| /external/tensorflow/tensorflow/stream_executor/ |
| D | kernel.cc | 60 KernelBase::KernelBase(KernelBase &&from) in KernelBase() function in stream_executor::KernelBase
69 KernelBase::KernelBase(StreamExecutor *parent) in KernelBase() function in stream_executor::KernelBase
73 KernelBase::KernelBase(StreamExecutor *parent, in KernelBase() function in stream_executor::KernelBase
77 KernelBase::~KernelBase() { in ~KernelBase()
83 unsigned KernelBase::Arity() const { return implementation_->Arity(); } in Arity()
85 void KernelBase::SetPreferredCacheConfig(KernelCacheConfig config) { in SetPreferredCacheConfig()
89 KernelCacheConfig KernelBase::GetPreferredCacheConfig() const { in GetPreferredCacheConfig()
93 void KernelBase::set_name(absl::string_view name) { in set_name()
|
| D | kernel.h | 136 class KernelBase {
138 KernelBase(KernelBase &&from);
144 explicit KernelBase(StreamExecutor *parent);
147 KernelBase(StreamExecutor *parent, internal::KernelInterface *implementation);
151 ~KernelBase();
197 SE_DISALLOW_COPY_AND_ASSIGN(KernelBase);
508 class TypedKernel : public KernelBase {
513 explicit TypedKernel(StreamExecutor *parent) : KernelBase(parent) {}
518 : KernelBase(parent, implementation) {}
|
| D | stream_executor_internal.h | 182 KernelBase *kernel) { in GetKernel()
191 const BlockDim &block_dims, const KernelBase &k, in Launch()
197 virtual void UnloadKernel(const KernelBase *kernel) {} in UnloadKernel()
|
| D | trace_listener.h | 51 const KernelBase& kernel, in LaunchSubmit()
|
| D | stream_executor_pimpl.h | 105 port::Status GetKernel(const MultiKernelLoaderSpec &spec, KernelBase *kernel);
108 void UnloadKernel(const KernelBase *kernel);
475 const BlockDim &block_dims, const KernelBase &kernel,
|
| D | stream_executor_pimpl.cc | 193 KernelBase *kernel) { in GetKernel()
197 void StreamExecutor::UnloadKernel(const KernelBase *kernel) { in UnloadKernel()
467 const KernelBase &kernel, in Launch()
|
| /external/gemmlowp/internal/ |
| D | kernel_neon.h | 34 struct NEON_32_Kernel12x4Depth2 : KernelBase {
294 struct NEON_32_Kernel12x4Depth2Assuming12BitProducts : KernelBase {
669 struct NEON_32bit_GEMM_Int8Operands_LhsNonzero : KernelBase {
943 struct NEON_64bit_GEMM_Int8Operands_LhsNonzero : KernelBase {
1291 struct NEON_64_Kernel12x8Depth2 : KernelBase {
1665 struct NEON_64_Kernel12x8Depth4_dotprod : KernelBase {
|
| D | compute.h | 34 const KernelBase& kernel_;
42 ComputeImpl(const KernelBase& _kernel, const BlockParams& _block_params, in ComputeImpl()
106 void Compute(const KernelBase& kernel, const BlockParams& block_params, in Compute()
|
| D | kernel_sse.h | 32 struct SSE4_32_Kernel4x4Depth2 : KernelBase {
207 struct SSE4_64_Kernel12x4Depth2 : KernelBase {
|
| D | kernel_msa.h | 44 struct MSA_Kernel12x8Depth2 : KernelBase {
281 struct MSA_GEMM_Int8Operands_LhsNonzero : KernelBase {
|
| D | kernel.h | 217 struct KernelBase { struct
228 virtual ~KernelBase() {} in ~KernelBase() argument
|
| D | kernel_avx.h | 31 struct AVX2_64_Kernel24x8Depth2 : KernelBase {
|
| D | multi_thread_gemm.h | 469 GemmWithPackedRhsTask(GemmContextType* _context, const KernelBase& _kernel, in GemmWithPackedRhsTask()
527 const KernelBase& kernel;
636 void MultiThreadGemm(GemmContextType* context, const KernelBase& kernel, in MultiThreadGemm()
|
| D | kernel_reference.h | 33 struct ReferenceKernel : KernelBase {
|
| D | single_thread_gemm.h | 67 const KernelBase& kernel, in SingleThreadGemm()
|
| /external/tensorflow/tensorflow/stream_executor/gpu/ |
| D | gpu_kernel.h | 92 inline const GpuKernel* AsGpuKernel(const KernelBase* kernel) { in AsGpuKernel()
98 inline GpuKernel* AsGpuKernel(KernelBase* kernel) { in AsGpuKernel()
|
| D | gpu_executor.h | 65 KernelBase* kernel) override;
67 void UnloadKernel(const KernelBase* kernel) override;
73 const BlockDim& block_dims, const KernelBase& k,
267 void VlogOccupancyInfo(const KernelBase& kernel, const ThreadDim& thread_dims,
303 std::unordered_map<const KernelBase*, const void*> kernel_to_gpu_binary_
|
| /external/tensorflow/tensorflow/compiler/xla/service/gpu/ |
| D | stream_executor_util.h | 69 StatusOr<std::unique_ptr<se::KernelBase>> CreateKernel(
74 Status ExecuteKernelOnStream(const se::KernelBase& kernel,
|
| D | kernel_thunk.cc | 55 std::unique_ptr<se::KernelBase> kernel, in Initialize()
91 const se::KernelBase* kernel = nullptr; in ExecuteOnStream()
|
| D | kernel_thunk.h | 79 std::unordered_map<se::StreamExecutor*, std::unique_ptr<se::KernelBase>>
|
| D | stream_executor_util.cc | 194 StatusOr<std::unique_ptr<se::KernelBase>> CreateKernel( in CreateKernel()
205 auto kernel_base = absl::make_unique<se::KernelBase>(stream_exec); in CreateKernel()
210 Status ExecuteKernelOnStream(const se::KernelBase& kernel, in ExecuteKernelOnStream()
|
| /external/tensorflow/tensorflow/stream_executor/host/ |
| D | host_gpu_executor.h | 54 KernelBase *kernel) override { in GetKernel()
58 const BlockDim &block_dims, const KernelBase &kernel, in Launch()
|
| /external/tensorflow/tensorflow/compiler/xla/service/interpreter/ |
| D | executor.h | 61 KernelBase *kernel) override { in GetKernel()
65 const BlockDim &block_dims, const KernelBase &kernel, in Launch()
|
| /external/tensorflow/tensorflow/stream_executor/rocm/ |
| D | rocm_gpu_executor.cc | 144 void GpuExecutor::UnloadKernel(const KernelBase* kernel) { in UnloadKernel()
239 KernelBase* kernel) { in GetKernel()
298 const KernelBase& kernel, in Launch()
419 void GpuExecutor::VlogOccupancyInfo(const KernelBase& kernel, in VlogOccupancyInfo()
|
| /external/tensorflow/tensorflow/stream_executor/cuda/ |
| D | cuda_gpu_executor.cc | 264 KernelBase* kernel) { in GetKernel()
332 void GpuExecutor::UnloadKernel(const KernelBase* kernel) { in UnloadKernel()
402 const KernelBase& kernel, in Launch()
439 void GpuExecutor::VlogOccupancyInfo(const KernelBase& kernel, in VlogOccupancyInfo()
|