gpu.h File Reference

Gpu operations for TIRx. More...

#include <tvm/ir/op.h>
#include <tvm/ir/prim/expr.h>

Include dependency graph for gpu.h:

Go to the source code of this file.

Namespaces
namespace	tvm
	An object that builds and maintains block scope and StmtSref mapping for Dependence analysis.
namespace	tvm::tirx
	Backend-defined execution hierarchy nodes shared by layouts and tile calls.

Functions
TVM_TIRX_DLL const Op &	tvm::tirx::gpu_thread_return_op ()
	Return from a GPU thread without returning a function value.
TVM_TIRX_DLL PrimExpr	tvm::tirx::gpu_thread_return (Location loc=UnknownLoc())
	Return from a GPU thread without returning a function value.
TVM_TIRX_DLL const Op &	tvm::tirx::gpu_thread_filter_op ()
	Mark an opaque predicate that selects one thread on an active-set axis.
TVM_TIRX_DLL const Op &	tvm::tirx::gpu_active_thread_selector_op ()
	Analysis-only active-thread selector.
TVM_TIRX_DLL const Op &	tvm::tirx::gpu_thread_invariant_op ()
	Mark a condition to be thread invariant.
TVM_TIRX_DLL const Op &	tvm::tirx::gpu_storage_sync_op ()
	Synchronize accesses in a storage scope.
TVM_TIRX_DLL const Op &	tvm::tirx::gpu_warp_shuffle_op ()
	Read the value from a selected lane in the warp.
TVM_TIRX_DLL const Op &	tvm::tirx::gpu_warp_shuffle_up_op ()
	Read the value from a lower lane in the warp.
TVM_TIRX_DLL const Op &	tvm::tirx::gpu_warp_shuffle_down_op ()
	Read the value from a higher lane in the warp.
TVM_TIRX_DLL const Op &	tvm::tirx::gpu_warp_shuffle_xor_op ()
	Read the value from the lane selected by an XOR mask.
TVM_TIRX_DLL const Op &	tvm::tirx::gpu_warp_activemask_op ()
	Return the bit mask of active lanes in the warp.
TVM_TIRX_DLL const Op &	tvm::tirx::gpu_thread_allreduce_op ()
	Cross-thread reduction with an explicit typed combiner and identities.
TVM_TIRX_DLL const Op &	tvm::tirx::gpu_dp4a_op ()
	Dot product of two int8x4 vectors and add an optional accumulator.

Detailed Description

Gpu operations for TIRx.