| /external/tensorflow/tensorflow/stream_executor/ |
| D | kernel.cc | 61 KernelBase::KernelBase(KernelBase &&from) in KernelBase() function in stream_executor::KernelBase
70 KernelBase::KernelBase(StreamExecutor *parent) in KernelBase() function in stream_executor::KernelBase
74 KernelBase::KernelBase(StreamExecutor *parent, in KernelBase() function in stream_executor::KernelBase
78 KernelBase::~KernelBase() { in ~KernelBase()
84 unsigned KernelBase::Arity() const { return implementation_->Arity(); } in Arity()
86 void KernelBase::SetPreferredCacheConfig(KernelCacheConfig config) { in SetPreferredCacheConfig()
90 KernelCacheConfig KernelBase::GetPreferredCacheConfig() const { in GetPreferredCacheConfig()
97 void KernelBase::set_name(absl::string_view name) { in set_name()
|
| D | kernel.h | 135 class KernelBase {
137 KernelBase(KernelBase &&from);
143 explicit KernelBase(StreamExecutor *parent);
146 KernelBase(StreamExecutor *parent, internal::KernelInterface *implementation);
150 ~KernelBase();
196 SE_DISALLOW_COPY_AND_ASSIGN(KernelBase);
497 class TypedKernel : public KernelBase {
502 explicit TypedKernel(StreamExecutor *parent) : KernelBase(parent) {}
507 : KernelBase(parent, implementation) {}
|
| D | stream_executor_internal.h | 183 KernelBase *kernel) { in GetKernel()
192 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 bool GetKernel(const MultiKernelLoaderSpec &spec, KernelBase *kernel);
108 void UnloadKernel(const KernelBase *kernel);
453 const BlockDim &block_dims, const KernelBase &kernel,
|
| D | stream_executor_pimpl.cc | 235 KernelBase *kernel) { in GetKernel()
239 void StreamExecutor::UnloadKernel(const KernelBase *kernel) { in UnloadKernel()
481 const KernelBase &kernel, in Launch()
|
| /external/gemmlowp/internal/ |
| D | kernel_neon.h | 34 struct NEON_32_Kernel12x4Depth2 : KernelBase {
293 struct NEON_32_Kernel12x4Depth2Assuming12BitProducts : KernelBase {
667 struct NEON_32bit_GEMM_Int8Operands_LhsNonzero : KernelBase {
929 struct NEON_64bit_GEMM_Int8Operands_LhsNonzero : KernelBase {
1265 struct NEON_64_Kernel12x8Depth2 : 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 {
206 struct SSE4_64_Kernel12x4Depth2 : KernelBase {
|
| D | kernel.h | 205 struct KernelBase { struct
216 virtual ~KernelBase() {} in ~KernelBase() argument
|
| D | kernel_msa.h | 44 struct MSA_Kernel12x8Depth2 : KernelBase {
|
| D | multi_thread_gemm.h | 441 GemmWithPackedRhsTask(GemmContextType* _context, const KernelBase& _kernel, in GemmWithPackedRhsTask()
499 const KernelBase& kernel;
608 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_executor.h | 65 KernelBase* kernel) override;
67 void UnloadKernel(const KernelBase* kernel) override;
73 const BlockDim& block_dims, const KernelBase& k,
269 void VlogOccupancyInfo(const KernelBase& kernel, const ThreadDim& thread_dims,
304 std::unordered_map<const KernelBase*, const void*> kernel_to_gpu_binary_
|
| D | gpu_kernel.h | 92 inline const GpuKernel* AsGpuKernel(const KernelBase* kernel) { in AsGpuKernel()
98 inline GpuKernel* AsGpuKernel(KernelBase* kernel) { in AsGpuKernel()
|
| /external/tensorflow/tensorflow/compiler/xla/service/gpu/ |
| D | kernel_thunk.cc | 60 it = kernel_cache_.emplace(executor, se::KernelBase(executor)).first; in Initialize()
80 const se::KernelBase* kernel = nullptr; in ExecuteOnStream()
|
| D | kernel_thunk.h | 91 std::unordered_map<se::StreamExecutor*, se::KernelBase> kernel_cache_
|
| /external/tensorflow/tensorflow/compiler/xla/service/interpreter/ |
| D | executor.h | 62 KernelBase *kernel) override { in GetKernel()
66 const BlockDim &block_dims, const KernelBase &kernel, in Launch()
|
| /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/stream_executor/rocm/ |
| D | rocm_gpu_executor.cc | 139 void GpuExecutor::UnloadKernel(const KernelBase* kernel) { in UnloadKernel()
222 KernelBase* kernel) { in GetKernel()
287 const BlockDim& block_dims, 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 | 268 KernelBase* kernel) { in GetKernel()
344 void GpuExecutor::UnloadKernel(const KernelBase* kernel) { in UnloadKernel()
421 const BlockDim& block_dims, const KernelBase& kernel, in Launch()
466 void GpuExecutor::VlogOccupancyInfo(const KernelBase& kernel, in VlogOccupancyInfo()
|