[mlir] Remove dialect specific bufferization passes (Reland) #93535
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
Groverkss
requested review from
aartbik,
PeimingLiu,
yinying-lisa-li,
matthias-springer,
hanhanW,
nicolasvasilache,
dcaballe and
rengolin
as code owners
llvmbot
added
mlir:linalg
mlir:vectorops
mlir:sparse
|
@llvm/pr-subscribers-mlir-arith @llvm/pr-subscribers-mlir Author: Kunwar Grover (Groverkss) ChangesThese passes have been depreciated for a long time and replaced by one-shot bufferization. These passes are also unsafe because they do not check for read-after-write conflicts. Relands #93488 which failed on buildbot. Fixes the failure by updating integration tests to use one-shot-bufferize instead. Patch is 44.19 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/93535.diff 45 Files Affected:
diff --git a/mlir/include/mlir/Dialect/Arith/Transforms/Passes.h b/mlir/include/mlir/Dialect/Arith/Transforms/Passes.h
index cbc6147cb81e2..9dc262cc72ed0 100644
--- a/mlir/include/mlir/Dialect/Arith/Transforms/Passes.h
+++ b/mlir/include/mlir/Dialect/Arith/Transforms/Passes.h
@@ -24,9 +24,6 @@ namespace arith {
class WideIntEmulationConverter;
class NarrowTypeEmulationConverter;
-/// Create a pass to bufferize arith.constant ops.
-std::unique_ptr<Pass> createConstantBufferizePass(uint64_t alignment = 0);
-
/// Adds patterns to emulate wide Arith and Function ops over integer
/// types into supported ones. This is done by splitting original power-of-two
/// i2N integer types into two iN halves.
diff --git a/mlir/include/mlir/Dialect/Arith/Transforms/Passes.td b/mlir/include/mlir/Dialect/Arith/Transforms/Passes.td
index 4096e309199e9..550c5c0cf4f60 100644
--- a/mlir/include/mlir/Dialect/Arith/Transforms/Passes.td
+++ b/mlir/include/mlir/Dialect/Arith/Transforms/Passes.td
@@ -11,22 +11,6 @@
include "mlir/Pass/PassBase.td"
-def ArithBufferizePass : Pass<"arith-bufferize", "ModuleOp"> {
- let summary = "Bufferize Arith dialect ops.";
- let description = [{
- This pass bufferizes arith dialect ops.
-
- This pass needs to be a module pass because it inserts memref.global
- ops into the module, which cannot be done safely from a function pass due to
- multi-threading. Most other bufferization passes can run in parallel at
- function granularity.
- }];
- let options = [
- Option<"alignment", "alignment", "unsigned", /*default=*/"0",
- "Create global memrefs with a specified alignment">,
- ];
-}
-
def ArithExpandOpsPass : Pass<"arith-expand"> {
let summary = "Legalize Arith ops to be convertible to LLVM.";
let dependentDialects = ["vector::VectorDialect"];
diff --git a/mlir/include/mlir/Dialect/Bufferization/Transforms/Passes.h b/mlir/include/mlir/Dialect/Bufferization/Transforms/Passes.h
index 459c252b70712..e053e6c97e143 100644
--- a/mlir/include/mlir/Dialect/Bufferization/Transforms/Passes.h
+++ b/mlir/include/mlir/Dialect/Bufferization/Transforms/Passes.h
@@ -221,9 +221,6 @@ createPromoteBuffersToStackPass(std::function<bool(Value)> isSmallAlloc);
/// insert_slice ops.
std::unique_ptr<Pass> createEmptyTensorEliminationPass();
-/// Create a pass that bufferizes ops from the bufferization dialect.
-std::unique_ptr<Pass> createBufferizationBufferizePass();
-
//===----------------------------------------------------------------------===//
// Registration
//===----------------------------------------------------------------------===//
diff --git a/mlir/include/mlir/Dialect/Bufferization/Transforms/Passes.td b/mlir/include/mlir/Dialect/Bufferization/Transforms/Passes.td
index 75ce85c9128c9..8f8826b9ad56b 100644
--- a/mlir/include/mlir/Dialect/Bufferization/Transforms/Passes.td
+++ b/mlir/include/mlir/Dialect/Bufferization/Transforms/Passes.td
@@ -350,11 +350,6 @@ def FinalizingBufferize : Pass<"finalizing-bufferize", "func::FuncOp"> {
let constructor = "mlir::bufferization::createFinalizingBufferizePass()";
}
-def BufferizationBufferize : Pass<"bufferization-bufferize", "func::FuncOp"> {
- let summary = "Bufferize the `bufferization` dialect";
- let constructor = "mlir::bufferization::createBufferizationBufferizePass()";
-}
-
def DropEquivalentBufferResults : Pass<"drop-equivalent-buffer-results", "ModuleOp"> {
let summary = "Remove MemRef return values that are equivalent to a bbArg";
let description = [{
diff --git a/mlir/include/mlir/Dialect/Linalg/Passes.h b/mlir/include/mlir/Dialect/Linalg/Passes.h
index d36d1e70f0b14..f2955d55e59ec 100644
--- a/mlir/include/mlir/Dialect/Linalg/Passes.h
+++ b/mlir/include/mlir/Dialect/Linalg/Passes.h
@@ -22,10 +22,6 @@ namespace func {
class FuncOp;
} // namespace func
-namespace bufferization {
-struct OneShotBufferizationOptions;
-} // namespace bufferization
-
#define GEN_PASS_DECL
#include "mlir/Dialect/Linalg/Passes.h.inc" // IWYU pragma: keep
diff --git a/mlir/include/mlir/Dialect/Linalg/Passes.td b/mlir/include/mlir/Dialect/Linalg/Passes.td
index 0a4ce8953136d..0621a9f33ba1e 100644
--- a/mlir/include/mlir/Dialect/Linalg/Passes.td
+++ b/mlir/include/mlir/Dialect/Linalg/Passes.td
@@ -89,16 +89,6 @@ def LinalgInlineScalarOperandsPass : Pass<"linalg-inline-scalar-operands"> {
];
}
-def LinalgBufferizePass : Pass<"linalg-bufferize"> {
- let summary = "Bufferize the linalg dialect";
- let dependentDialects = [
- "affine::AffineDialect",
- "bufferization::BufferizationDialect",
- "linalg::LinalgDialect",
- "memref::MemRefDialect",
- ];
-}
-
def LinalgGeneralizeNamedOpsPass : Pass<"linalg-generalize-named-ops"> {
let summary = "Convert named ops into generic ops";
let dependentDialects = ["linalg::LinalgDialect"];
diff --git a/mlir/include/mlir/Dialect/Shape/Transforms/Passes.h b/mlir/include/mlir/Dialect/Shape/Transforms/Passes.h
index cfb637f133f54..28e17459ff962 100644
--- a/mlir/include/mlir/Dialect/Shape/Transforms/Passes.h
+++ b/mlir/include/mlir/Dialect/Shape/Transforms/Passes.h
@@ -47,13 +47,6 @@ void populateShapeRewritePatterns(RewritePatternSet &patterns);
void populateRemoveShapeConstraintsPatterns(RewritePatternSet &patterns);
std::unique_ptr<OperationPass<func::FuncOp>> createRemoveShapeConstraintsPass();
-// Bufferizes shape dialect ops.
-//
-// Note that most shape dialect ops must be converted to std before
-// bufferization happens, as they are intended to be bufferized at the std
-// level.
-std::unique_ptr<OperationPass<func::FuncOp>> createShapeBufferizePass();
-
/// Outline the shape computation part by adding shape.func and populate
/// conrresponding mapping infomation into ShapeMappingAnalysis.
std::unique_ptr<OperationPass<ModuleOp>> createOutlineShapeComputationPass();
diff --git a/mlir/include/mlir/Dialect/Shape/Transforms/Passes.td b/mlir/include/mlir/Dialect/Shape/Transforms/Passes.td
index 9dfda9ea33615..83834509b4a35 100644
--- a/mlir/include/mlir/Dialect/Shape/Transforms/Passes.td
+++ b/mlir/include/mlir/Dialect/Shape/Transforms/Passes.td
@@ -103,11 +103,4 @@ def ShapeToShapeLowering : Pass<"shape-to-shape-lowering", "func::FuncOp"> {
let constructor = "mlir::createShapeToShapeLowering()";
}
-// TODO: Generalize this to allow any type conversions desired.
-def ShapeBufferize : Pass<"shape-bufferize", "func::FuncOp"> {
- let summary = "Bufferize the shape dialect.";
- let constructor = "mlir::createShapeBufferizePass()";
- let dependentDialects = ["bufferization::BufferizationDialect",
- "memref::MemRefDialect"];
-}
#endif // MLIR_DIALECT_SHAPE_TRANSFORMS_PASSES
diff --git a/mlir/include/mlir/Dialect/Tensor/Transforms/Passes.h b/mlir/include/mlir/Dialect/Tensor/Transforms/Passes.h
index 48f9066934a25..964c35b3f15b8 100644
--- a/mlir/include/mlir/Dialect/Tensor/Transforms/Passes.h
+++ b/mlir/include/mlir/Dialect/Tensor/Transforms/Passes.h
@@ -21,9 +21,6 @@ namespace tensor {
/// Creates an instance of the `tensor` subset folding pass.
std::unique_ptr<Pass> createFoldTensorSubsetOpsPass();
-/// Creates an instance of the `tensor` dialect bufferization pass.
-std::unique_ptr<Pass> createTensorBufferizePass();
-
//===----------------------------------------------------------------------===//
// Registration
//===----------------------------------------------------------------------===//
diff --git a/mlir/include/mlir/Dialect/Tensor/Transforms/Passes.td b/mlir/include/mlir/Dialect/Tensor/Transforms/Passes.td
index 4cc3844f29120..be4c333836ec0 100644
--- a/mlir/include/mlir/Dialect/Tensor/Transforms/Passes.td
+++ b/mlir/include/mlir/Dialect/Tensor/Transforms/Passes.td
@@ -27,9 +27,4 @@ def FoldTensorSubsetOps : Pass<"fold-tensor-subset-ops"> {
];
}
-def TensorBufferize : Pass<"tensor-bufferize", "func::FuncOp"> {
- let summary = "Bufferize the `tensor` dialect";
- let constructor = "mlir::tensor::createTensorBufferizePass()";
-}
-
#endif // MLIR_DIALECT_TENSOR_TRANSFORMS_PASSES
diff --git a/mlir/include/mlir/Dialect/Vector/Transforms/Passes.h b/mlir/include/mlir/Dialect/Vector/Transforms/Passes.h
index 911402551e14d..5667f4fa95ace 100644
--- a/mlir/include/mlir/Dialect/Vector/Transforms/Passes.h
+++ b/mlir/include/mlir/Dialect/Vector/Transforms/Passes.h
@@ -17,9 +17,6 @@ namespace vector {
#define GEN_PASS_DECL
#include "mlir/Dialect/Vector/Transforms/Passes.h.inc"
-/// Creates an instance of the `vector` dialect bufferization pass.
-std::unique_ptr<Pass> createVectorBufferizePass();
-
/// Creates an instance of the `vector.mask` lowering pass.
std::unique_ptr<Pass> createLowerVectorMaskPass();
diff --git a/mlir/include/mlir/Dialect/Vector/Transforms/Passes.td b/mlir/include/mlir/Dialect/Vector/Transforms/Passes.td
index 31a0b3b2f0c53..7436998749791 100644
--- a/mlir/include/mlir/Dialect/Vector/Transforms/Passes.td
+++ b/mlir/include/mlir/Dialect/Vector/Transforms/Passes.td
@@ -11,11 +11,6 @@
include "mlir/Pass/PassBase.td"
-def VectorBufferize : Pass<"vector-bufferize", "func::FuncOp"> {
- let summary = "Bufferize Vector dialect ops";
- let constructor = "mlir::vector::createVectorBufferizePass()";
-}
-
def LowerVectorMaskPass : Pass<"lower-vector-mask", "func::FuncOp"> {
let summary = "Lower 'vector.mask' operations";
let constructor = "mlir::vector::createLowerVectorMaskPass()";
diff --git a/mlir/lib/Dialect/Arith/Transforms/Bufferize.cpp b/mlir/lib/Dialect/Arith/Transforms/Bufferize.cpp
deleted file mode 100644
index 9a066756f429c..0000000000000
--- a/mlir/lib/Dialect/Arith/Transforms/Bufferize.cpp
+++ /dev/null
@@ -1,67 +0,0 @@
-//===- Bufferize.cpp - Bufferization for Arith ops ---------*- C++ -*-===//
-//
-// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
-// See https://llvm.org/LICENSE.txt for license information.
-// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
-//
-//===----------------------------------------------------------------------===//
-
-#include "mlir/Dialect/Arith/Transforms/Passes.h"
-
-#include "mlir/Dialect/Arith/IR/Arith.h"
-#include "mlir/Dialect/Arith/Transforms/BufferizableOpInterfaceImpl.h"
-#include "mlir/Dialect/Bufferization/IR/BufferizableOpInterface.h"
-#include "mlir/Dialect/Bufferization/IR/Bufferization.h"
-#include "mlir/Dialect/Bufferization/Transforms/Bufferize.h"
-#include "mlir/Dialect/MemRef/IR/MemRef.h"
-
-namespace mlir {
-namespace arith {
-#define GEN_PASS_DEF_ARITHBUFFERIZEPASS
-#include "mlir/Dialect/Arith/Transforms/Passes.h.inc"
-} // namespace arith
-} // namespace mlir
-
-using namespace mlir;
-using namespace bufferization;
-
-namespace {
-/// Pass to bufferize Arith ops.
-struct ArithBufferizePass
- : public arith::impl::ArithBufferizePassBase<ArithBufferizePass> {
- using ArithBufferizePassBase::ArithBufferizePassBase;
-
- ArithBufferizePass(uint64_t alignment = 0, bool constantOpOnly = false)
- : constantOpOnly(constantOpOnly) {
- this->alignment = alignment;
- }
-
- void runOnOperation() override {
- BufferizationOptions options = getPartialBufferizationOptions();
- if (constantOpOnly) {
- options.opFilter.allowOperation<arith::ConstantOp>();
- } else {
- options.opFilter.allowDialect<arith::ArithDialect>();
- }
- options.bufferAlignment = alignment;
-
- if (failed(bufferizeOp(getOperation(), options)))
- signalPassFailure();
- }
-
- void getDependentDialects(DialectRegistry ®istry) const override {
- registry.insert<bufferization::BufferizationDialect, memref::MemRefDialect,
- arith::ArithDialect>();
- arith::registerBufferizableOpInterfaceExternalModels(registry);
- }
-
-private:
- bool constantOpOnly;
-};
-} // namespace
-
-std::unique_ptr<Pass>
-mlir::arith::createConstantBufferizePass(uint64_t alignment) {
- return std::make_unique<ArithBufferizePass>(alignment,
- /*constantOpOnly=*/true);
-}
diff --git a/mlir/lib/Dialect/Arith/Transforms/CMakeLists.txt b/mlir/lib/Dialect/Arith/Transforms/CMakeLists.txt
index 12659eaba1fa5..6b8bde8dc2aaf 100644
--- a/mlir/lib/Dialect/Arith/Transforms/CMakeLists.txt
+++ b/mlir/lib/Dialect/Arith/Transforms/CMakeLists.txt
@@ -1,7 +1,6 @@
add_mlir_dialect_library(MLIRArithTransforms
BufferDeallocationOpInterfaceImpl.cpp
BufferizableOpInterfaceImpl.cpp
- Bufferize.cpp
BufferViewFlowOpInterfaceImpl.cpp
EmulateUnsupportedFloats.cpp
EmulateWideInt.cpp
diff --git a/mlir/lib/Dialect/Bufferization/Transforms/Bufferize.cpp b/mlir/lib/Dialect/Bufferization/Transforms/Bufferize.cpp
index 7ba347a1f15e4..0fddd60eb8140 100644
--- a/mlir/lib/Dialect/Bufferization/Transforms/Bufferize.cpp
+++ b/mlir/lib/Dialect/Bufferization/Transforms/Bufferize.cpp
@@ -320,29 +320,6 @@ struct OneShotBufferizePass
};
} // namespace
-namespace {
-struct BufferizationBufferizePass
- : public bufferization::impl::BufferizationBufferizeBase<
- BufferizationBufferizePass> {
- void runOnOperation() override {
- BufferizationOptions options = getPartialBufferizationOptions();
- options.opFilter.allowDialect<BufferizationDialect>();
-
- if (failed(bufferizeOp(getOperation(), options)))
- signalPassFailure();
- }
-
- void getDependentDialects(DialectRegistry ®istry) const override {
- registry
- .insert<bufferization::BufferizationDialect, memref::MemRefDialect>();
- }
-};
-} // namespace
-
-std::unique_ptr<Pass> mlir::bufferization::createBufferizationBufferizePass() {
- return std::make_unique<BufferizationBufferizePass>();
-}
-
std::unique_ptr<Pass> mlir::bufferization::createOneShotBufferizePass() {
return std::make_unique<OneShotBufferizePass>();
}
diff --git a/mlir/lib/Dialect/Linalg/Transforms/Bufferize.cpp b/mlir/lib/Dialect/Linalg/Transforms/Bufferize.cpp
deleted file mode 100644
index 8812ca14ba610..0000000000000
--- a/mlir/lib/Dialect/Linalg/Transforms/Bufferize.cpp
+++ /dev/null
@@ -1,52 +0,0 @@
-//===- Bufferize.cpp - Bufferization of linalg ops ------------------------===//
-//
-// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
-// See https://llvm.org/LICENSE.txt for license information.
-// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
-//
-//===----------------------------------------------------------------------===//
-
-#include "mlir/Dialect/Linalg/Passes.h"
-
-#include "mlir/Dialect/Affine/IR/AffineOps.h"
-#include "mlir/Dialect/Bufferization/IR/Bufferization.h"
-#include "mlir/Dialect/Bufferization/Transforms/Bufferize.h"
-#include "mlir/Dialect/Func/IR/FuncOps.h"
-#include "mlir/Dialect/Linalg/IR/Linalg.h"
-#include "mlir/Dialect/Linalg/Transforms/BufferizableOpInterfaceImpl.h"
-#include "mlir/Dialect/Linalg/Transforms/Transforms.h"
-#include "mlir/Dialect/Tensor/IR/Tensor.h"
-#include "mlir/IR/BuiltinDialect.h"
-#include "mlir/IR/Operation.h"
-#include "mlir/Pass/Pass.h"
-
-namespace mlir {
-#define GEN_PASS_DEF_LINALGBUFFERIZEPASS
-#include "mlir/Dialect/Linalg/Passes.h.inc"
-} // namespace mlir
-
-using namespace mlir;
-using namespace bufferization;
-
-namespace {
-/// Converts Linalg operations that work on tensor-type operands or results to
-/// work on buffers.
-struct LinalgBufferizePass
- : public impl::LinalgBufferizePassBase<LinalgBufferizePass> {
- using impl::LinalgBufferizePassBase<
- LinalgBufferizePass>::LinalgBufferizePassBase;
- void runOnOperation() override {
- BufferizationOptions options = getPartialBufferizationOptions();
- options.opFilter.allowDialect<linalg::LinalgDialect>();
-
- if (failed(bufferizeOp(getOperation(), options)))
- signalPassFailure();
- }
-
- void getDependentDialects(DialectRegistry ®istry) const override {
- registry.insert<bufferization::BufferizationDialect, memref::MemRefDialect,
- tensor::TensorDialect, linalg::LinalgDialect>();
- linalg::registerBufferizableOpInterfaceExternalModels(registry);
- }
-};
-} // namespace
diff --git a/mlir/lib/Dialect/Linalg/Transforms/CMakeLists.txt b/mlir/lib/Dialect/Linalg/Transforms/CMakeLists.txt
index ed9f40089282a..7e3dc56e0acdc 100644
--- a/mlir/lib/Dialect/Linalg/Transforms/CMakeLists.txt
+++ b/mlir/lib/Dialect/Linalg/Transforms/CMakeLists.txt
@@ -2,7 +2,6 @@ add_mlir_dialect_library(MLIRLinalgTransforms
AllInterfaces.cpp
BubbleUpExtractSlice.cpp
BufferizableOpInterfaceImpl.cpp
- Bufferize.cpp
ConstantFold.cpp
ConvertToDestinationStyle.cpp
ConvertConv2DToImg2Col.cpp
diff --git a/mlir/lib/Dialect/Shape/Transforms/Bufferize.cpp b/mlir/lib/Dialect/Shape/Transforms/Bufferize.cpp
deleted file mode 100644
index 9dadbdbc91eca..0000000000000
--- a/mlir/lib/Dialect/Shape/Transforms/Bufferize.cpp
+++ /dev/null
@@ -1,49 +0,0 @@
-//====----- Bufferize.cpp - Bufferization of shape ops ---------*- C++-*--===//
-//
-// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
-// See https://llvm.org/LICENSE.txt for license information.
-// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
-//
-//===----------------------------------------------------------------------===//
-
-#include "mlir/Dialect/Shape/Transforms/Passes.h"
-
-#include "mlir/Dialect/Bufferization/IR/BufferizableOpInterface.h"
-#include "mlir/Dialect/Bufferization/IR/Bufferization.h"
-#include "mlir/Dialect/Bufferization/Transforms/Bufferize.h"
-#include "mlir/Dialect/Func/IR/FuncOps.h"
-#include "mlir/Dialect/MemRef/IR/MemRef.h"
-#include "mlir/Dialect/Shape/IR/Shape.h"
-#include "mlir/Dialect/Shape/Transforms/BufferizableOpInterfaceImpl.h"
-#include "mlir/Pass/Pass.h"
-
-namespace mlir {
-#define GEN_PASS_DEF_SHAPEBUFFERIZE
-#include "mlir/Dialect/Shape/Transforms/Passes.h.inc"
-} // namespace mlir
-
-using namespace mlir;
-using namespace bufferization;
-
-namespace {
-struct ShapeBufferizePass
- : public impl::ShapeBufferizeBase<ShapeBufferizePass> {
- void runOnOperation() override {
- BufferizationOptions options = getPartialBufferizationOptions();
- options.opFilter.allowDialect<shape::ShapeDialect>();
-
- if (failed(bufferizeOp(getOperation(), options)))
- signalPassFailure();
- }
-
- void getDependentDialects(DialectRegistry ®istry) const override {
- registry.insert<bufferization::BufferizationDialect, memref::MemRefDialect,
- shape::ShapeDialect>();
- shape::registerBufferizableOpInterfaceExternalModels(registry);
- }
-};
-} // namespace
-
-std::unique_ptr<OperationPass<func::FuncOp>> mlir::createShapeBufferizePass() {
- return std::make_unique<ShapeBufferizePass>();
-}
diff --git a/mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt b/mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt
index 7c9b0d2e5e3a8..a51c6780c2866 100644
--- a/mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt
+++ b/mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt
@@ -1,6 +1,5 @@
add_mlir_dialect_library(MLIRShapeOpsTransforms
BufferizableOpInterfaceImpl.cpp
- Bufferize.cpp
OutlineShapeComputation.cpp
RemoveShapeConstraints.cpp
ShapeToShapeLowering.cpp
diff --git a/mlir/lib/Dialect/Tensor/Transforms/Bufferize.cpp b/mlir/lib/Dialect/Tensor/Transforms/Bufferize.cpp
deleted file mode 100644
index d27c4576a8b7a..0000000000000
--- a/mlir/lib/Dialect/Tensor/Transforms/Bufferize.cpp
+++ /dev/null
@@ -1,58 +0,0 @@
-//===- Bufferize.cpp - Bufferization for `tensor` dialect ops -------------===//
-//
-// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
-// See https://llvm.org/LICENSE.txt for license information.
-// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
-//
-//===----------------------------------------------------------------------===//
-//
-// This file implements bufferization of `tensor` dialect ops
-//
-//===----------------------------------------------------------------------===//
-
-#include "mlir/Dialect/Bufferization/Transforms/Bufferize.h"
-#include "mlir/Dialect/Arith/IR/Arith.h"
-#include "mlir/Dialect/Bufferization/IR/BufferizableOpInterface.h"
-#include "mlir/Dialect/Bufferization/IR/Bufferization.h"
-#include "mlir/Dialect/Func/IR/FuncOps.h"
-#include "mlir/Dialect/MemRef/IR/MemRef.h"
-#include "mlir/Dialect/SCF/IR/SCF.h"
-#include "mlir/Dialect/Tensor/IR/Tensor.h"
-#include "mlir/Dialect/Tensor/Transforms/BufferizableOpInterfaceImpl.h"
-#include "mlir/Dialect/Tensor/Transforms/Passes.h"
-#include "mlir/IR/ImplicitLocOpBuilder.h"
-#include "mlir/Transforms/DialectC...
[truncated]
|
matthias-springer
approved these changes
May 28, 2024
vg0204
pushed a commit
to vg0204/llvm-project
that referenced
this pull request
May 29, 2024
) These passes have been depreciated for a long time and replaced by one-shot bufferization. These passes are also unsafe because they do not check for read-after-write conflicts. Relands llvm#93488 which failed on buildbot. Fixes the failure by updating integration tests to use one-shot-bufferize instead.
tzunghanjuang
pushed a commit
to PennyLaneAI/catalyst
that referenced
this pull request
Jul 29, 2024
**Context:** We target at 0.4.28 instead of 0.4.30 because many bufferization passes are removed after [this llvm commit](llvm/llvm-project#93535). **Description of the Change:** ***Mandatory Updates:*** * Cmake - Remove `MhloShapeOpsToStandard` [link](tensorflow/mlir-hlo@57d2124) - Add `StablehloPasses` - Add `MhloQuantToIntConversion` (This will be removed after 0.4.29) - `EnzymeStatic-18` => `EnzymeStatic-19` - `RunnerUtils.h` requires `Float16Bits.h` [link](llvm/llvm-project@7bc6c4a) * LLVM - `setDataLayout` must happen before code generation or they will use the default one. [link](https://discord.com/channels/636084430946959380/636732535434510338/1265407221324451871) * MLIR - `updateRootInPlace` => `modifyOpInPlace` - `startRootUpdate` => `startOpModification` - `finalizeRootUpdate` => `finalizeOpModification` - The order of transformed mlir expressions is different and required fine-tuning for `CHECK-DAG`s. - Using rewriter's method to release MeasurementOps (Caused by LLVM commit-`b840d2968391dd610b792a65133a1edc1bcc397c`). [link](llvm/llvm-project@b840d29) - Allow `replaceTerminatorWithUnconditionalJumpToSuccessBlock` to accept `LLVM:br`. (New LLVM/MLIR will reuse `LLVM::UnreachableOp` and use `LLVM::br` to reach it.) * Frontend - `jax.linear_util` => `jax.extend.linear_util` [link](https://jax.readthedocs.io/en/latest/changelog.html#jax-0-4-24-feb-6-2024) - `jax_ctx.module_context.replace` => `jax_ctx.replace` - `gensym(jaxprs, suffix)` => `gensym(suffix)` [link](jax-ml/jax@67df647) - Move `name_stack` out of mlir.ModuleContext (functions like `lower_jaxpr_to_fun` is taking `name_stack` now). [link](jax-ml/jax#19856). - Pass `LoweringRuleContext.ModuleContext` instead of `LoweringRuleContext` to `jaxpr_subcomp` - Patch new `_sin_lowering` and `_cos_lowering` with `_nary_lower_hlo(sine/cosine)`. - Variable names (like `%0` => `%cst`) and orders in FileCheck. ***Deprecations/Warnings:*** * MLIR - `x.cast<T>()` => `mlir::cast<T>(x)` [link](https://discourse.llvm.org/t/preferred-casting-style-going-forward/68443) - `x.dyn_cast<T>()` => `mlir::dyn_cast<T>(x)` [link](https://discourse.llvm.org/t/preferred-casting-style-going-forward/68443) - `x.isa<T>()` => `mlir::isa<T>(x)` [link](https://discourse.llvm.org/t/preferred-casting-style-going-forward/68443) * Enzyme - Add `DCMAKE_POLICY_DEFAULT_CMP0116` **Related GitHub Issues:** #863 [sc-67111] --------- Co-authored-by: David Ittah <[email protected]> Co-authored-by: erick-xanadu <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
These passes have been depreciated for a long time and replaced by one-shot bufferization. These passes are also unsafe because they do not check for read-after-write conflicts.
Relands #93488 which failed on buildbot. Fixes the failure by updating integration tests to use one-shot-bufferize instead.