[mlir][linalg] Take artificial padding into account for pack/unpack folding. #150272
There are no files selected for viewing
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
|
|
@@ -4490,6 +4490,29 @@ Speculation::Speculatability ElementwiseOp::getSpeculatability() { | |
| //===----------------------------------------------------------------------===// | ||
| // PackOp/UnPackOp Common | ||
| //===----------------------------------------------------------------------===// | ||
|
|
||
| template <typename OpTy> | ||
| SmallVector<int64_t> | ||
| getPackedOuterShapeWithoutTransposition(OpTy packOrUnPack) { | ||
| RankedTensorType packedType = (std::is_same<OpTy, PackOp>::value) | ||
| ? packOrUnPack.getDestType() | ||
| : packOrUnPack.getSourceType(); | ||
| RankedTensorType unpackedType = (std::is_same<OpTy, PackOp>::value) | ||
| ? packOrUnPack.getSourceType() | ||
| : packOrUnPack.getDestType(); | ||
| SmallVector<int64_t> result( | ||
| packedType.getShape().take_front(unpackedType.getRank())); | ||
| if (!packOrUnPack.getOuterDimsPerm().empty()) { | ||
| applyPermutationToVector( | ||
| result, invertPermutationVector(packOrUnPack.getOuterDimsPerm())); | ||
| } | ||
| return result; | ||
| } | ||
| template SmallVector<int64_t> | ||
| getPackedOuterShapeWithoutTransposition<PackOp>(PackOp); | ||
| template SmallVector<int64_t> | ||
| getPackedOuterShapeWithoutTransposition<UnPackOp>(UnPackOp); | ||
|
|
||
| // Given the (potentially) updated packed type, `newPackedTy`, generates an | ||
| // updated mixed-tile-sizes attribute. A tile size is updated only | ||
| // when: | ||
|
|
@@ -5447,11 +5470,7 @@ LogicalResult UnPackOp::canonicalize(UnPackOp unPackOp, | |
| if (unPackOp->hasOneUse()) { | ||
| auto extractSliceUser = | ||
| dyn_cast<tensor::ExtractSliceOp>(*unPackOp->getUsers().begin()); | ||
| if (extractSliceUser && unPackOp.canFoldSliceOp(extractSliceUser)) { | ||
| OpBuilder::InsertionGuard g(rewriter); | ||
| rewriter.setInsertionPoint(unPackOp); | ||
| auto newDest = rewriter.create<tensor::ExtractSliceOp>( | ||
|
|
@@ -5494,6 +5513,32 @@ LogicalResult UnPackOp::canonicalize(UnPackOp unPackOp, | |
| return failure(); | ||
| } | ||
|
|
||
| bool UnPackOp::canFoldSliceOp(tensor::ExtractSliceOp sliceOp) { | ||
| // Rank-reduced folding is not supported. | ||
| if (sliceOp.getResultType().getRank() != this->getDestType().getRank()) | ||
| return false; | ||
| if (!areAllConstantIntValue(sliceOp.getMixedOffsets(), 0) || | ||
| !areAllConstantIntValue(sliceOp.getMixedStrides(), 1)) | ||
| return false; | ||
| RankedTensorType unpackedType = sliceOp.getResultType(); | ||
hanhanW marked this conversation as resolved.
Outdated
Show resolved
Hide resolved
|
||
| SmallVector<int64_t> outerShapeWithoutTranspose = | ||
| getPackedOuterShapeWithoutTransposition(*this); | ||
| for (auto [pos, tileSize] : | ||
| llvm::zip_equal(this->getInnerDimsPos(), this->getStaticInnerTiles())) { | ||
| if (unpackedType.isDynamicDim(pos)) | ||
| return false; | ||
| if (ShapedType::isDynamic(outerShapeWithoutTranspose[pos])) | ||
| return false; | ||
| if (ShapedType::isDynamic(tileSize)) | ||
| return false; | ||
| int64_t paddingSize = outerShapeWithoutTranspose[pos] * tileSize - | ||
| unpackedType.getDimSize(pos); | ||
| if (paddingSize >= tileSize) | ||
| return false; | ||
|
Contributor
There was a problem hiding this comment. It would be good to add a test where a non-trailing dim is "sliced", e.g. func.func @fold_extract_slice_into_unpack(
%src : tensor<28x2x1x16x16xf32>, %dest : tensor<28x28x15xf32>, %size : index
) -> tensor<28x16x15xf32> {
%unpack = linalg.unpack %src
outer_dims_perm = [0, 1, 2]
inner_dims_pos = [1, 2]
inner_tiles = [16, 16]
into %dest : tensor<28x2x1x16x16xf32> -> tensor<28x28x15xf32>
%extracted_slice = tensor.extract_slice %unpack
[0, 0, 0] [28, 16, 15] [1, 1, 1] : tensor<28x28x15xf32> to tensor<28x16x15xf32>
return %extracted_slice : tensor<28x16x15xf32>
}
Contributor
There was a problem hiding this comment. Also, if I'm not missing something, I don't see tests for when we hit this case where we would need artificial padding. It would be nice to add these negative cases :)
Contributor
Author
There was a problem hiding this comment. Thanks, I missed that! I think @banach-space's example is a negative test. Anyway, I added two more tests that should address your comments. |
||
| } | ||
| return true; | ||
| } | ||
|
|
||
| bool UnPackOp::isLikeUnPad() { | ||
| RankedTensorType packedTensorType = getSourceType(); | ||
| return isLikePadUnPad(*this, packedTensorType); | ||
|
|
||
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Uh oh!
There was an error while loading. Please reload this page.