[Triton-MLIR] Replace triton.extract_slice with tensor.extract_slice and support more general tensor slicing (#837)
## Features - Allow taking a block of tensor slice, as long as each dimension is contiguous (unit stride). - Fix some problems in `insert_slice_async`'s semantic. - More general verification for ops that return shared layout encoding. ## Known Limitations - `insert_slice_async` still uses the old semantic. May submit another PR later to support similar semantic like `tensor.extract_slice`. - No encoding verification for `tensor.extract_slice`. - 3d tensor ops are broken. - Strided accesses are not allowed. - May cause a little performance slowdown since we are passing strides as values but not constants (e.g., int). It would be difficult to pass strides as attributes when we have control flows. A block argument is possible to accept tensors with different strides.
This commit is contained in:
@@ -2,6 +2,7 @@
|
||||
#include "triton/Dialect/TritonGPU/IR/Dialect.h"
|
||||
|
||||
#include "mlir/Dialect/GPU/GPUDialect.h"
|
||||
#include "mlir/Dialect/Tensor/IR/Tensor.h"
|
||||
|
||||
namespace mlir {
|
||||
|
||||
@@ -43,8 +44,7 @@ void MembarAnalysis::dfsOperation(Operation *operation,
|
||||
void MembarAnalysis::transfer(Operation *op, RegionInfo *regionInfo,
|
||||
OpBuilder *builder) {
|
||||
if (isa<scf::ForOp>(op) || isa<scf::IfOp>(op) || isa<scf::YieldOp>(op) ||
|
||||
isa<triton::gpu::ExtractSliceOp>(op) ||
|
||||
isa<triton::gpu::AllocTensorOp>(op)) {
|
||||
isa<tensor::ExtractSliceOp>(op) || isa<triton::gpu::AllocTensorOp>(op)) {
|
||||
// Do not insert barriers before control flow operations and
|
||||
// alloc/extract/insert
|
||||
// alloc is an allocation op without memory write.
|
||||
|
Reference in New Issue
Block a user