/external/tensorflow/tensorflow/compiler/xla/service/interpreter/ |
D | executor.cc | 51 bool XlaInterpreterExecutor::Memcpy(Stream *stream, void *host_dst, in Memcpy() argument 54 AsExecutorStream(stream)->EnqueueTask([this, host_dst, dev_src, size]() { in Memcpy() 55 port::Status ok = SynchronousMemcpy(host_dst, dev_src, size); in Memcpy() 77 void *host_dst, const DeviceMemoryBase &dev_src, uint64 size) { in SynchronousMemcpy() argument 78 memcpy(host_dst, dev_src.opaque(), size); in SynchronousMemcpy()
|
D | executor.h | 82 bool Memcpy(Stream *stream, void *host_dst, const DeviceMemoryBase &dev_src, 119 port::Status SynchronousMemcpy(void *host_dst,
|
/external/tensorflow/tensorflow/stream_executor/host/ |
D | host_gpu_executor.cc | 99 bool HostExecutor::Memcpy(Stream *stream, void *host_dst, in Memcpy() argument 105 [host_dst, src_mem, size]() { memcpy(host_dst, src_mem, size); }); in Memcpy() 170 port::Status HostExecutor::SynchronousMemcpy(void *host_dst, in SynchronousMemcpy() argument 173 memcpy(host_dst, gpu_src.opaque(), size); in SynchronousMemcpy()
|
D | host_gpu_executor.h | 75 bool Memcpy(Stream *stream, void *host_dst, const DeviceMemoryBase &gpu_src, 100 port::Status SynchronousMemcpy(void *host_dst,
|
/external/tensorflow/tensorflow/stream_executor/tpu/ |
D | tpu_stream.h | 56 stream_executor::DeviceMemoryBase device_src, void* host_dst, in EnqueueTransferDeviceToHost() argument 60 stream_, ApiConverter::ToC(device_src), host_dst, size, in EnqueueTransferDeviceToHost()
|
D | tpu_executor.cc | 279 bool TpuExecutor::Memcpy(Stream* stream, void* host_dst, in Memcpy() argument 284 executor_, get_stream(stream->implementation()), host_dst, &se_base, in Memcpy() 308 void* host_dst, const ::stream_executor::DeviceMemoryBase& device_src, in SynchronousMemcpy() argument 313 executor_, host_dst, &se_base, size, status.c_status); in SynchronousMemcpy()
|
D | tpu_executor.h | 120 bool Memcpy(Stream* stream, void* host_dst, 138 void* host_dst, const ::stream_executor::DeviceMemoryBase& device_src,
|
D | tpu_executor_c_api.h | 89 void* host_dst, 97 void* host_dst, 136 void* host_dst, uint64_t size,
|
/external/tensorflow/tensorflow/stream_executor/ |
D | stream_executor_pimpl.cc | 625 bool StreamExecutor::SynchronousMemcpy(void *host_dst, in SynchronousMemcpy() argument 628 VLOG(1) << "Called StreamExecutor::SynchronousMemcpy(host_dst=" << host_dst in SynchronousMemcpy() 633 implementation_->SynchronousMemcpy(host_dst, device_src, size); in SynchronousMemcpy() 656 const DeviceMemoryBase &device_src, int64 size, void *host_dst) { in SynchronousMemcpyD2H() argument 659 << ", host_dst=" << host_dst << ")" << StackTraceIfVLOG10(); in SynchronousMemcpyD2H() 663 host_dst); in SynchronousMemcpyD2H() 665 result = implementation_->SynchronousMemcpy(host_dst, device_src, size); in SynchronousMemcpyD2H() 671 device_src.opaque(), host_dst, size, in SynchronousMemcpyD2H() 701 bool StreamExecutor::Memcpy(Stream *stream, void *host_dst, in Memcpy() argument 703 return implementation_->Memcpy(stream, host_dst, device_src, size); in Memcpy()
|
D | stream_executor_pimpl.h | 256 bool SynchronousMemcpy(void *host_dst, const DeviceMemoryBase &device_src, 276 int64 size, void *host_dst); 283 port::MutableArraySlice<T> host_dst) { in SynchronousMemcpyD2H() argument 284 auto host_size = host_dst.size() * sizeof(T); in SynchronousMemcpyD2H() 286 return SynchronousMemcpyD2H(device_src, host_size, host_dst.begin()); in SynchronousMemcpyD2H() 572 bool Memcpy(Stream *stream, void *host_dst,
|
D | trace_listener.h | 62 int64 size, void* host_dst) {} in SynchronousMemcpyD2HBegin() argument
|
D | stream_executor_internal.h | 225 virtual port::Status SynchronousMemcpy(void *host_dst, 239 virtual bool Memcpy(Stream *stream, void *host_dst,
|
D | stream.h | 701 void *host_dst, uint64 size); 709 port::MutableArraySlice<ElementType> host_dst) { in ThenMemcpyD2HQuantized() argument 712 host_dst.data(), host_dst.size() * sizeof(ElementType)); in ThenMemcpyD2HQuantized() 1722 Stream &ThenMemcpy(void *host_dst, const DeviceMemoryBase &gpu_src, 1737 port::MutableArraySlice<T> host_dst) { in ThenMemcpyD2H() argument 1738 auto host_size = host_dst.size() * sizeof(T); in ThenMemcpyD2H() 1740 return ThenMemcpy(host_dst.begin(), gpu_src, host_size); in ThenMemcpyD2H()
|
D | stream.cc | 1262 dnn::QuantizedActivationMode mode, void *host_dst, uint64 size) { in ThenMemcpyD2HQuantized() argument 1263 VLOG_CALL(PARAM(gpu_unquantized_src), PARAM(mode), PARAM(host_dst), in ThenMemcpyD2HQuantized() 1268 host_dst, size)); in ThenMemcpyD2HQuantized() 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()
|
/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() 843 << host_dst; in SynchronousMemcpyD2H() 878 void* host_dst, 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() 895 << host_dst << " on stream " << stream; in AsynchronousMemcpyD2H()
|
D | rocm_gpu_executor.cc | 490 port::Status GpuExecutor::SynchronousMemcpy(void* host_dst, in SynchronousMemcpy() argument 493 return GpuDriver::SynchronousMemcpyD2H(context_, host_dst, in SynchronousMemcpy() 535 bool GpuExecutor::Memcpy(Stream* stream, void* host_dst, in Memcpy() argument 537 return GpuDriver::AsynchronousMemcpyD2H(context_, host_dst, in Memcpy()
|
D | rocm_dnn.h | 562 dnn::QuantizedActivationMode mode, void* host_dst,
|
/external/tensorflow/tensorflow/c/experimental/stream_executor/ |
D | stream_executor.h | 341 void (*memcpy_dtoh)(const SP_Device* device, SP_Stream stream, void* host_dst, 360 void (*sync_memcpy_dtoh)(const SP_Device* device, void* host_dst,
|
D | stream_executor_test.cc | 95 void* host_dst, const SP_DeviceMemoryBase* const device_src, in memcpy_dtoh() argument 100 void sync_memcpy_dtoh(const SP_Device* const device, void* host_dst, in sync_memcpy_dtoh() argument 713 void* host_dst, in TEST_F() 718 std::memcpy(host_dst, device_src->opaque, size); in TEST_F() 775 se_.sync_memcpy_dtoh = [](const SP_Device* const device, void* host_dst, in TEST_F() 779 std::memcpy(host_dst, device_src->opaque, size); in TEST_F()
|
D | stream_executor.cc | 361 port::Status SynchronousMemcpy(void* host_dst, in SynchronousMemcpy() argument 366 stream_executor_->sync_memcpy_dtoh(&device_, host_dst, &device_memory_base, in SynchronousMemcpy() 395 bool Memcpy(Stream* stream, void* host_dst, const DeviceMemoryBase& gpu_src, in Memcpy() argument 401 stream_executor_->memcpy_dtoh(&device_, stream_handle, host_dst, in Memcpy()
|
/external/tensorflow/tensorflow/stream_executor/gpu/ |
D | gpu_executor.h | 131 port::Status SynchronousMemcpy(void* host_dst, 146 bool Memcpy(Stream* stream, void* host_dst, const DeviceMemoryBase& gpu_src,
|
D | gpu_driver.h | 339 static port::Status SynchronousMemcpyD2H(GpuContext* context, void* host_dst, 351 static bool AsynchronousMemcpyD2H(GpuContext* context, void* host_dst,
|
/external/tensorflow/tensorflow/stream_executor/cuda/ |
D | cuda_driver.cc | 1129 void* host_dst, 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() 1139 << host_dst; in SynchronousMemcpyD2H() 1175 void* host_dst, 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() 1190 << host_dst << " on stream " << stream; in AsynchronousMemcpyD2H()
|
D | cuda_gpu_executor.cc | 582 port::Status GpuExecutor::SynchronousMemcpy(void* host_dst, in SynchronousMemcpy() argument 585 return GpuDriver::SynchronousMemcpyD2H(context_, host_dst, in SynchronousMemcpy() 627 bool GpuExecutor::Memcpy(Stream* stream, void* host_dst, in Memcpy() argument 629 return GpuDriver::AsynchronousMemcpyD2H(context_, host_dst, in Memcpy()
|
D | cuda_dnn.h | 546 dnn::QuantizedActivationMode mode, void* host_dst,
|