Home
last modified time | relevance | path

Searched refs:num_threads_x (Results 1 – 3 of 3) sorted by relevance

/external/tensorflow/tensorflow/compiler/xla/service/gpu/
Dkernel_mapping_scheme.h92 int64 num_threads_x, IndexingOrder indexing_order,
96 num_threads_x_(num_threads_x),
Dir_emitter_unnested.cc4010 int64 num_threads_x = mapping_scheme.GetNumThreadsX(); in EmitTile() local
4014 int64 x_num_steps = tile_size_x / num_threads_x; in EmitTile()
4023 mapping_scheme.GetIndexingOrder() == kLinearIndexingX ? 1 : num_threads_x; in EmitTile()
4287 int64 num_threads_x = mapping_scheme.GetNumThreadsX(); in EmitPrologueForReduction() local
4301 CHECK_EQ(num_threads_x, mapping_scheme.GetNumThreadsY()); in EmitPrologueForReduction()
4304 llvm::ArrayType::get(primitive_type, num_threads_x + 1), in EmitPrologueForReduction()
4305 num_threads_x), in EmitPrologueForReduction()
4681 int64 threads_per_block, llvm::Type* index_ty, int64 num_threads_x) { in EmitThreadIdInfo() argument
4686 llvm::Value* num_threads_x_v = constant(num_threads_x); in EmitThreadIdInfo()
5321 int64 num_threads_x = [&] { in ComputeReductionCodegenInfo() local
[all …]
Dir_emitter_unnested.h710 int64 num_threads_x);