/external/tensorflow/tensorflow/stream_executor/host/ |
D | host_gpu_executor.cc | 100 const DeviceMemoryBase &gpu_src, uint64 size) { in Memcpy() argument 103 void *src_mem = const_cast<void *>(gpu_src.opaque()); in Memcpy() 121 const DeviceMemoryBase &gpu_src, in MemcpyDeviceToDevice() argument 124 void *src_mem = const_cast<void *>(gpu_src.opaque()); in MemcpyDeviceToDevice() 171 const DeviceMemoryBase &gpu_src, in SynchronousMemcpy() argument 173 memcpy(host_dst, gpu_src.opaque(), size); in SynchronousMemcpy() 178 DeviceMemoryBase *gpu_dst, const DeviceMemoryBase &gpu_src, uint64 size) { in SynchronousMemcpyDeviceToDevice() argument 179 memcpy(gpu_dst->opaque(), gpu_src.opaque(), size); in SynchronousMemcpyDeviceToDevice()
|
D | host_gpu_executor.h | 75 bool Memcpy(Stream *stream, void *host_dst, const DeviceMemoryBase &gpu_src, 80 const DeviceMemoryBase &gpu_src, 101 const DeviceMemoryBase &gpu_src, 104 const DeviceMemoryBase &gpu_src,
|
/external/tensorflow/tensorflow/stream_executor/rocm/ |
D | rocm_driver.cc | 835 GpuContext* context, void* host_dst, hipDeviceptr_t gpu_src, uint64 size) { in SynchronousMemcpyD2H() argument 838 tensorflow::wrap::hipMemcpyDtoH(host_dst, gpu_src, size), in SynchronousMemcpyD2H() 841 host_dst, absl::bit_cast<void*>(gpu_src), size, size)); in SynchronousMemcpyD2H() 863 GpuContext* context, hipDeviceptr_t gpu_dst, hipDeviceptr_t gpu_src, in SynchronousMemcpyD2D() argument 867 tensorflow::wrap::hipMemcpyDtoD(gpu_dst, gpu_src, size), in SynchronousMemcpyD2D() 871 absl::bit_cast<void*>(gpu_dst), absl::bit_cast<void*>(gpu_src), size, in SynchronousMemcpyD2D() 879 hipDeviceptr_t gpu_src, in AsynchronousMemcpyD2H() argument 884 tensorflow::wrap::hipMemcpyDtoHAsync(host_dst, gpu_src, size, stream); in AsynchronousMemcpyD2H() 889 ToString(res).c_str(), host_dst, absl::bit_cast<void*>(gpu_src), size, in AsynchronousMemcpyD2H() 894 << " bytes from " << absl::bit_cast<void*>(gpu_src) << " to " in AsynchronousMemcpyD2H() [all …]
|
D | rocm_gpu_executor.cc | 491 const DeviceMemoryBase& gpu_src, in SynchronousMemcpy() argument 494 AsROCmDevicePtr(gpu_src), size); in SynchronousMemcpy() 498 DeviceMemoryBase* gpu_dst, const DeviceMemoryBase& gpu_src, uint64 size) { in SynchronousMemcpyDeviceToDevice() argument 500 AsROCmDevicePtr(gpu_src), size); in SynchronousMemcpyDeviceToDevice() 536 const DeviceMemoryBase& gpu_src, uint64 size) { in Memcpy() argument 538 AsROCmDevicePtr(gpu_src), size, in Memcpy() 551 const DeviceMemoryBase& gpu_src, in MemcpyDeviceToDevice() argument 554 AsROCmDevicePtr(gpu_src), size, in MemcpyDeviceToDevice()
|
/external/tensorflow/tensorflow/stream_executor/ |
D | stream_executor_internal.h | 226 const DeviceMemoryBase &gpu_src, 229 DeviceMemoryBase *gpu_dst, const DeviceMemoryBase &gpu_src, 240 const DeviceMemoryBase &gpu_src, uint64 size) = 0; 244 const DeviceMemoryBase &gpu_src,
|
D | trace_listener.h | 61 const DeviceMemoryBase& gpu_src, in SynchronousMemcpyD2HBegin() argument
|
D | stream.h | 1722 Stream &ThenMemcpy(void *host_dst, const DeviceMemoryBase &gpu_src, 1736 Stream &ThenMemcpyD2H(const DeviceMemory<T> &gpu_src, in ThenMemcpyD2H() argument 1739 CHECK(gpu_src.size() == 0 || host_size >= gpu_src.size()); in ThenMemcpyD2H() 1740 return ThenMemcpy(host_dst.begin(), gpu_src, host_size); in ThenMemcpyD2H() 1757 Stream &ThenMemcpy(DeviceMemoryBase *gpu_dst, const DeviceMemoryBase &gpu_src, 1764 const DeviceMemoryBase &gpu_src, uint64 size) { in ThenMemcpyD2D() argument 1765 return ThenMemcpy(gpu_dst, gpu_src, size); in ThenMemcpyD2D()
|
D | stream.cc | 4499 Stream &Stream::ThenMemcpy(void *host_dst, const DeviceMemoryBase &gpu_src, in ThenMemcpy() argument 4501 VLOG_CALL(PARAM(host_dst), PARAM(gpu_src), PARAM(size)); in ThenMemcpy() 4503 CheckError(parent_->Memcpy(this, host_dst, gpu_src, size)); in ThenMemcpy() 4516 const DeviceMemoryBase &gpu_src, uint64 size) { in ThenMemcpy() argument 4517 VLOG_CALL(PARAM(gpu_dst), PARAM(gpu_src), PARAM(size)); in ThenMemcpy() 4519 CheckError(parent_->MemcpyDeviceToDevice(this, gpu_dst, gpu_src, size)); in ThenMemcpy()
|
/external/tensorflow/tensorflow/stream_executor/gpu/ |
D | gpu_executor.h | 132 const DeviceMemoryBase& gpu_src, 136 const DeviceMemoryBase& gpu_src, 146 bool Memcpy(Stream* stream, void* host_dst, const DeviceMemoryBase& gpu_src, 153 const DeviceMemoryBase& gpu_src,
|
D | gpu_driver.h | 340 GpuDevicePtr gpu_src, uint64 size); 346 GpuDevicePtr gpu_src, uint64 size); 352 GpuDevicePtr gpu_src, uint64 size, 358 GpuDevicePtr gpu_src, uint64 size,
|
/external/tensorflow/tensorflow/stream_executor/cuda/ |
D | cuda_driver.cc | 1130 CUdeviceptr gpu_src, in SynchronousMemcpyD2H() argument 1134 cuMemcpyDtoH(host_dst, gpu_src, size), in SynchronousMemcpyD2H() 1137 host_dst, absl::bit_cast<void*>(gpu_src), size, size)); in SynchronousMemcpyD2H() 1160 CUdeviceptr gpu_src, in SynchronousMemcpyD2D() argument 1164 cuMemcpyDtoD(gpu_dst, gpu_src, size), in SynchronousMemcpyD2D() 1168 absl::bit_cast<void*>(gpu_dst), absl::bit_cast<void*>(gpu_src), size, in SynchronousMemcpyD2D() 1176 CUdeviceptr gpu_src, in AsynchronousMemcpyD2H() argument 1180 CUresult res = cuMemcpyDtoHAsync(host_dst, gpu_src, size, stream); in AsynchronousMemcpyD2H() 1185 ToString(res), host_dst, absl::bit_cast<void*>(gpu_src), size, size); in AsynchronousMemcpyD2H() 1189 << " bytes from " << absl::bit_cast<void*>(gpu_src) << " to " in AsynchronousMemcpyD2H() [all …]
|
D | cuda_gpu_executor.cc | 583 const DeviceMemoryBase& gpu_src, in SynchronousMemcpy() argument 586 AsCudaDevicePtr(gpu_src), size); in SynchronousMemcpy() 590 DeviceMemoryBase* gpu_dst, const DeviceMemoryBase& gpu_src, uint64 size) { in SynchronousMemcpyDeviceToDevice() argument 592 AsCudaDevicePtr(gpu_src), size); in SynchronousMemcpyDeviceToDevice() 628 const DeviceMemoryBase& gpu_src, uint64 size) { in Memcpy() argument 630 AsCudaDevicePtr(gpu_src), size, in Memcpy() 643 const DeviceMemoryBase& gpu_src, in MemcpyDeviceToDevice() argument 646 AsCudaDevicePtr(gpu_src), size, in MemcpyDeviceToDevice()
|
/external/tensorflow/tensorflow/c/experimental/stream_executor/ |
D | stream_executor.cc | 362 const DeviceMemoryBase& gpu_src, in SynchronousMemcpy() argument 365 SP_DeviceMemoryBase device_memory_base = DeviceMemoryBaseToC(&gpu_src); in SynchronousMemcpy() 371 const DeviceMemoryBase& gpu_src, in SynchronousMemcpyDeviceToDevice() argument 375 SP_DeviceMemoryBase device_mem_src = DeviceMemoryBaseToC(&gpu_src); in SynchronousMemcpyDeviceToDevice() 395 bool Memcpy(Stream* stream, void* host_dst, const DeviceMemoryBase& gpu_src, in Memcpy() argument 400 SP_DeviceMemoryBase device_mem_src = DeviceMemoryBaseToC(&gpu_src); in Memcpy() 424 const DeviceMemoryBase& gpu_src, in MemcpyDeviceToDevice() argument 430 SP_DeviceMemoryBase device_mem_src = DeviceMemoryBaseToC(&gpu_src); in MemcpyDeviceToDevice()
|
/external/tensorflow/tensorflow/ |
D | tensorflow.bzl | 1553 for gpu_src in gpu_srcs: 1554 if gpu_src.endswith(".cc") and not gpu_src.endswith(".cu.cc"): 1556 .format(gpu_src))
|