en Copyright 2015 Java http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#6c3c5f80 [mlir][memref] Improve type inference for rank-reducing subviewsThe result shape of a rank-reducing subview cannot be inferred in the general case. Just the result rank is not enough. The only thing that we can infer is the layout map.This change also improves the bufferization patterns of tensor.extract_slice and tensor.insert_slice to fully support rank-reducing operations.Differential Revision: https://reviews.llvm.org/D129144 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Tue, 05 Jul 2022 14:39:29 +0000 Matthias Springer <springerm@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#cc6462a4 [mlir][tensor][bufferize][NFC] Clean up test caseInsert -split-input-file flag to make the test cases more stable.Differential Revision: https://reviews.llvm.org/D129143 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Tue, 05 Jul 2022 13:36:02 +0000 Matthias Springer <springerm@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#6eb0f8e2 [mlir][MemRef] Fix a crash when expanding a scalar shapeIn this case the reassociation is empty, yielding no strides for theresult type.Differential Revision: https://reviews.llvm.org/D127232 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Tue, 07 Jun 2022 17:30:10 +0000 Benjamin Kramer <benny.kra@googlemail.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#e287d647 [mlir] Add translation from tensor.reshape to memref.reshapeThis patch augments the `tensor-bufferize` pass by adding a conversionrule to translate ReshapeOp from the `tensor` dialect to the `memref`dialect, in addition to adding a unit test to validate the translation.Reviewed By: springermDifferential Revision: https://reviews.llvm.org/D125031 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Mon, 09 May 2022 15:41:21 +0000 Ashay Rane <ashay@users.noreply.github.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#1cddcfdc Fix CollapsedLayoutMap for dim size 1 caseThis change fixes `CollapsedLayoutMap` for cases where the collapseddims are size 1. The cases where inner most dims are size 1 andnoncontiguous can be represented by the strided form and therefore canbe allowed. For such cases, the new stride should be of the next entryin an association whose dimension is not size 1. If the next entry isdynamic, it's not possible to decide which stride to use at compilationtime and the stride is set to dynamic.Differential Revision: https://reviews.llvm.org/D124137 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Tue, 19 Apr 2022 00:50:30 +0000 Yi Zhang <cathyzhyi@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#d820acdd [mlir][bufferize][NFC] Use custom walk instead of GreedyPatternRewriterThe bufferization driver was previously using a GreedyPatternRewriter. This was problematic because bufferization must traverse ops top-to-bottom. The GreedyPatternRewriter was previously configured via `useTopDownTraversal`, but this was a hack; this API was just meant for performance improvements and should not affect the result of the rewrite.BEGIN_PUBLICNo public commit message needed.END_PUBLICDifferential Revision: https://reviews.llvm.org/D123618 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Fri, 22 Apr 2022 09:03:28 +0000 Matthias Springer <springerm@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#c48e3a13 [mlir][NFC] Update textual references of `func` to `func.func` in Tensor/Tosa/Vector testsThe special case parsing of `func` operations is being removed. List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Wed, 20 Apr 2022 23:19:48 +0000 River Riddle <riddleriver@gmail.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#d7a9bf91 [mlir][tensor] Fix verifier and bufferization of collapse_shapeInsert a buffer copy unless the dims are guaranteed to be collapsible. In the verifier, accept collapses unless they are guaranteed to be non-collapsible.Differential Revision: https://reviews.llvm.org/D123316 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Fri, 08 Apr 2022 09:04:00 +0000 Matthias Springer <springerm@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#af371f9f Reland [GreedPatternRewriter] Preprocess constants while building worklist when not processing top downReland Note: Adds a fix to properly mark a commutative operation as folded if we change the order of its operands. This was uncovered by the fact that we no longer re-process constants.This avoids accidentally reversing the order of constants during successiveapplication, e.g. when running the canonicalizer. This helps reduce the numberof iterations, and also avoids unnecessary changes to input IR.Fixes #51892Differential Revision: https://reviews.llvm.org/D122692 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Thu, 07 Apr 2022 01:52:39 +0000 River Riddle <riddleriver@gmail.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#fc8f465a [mlir][MemRef] Allow transposed layouts in ExpandShapeOp.https://reviews.llvm.org/D122641 introduced fixes to the ExpandShapeOp verifierbut also introduced an artificial layout limitation that prevents the consideration of transposed layouts.This revision fixes the omissions and reimplements the logic using saturated arithmetic which is moreidiomatic and avoids leaking internal implementation details.Tests cases are added for transposed layouts.Reviewed By: springermDifferential Revision: https://reviews.llvm.org/D122845 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Wed, 06 Apr 2022 07:57:03 +0000 Nicolas Vasilache <ntv@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#73c0333d [mlir][tensor][bufferize] Support 0-d collapse_shape with offsetDifferential Revision: https://reviews.llvm.org/D122901 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Fri, 01 Apr 2022 13:07:00 +0000 Matthias Springer <springerm@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#ba43d6f8 Revert "[GreedPatternRewriter] Preprocess constants while building worklist when not processing top down"This reverts commit 59bbc7a0851b6e0054bb3ed47df0958822f08880.This exposes an issue breaking the contract of`applyPatternsAndFoldGreedily` where we "converge" without applyingremaining patterns. List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Fri, 01 Apr 2022 06:16:00 +0000 Mehdi Amini <joker.eph@gmail.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#59bbc7a0 [GreedPatternRewriter] Preprocess constants while building worklist when not processing top downThis avoids accidentally reversing the order of constants during successiveapplication, e.g. when running the canonicalizer. This helps reduce the numberof iterations, and also avoids unnecessary changes to input IR.Fixes #51892Differential Revision: https://reviews.llvm.org/D122692 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Tue, 29 Mar 2022 23:28:38 +0000 River Riddle <riddleriver@gmail.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#51df6238 [mlir][tensor] Fix bufferization of CollapseShapeOp / ExpandShapeOpInfer a tighter MemRef type instead of always falling back to the most dynamic MemRef type. This is inefficient and caused op verification errors.Differential Revision: https://reviews.llvm.org/D122649 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Thu, 31 Mar 2022 08:04:38 +0000 Matthias Springer <springerm@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#39ec46bd [mlir][bufferize] Extract buffer hoisting into separate functionThis improves the modularity of the bufferization.From now on, all ops that do not implement BufferizableOpInterface are considered hoisting barriers. Previously, all ops that do not implement the interface were not considered barriers and such ops had to be marked as barriers explicitly. This was unsafe because we could've hoisted across unknown ops where it was not safe to hoist.As a side effect, this allows for cleaning up AffineBufferizableOpInterfaceImpl. This build unit no longer needed and can be deleted.Differential Revision: https://reviews.llvm.org/D121519 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Tue, 15 Mar 2022 12:22:21 +0000 Matthias Springer <springerm@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#e6f69161 [mlir][bufferize] Support tensor.expand_shape and tensor.collapse_shapeDifferential Revision: https://reviews.llvm.org/D112512 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Tue, 15 Feb 2022 10:44:31 +0000 Matthias Springer <springerm@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#daf18108 [mlir][tensor] Replace tensor-bufferize with BufferizableOpInterface implThis commit switches the `tensor-bufferize` pass over to BufferizableOpInterface-based bufferization.Differential Revision: https://reviews.llvm.org/D118246 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Thu, 27 Jan 2022 10:18:59 +0000 Matthias Springer <springerm@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#f77e9f87 [mlir] Extend `tensor.from_elements` to support N-D case.RFC: https://llvm.discourse.group/t/rfc-extend-tensor-fromelementsop-to-n-d/4715Differential Revision: https://reviews.llvm.org/D115821 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Thu, 16 Dec 2021 13:42:27 +0000 Alexander Belyaev <pifon@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#a82a19c1 [mlir] Add a missing pattern to bufferize tensor.rank.Differential Revision: https://reviews.llvm.org/D115745 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Tue, 14 Dec 2021 18:58:40 +0000 Alexander Belyaev <pifon@google.com> http://172.16.0.5:8080/history/llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir#57470abc [mlir] Move memref.[tensor_load|buffer_cast|clone] to "bufferization" dialect.https://llvm.discourse.group/t/rfc-dialect-for-bufferization-related-ops/4712Differential Revision: https://reviews.llvm.org/D114552 List of files: /llvm-project-15.0.7/mlir/test/Dialect/Tensor/bufferize.mlir Thu, 25 Nov 2021 10:42:16 +0000 Alexander Belyaev <pifon@google.com>