[mlir] add a fluent API to GreedyRewriterConfig #137122
Merged
Conversation
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
ftynse requested review from MaheshRavishankar, antiagainst, kuhar, matthias-springer and nicolasvasilache as code owners 
llvmbot added mlir:core 
Member
| @llvm/pr-subscribers-flang-fir-hlfir @llvm/pr-subscribers-mlir-core Author: Oleksandr "Alex" Zinenko (ftynse) ChangesThis is similar to other configuration objects used across MLIR. Rename some fields to better reflect that they are no longer booleans. Reland 04d2611 / #132253. Patch is 48.43 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/137122.diff 30 Files Affected:
diff --git a/flang/lib/Optimizer/CodeGen/LowerRepackArrays.cpp b/flang/lib/Optimizer/CodeGen/LowerRepackArrays.cpp index 7deed3d44ae5b..7fb713ff1a6c7 100644 --- a/flang/lib/Optimizer/CodeGen/LowerRepackArrays.cpp +++ b/flang/lib/Optimizer/CodeGen/LowerRepackArrays.cpp @@ -357,8 +357,8 @@ class LowerRepackArraysPass patterns.insert<PackArrayConversion>(context); patterns.insert<UnpackArrayConversion>(context); mlir::GreedyRewriteConfig config; - config.enableRegionSimplification = - mlir::GreedySimplifyRegionLevel::Disabled; + config.setRegionSimplificationLevel( + mlir::GreedySimplifyRegionLevel::Disabled); (void)applyPatternsGreedily(module, std::move(patterns), config); } diff --git a/flang/lib/Optimizer/HLFIR/Transforms/InlineElementals.cpp b/flang/lib/Optimizer/HLFIR/Transforms/InlineElementals.cpp index b68fe6ee0c747..c42b895946d19 100644 --- a/flang/lib/Optimizer/HLFIR/Transforms/InlineElementals.cpp +++ b/flang/lib/Optimizer/HLFIR/Transforms/InlineElementals.cpp @@ -119,8 +119,8 @@ class InlineElementalsPass mlir::GreedyRewriteConfig config; // Prevent the pattern driver from merging blocks. - config.enableRegionSimplification = - mlir::GreedySimplifyRegionLevel::Disabled; + config.setRegionSimplificationLevel( + mlir::GreedySimplifyRegionLevel::Disabled); mlir::RewritePatternSet patterns(context); patterns.insert<InlineElementalConversion>(context); diff --git a/flang/lib/Optimizer/HLFIR/Transforms/InlineHLFIRAssign.cpp b/flang/lib/Optimizer/HLFIR/Transforms/InlineHLFIRAssign.cpp index 249976d5509b0..6e209cce07ad4 100644 --- a/flang/lib/Optimizer/HLFIR/Transforms/InlineHLFIRAssign.cpp +++ b/flang/lib/Optimizer/HLFIR/Transforms/InlineHLFIRAssign.cpp @@ -135,8 +135,8 @@ class InlineHLFIRAssignPass mlir::GreedyRewriteConfig config; // Prevent the pattern driver from merging blocks. - config.enableRegionSimplification = - mlir::GreedySimplifyRegionLevel::Disabled; + config.setRegionSimplificationLevel( + mlir::GreedySimplifyRegionLevel::Disabled); mlir::RewritePatternSet patterns(context); patterns.insert<InlineHLFIRAssignConversion>(context); diff --git a/flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIRIntrinsics.cpp b/flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIRIntrinsics.cpp index 7c0fcba806869..31e5bc1193e22 100644 --- a/flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIRIntrinsics.cpp +++ b/flang/lib/Optimizer/HLFIR/Transforms/LowerHLFIRIntrinsics.cpp @@ -557,8 +557,8 @@ class LowerHLFIRIntrinsics // Pattern rewriting only requires that the resulting IR is still valid mlir::GreedyRewriteConfig config; // Prevent the pattern driver from merging blocks - config.enableRegionSimplification = - mlir::GreedySimplifyRegionLevel::Disabled; + config.setRegionSimplificationLevel( + mlir::GreedySimplifyRegionLevel::Disabled); if (mlir::failed( mlir::applyPatternsGreedily(module, std::move(patterns), config))) { diff --git a/flang/lib/Optimizer/HLFIR/Transforms/OptimizedBufferization.cpp b/flang/lib/Optimizer/HLFIR/Transforms/OptimizedBufferization.cpp index 79aabd2981e1a..2f6ee2592a84f 100644 --- a/flang/lib/Optimizer/HLFIR/Transforms/OptimizedBufferization.cpp +++ b/flang/lib/Optimizer/HLFIR/Transforms/OptimizedBufferization.cpp @@ -875,8 +875,8 @@ class OptimizedBufferizationPass mlir::GreedyRewriteConfig config; // Prevent the pattern driver from merging blocks - config.enableRegionSimplification = - mlir::GreedySimplifyRegionLevel::Disabled; + config.setRegionSimplificationLevel( + mlir::GreedySimplifyRegionLevel::Disabled); mlir::RewritePatternSet patterns(context); // TODO: right now the patterns are non-conflicting, diff --git a/flang/lib/Optimizer/HLFIR/Transforms/SimplifyHLFIRIntrinsics.cpp b/flang/lib/Optimizer/HLFIR/Transforms/SimplifyHLFIRIntrinsics.cpp index e9d820adbd22b..1dea7d89e180d 100644 --- a/flang/lib/Optimizer/HLFIR/Transforms/SimplifyHLFIRIntrinsics.cpp +++ b/flang/lib/Optimizer/HLFIR/Transforms/SimplifyHLFIRIntrinsics.cpp @@ -2132,8 +2132,8 @@ class SimplifyHLFIRIntrinsics mlir::GreedyRewriteConfig config; // Prevent the pattern driver from merging blocks - config.enableRegionSimplification = - mlir::GreedySimplifyRegionLevel::Disabled; + config.setRegionSimplificationLevel( + mlir::GreedySimplifyRegionLevel::Disabled); mlir::RewritePatternSet patterns(context); patterns.insert<TransposeAsElementalConversion>(context); diff --git a/flang/lib/Optimizer/Passes/Pipelines.cpp b/flang/lib/Optimizer/Passes/Pipelines.cpp index 81ff6bf9b2c6a..7a06a27748ebd 100644 --- a/flang/lib/Optimizer/Passes/Pipelines.cpp +++ b/flang/lib/Optimizer/Passes/Pipelines.cpp @@ -35,7 +35,8 @@ void addNestedPassToAllTopLevelOperationsConditionally( void addCanonicalizerPassWithoutRegionSimplification(mlir::OpPassManager &pm) { mlir::GreedyRewriteConfig config; - config.enableRegionSimplification = mlir::GreedySimplifyRegionLevel::Disabled; + config.setRegionSimplificationLevel( + mlir::GreedySimplifyRegionLevel::Disabled); pm.addPass(mlir::createCanonicalizerPass(config)); } @@ -163,7 +164,8 @@ void createDefaultFIROptimizerPassPipeline(mlir::PassManager &pm, // simplify the IR mlir::GreedyRewriteConfig config; - config.enableRegionSimplification = mlir::GreedySimplifyRegionLevel::Disabled; + config.setRegionSimplificationLevel( + mlir::GreedySimplifyRegionLevel::Disabled); pm.addPass(mlir::createCSEPass()); fir::addAVC(pm, pc.OptLevel); addNestedPassToAllTopLevelOperations<PassConstructor>( diff --git a/flang/lib/Optimizer/Transforms/AssumedRankOpConversion.cpp b/flang/lib/Optimizer/Transforms/AssumedRankOpConversion.cpp index d0bd67a236419..6af1cb988a4c1 100644 --- a/flang/lib/Optimizer/Transforms/AssumedRankOpConversion.cpp +++ b/flang/lib/Optimizer/Transforms/AssumedRankOpConversion.cpp @@ -152,8 +152,8 @@ class AssumedRankOpConversion patterns.insert<ReboxAssumedRankConv>(context, &symbolTable, kindMap); patterns.insert<IsAssumedSizeConv>(context, &symbolTable, kindMap); mlir::GreedyRewriteConfig config; - config.enableRegionSimplification = - mlir::GreedySimplifyRegionLevel::Disabled; + config.setRegionSimplificationLevel( + mlir::GreedySimplifyRegionLevel::Disabled); (void)applyPatternsGreedily(mod, std::move(patterns), config); } }; diff --git a/flang/lib/Optimizer/Transforms/ConstantArgumentGlobalisation.cpp b/flang/lib/Optimizer/Transforms/ConstantArgumentGlobalisation.cpp index 562f3058f20f3..239a7cdaa4cf2 100644 --- a/flang/lib/Optimizer/Transforms/ConstantArgumentGlobalisation.cpp +++ b/flang/lib/Optimizer/Transforms/ConstantArgumentGlobalisation.cpp @@ -168,9 +168,9 @@ class ConstantArgumentGlobalisationOpt auto *context = &getContext(); mlir::RewritePatternSet patterns(context); mlir::GreedyRewriteConfig config; - config.enableRegionSimplification = - mlir::GreedySimplifyRegionLevel::Disabled; - config.strictMode = mlir::GreedyRewriteStrictness::ExistingOps; + config.setRegionSimplificationLevel( + mlir::GreedySimplifyRegionLevel::Disabled); + config.setStrictness(mlir::GreedyRewriteStrictness::ExistingOps); patterns.insert<CallOpRewriter>(context, *di); if (mlir::failed( diff --git a/flang/lib/Optimizer/Transforms/SimplifyFIROperations.cpp b/flang/lib/Optimizer/Transforms/SimplifyFIROperations.cpp index 212de2f2286db..6d106046b70f2 100644 --- a/flang/lib/Optimizer/Transforms/SimplifyFIROperations.cpp +++ b/flang/lib/Optimizer/Transforms/SimplifyFIROperations.cpp @@ -205,7 +205,8 @@ void SimplifyFIROperationsPass::runOnOperation() { fir::populateSimplifyFIROperationsPatterns(patterns, preferInlineImplementation); mlir::GreedyRewriteConfig config; - config.enableRegionSimplification = mlir::GreedySimplifyRegionLevel::Disabled; + config.setRegionSimplificationLevel( + mlir::GreedySimplifyRegionLevel::Disabled); if (mlir::failed( mlir::applyPatternsGreedily(module, std::move(patterns), config))) { diff --git a/flang/lib/Optimizer/Transforms/StackArrays.cpp b/flang/lib/Optimizer/Transforms/StackArrays.cpp index 9a6566bef50f1..f9b9b4f4ff385 100644 --- a/flang/lib/Optimizer/Transforms/StackArrays.cpp +++ b/flang/lib/Optimizer/Transforms/StackArrays.cpp @@ -806,7 +806,8 @@ void StackArraysPass::runOnOperation() { mlir::RewritePatternSet patterns(&context); mlir::GreedyRewriteConfig config; // prevent the pattern driver form merging blocks - config.enableRegionSimplification = mlir::GreedySimplifyRegionLevel::Disabled; + config.setRegionSimplificationLevel( + mlir::GreedySimplifyRegionLevel::Disabled); patterns.insert<AllocMemConversion>(&context, *candidateOps); if (mlir::failed(mlir::applyOpPatternsGreedily( diff --git a/mlir/include/mlir/Transforms/GreedyPatternRewriteDriver.h b/mlir/include/mlir/Transforms/GreedyPatternRewriteDriver.h index 110b4f64856eb..45e61b68f5db2 100644 --- a/mlir/include/mlir/Transforms/GreedyPatternRewriteDriver.h +++ b/mlir/include/mlir/Transforms/GreedyPatternRewriteDriver.h @@ -49,25 +49,43 @@ class GreedyRewriteConfig { /// larger patterns when given an ambiguous pattern set. /// /// Note: Only applicable when simplifying entire regions. - bool useTopDownTraversal = false; + bool getUseTopDownTraversal() const { return useTopDownTraversal; } + GreedyRewriteConfig &setUseTopDownTraversal(bool use = true) { + useTopDownTraversal = use; + return *this; + } /// Perform control flow optimizations to the region tree after applying all /// patterns. /// /// Note: Only applicable when simplifying entire regions. - GreedySimplifyRegionLevel enableRegionSimplification = - GreedySimplifyRegionLevel::Aggressive; + GreedySimplifyRegionLevel getRegionSimplificationLevel() const { + return regionSimplificationLevel; + } + GreedyRewriteConfig & + setRegionSimplificationLevel(GreedySimplifyRegionLevel level) { + regionSimplificationLevel = level; + return *this; + } /// This specifies the maximum number of times the rewriter will iterate /// between applying patterns and simplifying regions. Use `kNoLimit` to /// disable this iteration limit. /// /// Note: Only applicable when simplifying entire regions. - int64_t maxIterations = 10; + int64_t getMaxIterations() const { return maxIterations; } + GreedyRewriteConfig &setMaxIterations(int64_t iterations) { + maxIterations = iterations; + return *this; + } /// This specifies the maximum number of rewrites within an iteration. Use /// `kNoLimit` to disable this limit. - int64_t maxNumRewrites = kNoLimit; + int64_t getMaxNumRewrites() const { return maxNumRewrites; } + GreedyRewriteConfig &setMaxNumRewrites(int64_t limit) { + maxNumRewrites = limit; + return *this; + } static constexpr int64_t kNoLimit = -1; @@ -75,7 +93,11 @@ class GreedyRewriteConfig { /// specified, the closest enclosing region around the initial list of ops /// (or the specified region, depending on which greedy rewrite entry point /// is used) is used as a scope. - Region *scope = nullptr; + Region *getScope() const { return scope; } + GreedyRewriteConfig &setScope(Region *scope) { + this->scope = scope; + return *this; + } /// Strict mode can restrict the ops that are added to the worklist during /// the rewrite. @@ -87,16 +109,44 @@ class GreedyRewriteConfig { /// * GreedyRewriteStrictness::ExistingOps: Only pre-existing ops (that were /// were on the worklist at the very beginning) enqueued. All other ops are /// excluded. - GreedyRewriteStrictness strictMode = GreedyRewriteStrictness::AnyOp; + GreedyRewriteStrictness getStrictness() const { return strictness; } + GreedyRewriteConfig &setStrictness(GreedyRewriteStrictness mode) { + strictness = mode; + return *this; + } /// An optional listener that should be notified about IR modifications. - RewriterBase::Listener *listener = nullptr; + RewriterBase::Listener *getListener() const { return listener; } + GreedyRewriteConfig &setListener(RewriterBase::Listener *listener) { + this->listener = listener; + return *this; + } /// Whether this should fold while greedily rewriting. - bool fold = true; + bool isFoldingEnabled() const { return fold; } + GreedyRewriteConfig &enableFolding(bool enable = true) { + fold = enable; + return *this; + } /// If set to "true", constants are CSE'd (even across multiple regions that /// are in a parent-ancestor relationship). + bool isConstantCSEEnabled() const { return cseConstants; } + GreedyRewriteConfig &enableConstantCSE(bool enable = true) { + cseConstants = enable; + return *this; + } + +private: + Region *scope = nullptr; + bool useTopDownTraversal = false; + GreedySimplifyRegionLevel regionSimplificationLevel = + GreedySimplifyRegionLevel::Aggressive; + int64_t maxIterations = 10; + int64_t maxNumRewrites = kNoLimit; + GreedyRewriteStrictness strictness = GreedyRewriteStrictness::AnyOp; + RewriterBase::Listener *listener = nullptr; + bool fold = true; bool cseConstants = true; }; @@ -128,14 +178,14 @@ applyPatternsGreedily(Region ®ion, const FrozenRewritePatternSet &patterns, GreedyRewriteConfig config = GreedyRewriteConfig(), bool *changed = nullptr); /// Same as `applyPatternsAndGreedily` above with folding. -/// FIXME: Remove this once transition to above is complieted. +/// FIXME: Remove this once transition to above is completed. LLVM_DEPRECATED("Use applyPatternsGreedily() instead", "applyPatternsGreedily") inline LogicalResult applyPatternsAndFoldGreedily(Region ®ion, const FrozenRewritePatternSet &patterns, GreedyRewriteConfig config = GreedyRewriteConfig(), bool *changed = nullptr) { - config.fold = true; + config.enableFolding(); return applyPatternsGreedily(region, patterns, config, changed); } @@ -187,7 +237,7 @@ applyPatternsAndFoldGreedily(Operation *op, const FrozenRewritePatternSet &patterns, GreedyRewriteConfig config = GreedyRewriteConfig(), bool *changed = nullptr) { - config.fold = true; + config.enableFolding(); return applyPatternsGreedily(op, patterns, config, changed); } @@ -233,7 +283,7 @@ applyOpPatternsAndFold(ArrayRef<Operation *> ops, const FrozenRewritePatternSet &patterns, GreedyRewriteConfig config = GreedyRewriteConfig(), bool *changed = nullptr, bool *allErased = nullptr) { - config.fold = true; + config.enableFolding(); return applyOpPatternsGreedily(ops, patterns, config, changed, allErased); } diff --git a/mlir/include/mlir/Transforms/Passes.td b/mlir/include/mlir/Transforms/Passes.td index a39ab77fc8fb3..1e89a78912e99 100644 --- a/mlir/include/mlir/Transforms/Passes.td +++ b/mlir/include/mlir/Transforms/Passes.td @@ -33,7 +33,7 @@ def Canonicalizer : Pass<"canonicalize"> { Option<"topDownProcessingEnabled", "top-down", "bool", /*default=*/"true", "Seed the worklist in general top-down order">, - Option<"enableRegionSimplification", "region-simplify", "mlir::GreedySimplifyRegionLevel", + Option<"regionSimplifyLevel", "region-simplify", "mlir::GreedySimplifyRegionLevel", /*default=*/"mlir::GreedySimplifyRegionLevel::Normal", "Perform control flow optimizations to the region tree", [{::llvm::cl::values( diff --git a/mlir/lib/Dialect/Affine/TransformOps/AffineTransformOps.cpp b/mlir/lib/Dialect/Affine/TransformOps/AffineTransformOps.cpp index 9f7df7823d997..43d37ee3332ef 100644 --- a/mlir/lib/Dialect/Affine/TransformOps/AffineTransformOps.cpp +++ b/mlir/lib/Dialect/Affine/TransformOps/AffineTransformOps.cpp @@ -127,12 +127,13 @@ SimplifyBoundedAffineOpsOp::apply(transform::TransformRewriter &rewriter, patterns.insert<SimplifyAffineMinMaxOp<AffineMinOp>, SimplifyAffineMinMaxOp<AffineMaxOp>>(getContext(), cstr); FrozenRewritePatternSet frozenPatterns(std::move(patterns)); - GreedyRewriteConfig config; - config.listener = - static_cast<RewriterBase::Listener *>(rewriter.getListener()); - config.strictMode = GreedyRewriteStrictness::ExistingAndNewOps; // Apply the simplification pattern to a fixpoint. - if (failed(applyOpPatternsGreedily(targets, frozenPatterns, config))) { + if (failed(applyOpPatternsGreedily( + targets, frozenPatterns, + GreedyRewriteConfig() + .setListener( + static_cast<RewriterBase::Listener *>(rewriter.getListener())) + .setStrictness(GreedyRewriteStrictness::ExistingAndNewOps)))) { auto diag = emitDefiniteFailure() << "affine.min/max simplification did not converge"; return diag; diff --git a/mlir/lib/Dialect/Affine/Transforms/AffineDataCopyGeneration.cpp b/mlir/lib/Dialect/Affine/Transforms/AffineDataCopyGeneration.cpp index 4d30213cc6ec2..62c1857e4b1da 100644 --- a/mlir/lib/Dialect/Affine/Transforms/AffineDataCopyGeneration.cpp +++ b/mlir/lib/Dialect/Affine/Transforms/AffineDataCopyGeneration.cpp @@ -237,7 +237,8 @@ void AffineDataCopyGeneration::runOnOperation() { AffineLoadOp::getCanonicalizationPatterns(patterns, &getContext()); AffineStoreOp::getCanonicalizationPatterns(patterns, &getContext()); FrozenRewritePatternSet frozenPatterns(std::move(patterns)); - GreedyRewriteConfig config; - config.strictMode = GreedyRewriteStrictness::ExistingAndNewOps; - (void)applyOpPatternsGreedily(copyOps, frozenPatterns, config); + (void)applyOpPatternsGreedily( + copyOps, frozenPatterns, + GreedyRewriteConfig().setStrictness( + GreedyRewriteStrictness::ExistingAndNewOps)); } diff --git a/mlir/lib/Dialect/Affine/Transforms/SimplifyAffineStructures.cpp b/mlir/lib/Dialect/Affine/Transforms/SimplifyAffineStructures.cpp index 31711ade3153b..9e9096c2e3186 100644 --- a/mlir/lib/Dialect/Affine/Transforms/SimplifyAffineStructures.cpp +++ b/mlir/lib/Dialect/Affine/Transforms/SimplifyAffineStructures.cpp @@ -109,7 +109,8 @@ void SimplifyAffineStructures::runOnOperation() { if (isa<AffineForOp, AffineIfOp, AffineApplyOp>(op)) opsToSimplify.push_back(op); }); - GreedyRewriteConfig config; - config.strictMode = GreedyRewriteStrictness::ExistingAndNewOps; - (void)applyOpPatternsGreedily(opsToSimplify, frozenPatterns, config); + (void)applyOpPatternsGreedily( + opsToSimplify, frozenPatterns, + GreedyRewriteConfig().setStrictness( + GreedyRewriteStrictness::ExistingAndNewOps)); } diff --git a/mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp b/mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp index dd539ff685653..0d4ba3940c48e 100644 --- a/mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp +++ b/mlir/lib/Dialect/Affine/Utils/LoopUtils.cpp @@ -315,11 +315,12 @@ LogicalResult mlir::affine::affineForOpBodySkew(AffineForOp forOp, // Simplify/canonicalize the affine.for. RewritePatternSet patterns(res.getContext()); AffineForOp::getCanonicalizationPatterns(patterns, res.getContext()); - GreedyRewriteConfig config; - config.strictMode = GreedyRewriteStrictness::ExistingOps; bool erased; - (void)applyOpPatternsGreedily(res.getOperation(), std::move(patterns), - config, /*changed=*/nullptr, &erased); + (void)applyOpPatternsGreedily( + res.getOperation(), std::move(patterns), + GreedyRewriteConfig().setStrictness( + GreedyRewriteStrictness::ExistingAndNewOps), + /*changed=*/nullptr, &erased); if (!erased && !prologue) prologue = res; if (!erased) diff --git a/mlir/lib/Dialect/Affine/Utils/Utils.cpp b/mlir/lib/Dialect/Affine/Utils/Utils.cpp index 2925aa918cb1c..11798b99fa879 100644 --- a/mlir/lib/Dialect/Affine/Utils/Utils.cpp +++ b/mlir/lib/Dialect/Affine/Utils/Utils.cpp @@ -426,11 +426,11 @@ LogicalResult mlir::affine::hoistAffineIfOp(AffineIfOp ifOp, bool *folded) { Rewr... [truncated] |
| ✅ With the latest revision this PR passed the C/C++ code formatter. |
This is similar to other configuration objects used across MLIR. Rename some fields to better reflect that they are no longer booleans. Reland 04d2611 / llvm#132253.
raikonenfnu added a commit to raikonenfnu/iree that referenced this pull request Apr 24, 2025
LIT change because of llvm/llvm-project#136640 which folds linalg.index who are unit dims. This patch carries revert of llvm/llvm-project#137122. StableHLO and Torch-mlir needs to update their usage of GreedyRewriteConfig to use fluent API. i.e enableRegionSimplification = VS setRegionSimplificationLevel llvm/llvm-project#135970. StableHLO has issue with VHLO_IntegerAttr and APInt not being updated. StableHLO needs to be updated with that PR's change for us to be able to integrate. llvm/llvm-project#121389. Torch-MLIR needs to be updated with that PR's change for us to be able to integrate. Signed-off-by: Stanley Winata <stanley.winata@amd.com>
raikonenfnu added a commit to iree-org/llvm-project that referenced this pull request Apr 24, 2025
This reverts commit 0c61b24.
raikonenfnu added a commit to raikonenfnu/iree that referenced this pull request Apr 24, 2025
LIT change because of llvm/llvm-project#136640 which folds linalg.index who are unit dims. This patch carries revert of: llvm/llvm-project#137122. StableHLO and Torch-mlir needs to update their usage of GreedyRewriteConfig to use fluent API. i.e enableRegionSimplification = VS setRegionSimplificationLevel llvm/llvm-project#135970. StableHLO has issue with VHLO_IntegerAttr and APInt not being updated. StableHLO needs to be updated with that PR's change for us to be able to integrate. llvm/llvm-project#121389. Torch-MLIR needs to be updated with that PR's change for us to be able to integrate. Signed-off-by: Stanley Winata <stanley.winata@amd.com>
raikonenfnu added a commit to raikonenfnu/iree that referenced this pull request Apr 25, 2025
This patch carries revert of llvm/llvm-project#136640 because linalg.index folder is segfaulting on inceptionv2 model llvm/llvm-project#133231. This PR breaks fp_to_subbytes and emulation_subbyte_types on llvm-cpu tests. iree-test-deps. llvm/llvm-project#137122. StableHLO and Torch-mlir needs to update their usage of GreedyRewriteConfig to use fluent API. i.e enableRegionSimplification = VS setRegionSimplificationLevel llvm/llvm-project#135970. StableHLO has issue with VHLO_IntegerAttr and APInt not being updated. StableHLO needs to be updated with that PR's change for us to be able to integrate. llvm/llvm-project#121389. Torch-MLIR needs to be updated with that PR's change for us to be able to integrate. Signed-off-by: Stanley Winata <stanley.winata@amd.com>
raikonenfnu added a commit to iree-org/llvm-project that referenced this pull request Apr 25, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to Muzammiluddin-Syed-ECE/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
Muzammiluddin-Syed-ECE pushed a commit to iree-org/llvm-project that referenced this pull request May 16, 2025
This reverts commit 0c61b24.
paul0403 added a commit to PennyLaneAI/catalyst that referenced this pull request Jul 28, 2025
**Context:** Update llvm, mhlo and enzyme, 2025 Q3. The latest pair of good versions, indicated by mhlo, is tensorflow/mlir-hlo@1dd2e71 ``` mhlo=1dd2e71331014ae0373f6bf900ce6be393357190 llvm=f8cb7987c64dcffb72414a40560055cb717dbf74 ``` For Enzyme, we go to the latest release https://github.com/EnzymeAD/Enzyme/releases/tag/v0.0.186 ``` enzyme=v0.0.186 ``` with commit `8c1a596158f6194f10e8ffd56a1660a61c54337e` **Description of the Change:** Miscellaneous: 1. `GreedyRewriteConfig.stuff = blah` -> `GreedyRewriteConfig.setStuff(blah)` llvm/llvm-project#137122 2. llvm gep op `inbounds` attribute is subsumed under a gep sign wrap enum flag llvm/llvm-project#137272 3. `arith::Constant[Int, Float]Op` builders now have the same argument order as other ops (output type first, then arguments) llvm/llvm-project#144636 (note that Enzyme also noticed this EnzymeAD/Enzyme#2379 😆 ) 4. The `lookupOrCreateFn` functions now take in a builder instead of instantiating a new one llvm/llvm-project#136421 5. `getStridedElementPtr` now takes in `rewriter` as the first argument (instead of the last), like all the other utils llvm/llvm-project#138984 6. The following functions now return a `LogicalResult`, and will be caught by warnings as errors as `-Wunused-result`: - `func::FuncOp.[insert, erase]Argument(s)` llvm/llvm-project#137130 - `getBackwardSlice()` llvm/llvm-project#140961 Things related to `transform.apply_registered_pass` op: 1. It now takes in a `dynamic_options` llvm/llvm-project#142683. We don't need to use this as all our pass options are static. 2. The options it takes in are now dictionaries instead of strings llvm/llvm-project#143159 Bufferization: 1. `bufferization.to_memref` op is renamed to `bufferization.to_buffer` llvm/llvm-project#137180 3. `bufferization.to_tensor` op's builder now needs the result type to be explicit llvm/llvm-project#142986. This is also needed by a patched mhlo pass. 4. The `getBuffer()` methods take in a new arg for `BufferizationState` llvm/llvm-project#141019, llvm/llvm-project#141466 5. `UnknownTypeConverterFn` in bufferization options now takes in just a type instead of a full value llvm/llvm-project#144658 **Related GitHub Issues:** [sc-95176] [sc-95664] --------- Co-authored-by: Mehrdad Malek <39844030+mehrdad2m@users.noreply.github.com>
AWoloszyn pushed a commit to iree-org/iree that referenced this pull request Dec 1, 2025
This patch carries revert of llvm/llvm-project#136640 because linalg.index folder is segfaulting on inceptionv2 model llvm/llvm-project#133231. This PR breaks fp_to_subbytes and emulation_subbyte_types on llvm-cpu tests. iree-test-deps. llvm/llvm-project#137122. StableHLO and Torch-mlir needs to update their usage of GreedyRewriteConfig to use fluent API. i.e enableRegionSimplification = VS setRegionSimplificationLevel llvm/llvm-project#135970. StableHLO has issue with VHLO_IntegerAttr and APInt not being updated. StableHLO needs to be updated with that PR's change for us to be able to integrate. llvm/llvm-project#121389. Torch-MLIR needs to be updated with that PR's change for us to be able to integrate. Signed-off-by: Stanley Winata <stanley.winata@amd.com>
AWoloszyn pushed a commit to iree-org/iree that referenced this pull request Dec 1, 2025
Updated LIT test from landing llvm/llvm-project#136640 which folds linalg.index when size is unit dim (1). Added chipSet argument into populateGpuToROCDLConversionPatterns based on changes in llvm/llvm-project#137360 This patch carries revert of llvm/llvm-project#133231. This PR breaks fp_to_subbytes and emulation_subbyte_types on llvm-cpu tests. iree-test-deps. tracker issue in #20645. llvm/llvm-project#137122. StableHLO and Torch-mlir needs to update their usage of GreedyRewriteConfig to use fluent API. i.e enableRegionSimplification = VS setRegionSimplificationLevel llvm/llvm-project#135970. StableHLO has issue with VHLO_IntegerAttr and APInt not being updated. StableHLO needs to be updated with that PR's change for us to be able to integrate. llvm/llvm-project#121389. Torch-MLIR needs to be updated with that PR's change for us to be able to integrate. --------- Signed-off-by: Stanley Winata <stanley.winata@amd.com>
AWoloszyn pushed a commit to iree-org/iree that referenced this pull request Dec 1, 2025
This patch carries 4 reverts from #20646: llvm/llvm-project#133231. This PR breaks fp_to_subbytes and emulation_subbyte_types on llvm-cpu tests. iree-test-deps. tracker issue in #20645. llvm/llvm-project#137122. StableHLO and Torch-mlir needs to update their usage of GreedyRewriteConfig to use fluent API. i.e enableRegionSimplification = VS setRegionSimplificationLevel llvm/llvm-project#135970. StableHLO has issue with VHLO_IntegerAttr and APInt not being updated. StableHLO needs to be updated with that PR's change for us to be able to integrate. llvm/llvm-project#121389. Torch-MLIR needs to be updated with that PR's change for us to be able to integrate. Signed-off-by: Ian Wood <ianwood2024@u.northwestern.edu>
AWoloszyn pushed a commit to iree-org/iree that referenced this pull request Dec 1, 2025
Carries the same 4 reverts as #20657: - llvm/llvm-project#133231. This PR breaks fp_to_subbytes and emulation_subbyte_types on llvm-cpu tests. iree-test-deps. tracker issue in #20645. - llvm/llvm-project#137122. StableHLO and Torch-mlir needs to update their usage of GreedyRewriteConfig to use fluent API. i.e enableRegionSimplification = VS setRegionSimplificationLevel - llvm/llvm-project#135970. StableHLO has issue with VHLO_IntegerAttr and APInt not being updated. StableHLO needs to be updated with that PR's change for us to be able to integrate. - llvm/llvm-project#121389. Torch-MLIR needs to be updated with that PR's change for us to be able to integrate. Signed-off-by: Ian Wood <ianwood2024@u.northwestern.edu>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
2 participants
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.
This is similar to other configuration objects used across MLIR.
Rename some fields to better reflect that they are no longer booleans.
Reland 04d2611 / #132253.