Home
last modified time | relevance | path

Searched defs:input_addr (Results 1 – 25 of 175) sorted by relevance

1234567

/third_party/mindspore/mindspore-src/source/mindspore/ccsrc/plugin/device/gpu/kernel/cuda_impl/cuda_ops/
Dmish_impl.cu21 __global__ void MishKernel(const size_t size, const T *input_addr, T *output_addr) { in MishKernel()
28 __global__ void MishKernel(const size_t size, const half *input_addr, half *output_addr) { in MishKernel()
35 __global__ void MishKernel(const size_t size, const double *input_addr, double *output_addr) { in MishKernel()
42 cudaError_t Mish(const size_t size, const T *input_addr, T *output_addr, const uint32_t &device_id, in Mish()
49 cudaError_t Mish(const size_t size, const half *input_addr, half *output_addr, const uint32_t &devi… in Mish()
57 cudaError_t Mish(const size_t size, const double *input_addr, double *output_addr, const uint32_t &… in Mish()
Dsoftsign_impl.cu21 __global__ void SoftsignKernel(const size_t size, const T *input_addr, T *output_addr) { in SoftsignKernel()
28 __global__ void SoftsignKernel(const size_t size, const half *input_addr, half *output_addr) { in SoftsignKernel()
35 __global__ void SoftsignKernel(const size_t size, const double *input_addr, double *output_addr) { in SoftsignKernel()
42 cudaError_t Softsign(const size_t size, const T *input_addr, T *output_addr, const uint32_t &device… in Softsign()
50 cudaError_t Softsign(const size_t size, const half *input_addr, half *output_addr, const uint32_t &… in Softsign()
58 cudaError_t Softsign(const size_t size, const double *input_addr, double *output_addr, const uint32… in Softsign()
Dsparse_matrix_transpose_impl.cu22 __global__ void ConjKernel(const size_t input_size, cuComplex *input_addr) { in ConjKernel()
29 __global__ void ConjKernel(const size_t input_size, cuDoubleComplex *input_addr) { in ConjKernel()
36 cudaError_t Conj(const size_t input_size, cuComplex *input_addr, cudaStream_t stream) { in Conj()
41 cudaError_t Conj(const size_t input_size, cuDoubleComplex *input_addr, cudaStream_t stream) { in Conj()
Dcorrelate_impl.cu21 __global__ void Conv1D(const T *input_addr, const T *kernel_addr, T *output_addr, const size_t out_… in Conv1D()
35 __global__ void Conj(const T *input_addr, T *output_addr, const size_t input_size) { in Conj()
43 __global__ void Reverse(const T *input_addr, T *output_addr, const size_t *input_size_d, size_t inp… in Reverse()
51 cudaError_t CorrelateCalc(const T *input_addr, const T *kernel_addr, T *output_addr, const size_t i… in CorrelateCalc()
77 cudaError_t CalConj(const T *input_addr, T *output_addr, const size_t input_size, const uint32_t &d… in CalConj()
84 cudaError_t CalReverse1D(const T *input_addr, T *output_addr, const size_t *input_size_d, size_t in… in CalReverse1D()
Dconvert_gradient_impl.cu21 … const size_t batchwidth, const size_t width, T *input_addr, T *output_addr) { in ConvertGradientKernel()
35 … const size_t batchwidth, const size_t width, T *input_addr, T *output_addr) { in ConvertGradientBackKernel()
50 const size_t width, T *input_addr, T *output_addr) { in ConvertGradientBackKernel()
68 … const size_t width, T *input_addr, T *output_addr, cudaStream_t cuda_stream) { in ConvertGradient()
76 … const size_t batchwidth, const size_t width, T *input_addr, T *output_addr, in ConvertGradientBack()
85 … const size_t ori_w, const size_t batchwidth, const size_t width, T *input_addr, in ConvertGradientBack()
Ddata_format_dim_map_impl.cu21 __global__ void DataFormatDimMapKernel(size_t size, T *input_addr, T *output_addr, int32_t *dim_map… in DataFormatDimMapKernel()
28 cudaError_t DataFormatDimMap(size_t size, T *input_addr, T *output_addr, int32_t *dim_map, cudaStre… in DataFormatDimMap()
Dmatrix_split_impl.cu20 … void MatrixSplitKernel(const size_t size, const size_t split_dim, const size_t dim, T *input_addr, in MatrixSplitKernel()
33 T *input_addr, T *output_addr) { in MatrixSplitKernel()
55 cudaError_t MatrixSplit(const size_t size, const size_t split_dim, const size_t dim, T *input_addr,… in MatrixSplit()
Dfast_gelu_impl.cu21 __global__ void FastGeluKernel(size_t size, T *input_addr, T *output_addr) { in FastGeluKernel()
33 __global__ void FastGeluKernel(size_t size, half *input_addr, half *output_addr) { in FastGeluKernel()
43 __global__ void FastGeluKernel(size_t size, half2 *input_addr, half2 *output_addr) { in FastGeluKernel()
60 cudaError_t FastGelu(size_t size, T *input_addr, T *output_addr, cudaStream_t cuda_stream) { in FastGelu()
66 cudaError_t FastGelu(size_t size, half *input_addr, half *output_addr, cudaStream_t cuda_stream) { in FastGelu()
Dgelu_impl.cu21 __global__ void GeluKernel(size_t size, const T *input_addr, T *output_addr) { in GeluKernel()
34 __global__ void GeluKernel(size_t size, const half *input_addr, half *output_addr) { in GeluKernel()
43 __global__ void GeluKernel(size_t size, const half2 *input_addr, half2 *output_addr) { in GeluKernel()
55 cudaError_t Gelu(size_t size, const T *input_addr, T *output_addr, cudaStream_t cuda_stream, const … in Gelu()
61 cudaError_t Gelu(size_t size, const half *input_addr, half *output_addr, cudaStream_t cuda_stream, in Gelu()
Dmatrix_combine_impl.cu21 const size_t dst_width, T *input_addr, T *output_addr) { in MatrixCombineKernel()
34 … const size_t dst_width, const size_t res_width, const size_t batch, T *input_addr, in MatrixCombineKernel()
57 … const size_t residual, const size_t res_width, const size_t batch, T *input_addr, in MatrixCombine()
Dsoftplus_impl.cu25 __global__ void SoftplusKernel(const size_t size, const T threshold, const T *input_addr, T *output… in SoftplusKernel()
33 __global__ void SoftplusKernel(const size_t size, const half threshold, const half *input_addr, hal… in SoftplusKernel()
43 cudaError_t Softplus(const size_t size, const T *input_addr, T *output_addr, cudaStream_t cuda_stre… in Softplus()
Dunsorted_segment_sum.cu22 T *input_addr, S *ids_addr, T *output_addr) { in UnsortedSegmentSumCal()
47 T *input_addr, S *ids_addr, T *output_addr, cudaStream_t stream, in UnsortedSegmentSum()
Dunsorted_segment_prod.cu22 T *input_addr, S *ids_addr, T *output_addr) { in UnsortedSegmentProdCal()
47 T *input_addr, S *ids_addr, T *output_addr, cudaStream_t stream, in UnsortedSegmentProd()
Dunsorted_segment_max.cu34 T *input_addr, S *ids_addr, T *output_addr) { in UnsortedSegmentMaxCal()
60 T *input_addr, S *ids_addr, T *output_addr, cudaStream_t stream, in UnsortedSegmentMax()
Dunsorted_segment_min.cu34 T *input_addr, S *ids_addr, T *output_addr) { in UnsortedSegmentMinCal()
60 T *input_addr, S *ids_addr, T *output_addr, cudaStream_t stream, in UnsortedSegmentMin()
/third_party/mindspore/mindspore-src/source/mindspore/lite/src/extendrt/delegate/tensorrt/cuda_impl/
Dcast.cu22 __device__ __forceinline__ void CastBase(const S *input_addr, T *output_addr) { in CastBase()
27 __global__ void CastKernel(const int input_size, const S *input_addr, T *output_addr) { in CastKernel()
34 void Cast(const int input_size, const S *input_addr, T *output_addr, cudaStream_t stream) { in Cast()
/third_party/mindspore/mindspore-src/source/mindspore/lite/src/extendrt/delegate/tensorrt/distribution/
Ddistribution_collective.cc27 int DistributionCollective::ReduceScatterWrapper(const void *input_addr, void *output_addr, size_t … in ReduceScatterWrapper()
33 int DistributionCollective::AllReduceWrapper(const void *input_addr, void *output_addr, size_t coun… in AllReduceWrapper()
39 int DistributionCollective::AllGatherWrapper(const void *input_addr, void *output_addr, size_t coun… in AllGatherWrapper()
Ddistribution_collective_impl.cc36 int DistributionCollective::ReduceScatterWrapper(const void *input_addr, void *output_addr, size_t … in ReduceScatterWrapper()
55 int DistributionCollective::AllReduceWrapper(const void *input_addr, void *output_addr, size_t coun… in AllReduceWrapper()
74 int DistributionCollective::AllGatherWrapper(const void *input_addr, void *output_addr, size_t coun… in AllGatherWrapper()
/third_party/mindspore/mindspore-src/source/mindspore/ccsrc/plugin/device/cpu/kernel/
Drolling_cpu_kernel.cc57 S Var(const T *input_addr, const size_t *ids, size_t start, size_t end) const { in Var() argument
162 reduceMethod_ = [](const T *input_addr, const size_t *ids, size_t start, size_t end) { in MethodSwitch()
173 reduceMethod_ = [](const T *input_addr, const size_t *ids, size_t start, size_t end) { in MethodSwitch()
184 reduceMethod_ = [](const T *input_addr, const size_t *ids, size_t start, size_t end) { in MethodSwitch()
193 reduceMethod_ = [](const T *input_addr, const size_t *ids, size_t start, size_t end) { in MethodSwitch()
202 reduceMethod_ = [this](const T *input_addr, const size_t *ids, size_t start, size_t end) { in MethodSwitch()
207 reduceMethod_ = [this](const T *input_addr, const size_t *ids, size_t start, size_t end) { in MethodSwitch()
221 auto input_addr = reinterpret_cast<T *>(inputs[kIndex0]->device_ptr()); in RunFunc() local
Dstridedslice_cpu_kernel.cc175 common::Status StridedSliceCpuKernelMod::RunTaskOnOuter(const uint8_t *input_addr, uint8_t *output_… in RunTaskOnOuter()
193 common::Status StridedSliceCpuKernelMod::RunTaskOnSplitAxis(const uint8_t *input_addr, uint8_t *out… in RunTaskOnSplitAxis()
209 void StridedSliceCpuKernelMod::ParallelRun(const uint8_t *input_addr, uint8_t *output_addr, int thr… in ParallelRun()
238 auto input_addr = reinterpret_cast<uint8_t *>(inputs[0]->device_ptr()); in LaunchKernel() local
/third_party/mindspore/mindspore-src/source/mindspore/ccsrc/plugin/device/gpu/kernel/nccl/
Dnccl_gpu_kernel.cc41 bool NcclGpuKernelMod::AllReduce(const void *input_addr, void *output_addr, size_t count, ncclDataT… in AllReduce()
49 bool NcclGpuKernelMod::AllGather(const void *input_addr, void *output_addr, size_t count, ncclDataT… in AllGather()
57 bool NcclGpuKernelMod::ReduceScatter(const void *input_addr, void *output_addr, size_t count, ncclD… in ReduceScatter()
66 bool NcclGpuKernelMod::Broadcast(const void *input_addr, void *output_addr, size_t count, ncclDataT… in Broadcast()
Dnccl_collective_gpu_kernel.h145 T *input_addr = GetDeviceAddress<T>(inputs, 0); in LaunchAllReduce() local
153 T *input_addr = GetDeviceAddress<T>(inputs, 0); in LaunchAllGather() local
161 T *input_addr = GetDeviceAddress<T>(inputs, 0); in LaunchReduceScatter() local
169 T *input_addr = nullptr; in LaunchBroadcast() local
/third_party/mindspore/mindspore-src/source/mindspore/ccsrc/plugin/device/gpu/hal/device/distribution/
Dcollective_wrapper.cc37 ncclResult_t AllReduce(const void *input_addr, void *output_addr, size_t count, ncclDataType_t data… in AllReduce()
42 ncclResult_t AllGather(const void *input_addr, void *output_addr, size_t count, ncclDataType_t data… in AllGather()
47 ncclResult_t ReduceScatter(const void *input_addr, void *output_addr, size_t count, ncclDataType_t … in ReduceScatter()
52 ncclResult_t Broadcast(const void *input_addr, void *output_addr, size_t count, ncclDataType_t data… in Broadcast()
Dnccl_wrapper.cc57 ncclResult_t NCCLWrapper::AllReduce(const void *input_addr, void *output_addr, size_t count, ncclDa… in AllReduce()
65 ncclResult_t NCCLWrapper::AllGather(const void *input_addr, void *output_addr, size_t count, ncclDa… in AllGather()
73 ncclResult_t NCCLWrapper::ReduceScatter(const void *input_addr, void *output_addr, size_t count, in ReduceScatter()
82 ncclResult_t NCCLWrapper::Broadcast(const void *input_addr, void *output_addr, size_t count, ncclDa… in Broadcast()
/third_party/mindspore/mindspore-src/source/mindspore/ccsrc/plugin/device/cpu/kernel/mkldnn/
Dpooling_cpu_kernel_nnacl.cc307 CTask PoolingCpuKernelNnaclMod::KernelAvgPool(T *input_addr, T *output_addr) { in KernelAvgPool()
361 CTask PoolingCpuKernelNnaclMod::KernelMaxPool(T *input_addr, T *output_addr) { in KernelMaxPool()
397 void PoolingCpuKernelNnaclMod::LaunchTransposeFp32(float *input_addr, float *output_addr, int plane… in LaunchTransposeFp32()
408 void PoolingCpuKernelNnaclMod::LaunchPoolingChannelLastFp32(float *input_addr, float *transpose_out… in LaunchPoolingChannelLastFp32()
434 T *input_addr = reinterpret_cast<T *>(inputs[kIndex0]->device_ptr()); in LaunchKernel() local
459 float *input_addr = reinterpret_cast<float *>(inputs[kIndex0]->device_ptr()); in Launch() local

1234567