iii: s_clause 0x1 s_load_b32 s2, s[0:1], 0x34 s_load_b64 s[12:13], s[0:1], 0x18 v_bfe_u32 v10, v0, 10, 10 v_dual_mov_b32 v1, 0 :: v_dual_and_b32 v14, 0x3ff, v0 s_lshl_b32 s19, s14, 1 s_mov_b32 s5, 0 s_delay_alu instid0(VALU_DEP_2) v_lshlrev_b32_e32 v5, 1, v10 s_mov_b32 s14, 0 s_waitcnt lgkmcnt(0) s_and_b32 s18, s2, 0xffff s_lshr_b32 s4, s2, 16 s_mul_i32 s17, s19, s18 s_mul_i32 s20, s15, s4 v_add_nc_u32_e32 v3, s18, v14 v_add_nc_u32_e32 v0, s17, v14 v_add_nc_u32_e32 v4, s20, v10 v_mad_u32_u24 v2, v5, s18, v14 s_add_i32 s6, s12, -1 v_mad_u32_u24 v7, v5, s18, v3 v_cmp_lt_i32_e64 s2, 0, v0 v_cmp_lt_i32_e32 vcc_lo, 0, v4 v_cmp_gt_i32_e64 s3, s6, v0 v_lshl_add_u32 v15, v2, 2, 0 v_lshl_add_u32 v16, v7, 2, 0 s_and_b32 s2, s2, vcc_lo s_delay_alu instid0(VALU_DEP_3) | instid1(SALU_CYCLE_1) s_and_b32 s2, s2, s3 ds_store_b32 v15, v1 ds_store_b32 v16, v1 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_and_saveexec_b32 s3, s2 s_add_i32 s2, s13, -1 s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cmp_gt_i32_e64 s2, s2, v4 s_and_b32 s14, s2, exec_lo s_or_b32 exec_lo, exec_lo, s3 v_add_nc_u32_e32 v1, s18, v0 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_cmp_gt_i32_e64 s2, s6, v1 s_and_b32 s3, vcc_lo, s2 s_delay_alu instid0(SALU_CYCLE_1) s_and_saveexec_b32 s2, s3 s_add_i32 s3, s13, -1 s_delay_alu instid0(SALU_CYCLE_1) v_cmp_gt_i32_e32 vcc_lo, s3, v4 s_and_b32 s5, vcc_lo, exec_lo s_or_b32 exec_lo, exec_lo, s2 s_load_b64 s[6:7], s[0:1], 0x8 v_cmp_gt_i32_e64 s2, s12, v0 v_cmp_gt_i32_e32 vcc_lo, s13, v4 v_mov_b32_e32 v17, 0 v_add_nc_u32_e32 v5, s13, v4 v_mov_b32_e32 v19, 0 s_and_b32 s15, s2, vcc_lo s_delay_alu instid0(SALU_CYCLE_1) s_and_saveexec_b32 s3, s15 s_cbranch_execz .LBB0_6 v_mad_u64_u32 v[8:9], null, v5, s12, v[0:1] v_mad_u64_u32 v[11:12], null, v4, s12, v[0:1] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_ashrrev_i32_e32 v9, 31, v8 v_ashrrev_i32_e32 v12, 31, v11 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_lshlrev_b64 v[8:9], 2, v[8:9] v_lshlrev_b64 v[11:12], 2, v[11:12] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v8, s2, s6, v8 v_add_co_ci_u32_e64 v9, s2, s7, v9, s2 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v11, s2, s6, v11 v_add_co_ci_u32_e64 v12, s2, s7, v12, s2 s_clause 0x1 global_load_b32 v0, v[8:9], off global_load_b32 v19, v[11:12], off s_waitcnt vmcnt(1) ds_store_b32 v15, v0 .LBB0_6: s_or_b32 exec_lo, exec_lo, s3 v_cmp_gt_i32_e64 s2, s12, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(SALU_CYCLE_1) s_and_b32 s16, s2, vcc_lo s_and_saveexec_b32 s2, s16 s_cbranch_execz .LBB0_8 v_mad_u64_u32 v[8:9], null, v5, s12, v[1:2] v_mad_u64_u32 v[5:6], null, v4, s12, v[1:2] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_ashrrev_i32_e32 v9, 31, v8 v_ashrrev_i32_e32 v6, 31, v5 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_lshlrev_b64 v[0:1], 2, v[8:9] v_lshlrev_b64 v[4:5], 2, v[5:6] s_waitcnt lgkmcnt(0) s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_3) v_add_co_u32 v0, vcc_lo, s6, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s7, v1, vcc_lo s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_4) v_add_co_u32 v4, vcc_lo, s6, v4 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v5, vcc_lo s_clause 0x1 global_load_b32 v0, v[0:1], off global_load_b32 v17, v[4:5], off s_waitcnt vmcnt(1) ds_store_b32 v16, v0 .LBB0_8: s_or_b32 exec_lo, exec_lo, s2 s_load_b32 s21, s[0:1], 0x20 s_waitcnt vmcnt(0) lgkmcnt(0) s_barrier buffer_gl0_inv s_cmp_lt_i32 s21, 3 s_cbranch_scc1 .LBB0_31 v_lshl_add_u32 v6, v2, 2, 0 s_lshl_b32 s24, s18, 1 s_clause 0x1 s_load_b64 s[8:9], s[0:1], 0x0 s_load_b64 s[10:11], s[0:1], 0x10 v_subrev_nc_u32_e32 v1, s24, v2 s_add_i32 s2, s24, -1 v_dual_mov_b32 v29, 0 :: v_dual_add_nc_u32 v0, -4, v6 v_cmp_eq_u32_e64 s2, s2, v3 s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_4) v_lshl_add_u32 v3, v1, 2, 0 v_lshl_add_u32 v12, v7, 2, 0 v_cmp_ne_u32_e64 s0, -1, v0 s_mov_b64 s[22:23], src_shared_base v_add_nc_u32_e32 v2, s24, v2 v_subrev_nc_u32_e32 v5, s24, v7 v_add_nc_u32_e32 v8, 4, v12 v_cndmask_b32_e64 v0, 0, v0, s0 v_cndmask_b32_e64 v1, 0, s23, s0 v_cmp_ne_u32_e64 s0, -1, v3 v_lshl_add_u32 v4, v2, 2, 0 v_add3_u32 v13, s20, s13, v10 v_add_nc_u32_e32 v18, 4, v6 v_dual_mov_b32 v30, 0 :: v_dual_add_nc_u32 v11, s24, v7 v_cndmask_b32_e64 v2, 0, v3, s0 v_cndmask_b32_e64 v3, 0, s23, s0 v_cmp_ne_u32_e64 s0, -1, v8 v_cmp_ne_u32_e64 s1, -1, v4 v_lshl_add_u32 v9, v5, 2, 0 v_mul_lo_u32 v25, s12, v13 v_add_nc_u32_e32 v13, -1, v13 v_cndmask_b32_e64 v6, 0, v8, s0 v_cndmask_b32_e64 v7, 0, s23, s0 s_lshl_b32 s0, s13, 1 v_lshl_add_u32 v11, v11, 2, 0 v_add3_u32 v20, s0, s20, v10 v_cndmask_b32_e64 v4, 0, v4, s1 v_cndmask_b32_e64 v5, 0, s23, s1 v_cmp_ne_u32_e64 s1, -1, v9 v_mul_lo_u32 v13, s12, v13 v_mul_lo_u32 v28, s12, v20 v_cmp_ne_u32_e64 s0, -1, v11 v_add_nc_u32_e32 v26, s12, v25 v_cndmask_b32_e64 v8, 0, v9, s1 v_cndmask_b32_e64 v9, 0, s23, s1 s_add_i32 s1, s19, 1 s_add_i32 s4, s4, -1 s_mul_i32 s1, s1, s18 v_cmp_eq_u32_e32 vcc_lo, 0, v14 v_cmp_eq_u32_e64 s3, 0, v10 v_cmp_eq_u32_e64 s4, s4, v10 v_cndmask_b32_e64 v10, 0, v11, s0 v_cndmask_b32_e64 v11, 0, s23, s0 v_add_nc_u32_e32 v20, -4, v12 v_add_nc_u32_e32 v21, s1, v25 v_add_nc_u32_e32 v22, s1, v26 v_add_nc_u32_e32 v23, s1, v13 v_add_nc_u32_e32 v24, s1, v28 v_add_nc_u32_e32 v25, s17, v25 v_add_nc_u32_e32 v26, s17, v26 v_add_nc_u32_e32 v27, s17, v13 v_add_nc_u32_e32 v28, s17, v28 s_add_i32 s1, s21, -2 s_mul_i32 s12, s13, s12 s_branch .LBB0_12 .LBB0_10: s_or_b32 exec_lo, exec_lo, s17 flat_load_b32 v12, v[12:13] ds_load_b32 v13, v20 ds_load_b32 v32, v16 s_waitcnt vmcnt(2) lgkmcnt(0) v_add_f32_e32 v13, v19, v13 s_waitcnt vmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_f32_e32 v12, v12, v13 v_add_f32_e32 v13, v31, v12 v_dual_mul_f32 v31, s8, v32 :: v_dual_add_nc_u32 v12, v21, v14 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f32_e32 v19, v29, v13 v_ashrrev_i32_e32 v13, 31, v12 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f32_e32 v17, v17, v19 v_lshlrev_b64 v[12:13], 2, v[12:13] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f32 v17, v17, s9, -v31 v_add_co_u32 v12, s0, s10, v12 s_delay_alu instid0(VALU_DEP_1) v_add_co_ci_u32_e64 v13, s0, s11, v13, s0 global_store_b32 v[12:13], v17, off .LBB0_11: s_or_b32 exec_lo, exec_lo, s13 s_waitcnt vmcnt(0) lgkmcnt(0) s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv ds_load_b32 v19, v15 ds_store_b32 v15, v30 ds_load_b32 v17, v16 v_add_nc_u32_e32 v14, s12, v14 s_add_i32 s1, s1, -1 ds_store_b32 v16, v29 s_cmp_eq_u32 s1, 0 s_waitcnt lgkmcnt(0) s_barrier buffer_gl0_inv s_cbranch_scc1 .LBB0_31 .LBB0_12: s_and_saveexec_b32 s13, s15 s_cbranch_execz .LBB0_14 v_add_nc_u32_e32 v12, v28, v14 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 global_load_b32 v30, v[12:13], off .LBB0_14: s_or_b32 exec_lo, exec_lo, s13 s_and_saveexec_b32 s13, s14 s_cbranch_execz .LBB0_22 v_dual_mov_b32 v13, v1 :: v_dual_mov_b32 v12, v0 s_and_saveexec_b32 s17, vcc_lo v_add3_u32 v12, v25, v14, -1 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 s_or_b32 exec_lo, exec_lo, s17 flat_load_b32 v31, v[12:13] v_dual_mov_b32 v13, v3 :: v_dual_mov_b32 v12, v2 s_and_saveexec_b32 s17, s3 v_add_nc_u32_e32 v12, v27, v14 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 s_or_b32 exec_lo, exec_lo, s17 flat_load_b32 v32, v[12:13] v_dual_mov_b32 v13, v5 :: v_dual_mov_b32 v12, v4 s_and_saveexec_b32 s17, s4 v_add_nc_u32_e32 v12, v26, v14 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 s_or_b32 exec_lo, exec_lo, s17 flat_load_b32 v12, v[12:13] ds_load_b32 v13, v18 ds_load_b32 v33, v15 s_waitcnt vmcnt(2) lgkmcnt(0) v_add_f32_e32 v13, v31, v13 s_waitcnt vmcnt(0) s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_f32_e32 v12, v12, v13 v_dual_add_f32 v13, v32, v12 :: v_dual_mul_f32 v32, s8, v33 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_dual_add_f32 v31, v30, v13 :: v_dual_add_nc_u32 v12, v25, v14 v_ashrrev_i32_e32 v13, 31, v12 s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_add_f32_e32 v19, v19, v31 v_lshlrev_b64 v[12:13], 2, v[12:13] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_fma_f32 v19, v19, s9, -v32 v_add_co_u32 v12, s0, s10, v12 s_delay_alu instid0(VALU_DEP_1) v_add_co_ci_u32_e64 v13, s0, s11, v13, s0 global_store_b32 v[12:13], v19, off .LBB0_22: s_or_b32 exec_lo, exec_lo, s13 s_and_saveexec_b32 s13, s16 s_cbranch_execz .LBB0_24 v_add_nc_u32_e32 v12, v24, v14 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 global_load_b32 v29, v[12:13], off .LBB0_24: s_or_b32 exec_lo, exec_lo, s13 s_and_saveexec_b32 s13, s5 s_cbranch_execz .LBB0_11 v_dual_mov_b32 v13, v7 :: v_dual_mov_b32 v12, v6 s_and_saveexec_b32 s17, s2 v_add3_u32 v12, v21, v14, 1 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 s_or_b32 exec_lo, exec_lo, s17 flat_load_b32 v19, v[12:13] v_dual_mov_b32 v13, v9 :: v_dual_mov_b32 v12, v8 s_and_saveexec_b32 s17, s3 v_add_nc_u32_e32 v12, v23, v14 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 s_or_b32 exec_lo, exec_lo, s17 flat_load_b32 v31, v[12:13] v_dual_mov_b32 v13, v11 :: v_dual_mov_b32 v12, v10 s_and_saveexec_b32 s17, s4 s_cbranch_execz .LBB0_10 v_add_nc_u32_e32 v12, v22, v14 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_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_add_co_u32 v12, s0, s6, v12 v_add_co_ci_u32_e64 v13, s0, s7, v13, s0 s_branch .LBB0_10 .LBB0_31: s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z24block2D_hybrid_coarsen_xffPfS_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 34 .amdhsa_next_free_sgpr 25 .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 _Z24block2D_hybrid_coarsen_xffPfS_iii, .Lfunc_end0-_Z24block2D_hybrid_coarsen_xffPfS_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: - .offset: 0 .size: 4 .value_kind: by_value - .offset: 4 .size: 4 .value_kind: by_value - .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 - .offset: 160 .size: 4 .value_kind: hidden_dynamic_lds_size .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: _Z24block2D_hybrid_coarsen_xffPfS_iii .private_segment_fixed_size: 0 .sgpr_count: 27 .sgpr_spill_count: 0 .symbol: _Z24block2D_hybrid_coarsen_xffPfS_iii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 34 .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_000601f4_00000000-6_kernels.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 _Z51__device_stub__Z24block2D_hybrid_coarsen_xffPfS_iiiffPfS_iii .type _Z51__device_stub__Z24block2D_hybrid_coarsen_xffPfS_iiiffPfS_iii, @function _Z51__device_stub__Z24block2D_hybrid_coarsen_xffPfS_iiiffPfS_iii: .LFB2051: .cfi_startproc endbr64 subq $184, %rsp .cfi_def_cfa_offset 192 movss %xmm0, 44(%rsp) movss %xmm1, 40(%rsp) movq %rdi, 32(%rsp) movq %rsi, 24(%rsp) movl %edx, 20(%rsp) movl %ecx, 16(%rsp) movl %r8d, 12(%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 32(%rsp), %rax movq %rax, 128(%rsp) leaq 24(%rsp), %rax movq %rax, 136(%rsp) leaq 20(%rsp), %rax movq %rax, 144(%rsp) leaq 16(%rsp), %rax movq %rax, 152(%rsp) leaq 12(%rsp), %rax movq %rax, 160(%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 _Z24block2D_hybrid_coarsen_xffPfS_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 _Z51__device_stub__Z24block2D_hybrid_coarsen_xffPfS_iiiffPfS_iii, .-_Z51__device_stub__Z24block2D_hybrid_coarsen_xffPfS_iiiffPfS_iii .globl _Z24block2D_hybrid_coarsen_xffPfS_iii .type _Z24block2D_hybrid_coarsen_xffPfS_iii, @function _Z24block2D_hybrid_coarsen_xffPfS_iii: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z51__device_stub__Z24block2D_hybrid_coarsen_xffPfS_iiiffPfS_iii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z24block2D_hybrid_coarsen_xffPfS_iii, .-_Z24block2D_hybrid_coarsen_xffPfS_iii .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC0: .string "_Z24block2D_hybrid_coarsen_xffPfS_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 _Z24block2D_hybrid_coarsen_xffPfS_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 "kernels.hip" .globl _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii # -- Begin function _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii .p2align 4, 0x90 .type _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii,@function _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii: # @_Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii .cfi_startproc # %bb.0: subq $152, %rsp .cfi_def_cfa_offset 160 movss %xmm0, 28(%rsp) movss %xmm1, 24(%rsp) movq %rdi, 88(%rsp) movq %rsi, 80(%rsp) movl %edx, 20(%rsp) movl %ecx, 16(%rsp) movl %r8d, 12(%rsp) leaq 28(%rsp), %rax movq %rax, 96(%rsp) leaq 24(%rsp), %rax movq %rax, 104(%rsp) leaq 88(%rsp), %rax movq %rax, 112(%rsp) leaq 80(%rsp), %rax movq %rax, 120(%rsp) leaq 20(%rsp), %rax movq %rax, 128(%rsp) leaq 16(%rsp), %rax movq %rax, 136(%rsp) leaq 12(%rsp), %rax movq %rax, 144(%rsp) leaq 64(%rsp), %rdi leaq 48(%rsp), %rsi leaq 40(%rsp), %rdx leaq 32(%rsp), %rcx callq __hipPopCallConfiguration movq 64(%rsp), %rsi movl 72(%rsp), %edx movq 48(%rsp), %rcx movl 56(%rsp), %r8d leaq 96(%rsp), %r9 movl $_Z24block2D_hybrid_coarsen_xffPfS_iii, %edi pushq 32(%rsp) .cfi_adjust_cfa_offset 8 pushq 48(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $168, %rsp .cfi_adjust_cfa_offset -168 retq .Lfunc_end0: .size _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii, .Lfunc_end0-_Z39__device_stub__block2D_hybrid_coarsen_xffPfS_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 $_Z24block2D_hybrid_coarsen_xffPfS_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 _Z24block2D_hybrid_coarsen_xffPfS_iii,@object # @_Z24block2D_hybrid_coarsen_xffPfS_iii .section .rodata,"a",@progbits .globl _Z24block2D_hybrid_coarsen_xffPfS_iii .p2align 3, 0x0 _Z24block2D_hybrid_coarsen_xffPfS_iii: .quad _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii .size _Z24block2D_hybrid_coarsen_xffPfS_iii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z24block2D_hybrid_coarsen_xffPfS_iii" .size .L__unnamed_1, 38 .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 _Z39__device_stub__block2D_hybrid_coarsen_xffPfS_iii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z24block2D_hybrid_coarsen_xffPfS_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.
#include <stdlib.h> #include <stdio.h> #include <cuda_runtime.h> #include <device_launch_parameters.h> __global__ void saxpy(int n, float a, float *x, float *y) { for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < n; i += blockDim.x * gridDim.x) { y[i] = a * x[i] + y[i]; __syncthreads(); } } int main(void) { int N = 1<<25; printf("%d\n", N); float *x, *y, *d_x, *d_y; cudaEvent_t start, stop; cudaEventCreate(&start); cudaEventCreate(&stop); x = (float*)malloc(N*sizeof(float)); y = (float*)malloc(N*sizeof(float)); cudaMalloc(&d_x, N*sizeof(float)); cudaMalloc(&d_y, N*sizeof(float)); for (int n = 0; n < N; n++) { x[n] = 1.0f; y[n] = 2.0f; } cudaMemcpy(d_x, x, N*sizeof(float), cudaMemcpyHostToDevice); cudaMemcpy(d_y, y, N*sizeof(float), cudaMemcpyHostToDevice); int numSMs; cudaDeviceGetAttribute(&numSMs, cudaDevAttrMultiProcessorCount, 0); printf("Blocks Threads Max error Elapsed time Eff. Memory Throughput " "Peak Memory Throughput Eff. Computational Throughput\n"); for(int j =1; j < 10; j++){ for(int i = 1; i < 10; i++){ int numThreads = i*32, numBlocks = j *32; printf("%d ", numBlocks); printf("%d ", numThreads); cudaEventRecord(start); saxpy<<<numBlocks, numThreads>>>(N, 2.0f, d_x, d_y); cudaEventRecord(stop); cudaMemcpy(y, d_y, N*sizeof(float), cudaMemcpyDeviceToHost); cudaEventSynchronize(stop); float milliseconds = 0; cudaEventElapsedTime(&milliseconds, start, stop); float maxError = 0.0f; for (int k = 0; k < N; k++) maxError = fmaxf(maxError, fabsf(y[k]-4.0f)); printf("%f ", maxError); printf("%f ", milliseconds); // for Memory Throughput Benchmarking: printf("%f GB/s ", 4.0*N*3.0/milliseconds/1.0e6); int clockRate; int busWidth; cudaDeviceGetAttribute(&clockRate, cudaDevAttrMemoryClockRate, 0); cudaDeviceGetAttribute(&busWidth, cudaDevAttrGlobalMemoryBusWidth, 0); printf("%f GB/s ", 2.0*clockRate*(busWidth/8.0)/1.0e6); // for Computational Throughput Benchmarking: // (1* multiply + 1*add) times the amount of vector entries printf("%f GFLOP/s \n",(2*N/milliseconds)/1.0e6); } } cudaEventDestroy(start); cudaEventDestroy(stop); cudaFree(d_x); cudaFree(d_y); free(x); free(y); }
code for sm_80 Function : _Z5saxpyifPfS_ .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 */ /* 0x001fca00078e0203 */ /*0040*/ ISETP.GE.AND P0, PT, R0, c[0x0][0x160], PT ; /* 0x0000580000007a0c */ /* 0x000fda0003f06270 */ /*0050*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0060*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe40000000a00 */ /*0070*/ HFMA2.MMA R5, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff057435 */ /* 0x001fd400000001ff */ /*0080*/ IMAD.WIDE R2, R0, R5, c[0x0][0x168] ; /* 0x00005a0000027625 */ /* 0x000fc800078e0205 */ /*0090*/ IMAD.WIDE R4, R0, R5, c[0x0][0x170] ; /* 0x00005c0000047625 */ /* 0x000fe400078e0205 */ /*00a0*/ LDG.E R2, [R2.64] ; /* 0x0000000402027981 */ /* 0x000ea8000c1e1900 */ /*00b0*/ LDG.E R7, [R4.64] ; /* 0x0000000404077981 */ /* 0x000ea2000c1e1900 */ /*00c0*/ MOV R9, c[0x0][0x0] ; /* 0x0000000000097a02 */ /* 0x000fe20000000f00 */ /*00d0*/ WARPSYNC 0xffffffff ; /* 0xffffffff00007948 */ /* 0x000fe80003800000 */ /*00e0*/ IMAD R0, R9, c[0x0][0xc], R0 ; /* 0x0000030009007a24 */ /* 0x000fca00078e0200 */ /*00f0*/ ISETP.GE.AND P0, PT, R0, c[0x0][0x160], PT ; /* 0x0000580000007a0c */ /* 0x000fe20003f06270 */ /*0100*/ FFMA R7, R2, c[0x0][0x164], R7 ; /* 0x0000590002077a23 */ /* 0x004fca0000000007 */ /*0110*/ STG.E [R4.64], R7 ; /* 0x0000000704007986 */ /* 0x0001e8000c101904 */ /*0120*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0130*/ @!P0 BRA 0x70 ; /* 0xffffff3000008947 */ /* 0x000fea000383ffff */ /*0140*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0150*/ BRA 0x150; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*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 <stdlib.h> #include <stdio.h> #include <cuda_runtime.h> #include <device_launch_parameters.h> __global__ void saxpy(int n, float a, float *x, float *y) { for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < n; i += blockDim.x * gridDim.x) { y[i] = a * x[i] + y[i]; __syncthreads(); } } int main(void) { int N = 1<<25; printf("%d\n", N); float *x, *y, *d_x, *d_y; cudaEvent_t start, stop; cudaEventCreate(&start); cudaEventCreate(&stop); x = (float*)malloc(N*sizeof(float)); y = (float*)malloc(N*sizeof(float)); cudaMalloc(&d_x, N*sizeof(float)); cudaMalloc(&d_y, N*sizeof(float)); for (int n = 0; n < N; n++) { x[n] = 1.0f; y[n] = 2.0f; } cudaMemcpy(d_x, x, N*sizeof(float), cudaMemcpyHostToDevice); cudaMemcpy(d_y, y, N*sizeof(float), cudaMemcpyHostToDevice); int numSMs; cudaDeviceGetAttribute(&numSMs, cudaDevAttrMultiProcessorCount, 0); printf("Blocks Threads Max error Elapsed time Eff. Memory Throughput " "Peak Memory Throughput Eff. Computational Throughput\n"); for(int j =1; j < 10; j++){ for(int i = 1; i < 10; i++){ int numThreads = i*32, numBlocks = j *32; printf("%d ", numBlocks); printf("%d ", numThreads); cudaEventRecord(start); saxpy<<<numBlocks, numThreads>>>(N, 2.0f, d_x, d_y); cudaEventRecord(stop); cudaMemcpy(y, d_y, N*sizeof(float), cudaMemcpyDeviceToHost); cudaEventSynchronize(stop); float milliseconds = 0; cudaEventElapsedTime(&milliseconds, start, stop); float maxError = 0.0f; for (int k = 0; k < N; k++) maxError = fmaxf(maxError, fabsf(y[k]-4.0f)); printf("%f ", maxError); printf("%f ", milliseconds); // for Memory Throughput Benchmarking: printf("%f GB/s ", 4.0*N*3.0/milliseconds/1.0e6); int clockRate; int busWidth; cudaDeviceGetAttribute(&clockRate, cudaDevAttrMemoryClockRate, 0); cudaDeviceGetAttribute(&busWidth, cudaDevAttrGlobalMemoryBusWidth, 0); printf("%f GB/s ", 2.0*clockRate*(busWidth/8.0)/1.0e6); // for Computational Throughput Benchmarking: // (1* multiply + 1*add) times the amount of vector entries printf("%f GFLOP/s \n",(2*N/milliseconds)/1.0e6); } } cudaEventDestroy(start); cudaEventDestroy(stop); cudaFree(d_x); cudaFree(d_y); free(x); free(y); }
.file "tmpxft_000e3121_00000000-6_SaxpyCudaBenchmarking.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 _Z28__device_stub__Z5saxpyifPfS_ifPfS_ .type _Z28__device_stub__Z5saxpyifPfS_ifPfS_, @function _Z28__device_stub__Z5saxpyifPfS_ifPfS_: .LFB2082: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movl %edi, 28(%rsp) movss %xmm0, 24(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax leaq 28(%rsp), %rax movq %rax, 96(%rsp) leaq 24(%rsp), %rax movq %rax, 104(%rsp) leaq 16(%rsp), %rax movq %rax, 112(%rsp) leaq 8(%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 .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 _Z5saxpyifPfS_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _Z28__device_stub__Z5saxpyifPfS_ifPfS_, .-_Z28__device_stub__Z5saxpyifPfS_ifPfS_ .globl _Z5saxpyifPfS_ .type _Z5saxpyifPfS_, @function _Z5saxpyifPfS_: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z28__device_stub__Z5saxpyifPfS_ifPfS_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _Z5saxpyifPfS_, .-_Z5saxpyifPfS_ .section .rodata.str1.1,"aMS",@progbits,1 .LC1: .string "%d\n" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC4: .string "Blocks\tThreads\tMax error\tElapsed time \tEff. Memory Throughput\tPeak Memory Throughput\tEff. Computational Throughput\n" .section .rodata.str1.1 .LC5: .string "%d\t" .LC8: .string "%f\t" .LC11: .string "%f GB/s\t\t" .LC14: .string "%f GFLOP/s\t\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 $104, %rsp .cfi_def_cfa_offset 160 movq %fs:40, %rax movq %rax, 88(%rsp) xorl %eax, %eax movl $33554432, %edx leaq .LC1(%rip), %rsi movl $2, %edi call __printf_chk@PLT leaq 48(%rsp), %rdi call cudaEventCreate@PLT leaq 56(%rsp), %rdi call cudaEventCreate@PLT movl $134217728, %edi call malloc@PLT movq %rax, %r14 movl $134217728, %edi call malloc@PLT movq %rax, %r13 leaq 32(%rsp), %rdi movl $134217728, %esi call cudaMalloc@PLT leaq 40(%rsp), %rdi movl $134217728, %esi call cudaMalloc@PLT movl $0, %eax movss .LC2(%rip), %xmm1 movss .LC3(%rip), %xmm0 .L12: movss %xmm1, (%r14,%rax) movss %xmm0, 0(%r13,%rax) addq $4, %rax cmpq $134217728, %rax jne .L12 movl $1, %ecx movl $134217728, %edx movq %r14, %rsi movq 32(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movl $134217728, %edx movq %r13, %rsi movq 40(%rsp), %rdi call cudaMemcpy@PLT leaq 24(%rsp), %rdi movl $0, %edx movl $16, %esi call cudaDeviceGetAttribute@PLT leaq .LC4(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $32, %r15d jmp .L13 .L14: movl $0, %esi movq 56(%rsp), %rdi call cudaEventRecord@PLT movl $2, %ecx movl $134217728, %edx movq 40(%rsp), %rsi movq %r13, %rdi call cudaMemcpy@PLT movq 56(%rsp), %rdi call cudaEventSynchronize@PLT movl $0x00000000, 28(%rsp) leaq 28(%rsp), %rdi movq 56(%rsp), %rdx movq 48(%rsp), %rsi call cudaEventElapsedTime@PLT movq %r13, %rbx leaq 134217728(%r13), %rbp pxor %xmm1, %xmm1 .L15: movss (%rbx), %xmm2 subss .LC6(%rip), %xmm2 andps .LC7(%rip), %xmm2 movaps %xmm2, %xmm0 call fmaxf@PLT movaps %xmm0, %xmm1 addq $4, %rbx cmpq %rbp, %rbx jne .L15 cvtss2sd %xmm0, %xmm0 leaq .LC8(%rip), %rbx movq %rbx, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT pxor %xmm0, %xmm0 cvtss2sd 28(%rsp), %xmm0 movq %rbx, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT pxor %xmm1, %xmm1 cvtss2sd 28(%rsp), %xmm1 movsd .LC9(%rip), %xmm0 divsd %xmm1, %xmm0 divsd .LC10(%rip), %xmm0 leaq .LC11(%rip), %rbx movq %rbx, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT leaq 64(%rsp), %rdi movl $0, %edx movl $36, %esi call cudaDeviceGetAttribute@PLT leaq 76(%rsp), %rdi movl $0, %edx movl $37, %esi call cudaDeviceGetAttribute@PLT pxor %xmm0, %xmm0 cvtsi2sdl 64(%rsp), %xmm0 addsd %xmm0, %xmm0 pxor %xmm1, %xmm1 cvtsi2sdl 76(%rsp), %xmm1 mulsd .LC12(%rip), %xmm1 mulsd %xmm1, %xmm0 divsd .LC10(%rip), %xmm0 movq %rbx, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movss .LC13(%rip), %xmm0 divss 28(%rsp), %xmm0 cvtss2sd %xmm0, %xmm0 divsd .LC10(%rip), %xmm0 leaq .LC14(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT addl $32, %r12d cmpl $320, %r12d je .L23 .L16: movl 12(%rsp), %edx leaq .LC5(%rip), %rbx movq %rbx, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl %r12d, %edx movq %rbx, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $0, %esi movq 48(%rsp), %rdi call cudaEventRecord@PLT movl %r12d, 76(%rsp) movl $1, 80(%rsp) movl $1, 84(%rsp) movl %r15d, 64(%rsp) movl $1, 68(%rsp) movl $1, 72(%rsp) movl $0, %r9d movl $0, %r8d movq 76(%rsp), %rdx movl $1, %ecx movq 64(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax jne .L14 movq 40(%rsp), %rdx movq 32(%rsp), %rsi movss .LC3(%rip), %xmm0 movl $33554432, %edi call _Z28__device_stub__Z5saxpyifPfS_ifPfS_ jmp .L14 .L23: addl $32, %r15d cmpl $320, %r15d je .L17 .L13: movl %r15d, 12(%rsp) movl $32, %r12d jmp .L16 .L17: movq 48(%rsp), %rdi call cudaEventDestroy@PLT movq 56(%rsp), %rdi call cudaEventDestroy@PLT movq 32(%rsp), %rdi call cudaFree@PLT movq 40(%rsp), %rdi call cudaFree@PLT movq %r14, %rdi call free@PLT movq %r13, %rdi call free@PLT movq 88(%rsp), %rax subq %fs:40, %rax jne .L24 movl $0, %eax addq $104, %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 .L24: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size main, .-main .section .rodata.str1.1 .LC15: .string "_Z5saxpyifPfS_" .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 .LC15(%rip), %rdx movq %rdx, %rcx leaq _Z5saxpyifPfS_(%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.cst4,"aM",@progbits,4 .align 4 .LC2: .long 1065353216 .align 4 .LC3: .long 1073741824 .align 4 .LC6: .long 1082130432 .section .rodata.cst16,"aM",@progbits,16 .align 16 .LC7: .long 2147483647 .long 0 .long 0 .long 0 .section .rodata.cst8,"aM",@progbits,8 .align 8 .LC9: .long 0 .long 1102577664 .align 8 .LC10: .long 0 .long 1093567616 .align 8 .LC12: .long 0 .long 1069547520 .section .rodata.cst4 .align 4 .LC13: .long 1283457024 .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 <stdlib.h> #include <stdio.h> #include <cuda_runtime.h> #include <device_launch_parameters.h> __global__ void saxpy(int n, float a, float *x, float *y) { for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < n; i += blockDim.x * gridDim.x) { y[i] = a * x[i] + y[i]; __syncthreads(); } } int main(void) { int N = 1<<25; printf("%d\n", N); float *x, *y, *d_x, *d_y; cudaEvent_t start, stop; cudaEventCreate(&start); cudaEventCreate(&stop); x = (float*)malloc(N*sizeof(float)); y = (float*)malloc(N*sizeof(float)); cudaMalloc(&d_x, N*sizeof(float)); cudaMalloc(&d_y, N*sizeof(float)); for (int n = 0; n < N; n++) { x[n] = 1.0f; y[n] = 2.0f; } cudaMemcpy(d_x, x, N*sizeof(float), cudaMemcpyHostToDevice); cudaMemcpy(d_y, y, N*sizeof(float), cudaMemcpyHostToDevice); int numSMs; cudaDeviceGetAttribute(&numSMs, cudaDevAttrMultiProcessorCount, 0); printf("Blocks Threads Max error Elapsed time Eff. Memory Throughput " "Peak Memory Throughput Eff. Computational Throughput\n"); for(int j =1; j < 10; j++){ for(int i = 1; i < 10; i++){ int numThreads = i*32, numBlocks = j *32; printf("%d ", numBlocks); printf("%d ", numThreads); cudaEventRecord(start); saxpy<<<numBlocks, numThreads>>>(N, 2.0f, d_x, d_y); cudaEventRecord(stop); cudaMemcpy(y, d_y, N*sizeof(float), cudaMemcpyDeviceToHost); cudaEventSynchronize(stop); float milliseconds = 0; cudaEventElapsedTime(&milliseconds, start, stop); float maxError = 0.0f; for (int k = 0; k < N; k++) maxError = fmaxf(maxError, fabsf(y[k]-4.0f)); printf("%f ", maxError); printf("%f ", milliseconds); // for Memory Throughput Benchmarking: printf("%f GB/s ", 4.0*N*3.0/milliseconds/1.0e6); int clockRate; int busWidth; cudaDeviceGetAttribute(&clockRate, cudaDevAttrMemoryClockRate, 0); cudaDeviceGetAttribute(&busWidth, cudaDevAttrGlobalMemoryBusWidth, 0); printf("%f GB/s ", 2.0*clockRate*(busWidth/8.0)/1.0e6); // for Computational Throughput Benchmarking: // (1* multiply + 1*add) times the amount of vector entries printf("%f GFLOP/s \n",(2*N/milliseconds)/1.0e6); } } cudaEventDestroy(start); cudaEventDestroy(stop); cudaFree(d_x); cudaFree(d_y); free(x); free(y); }
#include <stdlib.h> #include <stdio.h> #include <hip/hip_runtime.h> __global__ void saxpy(int n, float a, float *x, float *y) { for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < n; i += blockDim.x * gridDim.x) { y[i] = a * x[i] + y[i]; __syncthreads(); } } int main(void) { int N = 1<<25; printf("%d\n", N); float *x, *y, *d_x, *d_y; hipEvent_t start, stop; hipEventCreate(&start); hipEventCreate(&stop); x = (float*)malloc(N*sizeof(float)); y = (float*)malloc(N*sizeof(float)); hipMalloc(&d_x, N*sizeof(float)); hipMalloc(&d_y, N*sizeof(float)); for (int n = 0; n < N; n++) { x[n] = 1.0f; y[n] = 2.0f; } hipMemcpy(d_x, x, N*sizeof(float), hipMemcpyHostToDevice); hipMemcpy(d_y, y, N*sizeof(float), hipMemcpyHostToDevice); int numSMs; hipDeviceGetAttribute(&numSMs, hipDeviceAttributeMultiprocessorCount, 0); printf("Blocks Threads Max error Elapsed time Eff. Memory Throughput " "Peak Memory Throughput Eff. Computational Throughput\n"); for(int j =1; j < 10; j++){ for(int i = 1; i < 10; i++){ int numThreads = i*32, numBlocks = j *32; printf("%d ", numBlocks); printf("%d ", numThreads); hipEventRecord(start); saxpy<<<numBlocks, numThreads>>>(N, 2.0f, d_x, d_y); hipEventRecord(stop); hipMemcpy(y, d_y, N*sizeof(float), hipMemcpyDeviceToHost); hipEventSynchronize(stop); float milliseconds = 0; hipEventElapsedTime(&milliseconds, start, stop); float maxError = 0.0f; for (int k = 0; k < N; k++) maxError = fmaxf(maxError, fabsf(y[k]-4.0f)); printf("%f ", maxError); printf("%f ", milliseconds); // for Memory Throughput Benchmarking: printf("%f GB/s ", 4.0*N*3.0/milliseconds/1.0e6); int clockRate; int busWidth; hipDeviceGetAttribute(&clockRate, hipDeviceAttributeMemoryClockRate, 0); hipDeviceGetAttribute(&busWidth, hipDeviceAttributeMemoryBusWidth, 0); printf("%f GB/s ", 2.0*clockRate*(busWidth/8.0)/1.0e6); // for Computational Throughput Benchmarking: // (1* multiply + 1*add) times the amount of vector entries printf("%f GFLOP/s \n",(2*N/milliseconds)/1.0e6); } } hipEventDestroy(start); hipEventDestroy(stop); hipFree(d_x); hipFree(d_y); free(x); free(y); }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <stdlib.h> #include <stdio.h> #include <hip/hip_runtime.h> __global__ void saxpy(int n, float a, float *x, float *y) { for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < n; i += blockDim.x * gridDim.x) { y[i] = a * x[i] + y[i]; __syncthreads(); } } int main(void) { int N = 1<<25; printf("%d\n", N); float *x, *y, *d_x, *d_y; hipEvent_t start, stop; hipEventCreate(&start); hipEventCreate(&stop); x = (float*)malloc(N*sizeof(float)); y = (float*)malloc(N*sizeof(float)); hipMalloc(&d_x, N*sizeof(float)); hipMalloc(&d_y, N*sizeof(float)); for (int n = 0; n < N; n++) { x[n] = 1.0f; y[n] = 2.0f; } hipMemcpy(d_x, x, N*sizeof(float), hipMemcpyHostToDevice); hipMemcpy(d_y, y, N*sizeof(float), hipMemcpyHostToDevice); int numSMs; hipDeviceGetAttribute(&numSMs, hipDeviceAttributeMultiprocessorCount, 0); printf("Blocks Threads Max error Elapsed time Eff. Memory Throughput " "Peak Memory Throughput Eff. Computational Throughput\n"); for(int j =1; j < 10; j++){ for(int i = 1; i < 10; i++){ int numThreads = i*32, numBlocks = j *32; printf("%d ", numBlocks); printf("%d ", numThreads); hipEventRecord(start); saxpy<<<numBlocks, numThreads>>>(N, 2.0f, d_x, d_y); hipEventRecord(stop); hipMemcpy(y, d_y, N*sizeof(float), hipMemcpyDeviceToHost); hipEventSynchronize(stop); float milliseconds = 0; hipEventElapsedTime(&milliseconds, start, stop); float maxError = 0.0f; for (int k = 0; k < N; k++) maxError = fmaxf(maxError, fabsf(y[k]-4.0f)); printf("%f ", maxError); printf("%f ", milliseconds); // for Memory Throughput Benchmarking: printf("%f GB/s ", 4.0*N*3.0/milliseconds/1.0e6); int clockRate; int busWidth; hipDeviceGetAttribute(&clockRate, hipDeviceAttributeMemoryClockRate, 0); hipDeviceGetAttribute(&busWidth, hipDeviceAttributeMemoryBusWidth, 0); printf("%f GB/s ", 2.0*clockRate*(busWidth/8.0)/1.0e6); // for Computational Throughput Benchmarking: // (1* multiply + 1*add) times the amount of vector entries printf("%f GFLOP/s \n",(2*N/milliseconds)/1.0e6); } } hipEventDestroy(start); hipEventDestroy(stop); hipFree(d_x); hipFree(d_y); free(x); free(y); }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z5saxpyifPfS_ .globl _Z5saxpyifPfS_ .p2align 8 .type _Z5saxpyifPfS_,@function _Z5saxpyifPfS_: s_clause 0x1 s_load_b32 s5, s[0:1], 0x24 s_load_b32 s4, s[0:1], 0x0 s_add_u32 s2, s0, 24 s_addc_u32 s3, s1, 0 s_waitcnt lgkmcnt(0) s_and_b32 s6, s5, 0xffff s_mov_b32 s5, exec_lo v_mad_u64_u32 v[1:2], null, s15, s6, v[0:1] s_delay_alu instid0(VALU_DEP_1) v_cmpx_gt_i32_e64 s4, v1 s_cbranch_execz .LBB0_3 s_load_b32 s7, s[2:3], 0x0 s_clause 0x1 s_load_b32 s5, s[0:1], 0x4 s_load_b128 s[0:3], s[0:1], 0x8 s_waitcnt lgkmcnt(0) s_mul_i32 s6, s7, s6 s_mov_b32 s7, 0 .p2align 6 .LBB0_2: v_ashrrev_i32_e32 v2, 31, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[2:3], 2, v[1:2] v_add_co_u32 v4, vcc_lo, s0, v2 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v5, vcc_lo, s1, v3, vcc_lo v_add_co_u32 v2, vcc_lo, s2, v2 v_add_co_ci_u32_e32 v3, vcc_lo, s3, v3, vcc_lo global_load_b32 v0, v[4:5], off global_load_b32 v4, v[2:3], off s_waitcnt vmcnt(0) v_dual_fmac_f32 v4, s5, v0 :: v_dual_add_nc_u32 v1, s6, v1 s_delay_alu instid0(VALU_DEP_1) v_cmp_le_i32_e32 vcc_lo, s4, v1 global_store_b32 v[2:3], v4, off s_or_b32 s7, vcc_lo, s7 s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv s_and_not1_b32 exec_lo, exec_lo, s7 s_cbranch_execnz .LBB0_2 .LBB0_3: s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z5saxpyifPfS_ .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 280 .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 6 .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 _Z5saxpyifPfS_, .Lfunc_end0-_Z5saxpyifPfS_ .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 - .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: hidden_block_count_x - .offset: 28 .size: 4 .value_kind: hidden_block_count_y - .offset: 32 .size: 4 .value_kind: hidden_block_count_z - .offset: 36 .size: 2 .value_kind: hidden_group_size_x - .offset: 38 .size: 2 .value_kind: hidden_group_size_y - .offset: 40 .size: 2 .value_kind: hidden_group_size_z - .offset: 42 .size: 2 .value_kind: hidden_remainder_x - .offset: 44 .size: 2 .value_kind: hidden_remainder_y - .offset: 46 .size: 2 .value_kind: hidden_remainder_z - .offset: 64 .size: 8 .value_kind: hidden_global_offset_x - .offset: 72 .size: 8 .value_kind: hidden_global_offset_y - .offset: 80 .size: 8 .value_kind: hidden_global_offset_z - .offset: 88 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 280 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z5saxpyifPfS_ .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z5saxpyifPfS_.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 6 .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 <stdlib.h> #include <stdio.h> #include <hip/hip_runtime.h> __global__ void saxpy(int n, float a, float *x, float *y) { for (int i = blockIdx.x * blockDim.x + threadIdx.x; i < n; i += blockDim.x * gridDim.x) { y[i] = a * x[i] + y[i]; __syncthreads(); } } int main(void) { int N = 1<<25; printf("%d\n", N); float *x, *y, *d_x, *d_y; hipEvent_t start, stop; hipEventCreate(&start); hipEventCreate(&stop); x = (float*)malloc(N*sizeof(float)); y = (float*)malloc(N*sizeof(float)); hipMalloc(&d_x, N*sizeof(float)); hipMalloc(&d_y, N*sizeof(float)); for (int n = 0; n < N; n++) { x[n] = 1.0f; y[n] = 2.0f; } hipMemcpy(d_x, x, N*sizeof(float), hipMemcpyHostToDevice); hipMemcpy(d_y, y, N*sizeof(float), hipMemcpyHostToDevice); int numSMs; hipDeviceGetAttribute(&numSMs, hipDeviceAttributeMultiprocessorCount, 0); printf("Blocks Threads Max error Elapsed time Eff. Memory Throughput " "Peak Memory Throughput Eff. Computational Throughput\n"); for(int j =1; j < 10; j++){ for(int i = 1; i < 10; i++){ int numThreads = i*32, numBlocks = j *32; printf("%d ", numBlocks); printf("%d ", numThreads); hipEventRecord(start); saxpy<<<numBlocks, numThreads>>>(N, 2.0f, d_x, d_y); hipEventRecord(stop); hipMemcpy(y, d_y, N*sizeof(float), hipMemcpyDeviceToHost); hipEventSynchronize(stop); float milliseconds = 0; hipEventElapsedTime(&milliseconds, start, stop); float maxError = 0.0f; for (int k = 0; k < N; k++) maxError = fmaxf(maxError, fabsf(y[k]-4.0f)); printf("%f ", maxError); printf("%f ", milliseconds); // for Memory Throughput Benchmarking: printf("%f GB/s ", 4.0*N*3.0/milliseconds/1.0e6); int clockRate; int busWidth; hipDeviceGetAttribute(&clockRate, hipDeviceAttributeMemoryClockRate, 0); hipDeviceGetAttribute(&busWidth, hipDeviceAttributeMemoryBusWidth, 0); printf("%f GB/s ", 2.0*clockRate*(busWidth/8.0)/1.0e6); // for Computational Throughput Benchmarking: // (1* multiply + 1*add) times the amount of vector entries printf("%f GFLOP/s \n",(2*N/milliseconds)/1.0e6); } } hipEventDestroy(start); hipEventDestroy(stop); hipFree(d_x); hipFree(d_y); free(x); free(y); }
.text .file "SaxpyCudaBenchmarking.hip" .globl _Z20__device_stub__saxpyifPfS_ # -- Begin function _Z20__device_stub__saxpyifPfS_ .p2align 4, 0x90 .type _Z20__device_stub__saxpyifPfS_,@function _Z20__device_stub__saxpyifPfS_: # @_Z20__device_stub__saxpyifPfS_ .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movl %edi, 12(%rsp) movss %xmm0, 8(%rsp) movq %rsi, 72(%rsp) movq %rdx, 64(%rsp) leaq 12(%rsp), %rax movq %rax, 80(%rsp) leaq 8(%rsp), %rax movq %rax, 88(%rsp) leaq 72(%rsp), %rax movq %rax, 96(%rsp) leaq 64(%rsp), %rax movq %rax, 104(%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 $_Z5saxpyifPfS_, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $136, %rsp .cfi_adjust_cfa_offset -136 retq .Lfunc_end0: .size _Z20__device_stub__saxpyifPfS_, .Lfunc_end0-_Z20__device_stub__saxpyifPfS_ .cfi_endproc # -- End function .section .rodata.cst4,"aM",@progbits,4 .p2align 2, 0x0 # -- Begin function main .LCPI1_0: .long 0xc0800000 # float -4 .LCPI1_5: .long 0x4c800000 # float 67108864 .section .rodata.cst16,"aM",@progbits,16 .p2align 4, 0x0 .LCPI1_1: .long 0x7fffffff # float NaN .long 0x7fffffff # float NaN .long 0x7fffffff # float NaN .long 0x7fffffff # float NaN .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 .LCPI1_2: .quad 0x41b8000000000000 # double 402653184 .LCPI1_3: .quad 0x412e848000000000 # double 1.0E+6 .LCPI1_4: .quad 0x3fc0000000000000 # double 0.125 .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 $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 xorl %ebx, %ebx movl $.L.str, %edi movl $33554432, %esi # imm = 0x2000000 xorl %eax, %eax callq printf leaq 24(%rsp), %rdi callq hipEventCreate leaq 8(%rsp), %rdi callq hipEventCreate movl $134217728, %edi # imm = 0x8000000 callq malloc movq %rax, %r15 movl $134217728, %edi # imm = 0x8000000 callq malloc movq %rax, %r14 leaq 32(%rsp), %rdi movl $134217728, %esi # imm = 0x8000000 callq hipMalloc leaq 16(%rsp), %rdi movl $134217728, %esi # imm = 0x8000000 callq hipMalloc .p2align 4, 0x90 .LBB1_1: # =>This Inner Loop Header: Depth=1 movl $1065353216, (%r15,%rbx,4) # imm = 0x3F800000 movl $1073741824, (%r14,%rbx,4) # imm = 0x40000000 incq %rbx cmpq $33554432, %rbx # imm = 0x2000000 jne .LBB1_1 # %bb.2: movq 32(%rsp), %rdi movl $134217728, %edx # imm = 0x8000000 movq %r15, 112(%rsp) # 8-byte Spill movq %r15, %rsi movl $1, %ecx callq hipMemcpy movq 16(%rsp), %rdi movl $134217728, %edx # imm = 0x8000000 movq %r14, %rsi movl $1, %ecx callq hipMemcpy leaq 164(%rsp), %rdi movl $63, %esi xorl %edx, %edx callq hipDeviceGetAttribute movl $.Lstr, %edi callq puts@PLT movl $1, %r13d leaq 80(%rsp), %rbx .p2align 4, 0x90 .LBB1_3: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB1_4 Depth 2 # Child Loop BB1_7 Depth 3 movq %r13, 120(%rsp) # 8-byte Spill shlq $5, %r13 movq %r13, %rbp movabsq $4294967296, %rax # imm = 0x100000000 orq %rax, %rbp movl $1, %r12d .p2align 4, 0x90 .LBB1_4: # Parent Loop BB1_3 Depth=1 # => This Loop Header: Depth=2 # Child Loop BB1_7 Depth 3 movq %r12, %r15 shlq $5, %r15 movl $.L.str.2, %edi movl %r13d, %esi xorl %eax, %eax callq printf movl $.L.str.2, %edi movl %r15d, %esi xorl %eax, %eax callq printf movq 24(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movabsq $4294967296, %rax # imm = 0x100000000 orq %rax, %r15 movq %rbp, %rdi movl $1, %esi movq %r15, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_6 # %bb.5: # in Loop: Header=BB1_4 Depth=2 movq 32(%rsp), %rax movq 16(%rsp), %rcx movl $33554432, 44(%rsp) # imm = 0x2000000 movl $1073741824, 40(%rsp) # imm = 0x40000000 movq %rax, 152(%rsp) movq %rcx, 144(%rsp) leaq 44(%rsp), %rax movq %rax, 80(%rsp) leaq 40(%rsp), %rax movq %rax, 88(%rsp) leaq 152(%rsp), %rax movq %rax, 96(%rsp) leaq 144(%rsp), %rax movq %rax, 104(%rsp) leaq 64(%rsp), %rdi leaq 48(%rsp), %rsi leaq 136(%rsp), %rdx leaq 128(%rsp), %rcx callq __hipPopCallConfiguration movq 64(%rsp), %rsi movl 72(%rsp), %edx movq 48(%rsp), %rcx movl 56(%rsp), %r8d movl $_Z5saxpyifPfS_, %edi movq %rbx, %r9 pushq 128(%rsp) .cfi_adjust_cfa_offset 8 pushq 144(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_6: # in Loop: Header=BB1_4 Depth=2 movq 8(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movq 16(%rsp), %rsi movl $134217728, %edx # imm = 0x8000000 movq %r14, %rdi movl $2, %ecx callq hipMemcpy movq 8(%rsp), %rdi callq hipEventSynchronize movl $0, 80(%rsp) movq 24(%rsp), %rsi movq 8(%rsp), %rdx movq %rbx, %rdi callq hipEventElapsedTime xorps %xmm1, %xmm1 xorl %eax, %eax movaps %xmm1, %xmm0 movss .LCPI1_0(%rip), %xmm4 # xmm4 = mem[0],zero,zero,zero movaps .LCPI1_1(%rip), %xmm5 # xmm5 = [NaN,NaN,NaN,NaN] .p2align 4, 0x90 .LBB1_7: # Parent Loop BB1_3 Depth=1 # Parent Loop BB1_4 Depth=2 # => This Inner Loop Header: Depth=3 movss (%r14,%rax,4), %xmm2 # xmm2 = mem[0],zero,zero,zero addss %xmm4, %xmm2 andps %xmm5, %xmm2 cmpunordss %xmm0, %xmm0 movaps %xmm0, %xmm3 andps %xmm2, %xmm3 maxss %xmm1, %xmm2 andnps %xmm2, %xmm0 orps %xmm3, %xmm0 incq %rax movaps %xmm0, %xmm1 cmpq $33554432, %rax # imm = 0x2000000 jne .LBB1_7 # %bb.8: # in Loop: Header=BB1_4 Depth=2 cvtss2sd %xmm0, %xmm0 movl $.L.str.3, %edi movb $1, %al callq printf movss 80(%rsp), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str.3, %edi movb $1, %al callq printf movss 80(%rsp), %xmm0 # xmm0 = mem[0],zero,zero,zero xorps %xmm1, %xmm1 cvtss2sd %xmm0, %xmm1 movsd .LCPI1_2(%rip), %xmm0 # xmm0 = mem[0],zero divsd %xmm1, %xmm0 movsd .LCPI1_3(%rip), %xmm1 # xmm1 = mem[0],zero divsd %xmm1, %xmm0 movl $.L.str.4, %edi movb $1, %al callq printf leaq 64(%rsp), %rdi movl $60, %esi xorl %edx, %edx callq hipDeviceGetAttribute leaq 48(%rsp), %rdi movl $59, %esi xorl %edx, %edx callq hipDeviceGetAttribute xorps %xmm1, %xmm1 cvtsi2sdl 64(%rsp), %xmm1 xorps %xmm0, %xmm0 cvtsi2sdl 48(%rsp), %xmm0 addsd %xmm1, %xmm1 mulsd .LCPI1_4(%rip), %xmm0 mulsd %xmm1, %xmm0 divsd .LCPI1_3(%rip), %xmm0 movl $.L.str.4, %edi movb $1, %al callq printf movss .LCPI1_5(%rip), %xmm0 # xmm0 = mem[0],zero,zero,zero divss 80(%rsp), %xmm0 cvtss2sd %xmm0, %xmm0 divsd .LCPI1_3(%rip), %xmm0 movl $.L.str.5, %edi movb $1, %al callq printf incq %r12 cmpq $10, %r12 jne .LBB1_4 # %bb.9: # in Loop: Header=BB1_3 Depth=1 movq 120(%rsp), %r13 # 8-byte Reload incq %r13 cmpq $10, %r13 jne .LBB1_3 # %bb.10: movq 24(%rsp), %rdi callq hipEventDestroy movq 8(%rsp), %rdi callq hipEventDestroy movq 32(%rsp), %rdi callq hipFree movq 16(%rsp), %rdi callq hipFree movq 112(%rsp), %rdi # 8-byte Reload callq free movq %r14, %rdi callq free 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 $_Z5saxpyifPfS_, %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 _Z5saxpyifPfS_,@object # @_Z5saxpyifPfS_ .section .rodata,"a",@progbits .globl _Z5saxpyifPfS_ .p2align 3, 0x0 _Z5saxpyifPfS_: .quad _Z20__device_stub__saxpyifPfS_ .size _Z5saxpyifPfS_, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "%d\n" .size .L.str, 4 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "%d\t" .size .L.str.2, 4 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "%f\t" .size .L.str.3, 4 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "%f GB/s\t\t" .size .L.str.4, 10 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "%f GFLOP/s\t\n" .size .L.str.5, 13 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z5saxpyifPfS_" .size .L__unnamed_1, 15 .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 "Blocks\tThreads\tMax error\tElapsed time \tEff. Memory Throughput\tPeak Memory Throughput\tEff. Computational Throughput" .size .Lstr, 115 .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 _Z20__device_stub__saxpyifPfS_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z5saxpyifPfS_ .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 : _Z5saxpyifPfS_ .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 */ /* 0x001fca00078e0203 */ /*0040*/ ISETP.GE.AND P0, PT, R0, c[0x0][0x160], PT ; /* 0x0000580000007a0c */ /* 0x000fda0003f06270 */ /*0050*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0060*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fe40000000a00 */ /*0070*/ HFMA2.MMA R5, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff057435 */ /* 0x001fd400000001ff */ /*0080*/ IMAD.WIDE R2, R0, R5, c[0x0][0x168] ; /* 0x00005a0000027625 */ /* 0x000fc800078e0205 */ /*0090*/ IMAD.WIDE R4, R0, R5, c[0x0][0x170] ; /* 0x00005c0000047625 */ /* 0x000fe400078e0205 */ /*00a0*/ LDG.E R2, [R2.64] ; /* 0x0000000402027981 */ /* 0x000ea8000c1e1900 */ /*00b0*/ LDG.E R7, [R4.64] ; /* 0x0000000404077981 */ /* 0x000ea2000c1e1900 */ /*00c0*/ MOV R9, c[0x0][0x0] ; /* 0x0000000000097a02 */ /* 0x000fe20000000f00 */ /*00d0*/ WARPSYNC 0xffffffff ; /* 0xffffffff00007948 */ /* 0x000fe80003800000 */ /*00e0*/ IMAD R0, R9, c[0x0][0xc], R0 ; /* 0x0000030009007a24 */ /* 0x000fca00078e0200 */ /*00f0*/ ISETP.GE.AND P0, PT, R0, c[0x0][0x160], PT ; /* 0x0000580000007a0c */ /* 0x000fe20003f06270 */ /*0100*/ FFMA R7, R2, c[0x0][0x164], R7 ; /* 0x0000590002077a23 */ /* 0x004fca0000000007 */ /*0110*/ STG.E [R4.64], R7 ; /* 0x0000000704007986 */ /* 0x0001e8000c101904 */ /*0120*/ BAR.SYNC.DEFER_BLOCKING 0x0 ; /* 0x0000000000007b1d */ /* 0x000fec0000010000 */ /*0130*/ @!P0 BRA 0x70 ; /* 0xffffff3000008947 */ /* 0x000fea000383ffff */ /*0140*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0150*/ BRA 0x150; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*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 */ ..........
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z5saxpyifPfS_ .globl _Z5saxpyifPfS_ .p2align 8 .type _Z5saxpyifPfS_,@function _Z5saxpyifPfS_: s_clause 0x1 s_load_b32 s5, s[0:1], 0x24 s_load_b32 s4, s[0:1], 0x0 s_add_u32 s2, s0, 24 s_addc_u32 s3, s1, 0 s_waitcnt lgkmcnt(0) s_and_b32 s6, s5, 0xffff s_mov_b32 s5, exec_lo v_mad_u64_u32 v[1:2], null, s15, s6, v[0:1] s_delay_alu instid0(VALU_DEP_1) v_cmpx_gt_i32_e64 s4, v1 s_cbranch_execz .LBB0_3 s_load_b32 s7, s[2:3], 0x0 s_clause 0x1 s_load_b32 s5, s[0:1], 0x4 s_load_b128 s[0:3], s[0:1], 0x8 s_waitcnt lgkmcnt(0) s_mul_i32 s6, s7, s6 s_mov_b32 s7, 0 .p2align 6 .LBB0_2: v_ashrrev_i32_e32 v2, 31, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_lshlrev_b64 v[2:3], 2, v[1:2] v_add_co_u32 v4, vcc_lo, s0, v2 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v5, vcc_lo, s1, v3, vcc_lo v_add_co_u32 v2, vcc_lo, s2, v2 v_add_co_ci_u32_e32 v3, vcc_lo, s3, v3, vcc_lo global_load_b32 v0, v[4:5], off global_load_b32 v4, v[2:3], off s_waitcnt vmcnt(0) v_dual_fmac_f32 v4, s5, v0 :: v_dual_add_nc_u32 v1, s6, v1 s_delay_alu instid0(VALU_DEP_1) v_cmp_le_i32_e32 vcc_lo, s4, v1 global_store_b32 v[2:3], v4, off s_or_b32 s7, vcc_lo, s7 s_waitcnt_vscnt null, 0x0 s_barrier buffer_gl0_inv s_and_not1_b32 exec_lo, exec_lo, s7 s_cbranch_execnz .LBB0_2 .LBB0_3: s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z5saxpyifPfS_ .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 280 .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 6 .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 _Z5saxpyifPfS_, .Lfunc_end0-_Z5saxpyifPfS_ .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 - .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: hidden_block_count_x - .offset: 28 .size: 4 .value_kind: hidden_block_count_y - .offset: 32 .size: 4 .value_kind: hidden_block_count_z - .offset: 36 .size: 2 .value_kind: hidden_group_size_x - .offset: 38 .size: 2 .value_kind: hidden_group_size_y - .offset: 40 .size: 2 .value_kind: hidden_group_size_z - .offset: 42 .size: 2 .value_kind: hidden_remainder_x - .offset: 44 .size: 2 .value_kind: hidden_remainder_y - .offset: 46 .size: 2 .value_kind: hidden_remainder_z - .offset: 64 .size: 8 .value_kind: hidden_global_offset_x - .offset: 72 .size: 8 .value_kind: hidden_global_offset_y - .offset: 80 .size: 8 .value_kind: hidden_global_offset_z - .offset: 88 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 280 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z5saxpyifPfS_ .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z5saxpyifPfS_.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 6 .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_000e3121_00000000-6_SaxpyCudaBenchmarking.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 _Z28__device_stub__Z5saxpyifPfS_ifPfS_ .type _Z28__device_stub__Z5saxpyifPfS_ifPfS_, @function _Z28__device_stub__Z5saxpyifPfS_ifPfS_: .LFB2082: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movl %edi, 28(%rsp) movss %xmm0, 24(%rsp) movq %rsi, 16(%rsp) movq %rdx, 8(%rsp) movq %fs:40, %rax movq %rax, 136(%rsp) xorl %eax, %eax leaq 28(%rsp), %rax movq %rax, 96(%rsp) leaq 24(%rsp), %rax movq %rax, 104(%rsp) leaq 16(%rsp), %rax movq %rax, 112(%rsp) leaq 8(%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 .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 _Z5saxpyifPfS_(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L3 .L8: call __stack_chk_fail@PLT .cfi_endproc .LFE2082: .size _Z28__device_stub__Z5saxpyifPfS_ifPfS_, .-_Z28__device_stub__Z5saxpyifPfS_ifPfS_ .globl _Z5saxpyifPfS_ .type _Z5saxpyifPfS_, @function _Z5saxpyifPfS_: .LFB2083: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z28__device_stub__Z5saxpyifPfS_ifPfS_ addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2083: .size _Z5saxpyifPfS_, .-_Z5saxpyifPfS_ .section .rodata.str1.1,"aMS",@progbits,1 .LC1: .string "%d\n" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC4: .string "Blocks\tThreads\tMax error\tElapsed time \tEff. Memory Throughput\tPeak Memory Throughput\tEff. Computational Throughput\n" .section .rodata.str1.1 .LC5: .string "%d\t" .LC8: .string "%f\t" .LC11: .string "%f GB/s\t\t" .LC14: .string "%f GFLOP/s\t\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 $104, %rsp .cfi_def_cfa_offset 160 movq %fs:40, %rax movq %rax, 88(%rsp) xorl %eax, %eax movl $33554432, %edx leaq .LC1(%rip), %rsi movl $2, %edi call __printf_chk@PLT leaq 48(%rsp), %rdi call cudaEventCreate@PLT leaq 56(%rsp), %rdi call cudaEventCreate@PLT movl $134217728, %edi call malloc@PLT movq %rax, %r14 movl $134217728, %edi call malloc@PLT movq %rax, %r13 leaq 32(%rsp), %rdi movl $134217728, %esi call cudaMalloc@PLT leaq 40(%rsp), %rdi movl $134217728, %esi call cudaMalloc@PLT movl $0, %eax movss .LC2(%rip), %xmm1 movss .LC3(%rip), %xmm0 .L12: movss %xmm1, (%r14,%rax) movss %xmm0, 0(%r13,%rax) addq $4, %rax cmpq $134217728, %rax jne .L12 movl $1, %ecx movl $134217728, %edx movq %r14, %rsi movq 32(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movl $134217728, %edx movq %r13, %rsi movq 40(%rsp), %rdi call cudaMemcpy@PLT leaq 24(%rsp), %rdi movl $0, %edx movl $16, %esi call cudaDeviceGetAttribute@PLT leaq .LC4(%rip), %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $32, %r15d jmp .L13 .L14: movl $0, %esi movq 56(%rsp), %rdi call cudaEventRecord@PLT movl $2, %ecx movl $134217728, %edx movq 40(%rsp), %rsi movq %r13, %rdi call cudaMemcpy@PLT movq 56(%rsp), %rdi call cudaEventSynchronize@PLT movl $0x00000000, 28(%rsp) leaq 28(%rsp), %rdi movq 56(%rsp), %rdx movq 48(%rsp), %rsi call cudaEventElapsedTime@PLT movq %r13, %rbx leaq 134217728(%r13), %rbp pxor %xmm1, %xmm1 .L15: movss (%rbx), %xmm2 subss .LC6(%rip), %xmm2 andps .LC7(%rip), %xmm2 movaps %xmm2, %xmm0 call fmaxf@PLT movaps %xmm0, %xmm1 addq $4, %rbx cmpq %rbp, %rbx jne .L15 cvtss2sd %xmm0, %xmm0 leaq .LC8(%rip), %rbx movq %rbx, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT pxor %xmm0, %xmm0 cvtss2sd 28(%rsp), %xmm0 movq %rbx, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT pxor %xmm1, %xmm1 cvtss2sd 28(%rsp), %xmm1 movsd .LC9(%rip), %xmm0 divsd %xmm1, %xmm0 divsd .LC10(%rip), %xmm0 leaq .LC11(%rip), %rbx movq %rbx, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT leaq 64(%rsp), %rdi movl $0, %edx movl $36, %esi call cudaDeviceGetAttribute@PLT leaq 76(%rsp), %rdi movl $0, %edx movl $37, %esi call cudaDeviceGetAttribute@PLT pxor %xmm0, %xmm0 cvtsi2sdl 64(%rsp), %xmm0 addsd %xmm0, %xmm0 pxor %xmm1, %xmm1 cvtsi2sdl 76(%rsp), %xmm1 mulsd .LC12(%rip), %xmm1 mulsd %xmm1, %xmm0 divsd .LC10(%rip), %xmm0 movq %rbx, %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT movss .LC13(%rip), %xmm0 divss 28(%rsp), %xmm0 cvtss2sd %xmm0, %xmm0 divsd .LC10(%rip), %xmm0 leaq .LC14(%rip), %rsi movl $2, %edi movl $1, %eax call __printf_chk@PLT addl $32, %r12d cmpl $320, %r12d je .L23 .L16: movl 12(%rsp), %edx leaq .LC5(%rip), %rbx movq %rbx, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl %r12d, %edx movq %rbx, %rsi movl $2, %edi movl $0, %eax call __printf_chk@PLT movl $0, %esi movq 48(%rsp), %rdi call cudaEventRecord@PLT movl %r12d, 76(%rsp) movl $1, 80(%rsp) movl $1, 84(%rsp) movl %r15d, 64(%rsp) movl $1, 68(%rsp) movl $1, 72(%rsp) movl $0, %r9d movl $0, %r8d movq 76(%rsp), %rdx movl $1, %ecx movq 64(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax jne .L14 movq 40(%rsp), %rdx movq 32(%rsp), %rsi movss .LC3(%rip), %xmm0 movl $33554432, %edi call _Z28__device_stub__Z5saxpyifPfS_ifPfS_ jmp .L14 .L23: addl $32, %r15d cmpl $320, %r15d je .L17 .L13: movl %r15d, 12(%rsp) movl $32, %r12d jmp .L16 .L17: movq 48(%rsp), %rdi call cudaEventDestroy@PLT movq 56(%rsp), %rdi call cudaEventDestroy@PLT movq 32(%rsp), %rdi call cudaFree@PLT movq 40(%rsp), %rdi call cudaFree@PLT movq %r14, %rdi call free@PLT movq %r13, %rdi call free@PLT movq 88(%rsp), %rax subq %fs:40, %rax jne .L24 movl $0, %eax addq $104, %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 .L24: .cfi_restore_state call __stack_chk_fail@PLT .cfi_endproc .LFE2057: .size main, .-main .section .rodata.str1.1 .LC15: .string "_Z5saxpyifPfS_" .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 .LC15(%rip), %rdx movq %rdx, %rcx leaq _Z5saxpyifPfS_(%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.cst4,"aM",@progbits,4 .align 4 .LC2: .long 1065353216 .align 4 .LC3: .long 1073741824 .align 4 .LC6: .long 1082130432 .section .rodata.cst16,"aM",@progbits,16 .align 16 .LC7: .long 2147483647 .long 0 .long 0 .long 0 .section .rodata.cst8,"aM",@progbits,8 .align 8 .LC9: .long 0 .long 1102577664 .align 8 .LC10: .long 0 .long 1093567616 .align 8 .LC12: .long 0 .long 1069547520 .section .rodata.cst4 .align 4 .LC13: .long 1283457024 .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 "SaxpyCudaBenchmarking.hip" .globl _Z20__device_stub__saxpyifPfS_ # -- Begin function _Z20__device_stub__saxpyifPfS_ .p2align 4, 0x90 .type _Z20__device_stub__saxpyifPfS_,@function _Z20__device_stub__saxpyifPfS_: # @_Z20__device_stub__saxpyifPfS_ .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movl %edi, 12(%rsp) movss %xmm0, 8(%rsp) movq %rsi, 72(%rsp) movq %rdx, 64(%rsp) leaq 12(%rsp), %rax movq %rax, 80(%rsp) leaq 8(%rsp), %rax movq %rax, 88(%rsp) leaq 72(%rsp), %rax movq %rax, 96(%rsp) leaq 64(%rsp), %rax movq %rax, 104(%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 $_Z5saxpyifPfS_, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $136, %rsp .cfi_adjust_cfa_offset -136 retq .Lfunc_end0: .size _Z20__device_stub__saxpyifPfS_, .Lfunc_end0-_Z20__device_stub__saxpyifPfS_ .cfi_endproc # -- End function .section .rodata.cst4,"aM",@progbits,4 .p2align 2, 0x0 # -- Begin function main .LCPI1_0: .long 0xc0800000 # float -4 .LCPI1_5: .long 0x4c800000 # float 67108864 .section .rodata.cst16,"aM",@progbits,16 .p2align 4, 0x0 .LCPI1_1: .long 0x7fffffff # float NaN .long 0x7fffffff # float NaN .long 0x7fffffff # float NaN .long 0x7fffffff # float NaN .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 .LCPI1_2: .quad 0x41b8000000000000 # double 402653184 .LCPI1_3: .quad 0x412e848000000000 # double 1.0E+6 .LCPI1_4: .quad 0x3fc0000000000000 # double 0.125 .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 $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 xorl %ebx, %ebx movl $.L.str, %edi movl $33554432, %esi # imm = 0x2000000 xorl %eax, %eax callq printf leaq 24(%rsp), %rdi callq hipEventCreate leaq 8(%rsp), %rdi callq hipEventCreate movl $134217728, %edi # imm = 0x8000000 callq malloc movq %rax, %r15 movl $134217728, %edi # imm = 0x8000000 callq malloc movq %rax, %r14 leaq 32(%rsp), %rdi movl $134217728, %esi # imm = 0x8000000 callq hipMalloc leaq 16(%rsp), %rdi movl $134217728, %esi # imm = 0x8000000 callq hipMalloc .p2align 4, 0x90 .LBB1_1: # =>This Inner Loop Header: Depth=1 movl $1065353216, (%r15,%rbx,4) # imm = 0x3F800000 movl $1073741824, (%r14,%rbx,4) # imm = 0x40000000 incq %rbx cmpq $33554432, %rbx # imm = 0x2000000 jne .LBB1_1 # %bb.2: movq 32(%rsp), %rdi movl $134217728, %edx # imm = 0x8000000 movq %r15, 112(%rsp) # 8-byte Spill movq %r15, %rsi movl $1, %ecx callq hipMemcpy movq 16(%rsp), %rdi movl $134217728, %edx # imm = 0x8000000 movq %r14, %rsi movl $1, %ecx callq hipMemcpy leaq 164(%rsp), %rdi movl $63, %esi xorl %edx, %edx callq hipDeviceGetAttribute movl $.Lstr, %edi callq puts@PLT movl $1, %r13d leaq 80(%rsp), %rbx .p2align 4, 0x90 .LBB1_3: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB1_4 Depth 2 # Child Loop BB1_7 Depth 3 movq %r13, 120(%rsp) # 8-byte Spill shlq $5, %r13 movq %r13, %rbp movabsq $4294967296, %rax # imm = 0x100000000 orq %rax, %rbp movl $1, %r12d .p2align 4, 0x90 .LBB1_4: # Parent Loop BB1_3 Depth=1 # => This Loop Header: Depth=2 # Child Loop BB1_7 Depth 3 movq %r12, %r15 shlq $5, %r15 movl $.L.str.2, %edi movl %r13d, %esi xorl %eax, %eax callq printf movl $.L.str.2, %edi movl %r15d, %esi xorl %eax, %eax callq printf movq 24(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movabsq $4294967296, %rax # imm = 0x100000000 orq %rax, %r15 movq %rbp, %rdi movl $1, %esi movq %r15, %rdx movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB1_6 # %bb.5: # in Loop: Header=BB1_4 Depth=2 movq 32(%rsp), %rax movq 16(%rsp), %rcx movl $33554432, 44(%rsp) # imm = 0x2000000 movl $1073741824, 40(%rsp) # imm = 0x40000000 movq %rax, 152(%rsp) movq %rcx, 144(%rsp) leaq 44(%rsp), %rax movq %rax, 80(%rsp) leaq 40(%rsp), %rax movq %rax, 88(%rsp) leaq 152(%rsp), %rax movq %rax, 96(%rsp) leaq 144(%rsp), %rax movq %rax, 104(%rsp) leaq 64(%rsp), %rdi leaq 48(%rsp), %rsi leaq 136(%rsp), %rdx leaq 128(%rsp), %rcx callq __hipPopCallConfiguration movq 64(%rsp), %rsi movl 72(%rsp), %edx movq 48(%rsp), %rcx movl 56(%rsp), %r8d movl $_Z5saxpyifPfS_, %edi movq %rbx, %r9 pushq 128(%rsp) .cfi_adjust_cfa_offset 8 pushq 144(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB1_6: # in Loop: Header=BB1_4 Depth=2 movq 8(%rsp), %rdi xorl %esi, %esi callq hipEventRecord movq 16(%rsp), %rsi movl $134217728, %edx # imm = 0x8000000 movq %r14, %rdi movl $2, %ecx callq hipMemcpy movq 8(%rsp), %rdi callq hipEventSynchronize movl $0, 80(%rsp) movq 24(%rsp), %rsi movq 8(%rsp), %rdx movq %rbx, %rdi callq hipEventElapsedTime xorps %xmm1, %xmm1 xorl %eax, %eax movaps %xmm1, %xmm0 movss .LCPI1_0(%rip), %xmm4 # xmm4 = mem[0],zero,zero,zero movaps .LCPI1_1(%rip), %xmm5 # xmm5 = [NaN,NaN,NaN,NaN] .p2align 4, 0x90 .LBB1_7: # Parent Loop BB1_3 Depth=1 # Parent Loop BB1_4 Depth=2 # => This Inner Loop Header: Depth=3 movss (%r14,%rax,4), %xmm2 # xmm2 = mem[0],zero,zero,zero addss %xmm4, %xmm2 andps %xmm5, %xmm2 cmpunordss %xmm0, %xmm0 movaps %xmm0, %xmm3 andps %xmm2, %xmm3 maxss %xmm1, %xmm2 andnps %xmm2, %xmm0 orps %xmm3, %xmm0 incq %rax movaps %xmm0, %xmm1 cmpq $33554432, %rax # imm = 0x2000000 jne .LBB1_7 # %bb.8: # in Loop: Header=BB1_4 Depth=2 cvtss2sd %xmm0, %xmm0 movl $.L.str.3, %edi movb $1, %al callq printf movss 80(%rsp), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str.3, %edi movb $1, %al callq printf movss 80(%rsp), %xmm0 # xmm0 = mem[0],zero,zero,zero xorps %xmm1, %xmm1 cvtss2sd %xmm0, %xmm1 movsd .LCPI1_2(%rip), %xmm0 # xmm0 = mem[0],zero divsd %xmm1, %xmm0 movsd .LCPI1_3(%rip), %xmm1 # xmm1 = mem[0],zero divsd %xmm1, %xmm0 movl $.L.str.4, %edi movb $1, %al callq printf leaq 64(%rsp), %rdi movl $60, %esi xorl %edx, %edx callq hipDeviceGetAttribute leaq 48(%rsp), %rdi movl $59, %esi xorl %edx, %edx callq hipDeviceGetAttribute xorps %xmm1, %xmm1 cvtsi2sdl 64(%rsp), %xmm1 xorps %xmm0, %xmm0 cvtsi2sdl 48(%rsp), %xmm0 addsd %xmm1, %xmm1 mulsd .LCPI1_4(%rip), %xmm0 mulsd %xmm1, %xmm0 divsd .LCPI1_3(%rip), %xmm0 movl $.L.str.4, %edi movb $1, %al callq printf movss .LCPI1_5(%rip), %xmm0 # xmm0 = mem[0],zero,zero,zero divss 80(%rsp), %xmm0 cvtss2sd %xmm0, %xmm0 divsd .LCPI1_3(%rip), %xmm0 movl $.L.str.5, %edi movb $1, %al callq printf incq %r12 cmpq $10, %r12 jne .LBB1_4 # %bb.9: # in Loop: Header=BB1_3 Depth=1 movq 120(%rsp), %r13 # 8-byte Reload incq %r13 cmpq $10, %r13 jne .LBB1_3 # %bb.10: movq 24(%rsp), %rdi callq hipEventDestroy movq 8(%rsp), %rdi callq hipEventDestroy movq 32(%rsp), %rdi callq hipFree movq 16(%rsp), %rdi callq hipFree movq 112(%rsp), %rdi # 8-byte Reload callq free movq %r14, %rdi callq free 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 $_Z5saxpyifPfS_, %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 _Z5saxpyifPfS_,@object # @_Z5saxpyifPfS_ .section .rodata,"a",@progbits .globl _Z5saxpyifPfS_ .p2align 3, 0x0 _Z5saxpyifPfS_: .quad _Z20__device_stub__saxpyifPfS_ .size _Z5saxpyifPfS_, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "%d\n" .size .L.str, 4 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "%d\t" .size .L.str.2, 4 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "%f\t" .size .L.str.3, 4 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "%f GB/s\t\t" .size .L.str.4, 10 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "%f GFLOP/s\t\n" .size .L.str.5, 13 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z5saxpyifPfS_" .size .L__unnamed_1, 15 .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 "Blocks\tThreads\tMax error\tElapsed time \tEff. Memory Throughput\tPeak Memory Throughput\tEff. Computational Throughput" .size .Lstr, 115 .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 _Z20__device_stub__saxpyifPfS_ .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z5saxpyifPfS_ .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 colorToGray(unsigned char *input, unsigned char *output, int height, int width) { int col= blockDim.x * blockIdx.x + threadIdx.x; int row = blockDim.y * blockIdx.y + threadIdx.y; float scale[3] = {0.299, 0.587, 0.114}; if (row < height && col < width) { int pixelIndex = row * width + col; int rgbIndex = pixelIndex * 3; unsigned char r = input[rgbIndex]; // rgb rgb rgb rgb rgb unsigned char g = input[rgbIndex + 1]; unsigned char b = input[rgbIndex + 2]; output[pixelIndex] = r* scale[0] + g * scale[1] + b*scale[2]; } }
code for sm_80 Function : _Z11colorToGrayPhS_ii .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 R2, SR_CTAID.Y ; /* 0x0000000000027919 */ /* 0x000e680000002600 */ /*0040*/ S2R R5, SR_TID.Y ; /* 0x0000000000057919 */ /* 0x000e620000002200 */ /*0050*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x001fca00078e0203 */ /*0060*/ ISETP.GE.AND P0, PT, R0, c[0x0][0x174], PT ; /* 0x00005d0000007a0c */ /* 0x000fe20003f06270 */ /*0070*/ IMAD R3, R2, c[0x0][0x4], R5 ; /* 0x0000010002037a24 */ /* 0x002fca00078e0205 */ /*0080*/ ISETP.GE.OR P0, PT, R3, c[0x0][0x170], P0 ; /* 0x00005c0003007a0c */ /* 0x000fda0000706670 */ /*0090*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*00a0*/ IMAD R0, R3, c[0x0][0x174], R0 ; /* 0x00005d0003007a24 */ /* 0x000fe200078e0200 */ /*00b0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fc60000000a00 */ /*00c0*/ IMAD R3, R0, 0x3, RZ ; /* 0x0000000300037824 */ /* 0x000fca00078e02ff */ /*00d0*/ IADD3 R2, P0, R3, c[0x0][0x160], RZ ; /* 0x0000580003027a10 */ /* 0x000fc80007f1e0ff */ /*00e0*/ LEA.HI.X.SX32 R3, R3, c[0x0][0x164], 0x1, P0 ; /* 0x0000590003037a11 */ /* 0x000fca00000f0eff */ /*00f0*/ LDG.E.U8 R7, [R2.64+0x1] ; /* 0x0000010402077981 */ /* 0x000ea8000c1e1100 */ /*0100*/ LDG.E.U8 R6, [R2.64] ; /* 0x0000000402067981 */ /* 0x000ee8000c1e1100 */ /*0110*/ LDG.E.U8 R8, [R2.64+0x2] ; /* 0x0000020402087981 */ /* 0x000f22000c1e1100 */ /*0120*/ IADD3 R4, P0, R0, c[0x0][0x168], RZ ; /* 0x00005a0000047a10 */ /* 0x000fe20007f1e0ff */ /*0130*/ I2F.U16 R7, R7 ; /* 0x0000000700077306 */ /* 0x004e300000101000 */ /*0140*/ I2F.U16 R6, R6 ; /* 0x0000000600067306 */ /* 0x008e700000101000 */ /*0150*/ I2F.U16 R8, R8 ; /* 0x0000000800087306 */ /* 0x010ea20000101000 */ /*0160*/ FMUL R5, R7, 0.58700001239776611328 ; /* 0x3f1645a207057820 */ /* 0x001fc80000400000 */ /*0170*/ FFMA R5, R6, 0.29899999499320983887, R5 ; /* 0x3e99168706057823 */ /* 0x002fc80000000005 */ /*0180*/ FFMA R9, R8, 0.11400000005960464478, R5 ; /* 0x3de978d508097823 */ /* 0x004fe20000000005 */ /*0190*/ LEA.HI.X.SX32 R5, R0, c[0x0][0x16c], 0x1, P0 ; /* 0x00005b0000057a11 */ /* 0x000fca00000f0eff */ /*01a0*/ F2I.U32.TRUNC.NTZ R9, R9 ; /* 0x0000000900097305 */ /* 0x000e24000020f000 */ /*01b0*/ STG.E.U8 [R4.64], R9 ; /* 0x0000000904007986 */ /* 0x001fe2000c101104 */ /*01c0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*01d0*/ BRA 0x1d0; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*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 "includes.h" __global__ void colorToGray(unsigned char *input, unsigned char *output, int height, int width) { int col= blockDim.x * blockIdx.x + threadIdx.x; int row = blockDim.y * blockIdx.y + threadIdx.y; float scale[3] = {0.299, 0.587, 0.114}; if (row < height && col < width) { int pixelIndex = row * width + col; int rgbIndex = pixelIndex * 3; unsigned char r = input[rgbIndex]; // rgb rgb rgb rgb rgb unsigned char g = input[rgbIndex + 1]; unsigned char b = input[rgbIndex + 2]; output[pixelIndex] = r* scale[0] + g * scale[1] + b*scale[2]; } }
.file "tmpxft_00013430_00000000-6_colorToGray.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__Z11colorToGrayPhS_iiPhS_ii .type _Z35__device_stub__Z11colorToGrayPhS_iiPhS_ii, @function _Z35__device_stub__Z11colorToGrayPhS_iiPhS_ii: .LFB2051: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 24(%rsp) movq %rsi, 16(%rsp) movl %edx, 12(%rsp) movl %ecx, 8(%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 12(%rsp), %rax movq %rax, 112(%rsp) leaq 8(%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 .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 _Z11colorToGrayPhS_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 _Z35__device_stub__Z11colorToGrayPhS_iiPhS_ii, .-_Z35__device_stub__Z11colorToGrayPhS_iiPhS_ii .globl _Z11colorToGrayPhS_ii .type _Z11colorToGrayPhS_ii, @function _Z11colorToGrayPhS_ii: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z35__device_stub__Z11colorToGrayPhS_iiPhS_ii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z11colorToGrayPhS_ii, .-_Z11colorToGrayPhS_ii .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "_Z11colorToGrayPhS_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 _Z11colorToGrayPhS_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 colorToGray(unsigned char *input, unsigned char *output, int height, int width) { int col= blockDim.x * blockIdx.x + threadIdx.x; int row = blockDim.y * blockIdx.y + threadIdx.y; float scale[3] = {0.299, 0.587, 0.114}; if (row < height && col < width) { int pixelIndex = row * width + col; int rgbIndex = pixelIndex * 3; unsigned char r = input[rgbIndex]; // rgb rgb rgb rgb rgb unsigned char g = input[rgbIndex + 1]; unsigned char b = input[rgbIndex + 2]; output[pixelIndex] = r* scale[0] + g * scale[1] + b*scale[2]; } }
#include <hip/hip_runtime.h> #include "includes.h" __global__ void colorToGray(unsigned char *input, unsigned char *output, int height, int width) { int col= blockDim.x * blockIdx.x + threadIdx.x; int row = blockDim.y * blockIdx.y + threadIdx.y; float scale[3] = {0.299, 0.587, 0.114}; if (row < height && col < width) { int pixelIndex = row * width + col; int rgbIndex = pixelIndex * 3; unsigned char r = input[rgbIndex]; // rgb rgb rgb rgb rgb unsigned char g = input[rgbIndex + 1]; unsigned char b = input[rgbIndex + 2]; output[pixelIndex] = r* scale[0] + g * scale[1] + b*scale[2]; } }
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 colorToGray(unsigned char *input, unsigned char *output, int height, int width) { int col= blockDim.x * blockIdx.x + threadIdx.x; int row = blockDim.y * blockIdx.y + threadIdx.y; float scale[3] = {0.299, 0.587, 0.114}; if (row < height && col < width) { int pixelIndex = row * width + col; int rgbIndex = pixelIndex * 3; unsigned char r = input[rgbIndex]; // rgb rgb rgb rgb rgb unsigned char g = input[rgbIndex + 1]; unsigned char b = input[rgbIndex + 2]; output[pixelIndex] = r* scale[0] + g * scale[1] + b*scale[2]; } }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z11colorToGrayPhS_ii .globl _Z11colorToGrayPhS_ii .p2align 8 .type _Z11colorToGrayPhS_ii,@function _Z11colorToGrayPhS_ii: s_clause 0x1 s_load_b32 s4, s[0:1], 0x24 s_load_b64 s[2:3], s[0:1], 0x10 v_bfe_u32 v2, v0, 10, 10 v_and_b32_e32 v3, 0x3ff, v0 s_waitcnt lgkmcnt(0) s_and_b32 s5, s4, 0xffff s_lshr_b32 s4, s4, 16 s_delay_alu instid0(VALU_DEP_1) | instid1(SALU_CYCLE_1) v_mad_u64_u32 v[0:1], null, s15, s4, v[2:3] v_mad_u64_u32 v[1:2], null, s14, s5, v[3:4] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_cmp_gt_i32_e32 vcc_lo, s2, 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, s2, vcc_lo s_and_saveexec_b32 s4, s2 s_cbranch_execz .LBB0_2 s_load_b128 s[4:7], s[0:1], 0x0 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_mad_u64_u32 v[2:3], null, v0, s3, v[1:2] v_lshl_add_u32 v0, v2, 1, v2 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_2) v_ashrrev_i32_e32 v1, 31, v0 s_waitcnt lgkmcnt(0) v_add_co_u32 v0, vcc_lo, s4, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s5, v1, vcc_lo s_clause 0x2 global_load_u8 v3, v[0:1], off offset:1 global_load_u8 v4, v[0:1], off global_load_u8 v0, v[0:1], off offset:2 s_waitcnt vmcnt(2) v_cvt_f32_ubyte0_e32 v1, v3 s_waitcnt vmcnt(1) v_cvt_f32_ubyte0_e32 v3, v4 s_waitcnt vmcnt(0) v_cvt_f32_ubyte0_e32 v0, v0 v_mul_f32_e32 v1, 0x3f1645a2, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmamk_f32 v1, v3, 0x3e991687, v1 v_fmamk_f32 v0, v0, 0x3de978d5, v1 v_ashrrev_i32_e32 v1, 31, v2 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_3) v_cvt_i32_f32_e32 v3, v0 v_add_co_u32 v0, vcc_lo, s6, v2 v_add_co_ci_u32_e32 v1, vcc_lo, s7, v1, vcc_lo global_store_b8 v[0:1], v3, off .LBB0_2: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z11colorToGrayPhS_ii .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 280 .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 5 .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 _Z11colorToGrayPhS_ii, .Lfunc_end0-_Z11colorToGrayPhS_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 - .offset: 16 .size: 4 .value_kind: by_value - .offset: 20 .size: 4 .value_kind: by_value - .offset: 24 .size: 4 .value_kind: hidden_block_count_x - .offset: 28 .size: 4 .value_kind: hidden_block_count_y - .offset: 32 .size: 4 .value_kind: hidden_block_count_z - .offset: 36 .size: 2 .value_kind: hidden_group_size_x - .offset: 38 .size: 2 .value_kind: hidden_group_size_y - .offset: 40 .size: 2 .value_kind: hidden_group_size_z - .offset: 42 .size: 2 .value_kind: hidden_remainder_x - .offset: 44 .size: 2 .value_kind: hidden_remainder_y - .offset: 46 .size: 2 .value_kind: hidden_remainder_z - .offset: 64 .size: 8 .value_kind: hidden_global_offset_x - .offset: 72 .size: 8 .value_kind: hidden_global_offset_y - .offset: 80 .size: 8 .value_kind: hidden_global_offset_z - .offset: 88 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 280 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z11colorToGrayPhS_ii .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z11colorToGrayPhS_ii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 5 .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 colorToGray(unsigned char *input, unsigned char *output, int height, int width) { int col= blockDim.x * blockIdx.x + threadIdx.x; int row = blockDim.y * blockIdx.y + threadIdx.y; float scale[3] = {0.299, 0.587, 0.114}; if (row < height && col < width) { int pixelIndex = row * width + col; int rgbIndex = pixelIndex * 3; unsigned char r = input[rgbIndex]; // rgb rgb rgb rgb rgb unsigned char g = input[rgbIndex + 1]; unsigned char b = input[rgbIndex + 2]; output[pixelIndex] = r* scale[0] + g * scale[1] + b*scale[2]; } }
.text .file "colorToGray.hip" .globl _Z26__device_stub__colorToGrayPhS_ii # -- Begin function _Z26__device_stub__colorToGrayPhS_ii .p2align 4, 0x90 .type _Z26__device_stub__colorToGrayPhS_ii,@function _Z26__device_stub__colorToGrayPhS_ii: # @_Z26__device_stub__colorToGrayPhS_ii .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movl %edx, 12(%rsp) movl %ecx, 8(%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 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 $_Z11colorToGrayPhS_ii, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $136, %rsp .cfi_adjust_cfa_offset -136 retq .Lfunc_end0: .size _Z26__device_stub__colorToGrayPhS_ii, .Lfunc_end0-_Z26__device_stub__colorToGrayPhS_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 $_Z11colorToGrayPhS_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 _Z11colorToGrayPhS_ii,@object # @_Z11colorToGrayPhS_ii .section .rodata,"a",@progbits .globl _Z11colorToGrayPhS_ii .p2align 3, 0x0 _Z11colorToGrayPhS_ii: .quad _Z26__device_stub__colorToGrayPhS_ii .size _Z11colorToGrayPhS_ii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z11colorToGrayPhS_ii" .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 _Z26__device_stub__colorToGrayPhS_ii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z11colorToGrayPhS_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 : _Z11colorToGrayPhS_ii .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 R2, SR_CTAID.Y ; /* 0x0000000000027919 */ /* 0x000e680000002600 */ /*0040*/ S2R R5, SR_TID.Y ; /* 0x0000000000057919 */ /* 0x000e620000002200 */ /*0050*/ IMAD R0, R0, c[0x0][0x0], R3 ; /* 0x0000000000007a24 */ /* 0x001fca00078e0203 */ /*0060*/ ISETP.GE.AND P0, PT, R0, c[0x0][0x174], PT ; /* 0x00005d0000007a0c */ /* 0x000fe20003f06270 */ /*0070*/ IMAD R3, R2, c[0x0][0x4], R5 ; /* 0x0000010002037a24 */ /* 0x002fca00078e0205 */ /*0080*/ ISETP.GE.OR P0, PT, R3, c[0x0][0x170], P0 ; /* 0x00005c0003007a0c */ /* 0x000fda0000706670 */ /*0090*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*00a0*/ IMAD R0, R3, c[0x0][0x174], R0 ; /* 0x00005d0003007a24 */ /* 0x000fe200078e0200 */ /*00b0*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fc60000000a00 */ /*00c0*/ IMAD R3, R0, 0x3, RZ ; /* 0x0000000300037824 */ /* 0x000fca00078e02ff */ /*00d0*/ IADD3 R2, P0, R3, c[0x0][0x160], RZ ; /* 0x0000580003027a10 */ /* 0x000fc80007f1e0ff */ /*00e0*/ LEA.HI.X.SX32 R3, R3, c[0x0][0x164], 0x1, P0 ; /* 0x0000590003037a11 */ /* 0x000fca00000f0eff */ /*00f0*/ LDG.E.U8 R7, [R2.64+0x1] ; /* 0x0000010402077981 */ /* 0x000ea8000c1e1100 */ /*0100*/ LDG.E.U8 R6, [R2.64] ; /* 0x0000000402067981 */ /* 0x000ee8000c1e1100 */ /*0110*/ LDG.E.U8 R8, [R2.64+0x2] ; /* 0x0000020402087981 */ /* 0x000f22000c1e1100 */ /*0120*/ IADD3 R4, P0, R0, c[0x0][0x168], RZ ; /* 0x00005a0000047a10 */ /* 0x000fe20007f1e0ff */ /*0130*/ I2F.U16 R7, R7 ; /* 0x0000000700077306 */ /* 0x004e300000101000 */ /*0140*/ I2F.U16 R6, R6 ; /* 0x0000000600067306 */ /* 0x008e700000101000 */ /*0150*/ I2F.U16 R8, R8 ; /* 0x0000000800087306 */ /* 0x010ea20000101000 */ /*0160*/ FMUL R5, R7, 0.58700001239776611328 ; /* 0x3f1645a207057820 */ /* 0x001fc80000400000 */ /*0170*/ FFMA R5, R6, 0.29899999499320983887, R5 ; /* 0x3e99168706057823 */ /* 0x002fc80000000005 */ /*0180*/ FFMA R9, R8, 0.11400000005960464478, R5 ; /* 0x3de978d508097823 */ /* 0x004fe20000000005 */ /*0190*/ LEA.HI.X.SX32 R5, R0, c[0x0][0x16c], 0x1, P0 ; /* 0x00005b0000057a11 */ /* 0x000fca00000f0eff */ /*01a0*/ F2I.U32.TRUNC.NTZ R9, R9 ; /* 0x0000000900097305 */ /* 0x000e24000020f000 */ /*01b0*/ STG.E.U8 [R4.64], R9 ; /* 0x0000000904007986 */ /* 0x001fe2000c101104 */ /*01c0*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*01d0*/ BRA 0x1d0; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*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 _Z11colorToGrayPhS_ii .globl _Z11colorToGrayPhS_ii .p2align 8 .type _Z11colorToGrayPhS_ii,@function _Z11colorToGrayPhS_ii: s_clause 0x1 s_load_b32 s4, s[0:1], 0x24 s_load_b64 s[2:3], s[0:1], 0x10 v_bfe_u32 v2, v0, 10, 10 v_and_b32_e32 v3, 0x3ff, v0 s_waitcnt lgkmcnt(0) s_and_b32 s5, s4, 0xffff s_lshr_b32 s4, s4, 16 s_delay_alu instid0(VALU_DEP_1) | instid1(SALU_CYCLE_1) v_mad_u64_u32 v[0:1], null, s15, s4, v[2:3] v_mad_u64_u32 v[1:2], null, s14, s5, v[3:4] s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2) v_cmp_gt_i32_e32 vcc_lo, s2, 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, s2, vcc_lo s_and_saveexec_b32 s4, s2 s_cbranch_execz .LBB0_2 s_load_b128 s[4:7], s[0:1], 0x0 s_delay_alu instid0(VALU_DEP_3) | instskip(NEXT) | instid1(VALU_DEP_1) v_mad_u64_u32 v[2:3], null, v0, s3, v[1:2] v_lshl_add_u32 v0, v2, 1, v2 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_2) | instid1(VALU_DEP_2) v_ashrrev_i32_e32 v1, 31, v0 s_waitcnt lgkmcnt(0) v_add_co_u32 v0, vcc_lo, s4, v0 v_add_co_ci_u32_e32 v1, vcc_lo, s5, v1, vcc_lo s_clause 0x2 global_load_u8 v3, v[0:1], off offset:1 global_load_u8 v4, v[0:1], off global_load_u8 v0, v[0:1], off offset:2 s_waitcnt vmcnt(2) v_cvt_f32_ubyte0_e32 v1, v3 s_waitcnt vmcnt(1) v_cvt_f32_ubyte0_e32 v3, v4 s_waitcnt vmcnt(0) v_cvt_f32_ubyte0_e32 v0, v0 v_mul_f32_e32 v1, 0x3f1645a2, v1 s_delay_alu instid0(VALU_DEP_1) | instskip(NEXT) | instid1(VALU_DEP_1) v_fmamk_f32 v1, v3, 0x3e991687, v1 v_fmamk_f32 v0, v0, 0x3de978d5, v1 v_ashrrev_i32_e32 v1, 31, v2 s_delay_alu instid0(VALU_DEP_2) | instskip(SKIP_1) | instid1(VALU_DEP_3) v_cvt_i32_f32_e32 v3, v0 v_add_co_u32 v0, vcc_lo, s6, v2 v_add_co_ci_u32_e32 v1, vcc_lo, s7, v1, vcc_lo global_store_b8 v[0:1], v3, off .LBB0_2: s_nop 0 s_sendmsg sendmsg(MSG_DEALLOC_VGPRS) s_endpgm .section .rodata,"a",@progbits .p2align 6, 0x0 .amdhsa_kernel _Z11colorToGrayPhS_ii .amdhsa_group_segment_fixed_size 0 .amdhsa_private_segment_fixed_size 0 .amdhsa_kernarg_size 280 .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 5 .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 _Z11colorToGrayPhS_ii, .Lfunc_end0-_Z11colorToGrayPhS_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 - .offset: 16 .size: 4 .value_kind: by_value - .offset: 20 .size: 4 .value_kind: by_value - .offset: 24 .size: 4 .value_kind: hidden_block_count_x - .offset: 28 .size: 4 .value_kind: hidden_block_count_y - .offset: 32 .size: 4 .value_kind: hidden_block_count_z - .offset: 36 .size: 2 .value_kind: hidden_group_size_x - .offset: 38 .size: 2 .value_kind: hidden_group_size_y - .offset: 40 .size: 2 .value_kind: hidden_group_size_z - .offset: 42 .size: 2 .value_kind: hidden_remainder_x - .offset: 44 .size: 2 .value_kind: hidden_remainder_y - .offset: 46 .size: 2 .value_kind: hidden_remainder_z - .offset: 64 .size: 8 .value_kind: hidden_global_offset_x - .offset: 72 .size: 8 .value_kind: hidden_global_offset_y - .offset: 80 .size: 8 .value_kind: hidden_global_offset_z - .offset: 88 .size: 2 .value_kind: hidden_grid_dims .group_segment_fixed_size: 0 .kernarg_segment_align: 8 .kernarg_segment_size: 280 .language: OpenCL C .language_version: - 2 - 0 .max_flat_workgroup_size: 1024 .name: _Z11colorToGrayPhS_ii .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z11colorToGrayPhS_ii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 5 .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_00013430_00000000-6_colorToGray.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__Z11colorToGrayPhS_iiPhS_ii .type _Z35__device_stub__Z11colorToGrayPhS_iiPhS_ii, @function _Z35__device_stub__Z11colorToGrayPhS_iiPhS_ii: .LFB2051: .cfi_startproc endbr64 subq $152, %rsp .cfi_def_cfa_offset 160 movq %rdi, 24(%rsp) movq %rsi, 16(%rsp) movl %edx, 12(%rsp) movl %ecx, 8(%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 12(%rsp), %rax movq %rax, 112(%rsp) leaq 8(%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 .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 _Z11colorToGrayPhS_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 _Z35__device_stub__Z11colorToGrayPhS_iiPhS_ii, .-_Z35__device_stub__Z11colorToGrayPhS_iiPhS_ii .globl _Z11colorToGrayPhS_ii .type _Z11colorToGrayPhS_ii, @function _Z11colorToGrayPhS_ii: .LFB2052: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z35__device_stub__Z11colorToGrayPhS_iiPhS_ii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2052: .size _Z11colorToGrayPhS_ii, .-_Z11colorToGrayPhS_ii .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "_Z11colorToGrayPhS_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 _Z11colorToGrayPhS_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 "colorToGray.hip" .globl _Z26__device_stub__colorToGrayPhS_ii # -- Begin function _Z26__device_stub__colorToGrayPhS_ii .p2align 4, 0x90 .type _Z26__device_stub__colorToGrayPhS_ii,@function _Z26__device_stub__colorToGrayPhS_ii: # @_Z26__device_stub__colorToGrayPhS_ii .cfi_startproc # %bb.0: subq $120, %rsp .cfi_def_cfa_offset 128 movq %rdi, 72(%rsp) movq %rsi, 64(%rsp) movl %edx, 12(%rsp) movl %ecx, 8(%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 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 $_Z11colorToGrayPhS_ii, %edi pushq 16(%rsp) .cfi_adjust_cfa_offset 8 pushq 32(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $136, %rsp .cfi_adjust_cfa_offset -136 retq .Lfunc_end0: .size _Z26__device_stub__colorToGrayPhS_ii, .Lfunc_end0-_Z26__device_stub__colorToGrayPhS_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 $_Z11colorToGrayPhS_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 _Z11colorToGrayPhS_ii,@object # @_Z11colorToGrayPhS_ii .section .rodata,"a",@progbits .globl _Z11colorToGrayPhS_ii .p2align 3, 0x0 _Z11colorToGrayPhS_ii: .quad _Z26__device_stub__colorToGrayPhS_ii .size _Z11colorToGrayPhS_ii, 8 .type .L__unnamed_1,@object # @0 .section .rodata.str1.1,"aMS",@progbits,1 .L__unnamed_1: .asciz "_Z11colorToGrayPhS_ii" .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 _Z26__device_stub__colorToGrayPhS_ii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z11colorToGrayPhS_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 device assembly.
#include <stdio.h> #include <cuda.h> #include <cuda_runtime.h> #define LOG_INPUT if(0) #define LOG_OUTPUT if(1) #define LOG if(0) __global__ void hadamard(float *A, float *B, float *C, int M, int N) { // Complete the kernel code snippet int i = threadIdx.x + blockDim.x * blockIdx.x; if(i < M*N) C[i] = A[i]*B[i]; } /** * Host main routine */ void print_matrix(float *A,int m,int n) { for(int i =0;i<m;i++) { for(int j=0;j<n;j++) printf("%.2f ",A[i*n+j]); printf("\n"); } } int main(void) { // Error code to check return values for CUDA calls cudaError_t err = cudaSuccess; // Print the vector length to be used, and compute its size int t; //number of test cases scanf("%d",&t); while(t--) { int m,n; scanf("%d %d",&m,&n); size_t size = m*n * sizeof(float); LOG printf("[Hadamard product of two matrices ]\n"); // Allocate the host input vector A float *h_A = (float*)malloc(size); // Allocate the host input vector B float *h_B = (float*)malloc(size); // Allocate the host output vector C float *h_C = (float*)malloc(size); // Verify that allocations succeeded if (h_A == NULL || h_B == NULL || h_C == NULL) { fprintf(stderr, "Failed to allocate host vectors!\n"); exit(EXIT_FAILURE); } // Initialize the host input vectors for (int i = 0; i < n*m; ++i) { scanf("%f",&h_A[i]); scanf("%f",&h_B[i]); } float *d_A = NULL, *d_B = NULL, *d_C = NULL; // Allocate the device input vector A cudaMalloc((void**)&d_A, size); // Allocate the device input vector B cudaMalloc((void**)&d_B, size); // Allocate the device output vector C cudaMalloc((void**)&d_C, size); // Copy the host input vectors A and B in host memory to the device input vectors in // device memory cudaMemcpy(d_A, h_A, size, cudaMemcpyHostToDevice); cudaMemcpy(d_B, h_B, size, cudaMemcpyHostToDevice); cudaMemcpy(d_C, h_C, size, cudaMemcpyHostToDevice); // initialize blocksPerGrid and threads Per Block int threadsPerBlock = 256; int blocksPerGrid = ((m*n)+threadsPerBlock-1)/threadsPerBlock; hadamard<<<blocksPerGrid, threadsPerBlock>>>(d_A, d_B, d_C, m, n); err = cudaGetLastError(); if (err != cudaSuccess) { fprintf(stderr, "Failed to launch vectorAdd kernel (error code %s)!\n", cudaGetErrorString(err)); exit(EXIT_FAILURE); } // Copy the device result vector in device memory to the host result vector // in host memory. cudaMemcpy(h_C, d_C, size, cudaMemcpyDeviceToHost); // Verify that the result vector is correct for (int i = 0; i < n*m; ++i) { if (fabs(h_A[i] * h_B[i] - h_C[i]) > 1e-5) { fprintf(stderr, "Result verification failed at element %d!\n", i); exit(EXIT_FAILURE); } } LOG printf("Test PASSED\n"); // Free device global memory // Free host memory err = cudaDeviceReset(); if (err != cudaSuccess) { fprintf(stderr, "Failed to deinitialize the device! error=%s\n", cudaGetErrorString(err)); exit(EXIT_FAILURE); } print_matrix(h_C,m,n); LOG printf("Done\n"); } return 0; }
code for sm_80 Function : _Z8hadamardPfS_S_ii .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_TID.X ; /* 0x0000000000067919 */ /* 0x000e220000002100 */ /*0020*/ ULDC.64 UR4, c[0x0][0x178] ; /* 0x00005e0000047ab9 */ /* 0x000fe40000000a00 */ /*0030*/ UIMAD UR4, UR5, UR4, URZ ; /* 0x00000004050472a4 */ /* 0x000fe2000f8e023f */ /*0040*/ S2R R3, SR_CTAID.X ; /* 0x0000000000037919 */ /* 0x000e240000002500 */ /*0050*/ IMAD R6, R3, c[0x0][0x0], R6 ; /* 0x0000000003067a24 */ /* 0x001fca00078e0206 */ /*0060*/ ISETP.GE.AND P0, PT, R6, UR4, PT ; /* 0x0000000406007c0c */ /* 0x000fda000bf06270 */ /*0070*/ @P0 EXIT ; /* 0x000000000000094d */ /* 0x000fea0003800000 */ /*0080*/ HFMA2.MMA R7, -RZ, RZ, 0, 2.384185791015625e-07 ; /* 0x00000004ff077435 */ /* 0x000fe200000001ff */ /*0090*/ ULDC.64 UR4, c[0x0][0x118] ; /* 0x0000460000047ab9 */ /* 0x000fd20000000a00 */ /*00a0*/ IMAD.WIDE R4, R6, R7, c[0x0][0x168] ; /* 0x00005a0006047625 */ /* 0x000fc800078e0207 */ /*00b0*/ IMAD.WIDE R2, R6.reuse, R7.reuse, c[0x0][0x160] ; /* 0x0000580006027625 */ /* 0x0c0fe400078e0207 */ /*00c0*/ LDG.E R4, [R4.64] ; /* 0x0000000404047981 */ /* 0x000ea8000c1e1900 */ /*00d0*/ LDG.E R3, [R2.64] ; /* 0x0000000402037981 */ /* 0x000ea2000c1e1900 */ /*00e0*/ IMAD.WIDE R6, R6, R7, c[0x0][0x170] ; /* 0x00005c0006067625 */ /* 0x000fc800078e0207 */ /*00f0*/ FMUL R9, R4, R3 ; /* 0x0000000304097220 */ /* 0x004fca0000400000 */ /*0100*/ STG.E [R6.64], R9 ; /* 0x0000000906007986 */ /* 0x000fe2000c101904 */ /*0110*/ EXIT ; /* 0x000000000000794d */ /* 0x000fea0003800000 */ /*0120*/ BRA 0x120; /* 0xfffffff000007947 */ /* 0x000fc0000383ffff */ /*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 <stdio.h> #include <cuda.h> #include <cuda_runtime.h> #define LOG_INPUT if(0) #define LOG_OUTPUT if(1) #define LOG if(0) __global__ void hadamard(float *A, float *B, float *C, int M, int N) { // Complete the kernel code snippet int i = threadIdx.x + blockDim.x * blockIdx.x; if(i < M*N) C[i] = A[i]*B[i]; } /** * Host main routine */ void print_matrix(float *A,int m,int n) { for(int i =0;i<m;i++) { for(int j=0;j<n;j++) printf("%.2f ",A[i*n+j]); printf("\n"); } } int main(void) { // Error code to check return values for CUDA calls cudaError_t err = cudaSuccess; // Print the vector length to be used, and compute its size int t; //number of test cases scanf("%d",&t); while(t--) { int m,n; scanf("%d %d",&m,&n); size_t size = m*n * sizeof(float); LOG printf("[Hadamard product of two matrices ]\n"); // Allocate the host input vector A float *h_A = (float*)malloc(size); // Allocate the host input vector B float *h_B = (float*)malloc(size); // Allocate the host output vector C float *h_C = (float*)malloc(size); // Verify that allocations succeeded if (h_A == NULL || h_B == NULL || h_C == NULL) { fprintf(stderr, "Failed to allocate host vectors!\n"); exit(EXIT_FAILURE); } // Initialize the host input vectors for (int i = 0; i < n*m; ++i) { scanf("%f",&h_A[i]); scanf("%f",&h_B[i]); } float *d_A = NULL, *d_B = NULL, *d_C = NULL; // Allocate the device input vector A cudaMalloc((void**)&d_A, size); // Allocate the device input vector B cudaMalloc((void**)&d_B, size); // Allocate the device output vector C cudaMalloc((void**)&d_C, size); // Copy the host input vectors A and B in host memory to the device input vectors in // device memory cudaMemcpy(d_A, h_A, size, cudaMemcpyHostToDevice); cudaMemcpy(d_B, h_B, size, cudaMemcpyHostToDevice); cudaMemcpy(d_C, h_C, size, cudaMemcpyHostToDevice); // initialize blocksPerGrid and threads Per Block int threadsPerBlock = 256; int blocksPerGrid = ((m*n)+threadsPerBlock-1)/threadsPerBlock; hadamard<<<blocksPerGrid, threadsPerBlock>>>(d_A, d_B, d_C, m, n); err = cudaGetLastError(); if (err != cudaSuccess) { fprintf(stderr, "Failed to launch vectorAdd kernel (error code %s)!\n", cudaGetErrorString(err)); exit(EXIT_FAILURE); } // Copy the device result vector in device memory to the host result vector // in host memory. cudaMemcpy(h_C, d_C, size, cudaMemcpyDeviceToHost); // Verify that the result vector is correct for (int i = 0; i < n*m; ++i) { if (fabs(h_A[i] * h_B[i] - h_C[i]) > 1e-5) { fprintf(stderr, "Result verification failed at element %d!\n", i); exit(EXIT_FAILURE); } } LOG printf("Test PASSED\n"); // Free device global memory // Free host memory err = cudaDeviceReset(); if (err != cudaSuccess) { fprintf(stderr, "Failed to deinitialize the device! error=%s\n", cudaGetErrorString(err)); exit(EXIT_FAILURE); } print_matrix(h_C,m,n); LOG printf("Done\n"); } return 0; }
.file "tmpxft_001525a2_00000000-6_hadamard_product.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 .section .rodata.str1.1,"aMS",@progbits,1 .LC0: .string "%.2f " .LC1: .string "\n" .text .globl _Z12print_matrixPfii .type _Z12print_matrixPfii, @function _Z12print_matrixPfii: .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 $40, %rsp .cfi_def_cfa_offset 96 movq %rdi, 16(%rsp) movl %esi, 12(%rsp) testl %esi, %esi jle .L3 movl %edx, %r15d movl $0, %r14d movl $0, %r13d movslq %edx, %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 addl $1, %r13d addl %r15d, %r14d cmpl %r13d, 12(%rsp) je .L3 .L5: testl %r15d, %r15d jg .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 .LFE2057: .size _Z12print_matrixPfii, .-_Z12print_matrixPfii .globl _Z33__device_stub__Z8hadamardPfS_S_iiPfS_S_ii .type _Z33__device_stub__Z8hadamardPfS_S_iiPfS_S_ii, @function _Z33__device_stub__Z8hadamardPfS_S_iiPfS_S_ii: .LFB2083: .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 .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 _Z8hadamardPfS_S_ii(%rip), %rdi call cudaLaunchKernel@PLT addq $16, %rsp .cfi_def_cfa_offset 160 jmp .L11 .L16: call __stack_chk_fail@PLT .cfi_endproc .LFE2083: .size _Z33__device_stub__Z8hadamardPfS_S_iiPfS_S_ii, .-_Z33__device_stub__Z8hadamardPfS_S_iiPfS_S_ii .globl _Z8hadamardPfS_S_ii .type _Z8hadamardPfS_S_ii, @function _Z8hadamardPfS_S_ii: .LFB2084: .cfi_startproc endbr64 subq $8, %rsp .cfi_def_cfa_offset 16 call _Z33__device_stub__Z8hadamardPfS_S_iiPfS_S_ii addq $8, %rsp .cfi_def_cfa_offset 8 ret .cfi_endproc .LFE2084: .size _Z8hadamardPfS_S_ii, .-_Z8hadamardPfS_S_ii .section .rodata.str1.1 .LC2: .string "%d" .LC3: .string "%d %d" .section .rodata.str1.8,"aMS",@progbits,1 .align 8 .LC4: .string "Failed to allocate host vectors!\n" .section .rodata.str1.1 .LC5: .string "%f" .section .rodata.str1.8 .align 8 .LC6: .string "Failed to launch vectorAdd kernel (error code %s)!\n" .align 8 .LC9: .string "Result verification failed at element %d!\n" .align 8 .LC10: .string "Failed to deinitialize the device! error=%s\n" .text .globl main .type main, @function main: .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 $104, %rsp .cfi_def_cfa_offset 160 movq %fs:40, %rax movq %rax, 88(%rsp) xorl %eax, %eax leaq 28(%rsp), %rsi leaq .LC2(%rip), %rdi call __isoc23_scanf@PLT movl 28(%rsp), %eax leal -1(%rax), %edx movl %edx, 28(%rsp) testl %eax, %eax jne .L32 .L20: movq 88(%rsp), %rax subq %fs:40, %rax jne .L41 movl $0, %eax addq $104, %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 .L21: .cfi_restore_state leaq .LC4(%rip), %rdx movl $2, %esi movq stderr(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movl $1, %edi call exit@PLT .L25: call cudaGetLastError@PLT testl %eax, %eax jne .L42 movl $2, %ecx movq (%rsp), %rdx movq 56(%rsp), %rsi movq %r13, %rdi call cudaMemcpy@PLT movl 36(%rsp), %edx imull 32(%rsp), %edx testl %edx, %edx jle .L27 movslq %edx, %rdx movl $0, %eax .L30: movss (%r14,%rax,4), %xmm0 mulss (%r15,%rax,4), %xmm0 subss 0(%r13,%rax,4), %xmm0 andps .LC7(%rip), %xmm0 cvtss2sd %xmm0, %xmm0 comisd .LC8(%rip), %xmm0 ja .L43 addq $1, %rax cmpq %rdx, %rax jne .L30 .L27: call cudaDeviceReset@PLT testl %eax, %eax jne .L44 movl 36(%rsp), %edx movl 32(%rsp), %esi movq %r13, %rdi call _Z12print_matrixPfii movl 28(%rsp), %eax leal -1(%rax), %edx movl %edx, 28(%rsp) testl %eax, %eax je .L20 .L32: leaq 36(%rsp), %rdx leaq 32(%rsp), %rsi leaq .LC3(%rip), %rdi movl $0, %eax call __isoc23_scanf@PLT movl 32(%rsp), %eax imull 36(%rsp), %eax movl %eax, 12(%rsp) cltq leaq 0(,%rax,4), %rbx movq %rbx, (%rsp) movq %rbx, %rdi call malloc@PLT movq %rax, %r14 movq %rbx, %rdi call malloc@PLT movq %rax, %r15 movq %rbx, %rdi call malloc@PLT movq %rax, %r13 testq %r14, %r14 sete %al testq %r15, %r15 sete %dl orb %dl, %al jne .L21 testq %r13, %r13 je .L21 movq %r14, %r12 movq %r15, %rbp movl $0, %ebx cmpl $0, 12(%rsp) jle .L23 .L24: movq %r12, %rsi leaq .LC5(%rip), %rdi movl $0, %eax call __isoc23_scanf@PLT movq %rbp, %rsi leaq .LC5(%rip), %rdi movl $0, %eax call __isoc23_scanf@PLT addl $1, %ebx addq $4, %r12 addq $4, %rbp movl 36(%rsp), %eax imull 32(%rsp), %eax cmpl %ebx, %eax jg .L24 .L23: movq $0, 40(%rsp) movq $0, 48(%rsp) movq $0, 56(%rsp) leaq 40(%rsp), %rdi movq (%rsp), %rbx movq %rbx, %rsi call cudaMalloc@PLT leaq 48(%rsp), %rdi movq %rbx, %rsi call cudaMalloc@PLT leaq 56(%rsp), %rdi movq %rbx, %rsi call cudaMalloc@PLT movl $1, %ecx movq %rbx, %rdx movq %r14, %rsi movq 40(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movq %rbx, %rdx movq %r15, %rsi movq 48(%rsp), %rdi call cudaMemcpy@PLT movl $1, %ecx movq %rbx, %rdx movq %r13, %rsi movq 56(%rsp), %rdi call cudaMemcpy@PLT movl 32(%rsp), %eax imull 36(%rsp), %eax leal 510(%rax), %edx addl $255, %eax cmovs %edx, %eax sarl $8, %eax movl $256, 76(%rsp) movl $1, 80(%rsp) movl %eax, 64(%rsp) movl $1, 68(%rsp) movl $0, %r9d movl $0, %r8d movq 76(%rsp), %rdx movl $1, %ecx movq 64(%rsp), %rdi movl $1, %esi call __cudaPushCallConfiguration@PLT testl %eax, %eax jne .L25 movl 36(%rsp), %r8d movl 32(%rsp), %ecx movq 56(%rsp), %rdx movq 48(%rsp), %rsi movq 40(%rsp), %rdi call _Z33__device_stub__Z8hadamardPfS_S_iiPfS_S_ii jmp .L25 .L42: movl %eax, %edi call cudaGetErrorString@PLT movq %rax, %rcx leaq .LC6(%rip), %rdx movl $2, %esi movq stderr(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movl $1, %edi call exit@PLT .L43: movl %eax, %ecx leaq .LC9(%rip), %rdx movl $2, %esi movq stderr(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movl $1, %edi call exit@PLT .L44: movl %eax, %edi call cudaGetErrorString@PLT movq %rax, %rcx leaq .LC10(%rip), %rdx movl $2, %esi movq stderr(%rip), %rdi movl $0, %eax call __fprintf_chk@PLT movl $1, %edi call exit@PLT .L41: call __stack_chk_fail@PLT .cfi_endproc .LFE2058: .size main, .-main .section .rodata.str1.1 .LC11: .string "_Z8hadamardPfS_S_ii" .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 .LC11(%rip), %rdx movq %rdx, %rcx leaq _Z8hadamardPfS_S_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 .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.cst16,"aM",@progbits,16 .align 16 .LC7: .long 2147483647 .long 0 .long 0 .long 0 .section .rodata.cst8,"aM",@progbits,8 .align 8 .LC8: .long -1998362383 .long 1055193269 .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 <cuda.h> #include <cuda_runtime.h> #define LOG_INPUT if(0) #define LOG_OUTPUT if(1) #define LOG if(0) __global__ void hadamard(float *A, float *B, float *C, int M, int N) { // Complete the kernel code snippet int i = threadIdx.x + blockDim.x * blockIdx.x; if(i < M*N) C[i] = A[i]*B[i]; } /** * Host main routine */ void print_matrix(float *A,int m,int n) { for(int i =0;i<m;i++) { for(int j=0;j<n;j++) printf("%.2f ",A[i*n+j]); printf("\n"); } } int main(void) { // Error code to check return values for CUDA calls cudaError_t err = cudaSuccess; // Print the vector length to be used, and compute its size int t; //number of test cases scanf("%d",&t); while(t--) { int m,n; scanf("%d %d",&m,&n); size_t size = m*n * sizeof(float); LOG printf("[Hadamard product of two matrices ]\n"); // Allocate the host input vector A float *h_A = (float*)malloc(size); // Allocate the host input vector B float *h_B = (float*)malloc(size); // Allocate the host output vector C float *h_C = (float*)malloc(size); // Verify that allocations succeeded if (h_A == NULL || h_B == NULL || h_C == NULL) { fprintf(stderr, "Failed to allocate host vectors!\n"); exit(EXIT_FAILURE); } // Initialize the host input vectors for (int i = 0; i < n*m; ++i) { scanf("%f",&h_A[i]); scanf("%f",&h_B[i]); } float *d_A = NULL, *d_B = NULL, *d_C = NULL; // Allocate the device input vector A cudaMalloc((void**)&d_A, size); // Allocate the device input vector B cudaMalloc((void**)&d_B, size); // Allocate the device output vector C cudaMalloc((void**)&d_C, size); // Copy the host input vectors A and B in host memory to the device input vectors in // device memory cudaMemcpy(d_A, h_A, size, cudaMemcpyHostToDevice); cudaMemcpy(d_B, h_B, size, cudaMemcpyHostToDevice); cudaMemcpy(d_C, h_C, size, cudaMemcpyHostToDevice); // initialize blocksPerGrid and threads Per Block int threadsPerBlock = 256; int blocksPerGrid = ((m*n)+threadsPerBlock-1)/threadsPerBlock; hadamard<<<blocksPerGrid, threadsPerBlock>>>(d_A, d_B, d_C, m, n); err = cudaGetLastError(); if (err != cudaSuccess) { fprintf(stderr, "Failed to launch vectorAdd kernel (error code %s)!\n", cudaGetErrorString(err)); exit(EXIT_FAILURE); } // Copy the device result vector in device memory to the host result vector // in host memory. cudaMemcpy(h_C, d_C, size, cudaMemcpyDeviceToHost); // Verify that the result vector is correct for (int i = 0; i < n*m; ++i) { if (fabs(h_A[i] * h_B[i] - h_C[i]) > 1e-5) { fprintf(stderr, "Result verification failed at element %d!\n", i); exit(EXIT_FAILURE); } } LOG printf("Test PASSED\n"); // Free device global memory // Free host memory err = cudaDeviceReset(); if (err != cudaSuccess) { fprintf(stderr, "Failed to deinitialize the device! error=%s\n", cudaGetErrorString(err)); exit(EXIT_FAILURE); } print_matrix(h_C,m,n); LOG printf("Done\n"); } return 0; }
#include <stdio.h> #include <hip/hip_runtime.h> #define LOG_INPUT if(0) #define LOG_OUTPUT if(1) #define LOG if(0) __global__ void hadamard(float *A, float *B, float *C, int M, int N) { // Complete the kernel code snippet int i = threadIdx.x + blockDim.x * blockIdx.x; if(i < M*N) C[i] = A[i]*B[i]; } /** * Host main routine */ void print_matrix(float *A,int m,int n) { for(int i =0;i<m;i++) { for(int j=0;j<n;j++) printf("%.2f ",A[i*n+j]); printf("\n"); } } int main(void) { // Error code to check return values for CUDA calls hipError_t err = hipSuccess; // Print the vector length to be used, and compute its size int t; //number of test cases scanf("%d",&t); while(t--) { int m,n; scanf("%d %d",&m,&n); size_t size = m*n * sizeof(float); LOG printf("[Hadamard product of two matrices ]\n"); // Allocate the host input vector A float *h_A = (float*)malloc(size); // Allocate the host input vector B float *h_B = (float*)malloc(size); // Allocate the host output vector C float *h_C = (float*)malloc(size); // Verify that allocations succeeded if (h_A == NULL || h_B == NULL || h_C == NULL) { fprintf(stderr, "Failed to allocate host vectors!\n"); exit(EXIT_FAILURE); } // Initialize the host input vectors for (int i = 0; i < n*m; ++i) { scanf("%f",&h_A[i]); scanf("%f",&h_B[i]); } float *d_A = NULL, *d_B = NULL, *d_C = NULL; // Allocate the device input vector A hipMalloc((void**)&d_A, size); // Allocate the device input vector B hipMalloc((void**)&d_B, size); // Allocate the device output vector C hipMalloc((void**)&d_C, size); // Copy the host input vectors A and B in host memory to the device input vectors in // device memory hipMemcpy(d_A, h_A, size, hipMemcpyHostToDevice); hipMemcpy(d_B, h_B, size, hipMemcpyHostToDevice); hipMemcpy(d_C, h_C, size, hipMemcpyHostToDevice); // initialize blocksPerGrid and threads Per Block int threadsPerBlock = 256; int blocksPerGrid = ((m*n)+threadsPerBlock-1)/threadsPerBlock; hadamard<<<blocksPerGrid, threadsPerBlock>>>(d_A, d_B, d_C, m, n); err = hipGetLastError(); if (err != hipSuccess) { fprintf(stderr, "Failed to launch vectorAdd kernel (error code %s)!\n", hipGetErrorString(err)); exit(EXIT_FAILURE); } // Copy the device result vector in device memory to the host result vector // in host memory. hipMemcpy(h_C, d_C, size, hipMemcpyDeviceToHost); // Verify that the result vector is correct for (int i = 0; i < n*m; ++i) { if (fabs(h_A[i] * h_B[i] - h_C[i]) > 1e-5) { fprintf(stderr, "Result verification failed at element %d!\n", i); exit(EXIT_FAILURE); } } LOG printf("Test PASSED\n"); // Free device global memory // Free host memory err = hipDeviceReset(); if (err != hipSuccess) { fprintf(stderr, "Failed to deinitialize the device! error=%s\n", hipGetErrorString(err)); exit(EXIT_FAILURE); } print_matrix(h_C,m,n); LOG printf("Done\n"); } return 0; }
You are a helpful coding assistant specialized in converting HIP source code to AMD device assembly.
#include <stdio.h> #include <hip/hip_runtime.h> #define LOG_INPUT if(0) #define LOG_OUTPUT if(1) #define LOG if(0) __global__ void hadamard(float *A, float *B, float *C, int M, int N) { // Complete the kernel code snippet int i = threadIdx.x + blockDim.x * blockIdx.x; if(i < M*N) C[i] = A[i]*B[i]; } /** * Host main routine */ void print_matrix(float *A,int m,int n) { for(int i =0;i<m;i++) { for(int j=0;j<n;j++) printf("%.2f ",A[i*n+j]); printf("\n"); } } int main(void) { // Error code to check return values for CUDA calls hipError_t err = hipSuccess; // Print the vector length to be used, and compute its size int t; //number of test cases scanf("%d",&t); while(t--) { int m,n; scanf("%d %d",&m,&n); size_t size = m*n * sizeof(float); LOG printf("[Hadamard product of two matrices ]\n"); // Allocate the host input vector A float *h_A = (float*)malloc(size); // Allocate the host input vector B float *h_B = (float*)malloc(size); // Allocate the host output vector C float *h_C = (float*)malloc(size); // Verify that allocations succeeded if (h_A == NULL || h_B == NULL || h_C == NULL) { fprintf(stderr, "Failed to allocate host vectors!\n"); exit(EXIT_FAILURE); } // Initialize the host input vectors for (int i = 0; i < n*m; ++i) { scanf("%f",&h_A[i]); scanf("%f",&h_B[i]); } float *d_A = NULL, *d_B = NULL, *d_C = NULL; // Allocate the device input vector A hipMalloc((void**)&d_A, size); // Allocate the device input vector B hipMalloc((void**)&d_B, size); // Allocate the device output vector C hipMalloc((void**)&d_C, size); // Copy the host input vectors A and B in host memory to the device input vectors in // device memory hipMemcpy(d_A, h_A, size, hipMemcpyHostToDevice); hipMemcpy(d_B, h_B, size, hipMemcpyHostToDevice); hipMemcpy(d_C, h_C, size, hipMemcpyHostToDevice); // initialize blocksPerGrid and threads Per Block int threadsPerBlock = 256; int blocksPerGrid = ((m*n)+threadsPerBlock-1)/threadsPerBlock; hadamard<<<blocksPerGrid, threadsPerBlock>>>(d_A, d_B, d_C, m, n); err = hipGetLastError(); if (err != hipSuccess) { fprintf(stderr, "Failed to launch vectorAdd kernel (error code %s)!\n", hipGetErrorString(err)); exit(EXIT_FAILURE); } // Copy the device result vector in device memory to the host result vector // in host memory. hipMemcpy(h_C, d_C, size, hipMemcpyDeviceToHost); // Verify that the result vector is correct for (int i = 0; i < n*m; ++i) { if (fabs(h_A[i] * h_B[i] - h_C[i]) > 1e-5) { fprintf(stderr, "Result verification failed at element %d!\n", i); exit(EXIT_FAILURE); } } LOG printf("Test PASSED\n"); // Free device global memory // Free host memory err = hipDeviceReset(); if (err != hipSuccess) { fprintf(stderr, "Failed to deinitialize the device! error=%s\n", hipGetErrorString(err)); exit(EXIT_FAILURE); } print_matrix(h_C,m,n); LOG printf("Done\n"); } return 0; }
.text .amdgcn_target "amdgcn-amd-amdhsa--gfx1100" .protected _Z8hadamardPfS_S_ii .globl _Z8hadamardPfS_S_ii .p2align 8 .type _Z8hadamardPfS_S_ii,@function _Z8hadamardPfS_S_ii: s_clause 0x1 s_load_b32 s4, s[0:1], 0x2c s_load_b64 s[2:3], s[0:1], 0x18 s_waitcnt lgkmcnt(0) s_and_b32 s4, s4, 0xffff s_mul_i32 s2, s3, s2 v_mad_u64_u32 v[1:2], null, s15, s4, v[0:1] s_delay_alu instid0(VALU_DEP_1) v_cmp_gt_i32_e32 vcc_lo, s2, v1 s_and_saveexec_b32 s2, vcc_lo s_cbranch_execz .LBB0_2 s_load_b128 s[4:7], s[0:1], 0x0 v_ashrrev_i32_e32 v2, 31, v1 s_load_b64 s[0:1], s[0:1], 0x10 s_delay_alu instid0(VALU_DEP_1) | instskip(SKIP_1) | instid1(VALU_DEP_1) v_lshlrev_b64 v[0:1], 2, v[1:2] s_waitcnt lgkmcnt(0) v_add_co_u32 v2, vcc_lo, s4, v0 s_delay_alu instid0(VALU_DEP_2) v_add_co_ci_u32_e32 v3, vcc_lo, s5, v1, vcc_lo v_add_co_u32 v4, vcc_lo, s6, v0 v_add_co_ci_u32_e32 v5, vcc_lo, s7, v1, vcc_lo v_add_co_u32 v0, vcc_lo, s0, v0 global_load_b32 v2, v[2:3], off global_load_b32 v3, v[4:5], off v_add_co_ci_u32_e32 v1, vcc_lo, s1, v1, vcc_lo s_waitcnt vmcnt(0) v_mul_f32_e32 v2, v2, v3 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 _Z8hadamardPfS_S_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 6 .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 _Z8hadamardPfS_S_ii, .Lfunc_end0-_Z8hadamardPfS_S_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 .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: _Z8hadamardPfS_S_ii .private_segment_fixed_size: 0 .sgpr_count: 18 .sgpr_spill_count: 0 .symbol: _Z8hadamardPfS_S_ii.kd .uniform_work_group_size: 1 .uses_dynamic_stack: false .vgpr_count: 6 .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 <hip/hip_runtime.h> #define LOG_INPUT if(0) #define LOG_OUTPUT if(1) #define LOG if(0) __global__ void hadamard(float *A, float *B, float *C, int M, int N) { // Complete the kernel code snippet int i = threadIdx.x + blockDim.x * blockIdx.x; if(i < M*N) C[i] = A[i]*B[i]; } /** * Host main routine */ void print_matrix(float *A,int m,int n) { for(int i =0;i<m;i++) { for(int j=0;j<n;j++) printf("%.2f ",A[i*n+j]); printf("\n"); } } int main(void) { // Error code to check return values for CUDA calls hipError_t err = hipSuccess; // Print the vector length to be used, and compute its size int t; //number of test cases scanf("%d",&t); while(t--) { int m,n; scanf("%d %d",&m,&n); size_t size = m*n * sizeof(float); LOG printf("[Hadamard product of two matrices ]\n"); // Allocate the host input vector A float *h_A = (float*)malloc(size); // Allocate the host input vector B float *h_B = (float*)malloc(size); // Allocate the host output vector C float *h_C = (float*)malloc(size); // Verify that allocations succeeded if (h_A == NULL || h_B == NULL || h_C == NULL) { fprintf(stderr, "Failed to allocate host vectors!\n"); exit(EXIT_FAILURE); } // Initialize the host input vectors for (int i = 0; i < n*m; ++i) { scanf("%f",&h_A[i]); scanf("%f",&h_B[i]); } float *d_A = NULL, *d_B = NULL, *d_C = NULL; // Allocate the device input vector A hipMalloc((void**)&d_A, size); // Allocate the device input vector B hipMalloc((void**)&d_B, size); // Allocate the device output vector C hipMalloc((void**)&d_C, size); // Copy the host input vectors A and B in host memory to the device input vectors in // device memory hipMemcpy(d_A, h_A, size, hipMemcpyHostToDevice); hipMemcpy(d_B, h_B, size, hipMemcpyHostToDevice); hipMemcpy(d_C, h_C, size, hipMemcpyHostToDevice); // initialize blocksPerGrid and threads Per Block int threadsPerBlock = 256; int blocksPerGrid = ((m*n)+threadsPerBlock-1)/threadsPerBlock; hadamard<<<blocksPerGrid, threadsPerBlock>>>(d_A, d_B, d_C, m, n); err = hipGetLastError(); if (err != hipSuccess) { fprintf(stderr, "Failed to launch vectorAdd kernel (error code %s)!\n", hipGetErrorString(err)); exit(EXIT_FAILURE); } // Copy the device result vector in device memory to the host result vector // in host memory. hipMemcpy(h_C, d_C, size, hipMemcpyDeviceToHost); // Verify that the result vector is correct for (int i = 0; i < n*m; ++i) { if (fabs(h_A[i] * h_B[i] - h_C[i]) > 1e-5) { fprintf(stderr, "Result verification failed at element %d!\n", i); exit(EXIT_FAILURE); } } LOG printf("Test PASSED\n"); // Free device global memory // Free host memory err = hipDeviceReset(); if (err != hipSuccess) { fprintf(stderr, "Failed to deinitialize the device! error=%s\n", hipGetErrorString(err)); exit(EXIT_FAILURE); } print_matrix(h_C,m,n); LOG printf("Done\n"); } return 0; }
.text .file "hadamard_product.hip" .globl _Z23__device_stub__hadamardPfS_S_ii # -- Begin function _Z23__device_stub__hadamardPfS_S_ii .p2align 4, 0x90 .type _Z23__device_stub__hadamardPfS_S_ii,@function _Z23__device_stub__hadamardPfS_S_ii: # @_Z23__device_stub__hadamardPfS_S_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 $_Z8hadamardPfS_S_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 _Z23__device_stub__hadamardPfS_S_ii, .Lfunc_end0-_Z23__device_stub__hadamardPfS_S_ii .cfi_endproc # -- End function .globl _Z12print_matrixPfii # -- Begin function _Z12print_matrixPfii .p2align 4, 0x90 .type _Z12print_matrixPfii,@function _Z12print_matrixPfii: # @_Z12print_matrixPfii .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 testl %esi, %esi jle .LBB1_6 # %bb.1: # %.preheader.lr.ph movl %edx, %ebx movl %esi, %eax movq %rax, 16(%rsp) # 8-byte Spill movl %edx, %r12d xorl %r13d, %r13d xorl %ebp, %ebp jmp .LBB1_2 .p2align 4, 0x90 .LBB1_5: # %._crit_edge # in Loop: Header=BB1_2 Depth=1 movl $10, %edi callq putchar@PLT incq %rbp addl %ebx, %r13d cmpq 16(%rsp), %rbp # 8-byte Folded Reload je .LBB1_6 .LBB1_2: # %.preheader # =>This Loop Header: Depth=1 # Child Loop BB1_4 Depth 2 testl %ebx, %ebx jle .LBB1_5 # %bb.3: # %.lr.ph # in Loop: Header=BB1_2 Depth=1 movl %r13d, %eax movq 8(%rsp), %rcx # 8-byte Reload leaq (%rcx,%rax,4), %r14 xorl %r15d, %r15d .p2align 4, 0x90 .LBB1_4: # Parent Loop BB1_2 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, %edi movb $1, %al callq printf incq %r15 cmpq %r15, %r12 jne .LBB1_4 jmp .LBB1_5 .LBB1_6: # %._crit_edge13 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_end1: .size _Z12print_matrixPfii, .Lfunc_end1-_Z12print_matrixPfii .cfi_endproc # -- End function .section .rodata.cst16,"aM",@progbits,16 .p2align 4, 0x0 # -- Begin function main .LCPI2_0: .long 0x7fffffff # float NaN .long 0x7fffffff # float NaN .long 0x7fffffff # float NaN .long 0x7fffffff # float NaN .section .rodata.cst8,"aM",@progbits,8 .p2align 3, 0x0 .LCPI2_1: .quad 0x3ee4f8b588e368f1 # double 1.0000000000000001E-5 .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 $184, %rsp .cfi_def_cfa_offset 240 .cfi_offset %rbx, -56 .cfi_offset %r12, -48 .cfi_offset %r13, -40 .cfi_offset %r14, -32 .cfi_offset %r15, -24 .cfi_offset %rbp, -16 leaq 20(%rsp), %rsi movl $.L.str.2, %edi xorl %eax, %eax callq __isoc23_scanf subl $1, 20(%rsp) jae .LBB2_1 .LBB2_26: # %._crit_edge69 xorl %eax, %eax addq $184, %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 .LBB2_1: # %.lr.ph68 .cfi_def_cfa_offset 240 movabsq $4294967296, %rax # imm = 0x100000000 addq $256, %rax # imm = 0x100 movq %rax, 64(%rsp) # 8-byte Spill jmp .LBB2_2 .p2align 4, 0x90 .LBB2_25: # %_Z12print_matrixPfii.exit # in Loop: Header=BB2_2 Depth=1 subl $1, 20(%rsp) jb .LBB2_26 .LBB2_2: # =>This Loop Header: Depth=1 # Child Loop BB2_7 Depth 2 # Child Loop BB2_15 Depth 2 # Child Loop BB2_21 Depth 2 # Child Loop BB2_23 Depth 3 movl $.L.str.3, %edi leaq 16(%rsp), %rsi leaq 12(%rsp), %rdx xorl %eax, %eax callq __isoc23_scanf movslq 16(%rsp), %rax movslq 12(%rsp), %r12 imulq %rax, %r12 shlq $2, %r12 movq %r12, %rdi callq malloc movq %rax, %r14 movq %r12, %rdi callq malloc movq %rax, %rbp movq %r12, %rdi callq malloc movq %r14, 24(%rsp) # 8-byte Spill testq %r14, %r14 je .LBB2_28 # %bb.3: # in Loop: Header=BB2_2 Depth=1 testq %rbp, %rbp je .LBB2_28 # %bb.4: # in Loop: Header=BB2_2 Depth=1 movq %rax, %rbx testq %rax, %rax je .LBB2_28 # %bb.5: # %.preheader # in Loop: Header=BB2_2 Depth=1 movl 16(%rsp), %eax imull 12(%rsp), %eax testl %eax, %eax jle .LBB2_8 # %bb.6: # %.lr.ph.preheader # in Loop: Header=BB2_2 Depth=1 movq 24(%rsp), %r14 # 8-byte Reload movq %rbp, %r15 xorl %r13d, %r13d .p2align 4, 0x90 .LBB2_7: # %.lr.ph # Parent Loop BB2_2 Depth=1 # => This Inner Loop Header: Depth=2 movl $.L.str.5, %edi movq %r14, %rsi xorl %eax, %eax callq __isoc23_scanf movl $.L.str.5, %edi movq %r15, %rsi xorl %eax, %eax callq __isoc23_scanf incq %r13 movslq 12(%rsp), %rax movslq 16(%rsp), %rcx imulq %rax, %rcx addq $4, %r15 addq $4, %r14 cmpq %rcx, %r13 jl .LBB2_7 .LBB2_8: # %._crit_edge # in Loop: Header=BB2_2 Depth=1 movq $0, 48(%rsp) movq $0, 40(%rsp) movq $0, 32(%rsp) leaq 48(%rsp), %rdi movq %r12, %rsi callq hipMalloc leaq 40(%rsp), %rdi movq %r12, %rsi callq hipMalloc leaq 32(%rsp), %rdi movq %r12, %rsi callq hipMalloc movq 48(%rsp), %rdi movq 24(%rsp), %r14 # 8-byte Reload movq %r14, %rsi movq %r12, %rdx movl $1, %ecx callq hipMemcpy movq 40(%rsp), %rdi movq %rbp, %rsi movq %r12, %rdx movl $1, %ecx callq hipMemcpy movq 32(%rsp), %rdi movq %rbx, %rsi movq %r12, %rdx movl $1, %ecx callq hipMemcpy movl 12(%rsp), %edi imull 16(%rsp), %edi leal 255(%rdi), %eax addl $510, %edi # imm = 0x1FE testl %eax, %eax cmovnsl %eax, %edi sarl $8, %edi movabsq $4294967296, %rax # imm = 0x100000000 orq %rax, %rdi movl $1, %esi movq 64(%rsp), %rdx # 8-byte Reload movl $1, %ecx xorl %r8d, %r8d xorl %r9d, %r9d callq __hipPushCallConfiguration testl %eax, %eax jne .LBB2_10 # %bb.9: # in Loop: Header=BB2_2 Depth=1 movq 48(%rsp), %rax movq 40(%rsp), %rcx movq 32(%rsp), %rdx movl 16(%rsp), %esi movl 12(%rsp), %edi movq %rax, 136(%rsp) movq %rcx, 128(%rsp) movq %rdx, 120(%rsp) movl %esi, 60(%rsp) movl %edi, 56(%rsp) leaq 136(%rsp), %rax movq %rax, 144(%rsp) leaq 128(%rsp), %rax movq %rax, 152(%rsp) leaq 120(%rsp), %rax movq %rax, 160(%rsp) leaq 60(%rsp), %rax movq %rax, 168(%rsp) leaq 56(%rsp), %rax movq %rax, 176(%rsp) leaq 104(%rsp), %rdi leaq 88(%rsp), %rsi leaq 80(%rsp), %rdx leaq 72(%rsp), %rcx callq __hipPopCallConfiguration movq 104(%rsp), %rsi movl 112(%rsp), %edx movq 88(%rsp), %rcx movl 96(%rsp), %r8d movl $_Z8hadamardPfS_S_ii, %edi leaq 144(%rsp), %r9 pushq 72(%rsp) .cfi_adjust_cfa_offset 8 pushq 88(%rsp) .cfi_adjust_cfa_offset 8 callq hipLaunchKernel addq $16, %rsp .cfi_adjust_cfa_offset -16 .LBB2_10: # in Loop: Header=BB2_2 Depth=1 callq hipGetLastError testl %eax, %eax jne .LBB2_11 # %bb.13: # in Loop: Header=BB2_2 Depth=1 movq 32(%rsp), %rsi movq %rbx, %rdi movq %r12, %rdx movl $2, %ecx callq hipMemcpy movl 16(%rsp), %eax imull 12(%rsp), %eax testl %eax, %eax movaps .LCPI2_0(%rip), %xmm1 # xmm1 = [NaN,NaN,NaN,NaN] movsd .LCPI2_1(%rip), %xmm2 # xmm2 = mem[0],zero jle .LBB2_17 # %bb.14: # %.lr.ph64.preheader # in Loop: Header=BB2_2 Depth=1 movl %eax, %eax xorl %edx, %edx .p2align 4, 0x90 .LBB2_15: # %.lr.ph64 # Parent Loop BB2_2 Depth=1 # => This Inner Loop Header: Depth=2 movss (%r14,%rdx,4), %xmm0 # xmm0 = mem[0],zero,zero,zero mulss (%rbp,%rdx,4), %xmm0 subss (%rbx,%rdx,4), %xmm0 andps %xmm1, %xmm0 cvtss2sd %xmm0, %xmm0 ucomisd %xmm2, %xmm0 ja .LBB2_27 # %bb.16: # in Loop: Header=BB2_15 Depth=2 incq %rdx cmpq %rdx, %rax jne .LBB2_15 .LBB2_17: # %._crit_edge65 # in Loop: Header=BB2_2 Depth=1 movq %rbx, 24(%rsp) # 8-byte Spill callq hipDeviceReset testl %eax, %eax jne .LBB2_18 # %bb.19: # in Loop: Header=BB2_2 Depth=1 movl 16(%rsp), %r14d testl %r14d, %r14d jle .LBB2_25 # %bb.20: # %.preheader.lr.ph.i # in Loop: Header=BB2_2 Depth=1 movl 12(%rsp), %r15d xorl %r12d, %r12d xorl %r13d, %r13d jmp .LBB2_21 .p2align 4, 0x90 .LBB2_24: # %._crit_edge.i # in Loop: Header=BB2_21 Depth=2 movl $10, %edi callq putchar@PLT incq %r13 addl %r15d, %r12d cmpq %r14, %r13 je .LBB2_25 .LBB2_21: # %.preheader.i # Parent Loop BB2_2 Depth=1 # => This Loop Header: Depth=2 # Child Loop BB2_23 Depth 3 testl %r15d, %r15d jle .LBB2_24 # %bb.22: # %.lr.ph.i # in Loop: Header=BB2_21 Depth=2 movl %r12d, %eax movq 24(%rsp), %rcx # 8-byte Reload leaq (%rcx,%rax,4), %rbp xorl %ebx, %ebx .p2align 4, 0x90 .LBB2_23: # Parent Loop BB2_2 Depth=1 # Parent Loop BB2_21 Depth=2 # => This Inner Loop Header: Depth=3 movss (%rbp,%rbx,4), %xmm0 # xmm0 = mem[0],zero,zero,zero cvtss2sd %xmm0, %xmm0 movl $.L.str, %edi movb $1, %al callq printf incq %rbx cmpq %rbx, %r15 jne .LBB2_23 jmp .LBB2_24 .LBB2_27: movq stderr(%rip), %rdi movl $.L.str.7, %esi # kill: def $edx killed $edx killed $rdx xorl %eax, %eax callq fprintf movl $1, %edi callq exit .LBB2_28: movq stderr(%rip), %rcx movl $.L.str.4, %edi movl $33, %esi movl $1, %edx callq fwrite@PLT movl $1, %edi callq exit .LBB2_11: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.6, %esi jmp .LBB2_12 .LBB2_18: movq stderr(%rip), %rbx movl %eax, %edi callq hipGetErrorString movl $.L.str.8, %esi .LBB2_12: movq %rbx, %rdi movq %rax, %rdx xorl %eax, %eax callq fprintf 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 $_Z8hadamardPfS_S_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_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 _Z8hadamardPfS_S_ii,@object # @_Z8hadamardPfS_S_ii .section .rodata,"a",@progbits .globl _Z8hadamardPfS_S_ii .p2align 3, 0x0 _Z8hadamardPfS_S_ii: .quad _Z23__device_stub__hadamardPfS_S_ii .size _Z8hadamardPfS_S_ii, 8 .type .L.str,@object # @.str .section .rodata.str1.1,"aMS",@progbits,1 .L.str: .asciz "%.2f " .size .L.str, 6 .type .L.str.2,@object # @.str.2 .L.str.2: .asciz "%d" .size .L.str.2, 3 .type .L.str.3,@object # @.str.3 .L.str.3: .asciz "%d %d" .size .L.str.3, 6 .type .L.str.4,@object # @.str.4 .L.str.4: .asciz "Failed to allocate host vectors!\n" .size .L.str.4, 34 .type .L.str.5,@object # @.str.5 .L.str.5: .asciz "%f" .size .L.str.5, 3 .type .L.str.6,@object # @.str.6 .L.str.6: .asciz "Failed to launch vectorAdd kernel (error code %s)!\n" .size .L.str.6, 52 .type .L.str.7,@object # @.str.7 .L.str.7: .asciz "Result verification failed at element %d!\n" .size .L.str.7, 43 .type .L.str.8,@object # @.str.8 .L.str.8: .asciz "Failed to deinitialize the device! error=%s\n" .size .L.str.8, 45 .type .L__unnamed_1,@object # @0 .L__unnamed_1: .asciz "_Z8hadamardPfS_S_ii" .size .L__unnamed_1, 20 .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 _Z23__device_stub__hadamardPfS_S_ii .addrsig_sym __hip_module_ctor .addrsig_sym __hip_module_dtor .addrsig_sym _Z8hadamardPfS_S_ii .addrsig_sym __hip_fatbin .addrsig_sym __hip_fatbin_wrapper .addrsig_sym __hip_cuid_