[mlir][amdgpu] Remove shared memory optimization pass #88225
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
This reverts commit 87c0260. This implementation has a number of issues and ultimately does not work on gfx9. * It does not reduce bank conflicts with wide memory accesses. * It does not consider when LDS bank conflicts work on amdgpu. * The implementation is too fragile to be used on real-world code. For example, the code bails out on any `memref.subview` in the root op, even when the subview is not a user of any of the `memref.alloc` ops. I do not see how these can be easily fixed, therefore I think it's better to delete this code.
kuhar
requested review from
rupprecht,
ftynse and
nicolasvasilache
as code owners
llvmbot
added
backend:AMDGPU
mlir:gpu
mlir
bazel
|
@llvm/pr-subscribers-mlir-amdgpu @llvm/pr-subscribers-mlir Author: Jakub Kuderski (kuhar) ChangesThis implementation has a number of issues and ultimately does not work on gfx9.
I do not see how these can be easily fixed, therefore I think it's better to delete this code. Patch is 40.47 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/88225.diff 20 Files Affected:
diff --git a/mlir/include/mlir/Dialect/AMDGPU/CMakeLists.txt b/mlir/include/mlir/Dialect/AMDGPU/CMakeLists.txt
index 660deb21479d29..9f57627c321fb0 100644
--- a/mlir/include/mlir/Dialect/AMDGPU/CMakeLists.txt
+++ b/mlir/include/mlir/Dialect/AMDGPU/CMakeLists.txt
@@ -1,3 +1,2 @@
add_subdirectory(IR)
-add_subdirectory(TransformOps)
add_subdirectory(Transforms)
diff --git a/mlir/include/mlir/Dialect/AMDGPU/IR/AMDGPU.td b/mlir/include/mlir/Dialect/AMDGPU/IR/AMDGPU.td
index 21942b179a0013..3f27e1541cf38c 100644
--- a/mlir/include/mlir/Dialect/AMDGPU/IR/AMDGPU.td
+++ b/mlir/include/mlir/Dialect/AMDGPU/IR/AMDGPU.td
@@ -29,23 +29,6 @@ def AMDGPU_Dialect : Dialect {
"gpu::GPUDialect"
];
let useDefaultAttributePrinterParser = 1;
-
- let extraClassDeclaration = [{
- /// Return true if the given MemRefType has an integer address
- /// space that matches the ROCDL shared memory address space or
- /// is a gpu::AddressSpaceAttr attribute with value 'workgroup`.
- static bool hasSharedMemoryAddressSpace(MemRefType type);
-
- /// Return true if the given Attribute has an integer address
- /// space that matches the ROCDL shared memory address space or
- /// is a gpu::AddressSpaceAttr attribute with value 'workgroup`.
- static bool isSharedMemoryAddressSpace(Attribute type);
-
- /// Defines the MemRef memory space attribute numeric value that indicates
- /// a memref is located in shared memory. This should correspond to the
- /// value used in ROCDL.
- static constexpr unsigned kSharedMemoryAddressSpace = 3;
- }];
}
//===----------------------------------------------------------------------===//
diff --git a/mlir/include/mlir/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.h b/mlir/include/mlir/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.h
deleted file mode 100644
index dcf934c71dd1fc..00000000000000
--- a/mlir/include/mlir/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.h
+++ /dev/null
@@ -1,48 +0,0 @@
-//===- AMDGPUTransformOps.h - AMDGPU transform 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
-//
-//===----------------------------------------------------------------------===//
-
-#ifndef MLIR_DIALECT_AMDGPU_TRANSFORMOPS_AMDGPUTRANSFORMOPS_H
-#define MLIR_DIALECT_AMDGPU_TRANSFORMOPS_AMDGPUTRANSFORMOPS_H
-
-#include "mlir/Dialect/Func/IR/FuncOps.h"
-#include "mlir/Dialect/Transform/IR/TransformAttrs.h"
-#include "mlir/Dialect/Transform/IR/TransformDialect.h"
-#include "mlir/Dialect/Transform/Interfaces/TransformInterfaces.h"
-#include "mlir/IR/OpImplementation.h"
-#include "mlir/IR/RegionKindInterface.h"
-
-namespace mlir {
-namespace transform {
-class TransformHandleTypeInterface;
-} // namespace transform
-} // namespace mlir
-
-namespace mlir {
-class DialectRegistry;
-
-namespace linalg {
-class LinalgOp;
-} // namespace linalg
-
-namespace scf {
-class ForOp;
-} // namespace scf
-
-namespace amdgpu {
-void registerTransformDialectExtension(DialectRegistry ®istry);
-} // namespace amdgpu
-} // namespace mlir
-
-//===----------------------------------------------------------------------===//
-// AMDGPU Transform Operations
-//===----------------------------------------------------------------------===//
-
-#define GET_OP_CLASSES
-#include "mlir/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.h.inc"
-
-#endif // MLIR_DIALECT_AMDGPU_TRANSFORMOPS_AMDGPUTRANSFORMOPS_H
diff --git a/mlir/include/mlir/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.td b/mlir/include/mlir/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.td
deleted file mode 100644
index 8aaa87511a2be6..00000000000000
--- a/mlir/include/mlir/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.td
+++ /dev/null
@@ -1,47 +0,0 @@
-//===- AMDGPUTransformOps.td - AMDGPU transform ops --------*- tablegen -*-===//
-//
-// 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
-//
-//===----------------------------------------------------------------------===//
-
-#ifndef AMDGPU_TRANSFORM_OPS
-#define AMDGPU_TRANSFORM_OPS
-
-include "mlir/Dialect/Transform/IR/TransformAttrs.td"
-include "mlir/Dialect/Transform/IR/TransformDialect.td"
-include "mlir/Dialect/Transform/Interfaces/TransformInterfaces.td"
-include "mlir/Dialect/Transform/IR/TransformTypes.td"
-
-include "mlir/Interfaces/SideEffectInterfaces.td"
-//===----------------------------------------------------------------------===//
-// ApplyOptimizeSharedMemoryReadsAndWritesOp
-//===----------------------------------------------------------------------===//
-
-def ApplyOptimizeSharedMemoryReadsAndWritesOp :
- Op<Transform_Dialect, "amdgpu.optimize_shared_memory_reads_and_writes",
- [DeclareOpInterfaceMethods<MemoryEffectsOpInterface>,
- TransformOpInterface, TransformEachOpTrait]> {
- let summary = "Reduce shared memory bank conflicts";
- let description = [{ This op attempts to optimize GPU Shared memory
- reads/writes with the goal of avoiding bank conflicts.
- }];
-
- let arguments = (ins TransformHandleTypeInterface:$target,
- DefaultValuedOptionalAttr<I64Attr, "128">:$sharedMemoryLineSizeBytes,
- DefaultValuedOptionalAttr<I64Attr, "128">:$defaultVectorSizeBits);
- let results = (outs);
-
- let assemblyFormat = "$target attr-dict `:` functional-type(operands, results)";
-
- let extraClassDeclaration = [{
- ::mlir::DiagnosedSilenceableFailure applyToOne(
- ::mlir::transform::TransformRewriter &rewriter,
- ::mlir::func::FuncOp funcOp,
- ::mlir::transform::ApplyToEachResultList &results,
- ::mlir::transform::TransformState &state);
- }];
-}
-
-#endif // AMDGPU_TRANSFORM_OPS
diff --git a/mlir/include/mlir/Dialect/AMDGPU/TransformOps/CMakeLists.txt b/mlir/include/mlir/Dialect/AMDGPU/TransformOps/CMakeLists.txt
deleted file mode 100644
index 07bfebc9f96d2e..00000000000000
--- a/mlir/include/mlir/Dialect/AMDGPU/TransformOps/CMakeLists.txt
+++ /dev/null
@@ -1,4 +0,0 @@
-set(LLVM_TARGET_DEFINITIONS AMDGPUTransformOps.td)
-mlir_tablegen(AMDGPUTransformOps.h.inc -gen-op-decls)
-mlir_tablegen(AMDGPUTransformOps.cpp.inc -gen-op-defs)
-add_public_tablegen_target(MLIRAMDGPUTransformOpsIncGen)
diff --git a/mlir/include/mlir/Dialect/AMDGPU/Transforms/Passes.h b/mlir/include/mlir/Dialect/AMDGPU/Transforms/Passes.h
index ab695756d2a789..8dd5ff1a4b198a 100644
--- a/mlir/include/mlir/Dialect/AMDGPU/Transforms/Passes.h
+++ b/mlir/include/mlir/Dialect/AMDGPU/Transforms/Passes.h
@@ -20,7 +20,7 @@ namespace mlir {
class ConversionTarget;
namespace amdgpu {
-#define GEN_PASS_DECL
+#define GEN_PASS_DECL_AMDGPUEMULATEATOMICSPASS
#define GEN_PASS_REGISTRATION
#include "mlir/Dialect/AMDGPU/Transforms/Passes.h.inc"
diff --git a/mlir/include/mlir/Dialect/AMDGPU/Transforms/Passes.td b/mlir/include/mlir/Dialect/AMDGPU/Transforms/Passes.td
index 67f951fd19d172..e6b27aa842dfcd 100644
--- a/mlir/include/mlir/Dialect/AMDGPU/Transforms/Passes.td
+++ b/mlir/include/mlir/Dialect/AMDGPU/Transforms/Passes.td
@@ -30,24 +30,4 @@ def AmdgpuEmulateAtomicsPass : Pass<"amdgpu-emulate-atomics"> {
"Chipset that these operations will run on">];
}
-def OptimizeSharedMemory : Pass<"amdgpu-optimize-shared-memory"> {
- let summary = "Optimizes accesses to shared memory memrefs in order to reduce bank conflicts.";
- let description = [{
- This pass adds a transformation and pass to the AMDGPU dialect that
- attempts to optimize reads/writes from a memref representing GPU shared
- memory in order to avoid bank conflicts.
- }];
- let dependentDialects = [
- "memref::MemRefDialect", "vector::VectorDialect"
- ];
- let options = [
- Option<"sharedMemoryLineSizeBytes", "shared-memory-line-size-bytes", "int64_t",
- /*default=*/"128",
- "Shared memory line size in bytes">,
- Option<"defaultVectorSizeBits", "default-vector-size-bits", "int64_t",
- /*default=*/"128",
- "Default vector size in bits">,
- ];
-}
-
#endif // MLIR_DIALECT_AMDGPU_TRANSFORMS_PASSES_TD_
diff --git a/mlir/include/mlir/Dialect/AMDGPU/Transforms/Transforms.h b/mlir/include/mlir/Dialect/AMDGPU/Transforms/Transforms.h
deleted file mode 100644
index 843cea2c503b9a..00000000000000
--- a/mlir/include/mlir/Dialect/AMDGPU/Transforms/Transforms.h
+++ /dev/null
@@ -1,61 +0,0 @@
-//===- Transforms.h - AMDGPU Dialect transformations -------------*- 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
-//
-//===----------------------------------------------------------------------===//
-//
-// This file declares functions that assist transformations for the amdgpu
-// dialect.
-//
-//===----------------------------------------------------------------------===//
-#ifndef MLIR_DIALECT_AMDGPU_TRANSFORMS_TRANSFORMS_H_
-#define MLIR_DIALECT_AMDGPU_TRANSFORMS_TRANSFORMS_H_
-
-#include "mlir/Dialect/Func/IR/FuncOps.h"
-#include "mlir/IR/Operation.h"
-#include "mlir/Support/LogicalResult.h"
-
-namespace mlir {
-class RewriterBase;
-
-namespace amdgpu {
-
-///
-/// Passes
-///
-
-/// Optimizes vectorized accesses to a shared memory buffer specified by
-/// memrefValue. This transformation assumes the following:
-/// 1) All relevant accesses to `memrefValue` are contained with `parentOp`.
-/// 2) The function will fail precondition checks if any subviews are
-/// taken of `memrefValue`. All reads/writes to `memrefValue` should occur
-/// through `memrefValue` directly.
-///
-/// Shared memory bank conflicts occur when multiple threads attempt to read or
-/// write locations assigned to the same shared memory bank. For `2^N` byte
-/// vectorized accesses, we need to be concerned with conflicts among threads
-/// identified as `(tid) -> tid.floordiv(2^{7-N})`. As such, this transformation
-/// changes any indexed memory access (vector.load, memref.load, etc)
-/// such that the final dimension's index value is permuted such that
-/// `newColIndex = oldColIndex % vectorSize +
-/// perm[rowIndex](oldColIndex/vectorSize, rowIndex)` where `rowIndex` is the
-/// index for the second-to last dimension and `perm[rowIndex]` is a permutation
-/// function that depends on the row Index. The permutation function is chosen
-/// to ensure that sequential distributed+vectorized reads/writes down a single
-/// dimension of the memref have minimal conflicts.
-LogicalResult
-optimizeSharedMemoryReadsAndWrites(Operation *parentOp, Value memrefValue,
- int64_t sharedMemoryLineSizeBytes,
- int64_t defaultVectorSizeBits);
-
-std::optional<LogicalResult>
-optimizeSharedMemoryReadsAndWritesOp(func::FuncOp funcOp,
- int64_t sharedMemoryLineSizeBytes,
- int64_t defaultVectorSizeBits);
-
-} // namespace amdgpu
-} // namespace mlir
-
-#endif // MLIR_DIALECT_AMDGPU_TRANSFORMS_TRANSFORMS_H_
diff --git a/mlir/include/mlir/Dialect/AMDGPU/Transforms/Utils.h b/mlir/include/mlir/Dialect/AMDGPU/Transforms/Utils.h
deleted file mode 100644
index 9e5e9589d62f35..00000000000000
--- a/mlir/include/mlir/Dialect/AMDGPU/Transforms/Utils.h
+++ /dev/null
@@ -1,24 +0,0 @@
-//===- Utils.h - Transform utilities -----------------------------*- 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/IR/Operation.h"
-
-namespace mlir {
-namespace amdgpu {
-
-/// Get and set the indices that the given load/store operation is operating on.
-/// Preconditions:
-/// - The Op must have memory affects.
-/// - Considers memref::LoadOp, vector::LoadOp, and vector::TransferReadOp.
-/// - Considers memref::StoreOp, vector::StoreOp, and vector::TransferWriteOp.
-/// - Excludes subview op.
-std::optional<Operation::operand_range> getIndices(Operation *op);
-void setIndices(Operation *op, ArrayRef<Value> indices);
-
-} // namespace amdgpu
-} // namespace mlir
diff --git a/mlir/include/mlir/InitAllExtensions.h b/mlir/include/mlir/InitAllExtensions.h
index b31fb26f00f8f4..7708ca5571de3b 100644
--- a/mlir/include/mlir/InitAllExtensions.h
+++ b/mlir/include/mlir/InitAllExtensions.h
@@ -23,7 +23,6 @@
#include "mlir/Conversion/MemRefToLLVM/MemRefToLLVM.h"
#include "mlir/Conversion/NVVMToLLVM/NVVMToLLVM.h"
#include "mlir/Conversion/UBToLLVM/UBToLLVM.h"
-#include "mlir/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.h"
#include "mlir/Dialect/Affine/TransformOps/AffineTransformOps.h"
#include "mlir/Dialect/Bufferization/TransformOps/BufferizationTransformOps.h"
#include "mlir/Dialect/Func/Extensions/AllExtensions.h"
@@ -67,7 +66,6 @@ inline void registerAllExtensions(DialectRegistry ®istry) {
ub::registerConvertUBToLLVMInterface(registry);
// Register all transform dialect extensions.
- amdgpu::registerTransformDialectExtension(registry);
affine::registerTransformDialectExtension(registry);
bufferization::registerTransformDialectExtension(registry);
func::registerTransformDialectExtension(registry);
diff --git a/mlir/lib/Dialect/AMDGPU/CMakeLists.txt b/mlir/lib/Dialect/AMDGPU/CMakeLists.txt
index c47e4c5495c17b..31167e6af908b9 100644
--- a/mlir/lib/Dialect/AMDGPU/CMakeLists.txt
+++ b/mlir/lib/Dialect/AMDGPU/CMakeLists.txt
@@ -1,4 +1,3 @@
add_subdirectory(IR)
-add_subdirectory(TransformOps)
add_subdirectory(Transforms)
add_subdirectory(Utils)
diff --git a/mlir/lib/Dialect/AMDGPU/IR/AMDGPUDialect.cpp b/mlir/lib/Dialect/AMDGPU/IR/AMDGPUDialect.cpp
index 4e72fbf56b80a4..2575ad4984814b 100644
--- a/mlir/lib/Dialect/AMDGPU/IR/AMDGPUDialect.cpp
+++ b/mlir/lib/Dialect/AMDGPU/IR/AMDGPUDialect.cpp
@@ -43,21 +43,6 @@ void AMDGPUDialect::initialize() {
>();
}
-bool amdgpu::AMDGPUDialect::isSharedMemoryAddressSpace(Attribute memorySpace) {
- if (!memorySpace)
- return false;
- if (auto intAttr = llvm::dyn_cast<IntegerAttr>(memorySpace))
- return intAttr.getInt() == AMDGPUDialect::kSharedMemoryAddressSpace;
- if (auto gpuAttr = llvm::dyn_cast<gpu::AddressSpaceAttr>(memorySpace))
- return gpuAttr.getValue() == gpu::AddressSpace::Workgroup;
- return false;
-}
-
-bool amdgpu::AMDGPUDialect::hasSharedMemoryAddressSpace(MemRefType type) {
- Attribute memorySpace = type.getMemorySpace();
- return isSharedMemoryAddressSpace(memorySpace);
-}
-
//===----------------------------------------------------------------------===//
// 8-bit float ops
//===----------------------------------------------------------------------===//
diff --git a/mlir/lib/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.cpp b/mlir/lib/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.cpp
deleted file mode 100644
index b7e17a92897389..00000000000000
--- a/mlir/lib/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.cpp
+++ /dev/null
@@ -1,67 +0,0 @@
-//===- AMDGPUTransformOps.cpp - Implementation of AMDGPU transform 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/AMDGPU/TransformOps/AMDGPUTransformOps.h"
-
-#include "mlir/Dialect/AMDGPU/IR/AMDGPUDialect.h"
-#include "mlir/Dialect/AMDGPU/Transforms/Transforms.h"
-#include "mlir/Dialect/Affine/IR/AffineOps.h"
-#include "mlir/Dialect/Vector/IR/VectorOps.h"
-
-using namespace mlir;
-using namespace mlir::amdgpu;
-using namespace mlir::transform;
-using namespace mlir::func;
-
-#define DEBUG_TYPE "amdgpu-transforms"
-#define DBGS() (llvm::dbgs() << "[" DEBUG_TYPE "]: ")
-#define DBGSNL() (llvm::dbgs() << "\n")
-#define LDBG(X) LLVM_DEBUG(DBGS() << (X) << "\n")
-
-DiagnosedSilenceableFailure
-ApplyOptimizeSharedMemoryReadsAndWritesOp::applyToOne(
- TransformRewriter &rewriter, FuncOp funcOp, ApplyToEachResultList &results,
- TransformState &state) {
- optimizeSharedMemoryReadsAndWritesOp(funcOp, getSharedMemoryLineSizeBytes(),
- getDefaultVectorSizeBits());
- return DiagnosedSilenceableFailure::success();
-}
-
-void ApplyOptimizeSharedMemoryReadsAndWritesOp::getEffects(
- SmallVectorImpl<MemoryEffects::EffectInstance> &effects) {
- onlyReadsHandle(getTarget(), effects);
- modifiesPayload(effects);
-}
-
-//===----------------------------------------------------------------------===//
-// Transform op registration
-//===----------------------------------------------------------------------===//
-
-namespace {
-class AMDGPUTransformDialectExtension
- : public TransformDialectExtension<AMDGPUTransformDialectExtension> {
-public:
- AMDGPUTransformDialectExtension() {
- declareGeneratedDialect<arith::ArithDialect>();
- declareGeneratedDialect<affine::AffineDialect>();
- declareGeneratedDialect<amdgpu::AMDGPUDialect>();
- declareGeneratedDialect<vector::VectorDialect>();
- registerTransformOps<
-#define GET_OP_LIST
-#include "mlir/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.cpp.inc"
- >();
- }
-};
-} // namespace
-
-#define GET_OP_CLASSES
-#include "mlir/Dialect/AMDGPU/TransformOps/AMDGPUTransformOps.cpp.inc"
-
-void amdgpu::registerTransformDialectExtension(DialectRegistry ®istry) {
- registry.addExtensions<AMDGPUTransformDialectExtension>();
-}
diff --git a/mlir/lib/Dialect/AMDGPU/TransformOps/CMakeLists.txt b/mlir/lib/Dialect/AMDGPU/TransformOps/CMakeLists.txt
deleted file mode 100644
index c39a3b55eabca4..00000000000000
--- a/mlir/lib/Dialect/AMDGPU/TransformOps/CMakeLists.txt
+++ /dev/null
@@ -1,25 +0,0 @@
-add_mlir_dialect_library(MLIRAMDGPUTransformOps
- AMDGPUTransformOps.cpp
-
- ADDITIONAL_HEADER_DIRS
- ${MLIR_MAIN_INCLUDE_DIR}/mlir/Dialect/AMDGPU/TransformOps
-
- DEPENDS
- MLIRAMDGPUTransformOpsIncGen
-
- LINK_LIBS PUBLIC
- MLIRAffineDialect
- MLIRArithDialect
- MLIRIR
- MLIRLinalgDialect
- MLIRAMDGPUDialect
- MLIRAMDGPUTransforms
- MLIRParser
- MLIRSideEffectInterfaces
- MLIRSCFDialect
- MLIRSCFTransforms
- MLIRTransformDialect
- MLIRTransformDialectUtils
- MLIRVectorTransforms
-
- )
diff --git a/mlir/lib/Dialect/AMDGPU/Transforms/CMakeLists.txt b/mlir/lib/Dialect/AMDGPU/Transforms/CMakeLists.txt
index a955d585b9a1dc..0889a21bddc44c 100644
--- a/mlir/lib/Dialect/AMDGPU/Transforms/CMakeLists.txt
+++ b/mlir/lib/Dialect/AMDGPU/Transforms/CMakeLists.txt
@@ -1,7 +1,5 @@
add_mlir_dialect_library(MLIRAMDGPUTransforms
EmulateAtomics.cpp
- OptimizeSharedMemory.cpp
- Utils.cpp
ADDITIONAL_HEADER_DIRS
{$MLIR_MAIN_INCLUDE_DIR}/mlir/Dialect/AMDGPU/Transforms
@@ -19,5 +17,4 @@ add_mlir_dialect_library(MLIRAMDGPUTransforms
MLIRPass
MLIRTransforms
MLIRTransformUtils
- MLIRVectorDialect
)
diff --git a/mlir/lib/Dialect/AMDGPU/Transforms/OptimizeSharedMemory.cpp b/mlir/lib/Dialect/AMDGPU/Transforms/OptimizeSharedMemory.cpp
deleted file mode 100644
index 32fab265e03cc0..00000000000000
--- a/mlir/lib/Dialect/AMDGPU/Transforms/OptimizeSharedMemory.cpp
+++ /dev/null
@@ -1,261 +0,0 @@
-//===- OptimizeSharedMemory.cpp - MLIR AMDGPU pass implementation ---------===//
-//
-// 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 transforms to optimize accesses to shared memory.
-// It is inspired by
-// https://github.com/llvm/llvm-project/blob/main/mlir/lib/Dialect/NVGPU/Transforms/OptimizeSharedMemory.cpp
-//
-//===----------------------------------------------------------------------===//
-
-#include "mlir/Dialect/AMDGPU/Transforms/Passes.h"
-
-#include "mlir/Dialect/AMDGPU/IR/AMDGPUDialect.h"
-#include "mlir/Dialect/AMDGPU/Transforms/Transforms.h"
-#include ...
[truncated]
|
antiagainst
approved these changes
Apr 10, 2024
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.
This implementation has a number of issues and ultimately does not work on gfx9.
memref.subviewin the root op, even when the subview is not a user of any of thememref.allocops.I do not see how these can be easily fixed, therefore I think it's better to delete this code.