[mlir][sparse] move sparse tensor rewriting into its own pass

Makes individual testing and debugging easier.

Reviewed By: bixia

Differential Revision: https://reviews.llvm.org/D135319

mlir/include/mlir/Dialect/SparseTensor/Transforms/Passes.h

@@ -163,6 +163,10 @@ std::unique_ptr<Pass> createSparseTensorCodegenPass();
 
 void populateSparseTensorRewriting(RewritePatternSet &patterns, bool enableRT);
 
+std::unique_ptr<Pass> createSparseTensorRewritePass();
+std::unique_ptr<Pass>
+createSparseTensorRewritePass(const SparsificationOptions &options);
+
 std::unique_ptr<Pass> createDenseBufferizationPass(
     const bufferization::OneShotBufferizationOptions &options);
 

mlir/include/mlir/Dialect/SparseTensor/Transforms/Passes.td

@@ -11,6 +11,27 @@
 
 include "mlir/Pass/PassBase.td"
 
+def SparseTensorRewrite : Pass<"sparse-tensor-rewrite", "ModuleOp"> {
+  let summary = "Applies sparse tensor rewriting rules prior to sparsification";
+  let description = [{
+    A pass that applies rewriting rules to sparse tensor operations prior
+    to running the actual sparsification pass.
+  }];
+  let constructor = "mlir::createSparseTensorRewritePass()";
+  let dependentDialects = [
+    "arith::ArithDialect",
+    "bufferization::BufferizationDialect",
+    "linalg::LinalgDialect",
+    "memref::MemRefDialect",
+    "scf::SCFDialect",
+    "sparse_tensor::SparseTensorDialect",
+  ];
+  let options = [
+    Option<"enableRuntimeLibrary", "enable-runtime-library", "bool",
+           "true", "Enable runtime library for manipulating sparse tensors">
+  ];
+}
+
 def SparsificationPass : Pass<"sparsification", "ModuleOp"> {
   let summary = "Automatically generate sparse tensor code from sparse tensor types";
   let description = [{
@@ -57,6 +78,7 @@ def SparsificationPass : Pass<"sparsification", "ModuleOp"> {
     "arith::ArithDialect",
     "bufferization::BufferizationDialect",
     "LLVM::LLVMDialect",
+    "linalg::LinalgDialect",
     "memref::MemRefDialect",
     "scf::SCFDialect",
     "sparse_tensor::SparseTensorDialect",
@@ -193,4 +215,5 @@ def SparseBufferRewrite : Pass<"sparse-buffer-rewrite", "ModuleOp"> {
     "sparse_tensor::SparseTensorDialect",
   ];
 }
+
 #endif // MLIR_DIALECT_SPARSETENSOR_TRANSFORMS_PASSES

mlir/lib/Dialect/SparseTensor/Pipelines/SparseTensorPipelines.cpp

@@ -58,6 +58,7 @@ void mlir::sparse_tensor::buildSparseCompiler(
           /*analysisOnly=*/options.testBufferizationAnalysisOnly)));
   if (options.testBufferizationAnalysisOnly)
     return;
+  pm.addPass(createSparseTensorRewritePass(options.sparsificationOptions()));
   pm.addPass(createSparsificationPass(options.sparsificationOptions()));
   if (options.enableRuntimeLibrary)
     pm.addPass(createSparseTensorConversionPass(

mlir/lib/Dialect/SparseTensor/Transforms/SparseTensorPasses.cpp

@@ -21,6 +21,7 @@
 #include "mlir/Transforms/GreedyPatternRewriteDriver.h"
 
 namespace mlir {
+#define GEN_PASS_DEF_SPARSETENSORREWRITE
 #define GEN_PASS_DEF_SPARSIFICATIONPASS
 #define GEN_PASS_DEF_SPARSETENSORCONVERSIONPASS
 #define GEN_PASS_DEF_SPARSETENSORCODEGEN
@@ -37,6 +38,23 @@ namespace {
 // Passes implementation.
 //===----------------------------------------------------------------------===//
 
+struct SparseTensorRewritePass
+    : public impl::SparseTensorRewriteBase<SparseTensorRewritePass> {
+
+  SparseTensorRewritePass() = default;
+  SparseTensorRewritePass(const SparseTensorRewritePass &pass) = default;
+  SparseTensorRewritePass(const SparsificationOptions &options) {
+    enableRuntimeLibrary = options.enableRuntimeLibrary;
+  }
+
+  void runOnOperation() override {
+    auto *ctx = &getContext();
+    RewritePatternSet patterns(ctx);
+    populateSparseTensorRewriting(patterns, enableRuntimeLibrary);
+    (void)applyPatternsAndFoldGreedily(getOperation(), std::move(patterns));
+  }
+};
+
 struct SparsificationPass
     : public impl::SparsificationPassBase<SparsificationPass> {
 
@@ -53,14 +71,10 @@ struct SparsificationPass
 
   void runOnOperation() override {
     auto *ctx = &getContext();
-    RewritePatternSet prePatterns(ctx);
     // Translate strategy flags to strategy options.
     SparsificationOptions options(parallelization, vectorization, vectorLength,
                                   enableSIMDIndex32, enableVLAVectorization,
                                   enableRuntimeLibrary);
-    // Apply pre-rewriting.
-    populateSparseTensorRewriting(prePatterns, options.enableRuntimeLibrary);
-    (void)applyPatternsAndFoldGreedily(getOperation(), std::move(prePatterns));
     // Apply sparsification and vector cleanup rewriting.
     RewritePatternSet patterns(ctx);
     populateSparsificationPatterns(patterns, options);
@@ -236,6 +250,15 @@ mlir::sparseToSparseConversionStrategy(int32_t flag) {
 // Pass creation methods.
 //===----------------------------------------------------------------------===//
 
+std::unique_ptr<Pass> mlir::createSparseTensorRewritePass() {
+  return std::make_unique<SparseTensorRewritePass>();
+}
+
+std::unique_ptr<Pass>
+mlir::createSparseTensorRewritePass(const SparsificationOptions &options) {
+  return std::make_unique<SparseTensorRewritePass>(options);
+}
+
 std::unique_ptr<Pass> mlir::createSparsificationPass() {
   return std::make_unique<SparsificationPass>();
 }

mlir/test/Dialect/SparseTensor/rewriting.mlir

@@ -1,4 +1,4 @@
-// RUN: mlir-opt %s -sparsification | FileCheck %s
+// RUN: mlir-opt %s -sparse-tensor-rewrite | FileCheck %s
 
 #SparseVector = #sparse_tensor.encoding<{
   dimLevelType = ["compressed"]

mlir/test/Dialect/SparseTensor/sparse_concat_codegen.mlir

@@ -1,4 +1,4 @@
-// RUN: mlir-opt %s --sparsification=enable-runtime-library=false | FileCheck %s
+// RUN: mlir-opt %s --sparse-tensor-rewrite=enable-runtime-library=false --sparsification | FileCheck %s
 
 #DCSR = #sparse_tensor.encoding<{dimLevelType = ["compressed", "compressed"]}>
 

mlir/test/Dialect/SparseTensor/sparse_fill_zero.mlir

@@ -1,4 +1,4 @@
-// RUN: mlir-opt %s --linalg-generalize-named-ops --sparsification --sparse-tensor-conversion --canonicalize --cse | FileCheck %s
+// RUN: mlir-opt %s --linalg-generalize-named-ops --sparse-tensor-rewrite --sparsification --sparse-tensor-conversion --canonicalize --cse | FileCheck %s
 
 #DCSR = #sparse_tensor.encoding<{ dimLevelType = [ "compressed", "compressed" ] }>
 

mlir/test/Dialect/SparseTensor/sparse_sddmm.mlir

@@ -1,4 +1,4 @@
-// RUN: mlir-opt %s  --tensor-copy-insertion --sparsification --cse | FileCheck %s
+// RUN: mlir-opt %s  --tensor-copy-insertion --sparse-tensor-rewrite --sparsification --cse | FileCheck %s
 
 #SM = #sparse_tensor.encoding<{ dimLevelType = [ "compressed", "compressed" ] }>