[Mlir-commits] [mlir] [MLIR] Implement emulation of static indexing subbyte type vector stores (PR #115922)

llvmlistbot at llvm.org llvmlistbot at llvm.org
Tue Dec 10 06:01:53 PST 2024


https://github.com/lialan updated https://github.com/llvm/llvm-project/pull/115922

>From 72bdea0cc8c0d8975c89588e404245a82a9e35c0 Mon Sep 17 00:00:00 2001
From: Ubuntu <450283+lialan at users.noreply.github.com>
Date: Fri, 25 Oct 2024 15:19:42 +0000
Subject: [PATCH 01/10] Implement vector stores

---
 .../Vector/Transforms/VectorRewritePatterns.h |   5 +-
 .../Transforms/VectorEmulateNarrowType.cpp    | 265 +++++++++++++++---
 ...tor-emulate-narrow-type-unaligned-rmw.mlir | 104 +++++++
 .../vector-emulate-narrow-type-unaligned.mlir |  90 ++++++
 .../Dialect/MemRef/TestEmulateNarrowType.cpp  |   8 +-
 5 files changed, 436 insertions(+), 36 deletions(-)
 create mode 100644 mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned-rmw.mlir

diff --git a/mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h b/mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h
index a59f06f3c1ef1b..64bb3a2204cfdc 100644
--- a/mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h
+++ b/mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h
@@ -364,10 +364,11 @@ void populateVectorMaskMaterializationPatterns(RewritePatternSet &patterns,
                                                PatternBenefit benefit = 1);
 
 /// Appends patterns for emulating vector operations over narrow types with ops
-/// over wider types.
+/// over wider types. `useAtomicWrites` indicates whether to use atomic
+/// operations in the places where thread contention is possible.
 void populateVectorNarrowTypeEmulationPatterns(
     const arith::NarrowTypeEmulationConverter &typeConverter,
-    RewritePatternSet &patterns);
+    RewritePatternSet &patterns, bool useAtomicWrites = true);
 
 /// Rewrite a vector `bitcast(trunci)` to use a more efficient sequence of
 /// vector operations comprising `shuffle` and `bitwise` ops.
diff --git a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
index 87c30a733c363e..278b42a5b7c104 100644
--- a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
+++ b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
@@ -33,6 +33,7 @@
 #include "mlir/Transforms/DialectConversion.h"
 #include "llvm/ADT/SmallVector.h"
 #include "llvm/Support/Debug.h"
+#include "llvm/Support/LogicalResult.h"
 #include "llvm/Support/MathExtras.h"
 #include "llvm/Support/raw_ostream.h"
 #include <cstdint>
@@ -211,13 +212,10 @@ static FailureOr<Operation *> getCompressedMaskOp(OpBuilder &rewriter,
 /// Extracts 1-D subvector from a 1-D vector. It is a wrapper function for
 /// emitting `vector.extract_strided_slice`.
 static Value staticallyExtractSubvector(OpBuilder &rewriter, Location loc,
-                                        VectorType extractType, Value source,
-                                        int64_t frontOffset,
+                                        Value source, int64_t frontOffset,
                                         int64_t subvecSize) {
   auto vectorType = cast<VectorType>(source.getType());
-  assert((vectorType.getRank() == 1 && extractType.getRank() == 1) &&
-         "expected 1-D source and destination types");
-  (void)vectorType;
+  assert(vectorType.getRank() == 1 && "expected 1-D source types");
   assert(frontOffset + subvecSize <= vectorType.getNumElements() &&
          "subvector out of bounds");
 
@@ -228,9 +226,12 @@ static Value staticallyExtractSubvector(OpBuilder &rewriter, Location loc,
   auto offsets = rewriter.getI64ArrayAttr({frontOffset});
   auto sizes = rewriter.getI64ArrayAttr({subvecSize});
   auto strides = rewriter.getI64ArrayAttr({1});
+
+  auto resultVectorType =
+      VectorType::get({subvecSize}, vectorType.getElementType());
   return rewriter
-      .create<vector::ExtractStridedSliceOp>(loc, extractType, source, offsets,
-                                             sizes, strides)
+      .create<vector::ExtractStridedSliceOp>(loc, resultVectorType, source,
+                                             offsets, sizes, strides)
       ->getResult(0);
 }
 
@@ -309,6 +310,76 @@ emulatedVectorLoad(OpBuilder &rewriter, Location loc, Value base,
       newLoad);
 }
 
+/// Atomically store a subbyte-sized value to memory, with a mask.
+static void atomicStore(OpBuilder &builder, Location loc,
+                        TypedValue<MemRefType> emulatedMemref,
+                        Value linearizedIndex, TypedValue<VectorType> value,
+                        Value mask, int64_t) {
+  auto atomicOp = builder.create<memref::GenericAtomicRMWOp>(
+      loc, emulatedMemref, ValueRange{linearizedIndex});
+  Value origValue = atomicOp.getCurrentValue();
+
+  OpBuilder::InsertionGuard guard(builder);
+  builder.setInsertionPointToStart(atomicOp.getBody());
+
+  // i8 -> <1xi8> -> <numSrcElemsPerDest x i.>
+  auto oneVectorType = VectorType::get({1}, origValue.getType());
+  auto fromElem = builder.create<vector::FromElementsOp>(loc, oneVectorType,
+                                                         ValueRange{origValue});
+  auto vectorBitCast =
+      builder.create<vector::BitCastOp>(loc, value.getType(), fromElem);
+
+  auto select =
+      builder.create<arith::SelectOp>(loc, mask, value, vectorBitCast);
+  auto bitcast2 = builder.create<vector::BitCastOp>(loc, oneVectorType, select);
+  auto extract = builder.create<vector::ExtractOp>(loc, bitcast2, 0);
+  builder.create<memref::AtomicYieldOp>(loc, extract.getResult());
+}
+
+/// Generate a non-atomic read-modify-write sequence for subbyte storing.
+static void rmwStore(OpBuilder &rewriter, Location loc,
+                     TypedValue<MemRefType> emulatedMemref,
+                     Value linearizedIndex, TypedValue<VectorType> value,
+                     Value mask, int64_t numSrcElemsPerDest) {
+  auto emulatedIOType =
+      VectorType::get({1}, emulatedMemref.getType().getElementType());
+  auto elemLoad = rewriter.create<vector::LoadOp>(
+      loc, emulatedIOType, emulatedMemref, ValueRange{linearizedIndex});
+  auto fromBitcast = rewriter.create<vector::BitCastOp>(
+      loc,
+      VectorType::get({numSrcElemsPerDest}, value.getType().getElementType()),
+      elemLoad);
+  auto select = rewriter.create<arith::SelectOp>(loc, mask, fromBitcast, value);
+  auto toBitcast =
+      rewriter.create<vector::BitCastOp>(loc, emulatedIOType, select);
+  rewriter.create<vector::StoreOp>(loc, toBitcast, emulatedMemref,
+                                   linearizedIndex);
+}
+
+static_assert(std::is_same_v<decltype(atomicStore), decltype(rmwStore)> &&
+              "`atomicStore` and `rmwStore` must have same signature, as per "
+              "the design to keep the code clean, which one to call is "
+              "determined by the `useAtomicWrites` flag.");
+
+// Extract a slice of a vector, and insert it into a byte vector.
+static Value extractSliceIntoByte(ConversionPatternRewriter &rewriter,
+                                  Location loc, TypedValue<VectorType> vector,
+                                  int64_t sliceOffset, int64_t sliceNumElements,
+                                  int64_t byteOffset) {
+  auto vectorElementType = vector.getType().getElementType();
+  assert(8 % vectorElementType.getIntOrFloatBitWidth() == 0 &&
+         "vector element must be a valid sub-byte type");
+  auto scale = 8 / vectorElementType.getIntOrFloatBitWidth();
+  auto emptyByteVector = rewriter.create<arith::ConstantOp>(
+      loc, VectorType::get({scale}, vectorElementType),
+      rewriter.getZeroAttr(VectorType::get({scale}, vectorElementType)));
+  auto extracted = staticallyExtractSubvector(rewriter, loc, vector,
+                                              sliceOffset, sliceNumElements);
+  auto inserted = staticallyInsertSubvector(rewriter, loc, extracted,
+                                            emptyByteVector, byteOffset);
+  return inserted;
+}
+
 namespace {
 
 //===----------------------------------------------------------------------===//
@@ -318,6 +389,10 @@ namespace {
 struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
   using OpConversionPattern::OpConversionPattern;
 
+  ConvertVectorStore(MLIRContext *context, bool useAtomicWrites)
+      : OpConversionPattern<vector::StoreOp>(context),
+        useAtomicWrites_(useAtomicWrites) {}
+
   LogicalResult
   matchAndRewrite(vector::StoreOp op, OpAdaptor adaptor,
                   ConversionPatternRewriter &rewriter) const override {
@@ -329,8 +404,9 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
 
     auto loc = op.getLoc();
     auto convertedType = cast<MemRefType>(adaptor.getBase().getType());
-    Type oldElementType = op.getValueToStore().getType().getElementType();
-    Type newElementType = convertedType.getElementType();
+    auto valueToStore = cast<TypedValue<VectorType>>(op.getValueToStore());
+    auto oldElementType = valueToStore.getType().getElementType();
+    auto newElementType = convertedType.getElementType();
     int srcBits = oldElementType.getIntOrFloatBitWidth();
     int dstBits = newElementType.getIntOrFloatBitWidth();
 
@@ -338,7 +414,7 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
       return rewriter.notifyMatchFailure(
           op, "only dstBits % srcBits == 0 supported");
     }
-    int scale = dstBits / srcBits;
+    int numSrcElemsPerDest = dstBits / srcBits;
 
     // Adjust the number of elements to store when emulating narrow types.
     // Here only the 1-D vector store is considered, and the N-D memref types
@@ -353,15 +429,15 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
     // vector.store %bitcast, %alloc[%linear_index] : memref<16xi8>,
     // vector<4xi8>
 
-    auto origElements = op.getValueToStore().getType().getNumElements();
-    if (origElements % scale != 0)
-      return failure();
+    auto origElements = valueToStore.getType().getNumElements();
+    bool isUnalignedEmulation = origElements % numSrcElemsPerDest != 0;
 
     auto stridedMetadata =
         rewriter.create<memref::ExtractStridedMetadataOp>(loc, op.getBase());
 
     OpFoldResult linearizedIndices;
-    std::tie(std::ignore, linearizedIndices) =
+    memref::LinearizedMemRefInfo linearizedInfo;
+    std::tie(linearizedInfo, linearizedIndices) =
         memref::getLinearizedMemRefOffsetAndSize(
             rewriter, loc, srcBits, dstBits,
             stridedMetadata.getConstifiedMixedOffset(),
@@ -369,16 +445,137 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
             stridedMetadata.getConstifiedMixedStrides(),
             getAsOpFoldResult(adaptor.getIndices()));
 
-    auto numElements = origElements / scale;
-    auto bitCast = rewriter.create<vector::BitCastOp>(
-        loc, VectorType::get(numElements, newElementType),
-        op.getValueToStore());
+    auto foldedNumFrontPadElems =
+        isUnalignedEmulation
+            ? getConstantIntValue(linearizedInfo.intraDataOffset)
+            : 0;
 
-    rewriter.replaceOpWithNewOp<vector::StoreOp>(
-        op, bitCast.getResult(), adaptor.getBase(),
-        getValueOrCreateConstantIndexOp(rewriter, loc, linearizedIndices));
+    if (!foldedNumFrontPadElems) {
+      // Unimplemented case for dynamic front padding size != 0
+      return failure();
+    }
+
+    auto emulatedMemref = cast<TypedValue<MemRefType>>(adaptor.getBase());
+
+    // Shortcut: conditions when subbyte store at the front is not needed:
+    // 1. The source vector size is multiple of byte size
+    // 2. The address of the store is aligned to the emulated width boundary
+    if (!isUnalignedEmulation && *foldedNumFrontPadElems == 0) {
+      auto numElements = origElements / numSrcElemsPerDest;
+      auto bitCast = rewriter.create<vector::BitCastOp>(
+          loc, VectorType::get(numElements, newElementType),
+          op.getValueToStore());
+      rewriter.replaceOpWithNewOp<vector::StoreOp>(
+          op, bitCast.getResult(), emulatedMemref,
+          getValueOrCreateConstantIndexOp(rewriter, loc, linearizedIndices));
+      return success();
+    }
+
+    // The index into the target memref we are storing to
+    Value currentDestIndex =
+        getValueOrCreateConstantIndexOp(rewriter, loc, linearizedIndices);
+    auto constantOne = rewriter.create<arith::ConstantIndexOp>(loc, 1);
+    auto subWidthStoreMaskType =
+        VectorType::get({numSrcElemsPerDest}, rewriter.getI1Type());
+    // The index into the source vector we are currently processing
+    auto currentSourceIndex = 0;
+
+    // 1. Partial width store for the first byte, when the store address is not
+    // aligned to emulated width boundary, deal with the unaligned part so that
+    // the rest elements are aligned to width boundary.
+    auto frontSubWidthStoreElem =
+        (numSrcElemsPerDest - *foldedNumFrontPadElems) % numSrcElemsPerDest;
+    if (frontSubWidthStoreElem != 0) {
+      SmallVector<bool> frontMaskValues(numSrcElemsPerDest, false);
+      if (*foldedNumFrontPadElems + origElements < numSrcElemsPerDest) {
+        std::fill_n(frontMaskValues.begin() + *foldedNumFrontPadElems,
+                    origElements, true);
+        frontSubWidthStoreElem = origElements;
+      } else {
+        std::fill_n(frontMaskValues.end() - frontSubWidthStoreElem,
+                    *foldedNumFrontPadElems, true);
+      }
+      auto frontMask = rewriter.create<arith::ConstantOp>(
+          loc, DenseElementsAttr::get(subWidthStoreMaskType, frontMaskValues));
+
+      currentSourceIndex = numSrcElemsPerDest - (*foldedNumFrontPadElems);
+      auto value =
+          extractSliceIntoByte(rewriter, loc, valueToStore, 0,
+                               frontSubWidthStoreElem, *foldedNumFrontPadElems);
+
+      subEmulatedWidthStore(rewriter, loc, emulatedMemref, currentDestIndex,
+                            cast<TypedValue<VectorType>>(value),
+                            frontMask.getResult(), numSrcElemsPerDest);
+
+      currentDestIndex = rewriter.create<arith::AddIOp>(
+          loc, rewriter.getIndexType(), currentDestIndex, constantOne);
+    }
+
+    if (currentSourceIndex >= origElements) {
+      rewriter.eraseOp(op);
+      return success();
+    }
+
+    // 2. Full width store. After the previous step, the store address is
+    // aligned to the emulated width boundary.
+    int64_t fullWidthStoreSize =
+        (origElements - currentSourceIndex) / numSrcElemsPerDest;
+    int64_t numNonFullWidthElements = fullWidthStoreSize * numSrcElemsPerDest;
+    if (fullWidthStoreSize != 0) {
+      auto fullWidthStorePart = staticallyExtractSubvector(
+          rewriter, loc, valueToStore, currentSourceIndex,
+          numNonFullWidthElements);
+
+      auto originType = dyn_cast<VectorType>(fullWidthStorePart.getType());
+      auto memrefElemType =
+          dyn_cast<MemRefType>(emulatedMemref.getType()).getElementType();
+      auto storeType = VectorType::get(
+          {originType.getNumElements() / numSrcElemsPerDest}, memrefElemType);
+      auto bitCast = rewriter.create<vector::BitCastOp>(loc, storeType,
+                                                        fullWidthStorePart);
+      rewriter.create<vector::StoreOp>(loc, bitCast.getResult(), emulatedMemref,
+                                       currentDestIndex);
+
+      currentSourceIndex += numNonFullWidthElements;
+      currentDestIndex = rewriter.create<arith::AddIOp>(
+          loc, rewriter.getIndexType(), currentDestIndex,
+          rewriter.create<arith::ConstantIndexOp>(loc, fullWidthStoreSize));
+    }
+
+    // 3. Deal with trailing elements that are aligned to the emulated width,
+    // but their length is smaller than the emulated width.
+    auto remainingElements = origElements - currentSourceIndex;
+    if (remainingElements != 0) {
+      auto subWidthStorePart = extractSliceIntoByte(
+          rewriter, loc, cast<TypedValue<VectorType>>(valueToStore),
+          currentSourceIndex, remainingElements, 0);
+
+      // Generate back mask
+      auto maskValues = SmallVector<bool>(numSrcElemsPerDest, 0);
+      std::fill_n(maskValues.begin(), remainingElements, 1);
+      auto backMask = rewriter.create<arith::ConstantOp>(
+          loc, DenseElementsAttr::get(subWidthStoreMaskType, maskValues));
+
+      subEmulatedWidthStore(rewriter, loc, emulatedMemref, currentDestIndex,
+                            cast<TypedValue<VectorType>>(subWidthStorePart),
+                            backMask.getResult(), numSrcElemsPerDest);
+    }
+
+    rewriter.eraseOp(op);
     return success();
   }
+
+  /// Store a subbyte-sized value to memory, with a mask. Depending on the
+  /// configuration, it could be an atomic store or an RMW sequence.
+  template <typename... Args>
+  void subEmulatedWidthStore(Args &&...args) const {
+    std::function<decltype(atomicStore)> storeFunc =
+        useAtomicWrites_ ? atomicStore : rmwStore;
+    storeFunc(std::forward<Args>(args)...);
+  }
+
+private:
+  const bool useAtomicWrites_;
 };
 
 //===----------------------------------------------------------------------===//
@@ -584,9 +781,8 @@ struct ConvertVectorLoad final : OpConversionPattern<vector::LoadOp> {
           rewriter, loc, dyn_cast<TypedValue<VectorType>>(result), resultVector,
           linearizedInfo.intraDataOffset, origElements);
     } else if (isUnalignedEmulation) {
-      result =
-          staticallyExtractSubvector(rewriter, loc, op.getType(), result,
-                                     *foldedIntraVectorOffset, origElements);
+      result = staticallyExtractSubvector(
+          rewriter, loc, result, *foldedIntraVectorOffset, origElements);
     }
     rewriter.replaceOp(op, result);
     return success();
@@ -745,9 +941,8 @@ struct ConvertVectorMaskedLoad final
           rewriter, loc, dyn_cast<TypedValue<VectorType>>(result),
           op.getPassThru(), linearizedInfo.intraDataOffset, origElements);
     } else if (isUnalignedEmulation) {
-      result =
-          staticallyExtractSubvector(rewriter, loc, op.getType(), result,
-                                     *foldedIntraVectorOffset, origElements);
+      result = staticallyExtractSubvector(
+          rewriter, loc, result, *foldedIntraVectorOffset, origElements);
     }
     rewriter.replaceOp(op, result);
 
@@ -830,9 +1025,8 @@ struct ConvertVectorTransferRead final
                                            linearizedInfo.intraDataOffset,
                                            origElements);
     } else if (isUnalignedEmulation) {
-      result =
-          staticallyExtractSubvector(rewriter, loc, op.getType(), result,
-                                     *foldedIntraVectorOffset, origElements);
+      result = staticallyExtractSubvector(
+          rewriter, loc, result, *foldedIntraVectorOffset, origElements);
     }
     rewriter.replaceOp(op, result);
 
@@ -1577,12 +1771,17 @@ struct RewriteVectorTranspose : OpRewritePattern<vector::TransposeOp> {
 
 void vector::populateVectorNarrowTypeEmulationPatterns(
     const arith::NarrowTypeEmulationConverter &typeConverter,
-    RewritePatternSet &patterns) {
+    RewritePatternSet &patterns, bool useAtomicWrites) {
 
-  // Populate `vector.*` conversion patterns.
-  patterns.add<ConvertVectorLoad, ConvertVectorMaskedLoad, ConvertVectorStore,
+  // Populate `vector.*` load conversion patterns.
+  patterns.add<ConvertVectorLoad, ConvertVectorMaskedLoad,
                ConvertVectorMaskedStore, ConvertVectorTransferRead>(
       typeConverter, patterns.getContext());
+
+  // Populate `vector.*` store conversion patterns. The caller can choose
+  // to avoid emitting atomic operations and reduce it to load-modify-write
+  // sequence for stores if it is known there are no thread contentions.
+  patterns.insert<ConvertVectorStore>(patterns.getContext(), useAtomicWrites);
 }
 
 void vector::populateVectorNarrowTypeRewritePatterns(
diff --git a/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned-rmw.mlir b/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned-rmw.mlir
new file mode 100644
index 00000000000000..fa4d9cb5e4d4c7
--- /dev/null
+++ b/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned-rmw.mlir
@@ -0,0 +1,104 @@
+// RUN: mlir-opt --test-emulate-narrow-int="arith-compute-bitwidth=1 memref-load-bitwidth=8 atomic-store=false" --cse --split-input-file %s | FileCheck %s
+
+// TODO: remove memref.alloc() in the tests to eliminate noises.
+// memref.alloc exists here because sub-byte vector data types such as i2
+// are currently not supported as input arguments.
+
+func.func @vector_store_i2_const_rmw(%arg0: vector<3xi2>) {
+    %0 = memref.alloc() : memref<3x3xi2>
+    %c0 = arith.constant 0 : index
+    %c2 = arith.constant 2 : index
+    vector.store %arg0, %0[%c2, %c0] :memref<3x3xi2>, vector<3xi2>
+    return
+}
+// CHECK: func @vector_store_i2_const_rmw(
+// CHECK-SAME: %[[ARG0:.+]]: vector<3xi2>)
+// CHECK: %[[ALLOC:.+]] = memref.alloc() : memref<3xi8>
+// CHECK: %[[C1:.+]] = arith.constant 1 : index
+// CHECK: %[[CST:.+]] = arith.constant dense<[false, false, true, true]>
+// CHECK: %[[CST0:.+]] = arith.constant dense<0> : vector<4xi2>
+// CHECK: %[[EXTRACT:.+]] = vector.extract_strided_slice %[[ARG0]]
+// CHECK-SAME: {offsets = [0], sizes = [2], strides = [1]} : vector<3xi2> to vector<2xi2>
+// CHECK: %[[INSERT:.+]] = vector.insert_strided_slice %[[EXTRACT]], %[[CST0]]
+// CHECK-SAME: {offsets = [2], strides = [1]} : vector<2xi2> into vector<4xi2>
+// CHECK: %[[LOAD:.+]] = vector.load
+
+// Actual part to do RMW sequence
+// CHECK: %[[UPCAST:.+]] = vector.bitcast %[[LOAD]] : vector<1xi8> to vector<4xi2>
+// CHECK: %[[SELECT:.+]] = arith.select %[[CST]], %[[UPCAST]], %[[INSERT]]
+// CHECK: %[[DOWNCAST:.+]] = vector.bitcast %[[SELECT]]
+// CHECK: vector.store %[[DOWNCAST]], %[[ALLOC]][%[[C1]]]
+
+// -----
+
+func.func @vector_store_i2_atomic(%arg0: vector<7xi2>) {
+    %0 = memref.alloc() : memref<3x7xi2>
+    %c0 = arith.constant 0 : index
+    %c1 = arith.constant 1 : index
+    vector.store %arg0, %0[%c1, %c0] :memref<3x7xi2>, vector<7xi2>
+    return
+}
+
+// CHECK: func @vector_store_i2_atomic(
+// CHECK-SAME: %[[ARG0:.+]]:
+// CHECK: %[[ALLOC:.+]] = memref.alloc() : memref<6xi8>
+// CHECK: %[[C1:.+]] = arith.constant 1 : index
+// CHECK: %[[CST:.+]] = arith.constant dense<[false, false, false, true]>
+// CHECK: %[[CST0:.+]] = arith.constant dense<0> : vector<4xi2>
+// CHECK: %[[EXTRACT:.+]] = vector.extract_strided_slice %[[ARG0]]
+// CHECK-SAME: {offsets = [0], sizes = [1], strides = [1]}
+// CHECK: %[[INSERT:.+]] = vector.insert_strided_slice %[[EXTRACT]], %[[CST0]]
+// CHECK-SAME: {offsets = [3], strides = [1]}
+// First sub-width RMW:
+// CHECK: %[[LOAD:.+]] = vector.load %[[ALLOC]][%[[C1]]]
+// CHECK: %[[UPCAST:.+]] = vector.bitcast %[[LOAD]] : vector<1xi8> to vector<4xi2>
+// CHECK: %[[SELECT:.+]] = arith.select %[[CST]], %[[UPCAST]], %[[INSERT]]
+// CHECK: %[[DOWNCAST:.+]] = vector.bitcast %[[SELECT]]
+// CHECK: vector.store %[[DOWNCAST]], %[[ALLOC]][%[[C1]]]
+
+// Full-width store:
+// CHECK: %[[INDEX:.+]] = arith.addi %[[C1]], %[[C1]]
+// CHECK: %[[EXTRACT1:.+]] = vector.extract_strided_slice %[[ARG0]]
+// CHECK-SAME: {offsets = [1], sizes = [4], strides = [1]}
+// CHECK: %[[BITCAST:.+]] = vector.bitcast %[[EXTRACT1]]
+// CHECK: vector.store %[[BITCAST]], %[[ALLOC]][%[[INDEX]]]
+
+// Second sub-width RMW:
+// CHECK: %[[INDEX2:.+]] = arith.addi %[[INDEX]], %[[C1]]
+// CHECK: %[[EXTRACT2:.+]] = vector.extract_strided_slice %[[ARG0]]
+// CHECK-SAME: {offsets = [5], sizes = [2], strides = [1]}
+// CHECK: %[[INSERT2:.+]] = vector.insert_strided_slice %[[EXTRACT2]]
+// CHECK-SAME: {offsets = [0], strides = [1]}
+// CHECK: %[[CST1:.+]] = arith.constant dense<[true, true, false, false]>
+// CHECK: %[[LOAD1:.+]] = vector.load %[[ALLOC]][%[[INDEX2]]]
+// CHECK: %[[UPCAST1:.+]] = vector.bitcast %[[LOAD1]]
+// CHECK: %[[SELECT1:.+]] = arith.select %[[CST1]], %[[UPCAST1]], %[[INSERT2]]
+// CHECK: %[[DOWNCAST1:.+]] = vector.bitcast %[[SELECT1]]
+// CHECK: vector.store %[[DOWNCAST1]], %[[ALLOC]][%[[INDEX2]]]
+
+// -----
+
+func.func @vector_store_i2_single_rmw(%arg0: vector<1xi2>) {
+    %0 = memref.alloc() : memref<4x1xi2>
+    %c0 = arith.constant 0 : index
+    %c1 = arith.constant 1 : index
+    vector.store %arg0, %0[%c1, %c0] :memref<4x1xi2>, vector<1xi2>
+    return
+}
+
+// in this test, only emit 1 rmw store
+// CHECK: func @vector_store_i2_single_rmw(
+// CHECK-SAME: %[[ARG0:.+]]: vector<1xi2>)
+// CHECK: %[[ALLOC:.+]] = memref.alloc() : memref<1xi8>
+// CHECK: %[[C0:.+]] = arith.constant 0 : index
+// CHECK: %[[C1:.+]] = arith.constant 1 : index
+// CHECK: %[[CST:.+]] = arith.constant dense<[false, true, false, false]>
+// CHECK: %[[CST0:.+]] = arith.constant dense<0> : vector<4xi2>
+// CHECK: %[[INSERT:.+]] = vector.insert_strided_slice %[[ARG0]], %[[CST0]]
+// CHECK-SAME: {offsets = [1], strides = [1]} : vector<1xi2> into vector<4xi2>
+// CHECK: %[[LOAD:.+]] = vector.load %[[ALLOC]][%[[C0]]] : memref<1xi8>, vector<1xi8>
+// CHECK: %[[UPCAST:.+]] = vector.bitcast %[[LOAD]] : vector<1xi8> to vector<4xi2>
+// CHECK: %[[SELECT:.+]] = arith.select %[[CST]], %[[UPCAST]], %[[INSERT]]
+// CHECK: %[[DOWNCAST:.+]] = vector.bitcast %[[SELECT]]
+// CHECK: vector.store %[[DOWNCAST]], %[[ALLOC]][%[[C0]]]
+
diff --git a/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned.mlir b/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned.mlir
index 721c8a8d5d2034..fd526ada6cb7b2 100644
--- a/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned.mlir
+++ b/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned.mlir
@@ -336,3 +336,93 @@ func.func @vector_maskedload_i4_constant_mask_unaligned(%passthru: vector<5xi2>)
 // CHECK: %[[RESULT:.+]] = vector.extract_strided_slice %[[SELECT]]
 // CHECK-SAME: {offsets = [1], sizes = [5], strides = [1]} : vector<8xi2> to vector<5xi2>
 // CHECK: return %[[RESULT]] : vector<5xi2>
+
+///----------------------------------------------------------------------------------------
+/// vector.store
+///----------------------------------------------------------------------------------------
+
+// -----
+
+func.func @vector_store_i2_atomic(%arg0: vector<7xi2>) {
+    %0 = memref.alloc() : memref<3x7xi2>
+    %c0 = arith.constant 0 : index
+    %c1 = arith.constant 1 : index
+    vector.store %arg0, %0[%c1, %c0] :memref<3x7xi2>, vector<7xi2>
+    return
+}
+
+// In this example, emit 2 atomic RMWs and 1 non-atomic store:
+// CHECK-LABEL: func @vector_store_i2_atomic(
+// CHECK-SAME: %[[ARG0:.+]]: vector<7xi2>)
+// CHECK: %[[ALLOC:.+]] = memref.alloc() : memref<6xi8>
+// CHECK: %[[C1:.+]] = arith.constant 1 : index
+// CHECK: %[[CST:.+]] = arith.constant dense<[false, false, false, true]> : vector<4xi1>
+// CHECK: %[[CST0:.+]] = arith.constant dense<0> : vector<4xi2>
+
+// First atomic RMW:
+// CHECK: %[[EXTRACT:.+]] = vector.extract_strided_slice %[[ARG0]]
+// CHECK-SAME: {offsets = [0], sizes = [1], strides = [1]} : vector<7xi2> to vector<1xi2>
+// CHECK: %[[INSERT:.+]] = vector.insert_strided_slice %[[EXTRACT]], %[[CST0]]
+// CHECK-SAME: {offsets = [3], strides = [1]} : vector<1xi2> into vector<4xi2>
+// CHECK: %[[ATOMIC_RMW:.+]] = memref.generic_atomic_rmw %[[ALLOC]][%[[C1]]] : memref<6xi8> {
+// CHECK: %[[ARG:.+]]: i8):
+// CHECK: %[[FROM_ELEM:.+]] = vector.from_elements %[[ARG]] : vector<1xi8>
+// CHECK: %[[BITCAST:.+]] = vector.bitcast %[[FROM_ELEM]] : vector<1xi8> to vector<4xi2>
+// CHECK: %[[SELECT:.+]] = arith.select %[[CST]], %[[INSERT]], %[[BITCAST]] : vector<4xi1>, vector<4xi2>
+// CHECK: %[[BITCAST2:.+]] = vector.bitcast %[[SELECT]] : vector<4xi2> to vector<1xi8>
+// CHECK: %[[EXTRACT2:.+]] = vector.extract %[[BITCAST2]][0] : i8 from vector<1xi8>
+// CHECK: memref.atomic_yield %[[EXTRACT2]] : i8
+
+// Non-atomic store:
+// CHECK: %[[ADDR:.+]] = arith.addi %[[C1]], %[[C1]] : index
+// CHECK: %[[EXTRACT2:.+]] = vector.extract_strided_slice %[[ARG0]]
+// CHECK-SAME: {offsets = [1], sizes = [4], strides = [1]} : vector<7xi2> to vector<4xi2>
+// CHECK: %[[BITCAST3:.+]] = vector.bitcast %[[EXTRACT2]] : vector<4xi2> to vector<1xi8>
+// CHECK: vector.store %[[BITCAST3]], %[[ALLOC]][%[[ADDR]]] : memref<6xi8>, vector<1xi8>
+
+// Second atomic RMW:
+// CHECK: %[[ADDR2:.+]] = arith.addi %[[ADDR]], %[[C1]] : index
+// CHECK: %[[EXTRACT3:.+]] = vector.extract_strided_slice %[[ARG0]]
+// CHECK-SAME: {offsets = [5], sizes = [2], strides = [1]} : vector<7xi2> to vector<2xi2>
+// CHECK: %[[INSERT2:.+]] = vector.insert_strided_slice %[[EXTRACT3]], %[[CST0]]
+// CHECK-SAME: {offsets = [0], strides = [1]} : vector<2xi2> into vector<4xi2>
+// CHECK: %[[CST1:.+]] = arith.constant dense<[true, true, false, false]> : vector<4xi1> 
+// CHECK: %[[ATOMIC_RMW2:.+]] = memref.generic_atomic_rmw %[[ALLOC]][%[[ADDR2]]] : memref<6xi8> {
+// CHECK: %[[ARG2:.+]]: i8):
+// CHECK: %[[FROM_ELEM2:.+]] = vector.from_elements %[[ARG2]] : vector<1xi8>
+// CHECK: %[[BITCAST4:.+]] = vector.bitcast %[[FROM_ELEM2]] : vector<1xi8> to vector<4xi2>
+// CHECK: %[[SELECT2:.+]] = arith.select %[[CST1]], %[[INSERT2]], %[[BITCAST4]] :
+// CHECK-SAME: vector<4xi1>, vector<4xi2>
+// CHECK: %[[BITCAST5:.+]] = vector.bitcast %[[SELECT2]] : vector<4xi2> to vector<1xi8>
+// CHECK: %[[EXTRACT4:.+]] = vector.extract %[[BITCAST5]][0] : i8 from vector<1xi8>
+// CHECK: memref.atomic_yield %[[EXTRACT4]] : i8    
+
+// -----
+
+func.func @vector_store_i2_single_atomic(%arg0: vector<1xi2>) {
+    %0 = memref.alloc() : memref<4x1xi2>
+    %c0 = arith.constant 0 : index
+    %c1 = arith.constant 1 : index
+    vector.store %arg0, %0[%c1, %c0] :memref<4x1xi2>, vector<1xi2>
+    return
+}
+
+// In this example, only emit 1 atomic store
+// CHECK-LABEL: func @vector_store_i2_single_atomic(
+// CHECK-SAME: %[[ARG0:.+]]: vector<1xi2>)
+// CHECK: %[[ALLOC:.+]] = memref.alloc() : memref<1xi8>
+// CHECK: %[[C0:.+]] = arith.constant 0 : index
+// CHECK: %[[C1:.+]] = arith.constant 1 : index
+// CHECK: %[[CST:.+]] = arith.constant dense<[false, true, false, false]> : vector<4xi1>
+// CHECK: %[[CST0:.+]] = arith.constant dense<0> : vector<4xi2>
+// CHECK: %[[INSERT:.+]] = vector.insert_strided_slice %[[ARG0]], %[[CST0]]
+// CHECK-SAME: {offsets = [1], strides = [1]} : vector<1xi2> into vector<4xi2>
+
+// CHECK: %[[ATOMIC_RMW:.+]] = memref.generic_atomic_rmw %[[ALLOC]][%[[C0]]] : memref<1xi8> {
+// CHECK: %[[ARG:.+]]: i8):
+// CHECK: %[[FROM_ELEM:.+]] = vector.from_elements %[[ARG]] : vector<1xi8>
+// CHECK: %[[BITCAST:.+]] = vector.bitcast %[[FROM_ELEM]] : vector<1xi8> to vector<4xi2>
+// CHECK: %[[SELECT:.+]] = arith.select %[[CST]], %[[INSERT]], %[[BITCAST]] : vector<4xi1>, vector<4xi2>
+// CHECK: %[[BITCAST2:.+]] = vector.bitcast %[[SELECT]] : vector<4xi2> to vector<1xi8>
+// CHECK: %[[EXTRACT2:.+]] = vector.extract %[[BITCAST2]][0] : i8 from vector<1xi8>
+// CHECK: memref.atomic_yield %[[EXTRACT2]] : i8
diff --git a/mlir/test/lib/Dialect/MemRef/TestEmulateNarrowType.cpp b/mlir/test/lib/Dialect/MemRef/TestEmulateNarrowType.cpp
index 7401e470ed4f2c..9a3fac623fbd7d 100644
--- a/mlir/test/lib/Dialect/MemRef/TestEmulateNarrowType.cpp
+++ b/mlir/test/lib/Dialect/MemRef/TestEmulateNarrowType.cpp
@@ -99,7 +99,8 @@ struct TestEmulateNarrowTypePass
 
     arith::populateArithNarrowTypeEmulationPatterns(typeConverter, patterns);
     memref::populateMemRefNarrowTypeEmulationPatterns(typeConverter, patterns);
-    vector::populateVectorNarrowTypeEmulationPatterns(typeConverter, patterns);
+    vector::populateVectorNarrowTypeEmulationPatterns(typeConverter, patterns,
+                                                      atomicStore);
 
     if (failed(applyPartialConversion(op, target, std::move(patterns))))
       signalPassFailure();
@@ -118,6 +119,11 @@ struct TestEmulateNarrowTypePass
       *this, "skip-memref-type-conversion",
       llvm::cl::desc("disable memref type conversion (to test failures)"),
       llvm::cl::init(false)};
+
+  Option<bool> atomicStore{
+      *this, "atomic-store",
+      llvm::cl::desc("use atomic store instead of load-modify-write"),
+      llvm::cl::init(true)};
 };
 } // namespace
 

>From 99ee5d01675738f24cadb1e2c5e123a4a0917797 Mon Sep 17 00:00:00 2001
From: Alan Li <me at alanli.org>
Date: Tue, 3 Dec 2024 01:44:40 +0800
Subject: [PATCH 02/10] Refactoring

---
 .../Transforms/VectorEmulateNarrowType.cpp    | 64 ++++++++++---------
 1 file changed, 34 insertions(+), 30 deletions(-)

diff --git a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
index 278b42a5b7c104..d5012e387944f8 100644
--- a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
+++ b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
@@ -46,6 +46,9 @@ using namespace mlir;
 #define DBGSNL() (llvm::dbgs() << "\n")
 #define LDBG(X) LLVM_DEBUG(DBGS() << X << "\n")
 
+using VectorValue = TypedValue<VectorType>;
+using MemRefValue = TypedValue<MemRefType>;
+
 /// Returns a compressed mask for the emulated vector. For example, when
 /// emulating an eight-element `i8` vector with `i32` (i.e. when the source
 /// elements span two dest elements), this method compresses `vector<8xi1>`
@@ -255,8 +258,8 @@ static Value staticallyInsertSubvector(OpBuilder &rewriter, Location loc,
 /// function emits multiple `vector.extract` and `vector.insert` ops, so only
 /// use it when `offset` cannot be folded into a constant value.
 static Value dynamicallyExtractSubVector(OpBuilder &rewriter, Location loc,
-                                         TypedValue<VectorType> source,
-                                         Value dest, OpFoldResult offset,
+                                         VectorValue source, Value dest,
+                                         OpFoldResult offset,
                                          int64_t numElementsToExtract) {
   for (int i = 0; i < numElementsToExtract; ++i) {
     Value extractLoc =
@@ -273,8 +276,8 @@ static Value dynamicallyExtractSubVector(OpBuilder &rewriter, Location loc,
 
 /// Inserts a 1-D subvector into a 1-D `dest` vector at index `destOffsetVar`.
 static Value dynamicallyInsertSubVector(RewriterBase &rewriter, Location loc,
-                                        TypedValue<VectorType> source,
-                                        Value dest, OpFoldResult destOffsetVar,
+                                        VectorValue source, Value dest,
+                                        OpFoldResult destOffsetVar,
                                         size_t length) {
   assert(length > 0 && "length must be greater than 0");
   Value destOffsetVal =
@@ -295,11 +298,12 @@ static Value dynamicallyInsertSubVector(RewriterBase &rewriter, Location loc,
 /// specifically, use `emulatedElemType` for loading a vector of `origElemType`.
 /// The load location is given by `base` and `linearizedIndices`, and the
 /// load size is given by `numEmulatedElementsToLoad`.
-static TypedValue<VectorType>
-emulatedVectorLoad(OpBuilder &rewriter, Location loc, Value base,
-                   OpFoldResult linearizedIndices,
-                   int64_t numEmultedElementsToLoad, Type origElemType,
-                   Type emulatedElemType) {
+static VectorValue emulatedVectorLoad(OpBuilder &rewriter, Location loc,
+                                      Value base,
+                                      OpFoldResult linearizedIndices,
+                                      int64_t numEmultedElementsToLoad,
+                                      Type origElemType,
+                                      Type emulatedElemType) {
   auto scale = emulatedElemType.getIntOrFloatBitWidth() /
                origElemType.getIntOrFloatBitWidth();
   auto newLoad = rewriter.create<vector::LoadOp>(
@@ -312,9 +316,9 @@ emulatedVectorLoad(OpBuilder &rewriter, Location loc, Value base,
 
 /// Atomically store a subbyte-sized value to memory, with a mask.
 static void atomicStore(OpBuilder &builder, Location loc,
-                        TypedValue<MemRefType> emulatedMemref,
-                        Value linearizedIndex, TypedValue<VectorType> value,
-                        Value mask, int64_t) {
+                        MemRefValue emulatedMemref, Value linearizedIndex,
+                        VectorValue value, Value mask,
+                        int64_t numSrcElemsPerDest) {
   auto atomicOp = builder.create<memref::GenericAtomicRMWOp>(
       loc, emulatedMemref, ValueRange{linearizedIndex});
   Value origValue = atomicOp.getCurrentValue();
@@ -338,9 +342,9 @@ static void atomicStore(OpBuilder &builder, Location loc,
 
 /// Generate a non-atomic read-modify-write sequence for subbyte storing.
 static void rmwStore(OpBuilder &rewriter, Location loc,
-                     TypedValue<MemRefType> emulatedMemref,
-                     Value linearizedIndex, TypedValue<VectorType> value,
-                     Value mask, int64_t numSrcElemsPerDest) {
+                     MemRefValue emulatedMemref, Value linearizedIndex,
+                     VectorValue value, Value mask,
+                     int64_t numSrcElemsPerDest) {
   auto emulatedIOType =
       VectorType::get({1}, emulatedMemref.getType().getElementType());
   auto elemLoad = rewriter.create<vector::LoadOp>(
@@ -363,7 +367,7 @@ static_assert(std::is_same_v<decltype(atomicStore), decltype(rmwStore)> &&
 
 // Extract a slice of a vector, and insert it into a byte vector.
 static Value extractSliceIntoByte(ConversionPatternRewriter &rewriter,
-                                  Location loc, TypedValue<VectorType> vector,
+                                  Location loc, VectorValue vector,
                                   int64_t sliceOffset, int64_t sliceNumElements,
                                   int64_t byteOffset) {
   auto vectorElementType = vector.getType().getElementType();
@@ -404,7 +408,7 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
 
     auto loc = op.getLoc();
     auto convertedType = cast<MemRefType>(adaptor.getBase().getType());
-    auto valueToStore = cast<TypedValue<VectorType>>(op.getValueToStore());
+    auto valueToStore = cast<VectorValue>(op.getValueToStore());
     auto oldElementType = valueToStore.getType().getElementType();
     auto newElementType = convertedType.getElementType();
     int srcBits = oldElementType.getIntOrFloatBitWidth();
@@ -455,7 +459,7 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
       return failure();
     }
 
-    auto emulatedMemref = cast<TypedValue<MemRefType>>(adaptor.getBase());
+    auto emulatedMemref = cast<MemRefValue>(adaptor.getBase());
 
     // Shortcut: conditions when subbyte store at the front is not needed:
     // 1. The source vector size is multiple of byte size
@@ -504,8 +508,8 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
                                frontSubWidthStoreElem, *foldedNumFrontPadElems);
 
       subEmulatedWidthStore(rewriter, loc, emulatedMemref, currentDestIndex,
-                            cast<TypedValue<VectorType>>(value),
-                            frontMask.getResult(), numSrcElemsPerDest);
+                            cast<VectorValue>(value), frontMask.getResult(),
+                            numSrcElemsPerDest);
 
       currentDestIndex = rewriter.create<arith::AddIOp>(
           loc, rewriter.getIndexType(), currentDestIndex, constantOne);
@@ -546,9 +550,9 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
     // but their length is smaller than the emulated width.
     auto remainingElements = origElements - currentSourceIndex;
     if (remainingElements != 0) {
-      auto subWidthStorePart = extractSliceIntoByte(
-          rewriter, loc, cast<TypedValue<VectorType>>(valueToStore),
-          currentSourceIndex, remainingElements, 0);
+      auto subWidthStorePart =
+          extractSliceIntoByte(rewriter, loc, cast<VectorValue>(valueToStore),
+                               currentSourceIndex, remainingElements, 0);
 
       // Generate back mask
       auto maskValues = SmallVector<bool>(numSrcElemsPerDest, 0);
@@ -557,7 +561,7 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
           loc, DenseElementsAttr::get(subWidthStoreMaskType, maskValues));
 
       subEmulatedWidthStore(rewriter, loc, emulatedMemref, currentDestIndex,
-                            cast<TypedValue<VectorType>>(subWidthStorePart),
+                            cast<VectorValue>(subWidthStorePart),
                             backMask.getResult(), numSrcElemsPerDest);
     }
 
@@ -778,7 +782,7 @@ struct ConvertVectorLoad final : OpConversionPattern<vector::LoadOp> {
       auto resultVector = rewriter.create<arith::ConstantOp>(
           loc, op.getType(), rewriter.getZeroAttr(op.getType()));
       result = dynamicallyExtractSubVector(
-          rewriter, loc, dyn_cast<TypedValue<VectorType>>(result), resultVector,
+          rewriter, loc, cast<VectorValue>(result), resultVector,
           linearizedInfo.intraDataOffset, origElements);
     } else if (isUnalignedEmulation) {
       result = staticallyExtractSubvector(
@@ -899,8 +903,8 @@ struct ConvertVectorMaskedLoad final
         loc, newBitcastType, rewriter.getZeroAttr(newBitcastType));
     if (!foldedIntraVectorOffset) {
       passthru = dynamicallyInsertSubVector(
-          rewriter, loc, dyn_cast<TypedValue<VectorType>>(passthru),
-          emptyVector, linearizedInfo.intraDataOffset, origElements);
+          rewriter, loc, cast<VectorValue>(passthru), emptyVector,
+          linearizedInfo.intraDataOffset, origElements);
     } else if (isUnalignedEmulation) {
       passthru = staticallyInsertSubvector(rewriter, loc, passthru, emptyVector,
                                            *foldedIntraVectorOffset);
@@ -927,7 +931,7 @@ struct ConvertVectorMaskedLoad final
         loc, newSelectMaskType, rewriter.getZeroAttr(newSelectMaskType));
     if (!foldedIntraVectorOffset) {
       mask = dynamicallyInsertSubVector(
-          rewriter, loc, dyn_cast<TypedValue<VectorType>>(mask), emptyMask,
+          rewriter, loc, cast<VectorValue>(mask), emptyMask,
           linearizedInfo.intraDataOffset, origElements);
     } else if (isUnalignedEmulation) {
       mask = staticallyInsertSubvector(rewriter, loc, op.getMask(), emptyMask,
@@ -938,8 +942,8 @@ struct ConvertVectorMaskedLoad final
         rewriter.create<arith::SelectOp>(loc, mask, bitCast, passthru);
     if (!foldedIntraVectorOffset) {
       result = dynamicallyExtractSubVector(
-          rewriter, loc, dyn_cast<TypedValue<VectorType>>(result),
-          op.getPassThru(), linearizedInfo.intraDataOffset, origElements);
+          rewriter, loc, cast<VectorValue>(result), op.getPassThru(),
+          linearizedInfo.intraDataOffset, origElements);
     } else if (isUnalignedEmulation) {
       result = staticallyExtractSubvector(
           rewriter, loc, result, *foldedIntraVectorOffset, origElements);

>From c329d898bcf941eb9fecb208f18759a844b40319 Mon Sep 17 00:00:00 2001
From: Alan Li <me at alanli.org>
Date: Wed, 4 Dec 2024 18:41:48 +0800
Subject: [PATCH 03/10] updates

---
 .../Transforms/VectorEmulateNarrowType.cpp    | 47 +++++++++----------
 .../vector-emulate-narrow-type-unaligned.mlir |  2 -
 2 files changed, 23 insertions(+), 26 deletions(-)

diff --git a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
index d5012e387944f8..2aca8850d075c2 100644
--- a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
+++ b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
@@ -316,11 +316,11 @@ static VectorValue emulatedVectorLoad(OpBuilder &rewriter, Location loc,
 
 /// Atomically store a subbyte-sized value to memory, with a mask.
 static void atomicStore(OpBuilder &builder, Location loc,
-                        MemRefValue emulatedMemref, Value linearizedIndex,
-                        VectorValue value, Value mask,
+                        MemRefValue linearizedMemref, Value linearizedIndex,
+                        VectorValue valueToStore, Value mask,
                         int64_t numSrcElemsPerDest) {
   auto atomicOp = builder.create<memref::GenericAtomicRMWOp>(
-      loc, emulatedMemref, ValueRange{linearizedIndex});
+      loc, linearizedMemref, ValueRange{linearizedIndex});
   Value origValue = atomicOp.getCurrentValue();
 
   OpBuilder::InsertionGuard guard(builder);
@@ -331,10 +331,10 @@ static void atomicStore(OpBuilder &builder, Location loc,
   auto fromElem = builder.create<vector::FromElementsOp>(loc, oneVectorType,
                                                          ValueRange{origValue});
   auto vectorBitCast =
-      builder.create<vector::BitCastOp>(loc, value.getType(), fromElem);
+      builder.create<vector::BitCastOp>(loc, valueToStore.getType(), fromElem);
 
   auto select =
-      builder.create<arith::SelectOp>(loc, mask, value, vectorBitCast);
+      builder.create<arith::SelectOp>(loc, mask, valueToStore, vectorBitCast);
   auto bitcast2 = builder.create<vector::BitCastOp>(loc, oneVectorType, select);
   auto extract = builder.create<vector::ExtractOp>(loc, bitcast2, 0);
   builder.create<memref::AtomicYieldOp>(loc, extract.getResult());
@@ -342,13 +342,13 @@ static void atomicStore(OpBuilder &builder, Location loc,
 
 /// Generate a non-atomic read-modify-write sequence for subbyte storing.
 static void rmwStore(OpBuilder &rewriter, Location loc,
-                     MemRefValue emulatedMemref, Value linearizedIndex,
+                     MemRefValue linearizedMemref, Value linearizedIndex,
                      VectorValue value, Value mask,
                      int64_t numSrcElemsPerDest) {
   auto emulatedIOType =
-      VectorType::get({1}, emulatedMemref.getType().getElementType());
+      VectorType::get({1}, linearizedMemref.getType().getElementType());
   auto elemLoad = rewriter.create<vector::LoadOp>(
-      loc, emulatedIOType, emulatedMemref, ValueRange{linearizedIndex});
+      loc, emulatedIOType, linearizedMemref, ValueRange{linearizedIndex});
   auto fromBitcast = rewriter.create<vector::BitCastOp>(
       loc,
       VectorType::get({numSrcElemsPerDest}, value.getType().getElementType()),
@@ -356,15 +356,10 @@ static void rmwStore(OpBuilder &rewriter, Location loc,
   auto select = rewriter.create<arith::SelectOp>(loc, mask, fromBitcast, value);
   auto toBitcast =
       rewriter.create<vector::BitCastOp>(loc, emulatedIOType, select);
-  rewriter.create<vector::StoreOp>(loc, toBitcast, emulatedMemref,
+  rewriter.create<vector::StoreOp>(loc, toBitcast, linearizedMemref,
                                    linearizedIndex);
 }
 
-static_assert(std::is_same_v<decltype(atomicStore), decltype(rmwStore)> &&
-              "`atomicStore` and `rmwStore` must have same signature, as per "
-              "the design to keep the code clean, which one to call is "
-              "determined by the `useAtomicWrites` flag.");
-
 // Extract a slice of a vector, and insert it into a byte vector.
 static Value extractSliceIntoByte(ConversionPatternRewriter &rewriter,
                                   Location loc, VectorValue vector,
@@ -459,7 +454,7 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
       return failure();
     }
 
-    auto emulatedMemref = cast<MemRefValue>(adaptor.getBase());
+    auto linearizedMemref = cast<MemRefValue>(adaptor.getBase());
 
     // Shortcut: conditions when subbyte store at the front is not needed:
     // 1. The source vector size is multiple of byte size
@@ -470,7 +465,7 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
           loc, VectorType::get(numElements, newElementType),
           op.getValueToStore());
       rewriter.replaceOpWithNewOp<vector::StoreOp>(
-          op, bitCast.getResult(), emulatedMemref,
+          op, bitCast.getResult(), linearizedMemref,
           getValueOrCreateConstantIndexOp(rewriter, loc, linearizedIndices));
       return success();
     }
@@ -507,7 +502,7 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
           extractSliceIntoByte(rewriter, loc, valueToStore, 0,
                                frontSubWidthStoreElem, *foldedNumFrontPadElems);
 
-      subEmulatedWidthStore(rewriter, loc, emulatedMemref, currentDestIndex,
+      subEmulatedWidthStore(rewriter, loc, linearizedMemref, currentDestIndex,
                             cast<VectorValue>(value), frontMask.getResult(),
                             numSrcElemsPerDest);
 
@@ -525,20 +520,19 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
     int64_t fullWidthStoreSize =
         (origElements - currentSourceIndex) / numSrcElemsPerDest;
     int64_t numNonFullWidthElements = fullWidthStoreSize * numSrcElemsPerDest;
-    if (fullWidthStoreSize != 0) {
+    if (fullWidthStoreSize > 0) {
       auto fullWidthStorePart = staticallyExtractSubvector(
           rewriter, loc, valueToStore, currentSourceIndex,
           numNonFullWidthElements);
 
-      auto originType = dyn_cast<VectorType>(fullWidthStorePart.getType());
-      auto memrefElemType =
-          dyn_cast<MemRefType>(emulatedMemref.getType()).getElementType();
+      auto originType = cast<VectorType>(fullWidthStorePart.getType());
+      auto memrefElemType = getElementTypeOrSelf(linearizedMemref.getType());
       auto storeType = VectorType::get(
           {originType.getNumElements() / numSrcElemsPerDest}, memrefElemType);
       auto bitCast = rewriter.create<vector::BitCastOp>(loc, storeType,
                                                         fullWidthStorePart);
-      rewriter.create<vector::StoreOp>(loc, bitCast.getResult(), emulatedMemref,
-                                       currentDestIndex);
+      rewriter.create<vector::StoreOp>(loc, bitCast.getResult(),
+                                       linearizedMemref, currentDestIndex);
 
       currentSourceIndex += numNonFullWidthElements;
       currentDestIndex = rewriter.create<arith::AddIOp>(
@@ -560,7 +554,7 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
       auto backMask = rewriter.create<arith::ConstantOp>(
           loc, DenseElementsAttr::get(subWidthStoreMaskType, maskValues));
 
-      subEmulatedWidthStore(rewriter, loc, emulatedMemref, currentDestIndex,
+      subEmulatedWidthStore(rewriter, loc, linearizedMemref, currentDestIndex,
                             cast<VectorValue>(subWidthStorePart),
                             backMask.getResult(), numSrcElemsPerDest);
     }
@@ -573,6 +567,11 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
   /// configuration, it could be an atomic store or an RMW sequence.
   template <typename... Args>
   void subEmulatedWidthStore(Args &&...args) const {
+    static_assert(
+        std::is_same_v<decltype(atomicStore), decltype(rmwStore)> &&
+        "`atomicStore` and `rmwStore` must have same signature, as per "
+        "the design to keep the code clean, which one to call is "
+        "determined by the `useAtomicWrites` flag.");
     std::function<decltype(atomicStore)> storeFunc =
         useAtomicWrites_ ? atomicStore : rmwStore;
     storeFunc(std::forward<Args>(args)...);
diff --git a/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned.mlir b/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned.mlir
index fd526ada6cb7b2..e37053c25ff066 100644
--- a/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned.mlir
+++ b/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned.mlir
@@ -341,8 +341,6 @@ func.func @vector_maskedload_i4_constant_mask_unaligned(%passthru: vector<5xi2>)
 /// vector.store
 ///----------------------------------------------------------------------------------------
 
-// -----
-
 func.func @vector_store_i2_atomic(%arg0: vector<7xi2>) {
     %0 = memref.alloc() : memref<3x7xi2>
     %c0 = arith.constant 0 : index

>From c84bc0ea5db3003753b982660e215e92eecd844f Mon Sep 17 00:00:00 2001
From: Alan Li <me at alanli.org>
Date: Wed, 4 Dec 2024 23:46:43 +0800
Subject: [PATCH 04/10] update commnets

---
 .../Transforms/VectorEmulateNarrowType.cpp    | 27 ++++++++++++++++---
 1 file changed, 23 insertions(+), 4 deletions(-)

diff --git a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
index 2aca8850d075c2..b04a76c0c1c43e 100644
--- a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
+++ b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
@@ -314,7 +314,17 @@ static VectorValue emulatedVectorLoad(OpBuilder &rewriter, Location loc,
       newLoad);
 }
 
-/// Atomically store a subbyte-sized value to memory, with a mask.
+/// Emits `memref.generic_atomic_rmw` op to store a subbyte-sized value to a
+/// byte in memory, with a mask. The `valueToStore` is a vector of subbyte-sized
+/// elements, with size of 8 bits, and the mask is used to select which elements
+/// to store.
+///
+/// Before:
+///   memory = |ab|cd|ef|12| : <4xi2> (<1xi8>)
+///   valueToStore = |01|23|45|67| : vector<4xi2>
+///   mask = |0|0|1|1| : vector<4xi1>
+/// After:
+///   memory = |ab|cd|45|67| : <4xi2> (<1xi8>)
 static void atomicStore(OpBuilder &builder, Location loc,
                         MemRefValue linearizedMemref, Value linearizedIndex,
                         VectorValue valueToStore, Value mask,
@@ -326,7 +336,7 @@ static void atomicStore(OpBuilder &builder, Location loc,
   OpBuilder::InsertionGuard guard(builder);
   builder.setInsertionPointToStart(atomicOp.getBody());
 
-  // i8 -> <1xi8> -> <numSrcElemsPerDest x i.>
+  // i8 -> <1xi8> -> <numSrcElemsPerDest x i.>:
   auto oneVectorType = VectorType::get({1}, origValue.getType());
   auto fromElem = builder.create<vector::FromElementsOp>(loc, oneVectorType,
                                                          ValueRange{origValue});
@@ -341,6 +351,7 @@ static void atomicStore(OpBuilder &builder, Location loc,
 }
 
 /// Generate a non-atomic read-modify-write sequence for subbyte storing.
+/// It has similar logic to `atomicStore`, but without the atomicity.
 static void rmwStore(OpBuilder &rewriter, Location loc,
                      MemRefValue linearizedMemref, Value linearizedIndex,
                      VectorValue value, Value mask,
@@ -360,7 +371,15 @@ static void rmwStore(OpBuilder &rewriter, Location loc,
                                    linearizedIndex);
 }
 
-// Extract a slice of a vector, and insert it into a byte vector.
+/// Extract a slice from vector `vector`, with the size of `sliceNumElements`,
+/// and insert it into a zero, byte vector at offset `byteOffset`. For example:
+/// Inputs:
+///   vector = |01|23|45|67| : vector<4xi2>
+///   sliceOffset = 1
+///   sliceNumElements = 2
+///   byteOffset = 2
+/// Output:
+///   vector = |00|00|23|45| : vector<4xi2>
 static Value extractSliceIntoByte(ConversionPatternRewriter &rewriter,
                                   Location loc, VectorValue vector,
                                   int64_t sliceOffset, int64_t sliceNumElements,
@@ -484,7 +503,7 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
     // the rest elements are aligned to width boundary.
     auto frontSubWidthStoreElem =
         (numSrcElemsPerDest - *foldedNumFrontPadElems) % numSrcElemsPerDest;
-    if (frontSubWidthStoreElem != 0) {
+    if (frontSubWidthStoreElem > 0) {
       SmallVector<bool> frontMaskValues(numSrcElemsPerDest, false);
       if (*foldedNumFrontPadElems + origElements < numSrcElemsPerDest) {
         std::fill_n(frontMaskValues.begin() + *foldedNumFrontPadElems,

>From 11be4862edc8c2f7c11faac148714a8e84ddeaf7 Mon Sep 17 00:00:00 2001
From: Alan Li <me at alanli.org>
Date: Thu, 5 Dec 2024 00:27:09 +0800
Subject: [PATCH 05/10] updates

---
 .../Vector/Transforms/VectorEmulateNarrowType.cpp | 15 +++++++++++----
 1 file changed, 11 insertions(+), 4 deletions(-)

diff --git a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
index b04a76c0c1c43e..4ca492bd1b58bb 100644
--- a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
+++ b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
@@ -329,6 +329,7 @@ static void atomicStore(OpBuilder &builder, Location loc,
                         MemRefValue linearizedMemref, Value linearizedIndex,
                         VectorValue valueToStore, Value mask,
                         int64_t numSrcElemsPerDest) {
+  assert(valueToStore.getType().getRank() == 1 && "expected 1-D vector");
   auto atomicOp = builder.create<memref::GenericAtomicRMWOp>(
       loc, linearizedMemref, ValueRange{linearizedIndex});
   Value origValue = atomicOp.getCurrentValue();
@@ -336,7 +337,6 @@ static void atomicStore(OpBuilder &builder, Location loc,
   OpBuilder::InsertionGuard guard(builder);
   builder.setInsertionPointToStart(atomicOp.getBody());
 
-  // i8 -> <1xi8> -> <numSrcElemsPerDest x i.>:
   auto oneVectorType = VectorType::get({1}, origValue.getType());
   auto fromElem = builder.create<vector::FromElementsOp>(loc, oneVectorType,
                                                          ValueRange{origValue});
@@ -354,17 +354,20 @@ static void atomicStore(OpBuilder &builder, Location loc,
 /// It has similar logic to `atomicStore`, but without the atomicity.
 static void rmwStore(OpBuilder &rewriter, Location loc,
                      MemRefValue linearizedMemref, Value linearizedIndex,
-                     VectorValue value, Value mask,
+                     VectorValue valueToStore, Value mask,
                      int64_t numSrcElemsPerDest) {
+  assert(valueToStore.getType().getRank() == 1 && "expected 1-D vector");
   auto emulatedIOType =
       VectorType::get({1}, linearizedMemref.getType().getElementType());
   auto elemLoad = rewriter.create<vector::LoadOp>(
       loc, emulatedIOType, linearizedMemref, ValueRange{linearizedIndex});
   auto fromBitcast = rewriter.create<vector::BitCastOp>(
       loc,
-      VectorType::get({numSrcElemsPerDest}, value.getType().getElementType()),
+      VectorType::get({numSrcElemsPerDest},
+                      valueToStore.getType().getElementType()),
       elemLoad);
-  auto select = rewriter.create<arith::SelectOp>(loc, mask, fromBitcast, value);
+  auto select =
+      rewriter.create<arith::SelectOp>(loc, mask, fromBitcast, valueToStore);
   auto toBitcast =
       rewriter.create<vector::BitCastOp>(loc, emulatedIOType, select);
   rewriter.create<vector::StoreOp>(loc, toBitcast, linearizedMemref,
@@ -384,7 +387,11 @@ static Value extractSliceIntoByte(ConversionPatternRewriter &rewriter,
                                   Location loc, VectorValue vector,
                                   int64_t sliceOffset, int64_t sliceNumElements,
                                   int64_t byteOffset) {
+  assert(vector.getType().getRank() == 1 && "expected 1-D vector");
   auto vectorElementType = vector.getType().getElementType();
+  assert(
+      sliceNumElements * vectorElementType.getIntOrFloatBitWidth() <= 8 &&
+      "sliceNumElements * vector element size must be less than or equal to 8");
   assert(8 % vectorElementType.getIntOrFloatBitWidth() == 0 &&
          "vector element must be a valid sub-byte type");
   auto scale = 8 / vectorElementType.getIntOrFloatBitWidth();

>From 6c05b72096ddd96d7e616c82ec06ba0e078139ab Mon Sep 17 00:00:00 2001
From: Alan Li <me at alanli.org>
Date: Thu, 5 Dec 2024 08:50:50 +0800
Subject: [PATCH 06/10] fix according to comments

---
 .../Vector/Transforms/VectorRewritePatterns.h    |  5 +++--
 .../Transforms/VectorEmulateNarrowType.cpp       | 16 +++++++++-------
 2 files changed, 12 insertions(+), 9 deletions(-)

diff --git a/mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h b/mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h
index 64bb3a2204cfdc..43478aacb50a14 100644
--- a/mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h
+++ b/mlir/include/mlir/Dialect/Vector/Transforms/VectorRewritePatterns.h
@@ -364,8 +364,9 @@ void populateVectorMaskMaterializationPatterns(RewritePatternSet &patterns,
                                                PatternBenefit benefit = 1);
 
 /// Appends patterns for emulating vector operations over narrow types with ops
-/// over wider types. `useAtomicWrites` indicates whether to use atomic
-/// operations in the places where thread contention is possible.
+/// over wider types. The `useAtomicWrites` indicates whether to use
+/// op `memref.generic_atomic_rmw` to perform atomic subbyte storing, or just a
+/// rmw sequence otherwise.
 void populateVectorNarrowTypeEmulationPatterns(
     const arith::NarrowTypeEmulationConverter &typeConverter,
     RewritePatternSet &patterns, bool useAtomicWrites = true);
diff --git a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
index 4ca492bd1b58bb..3c5534c4a932ad 100644
--- a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
+++ b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
@@ -319,12 +319,14 @@ static VectorValue emulatedVectorLoad(OpBuilder &rewriter, Location loc,
 /// elements, with size of 8 bits, and the mask is used to select which elements
 /// to store.
 ///
-/// Before:
-///   memory = |ab|cd|ef|12| : <4xi2> (<1xi8>)
-///   valueToStore = |01|23|45|67| : vector<4xi2>
+/// Inputs:
+///   linearizedMemref = |a|b|c|d| : <4xi2> (<1xi8>)
+///   linearizedIndex = 2
+///   valueToStore = |e|f|g|h| : vector<4xi2>
 ///   mask = |0|0|1|1| : vector<4xi1>
-/// After:
-///   memory = |ab|cd|45|67| : <4xi2> (<1xi8>)
+///
+/// Result:
+///   linearizedMemref = |a|b|g|h| : <4xi2> (<1xi8>)
 static void atomicStore(OpBuilder &builder, Location loc,
                         MemRefValue linearizedMemref, Value linearizedIndex,
                         VectorValue valueToStore, Value mask,
@@ -374,8 +376,8 @@ static void rmwStore(OpBuilder &rewriter, Location loc,
                                    linearizedIndex);
 }
 
-/// Extract a slice from vector `vector`, with the size of `sliceNumElements`,
-/// and insert it into a zero, byte vector at offset `byteOffset`. For example:
+/// Extract `sliceNumElements` from source `vector` at `sliceOffset`,
+/// and insert it into an empty vector at offset `byteOffset`.
 /// Inputs:
 ///   vector = |01|23|45|67| : vector<4xi2>
 ///   sliceOffset = 1

>From 40acc9984e7b672730f778ec0c83dc1565021518 Mon Sep 17 00:00:00 2001
From: Alan Li <me at alanli.org>
Date: Sat, 7 Dec 2024 23:35:07 +0800
Subject: [PATCH 07/10] Updates according to comments

---
 ....mlir => vector-emulate-narrow-type-unaligned-non-atomic.mlir} | 0
 1 file changed, 0 insertions(+), 0 deletions(-)
 rename mlir/test/Dialect/Vector/{vector-emulate-narrow-type-unaligned-rmw.mlir => vector-emulate-narrow-type-unaligned-non-atomic.mlir} (100%)

diff --git a/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned-rmw.mlir b/mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned-non-atomic.mlir
similarity index 100%
rename from mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned-rmw.mlir
rename to mlir/test/Dialect/Vector/vector-emulate-narrow-type-unaligned-non-atomic.mlir

>From 62f9b406bf4530a7e751c7acc1edc4900a0bf0fa Mon Sep 17 00:00:00 2001
From: Alan Li <me at alanli.org>
Date: Tue, 10 Dec 2024 21:43:23 +0800
Subject: [PATCH 08/10] update for comments

---
 .../Transforms/VectorEmulateNarrowType.cpp    | 57 +++++++++----------
 1 file changed, 27 insertions(+), 30 deletions(-)

diff --git a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
index 3c5534c4a932ad..1d4c546db0bf0e 100644
--- a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
+++ b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
@@ -320,17 +320,20 @@ static VectorValue emulatedVectorLoad(OpBuilder &rewriter, Location loc,
 /// to store.
 ///
 /// Inputs:
-///   linearizedMemref = |a|b|c|d| : <4xi2> (<1xi8>)
+///   linearizedMemref = |2|2|2|2| : <4xi2> (<1xi8>)
 ///   linearizedIndex = 2
-///   valueToStore = |e|f|g|h| : vector<4xi2>
+///   valueToStore = |3|3|3|3| : vector<4xi2>
 ///   mask = |0|0|1|1| : vector<4xi1>
 ///
 /// Result:
-///   linearizedMemref = |a|b|g|h| : <4xi2> (<1xi8>)
+///   linearizedMemref = |2|2|3|3| : <4xi2> (<1xi8>)
 static void atomicStore(OpBuilder &builder, Location loc,
                         MemRefValue linearizedMemref, Value linearizedIndex,
-                        VectorValue valueToStore, Value mask,
-                        int64_t numSrcElemsPerDest) {
+                        VectorValue valueToStore, Value mask) {
+  // `numSrcElemsPerDest` is not used in this function, but to keep the function
+  // signature consistent with `rmwStore` so as to simplify the pattern to
+  // invoke.
+
   assert(valueToStore.getType().getRank() == 1 && "expected 1-D vector");
   auto atomicOp = builder.create<memref::GenericAtomicRMWOp>(
       loc, linearizedMemref, ValueRange{linearizedIndex});
@@ -354,37 +357,33 @@ static void atomicStore(OpBuilder &builder, Location loc,
 
 /// Generate a non-atomic read-modify-write sequence for subbyte storing.
 /// It has similar logic to `atomicStore`, but without the atomicity.
-static void rmwStore(OpBuilder &rewriter, Location loc,
+static void rmwStore(OpBuilder &builder, Location loc,
                      MemRefValue linearizedMemref, Value linearizedIndex,
-                     VectorValue valueToStore, Value mask,
-                     int64_t numSrcElemsPerDest) {
+                     VectorValue valueToStore, Value mask) {
   assert(valueToStore.getType().getRank() == 1 && "expected 1-D vector");
-  auto emulatedIOType =
+  auto oneElemVecType =
       VectorType::get({1}, linearizedMemref.getType().getElementType());
-  auto elemLoad = rewriter.create<vector::LoadOp>(
-      loc, emulatedIOType, linearizedMemref, ValueRange{linearizedIndex});
-  auto fromBitcast = rewriter.create<vector::BitCastOp>(
-      loc,
-      VectorType::get({numSrcElemsPerDest},
-                      valueToStore.getType().getElementType()),
-      elemLoad);
-  auto select =
-      rewriter.create<arith::SelectOp>(loc, mask, fromBitcast, valueToStore);
-  auto toBitcast =
-      rewriter.create<vector::BitCastOp>(loc, emulatedIOType, select);
-  rewriter.create<vector::StoreOp>(loc, toBitcast, linearizedMemref,
-                                   linearizedIndex);
+  auto origValue = builder.create<vector::LoadOp>(
+      loc, oneElemVecType, linearizedMemref, ValueRange{linearizedIndex});
+  auto castedValue =
+      builder.create<vector::BitCastOp>(loc, valueToStore.getType(), origValue);
+  auto result =
+      builder.create<arith::SelectOp>(loc, mask, castedValue, valueToStore);
+  auto resultBitcast =
+      builder.create<vector::BitCastOp>(loc, oneElemVecType, result);
+  builder.create<vector::StoreOp>(loc, resultBitcast, linearizedMemref,
+                                  linearizedIndex);
 }
 
 /// Extract `sliceNumElements` from source `vector` at `sliceOffset`,
 /// and insert it into an empty vector at offset `byteOffset`.
 /// Inputs:
-///   vector = |01|23|45|67| : vector<4xi2>
+///   vector = |1|2|3|4| : vector<4xi2>
 ///   sliceOffset = 1
 ///   sliceNumElements = 2
 ///   byteOffset = 2
 /// Output:
-///   vector = |00|00|23|45| : vector<4xi2>
+///   vector = |0|0|2|3| : vector<4xi2>
 static Value extractSliceIntoByte(ConversionPatternRewriter &rewriter,
                                   Location loc, VectorValue vector,
                                   int64_t sliceOffset, int64_t sliceNumElements,
@@ -402,9 +401,8 @@ static Value extractSliceIntoByte(ConversionPatternRewriter &rewriter,
       rewriter.getZeroAttr(VectorType::get({scale}, vectorElementType)));
   auto extracted = staticallyExtractSubvector(rewriter, loc, vector,
                                               sliceOffset, sliceNumElements);
-  auto inserted = staticallyInsertSubvector(rewriter, loc, extracted,
-                                            emptyByteVector, byteOffset);
-  return inserted;
+  return staticallyInsertSubvector(rewriter, loc, extracted, emptyByteVector,
+                                   byteOffset);
 }
 
 namespace {
@@ -531,8 +529,7 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
                                frontSubWidthStoreElem, *foldedNumFrontPadElems);
 
       subEmulatedWidthStore(rewriter, loc, linearizedMemref, currentDestIndex,
-                            cast<VectorValue>(value), frontMask.getResult(),
-                            numSrcElemsPerDest);
+                            cast<VectorValue>(value), frontMask.getResult());
 
       currentDestIndex = rewriter.create<arith::AddIOp>(
           loc, rewriter.getIndexType(), currentDestIndex, constantOne);
@@ -584,7 +581,7 @@ struct ConvertVectorStore final : OpConversionPattern<vector::StoreOp> {
 
       subEmulatedWidthStore(rewriter, loc, linearizedMemref, currentDestIndex,
                             cast<VectorValue>(subWidthStorePart),
-                            backMask.getResult(), numSrcElemsPerDest);
+                            backMask.getResult());
     }
 
     rewriter.eraseOp(op);

>From 15d2ad3ea270f1eba314a707749e1b8ea78a0159 Mon Sep 17 00:00:00 2001
From: Alan Li <me at alanli.org>
Date: Tue, 10 Dec 2024 21:53:55 +0800
Subject: [PATCH 09/10] checkpoint

---
 .../Transforms/VectorEmulateNarrowType.cpp    | 24 +++++++++----------
 1 file changed, 11 insertions(+), 13 deletions(-)

diff --git a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
index 1d4c546db0bf0e..c091aec6246988 100644
--- a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
+++ b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
@@ -330,10 +330,6 @@ static VectorValue emulatedVectorLoad(OpBuilder &rewriter, Location loc,
 static void atomicStore(OpBuilder &builder, Location loc,
                         MemRefValue linearizedMemref, Value linearizedIndex,
                         VectorValue valueToStore, Value mask) {
-  // `numSrcElemsPerDest` is not used in this function, but to keep the function
-  // signature consistent with `rmwStore` so as to simplify the pattern to
-  // invoke.
-
   assert(valueToStore.getType().getRank() == 1 && "expected 1-D vector");
   auto atomicOp = builder.create<memref::GenericAtomicRMWOp>(
       loc, linearizedMemref, ValueRange{linearizedIndex});
@@ -342,17 +338,19 @@ static void atomicStore(OpBuilder &builder, Location loc,
   OpBuilder::InsertionGuard guard(builder);
   builder.setInsertionPointToStart(atomicOp.getBody());
 
+  // Construct the vector value from the scalar.
   auto oneVectorType = VectorType::get({1}, origValue.getType());
-  auto fromElem = builder.create<vector::FromElementsOp>(loc, oneVectorType,
+  Value origVecValue = builder.create<vector::FromElementsOp>(loc, oneVectorType,
                                                          ValueRange{origValue});
-  auto vectorBitCast =
-      builder.create<vector::BitCastOp>(loc, valueToStore.getType(), fromElem);
-
-  auto select =
-      builder.create<arith::SelectOp>(loc, mask, valueToStore, vectorBitCast);
-  auto bitcast2 = builder.create<vector::BitCastOp>(loc, oneVectorType, select);
-  auto extract = builder.create<vector::ExtractOp>(loc, bitcast2, 0);
-  builder.create<memref::AtomicYieldOp>(loc, extract.getResult());
+  origVecValue =
+      builder.create<vector::BitCastOp>(loc, valueToStore.getType(), origVecValue);
+
+  // Construct the final masked value and yield it.
+  Value maskedValue =
+      builder.create<arith::SelectOp>(loc, mask, valueToStore, origVecValue);
+  maskedValue = builder.create<vector::BitCastOp>(loc, oneVectorType, maskedValue);
+  auto scalarMaskedValue = builder.create<vector::ExtractOp>(loc, maskedValue, 0);
+  builder.create<memref::AtomicYieldOp>(loc, scalarMaskedValue);
 }
 
 /// Generate a non-atomic read-modify-write sequence for subbyte storing.

>From 2b90bf73f90cb846f1f881a4dfa2fcce12179ee1 Mon Sep 17 00:00:00 2001
From: Alan Li <me at alanli.org>
Date: Tue, 10 Dec 2024 22:01:30 +0800
Subject: [PATCH 10/10] updates

---
 .../Transforms/VectorEmulateNarrowType.cpp    | 37 ++++++++++++-------
 1 file changed, 23 insertions(+), 14 deletions(-)

diff --git a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
index c091aec6246988..95eab0998ee9df 100644
--- a/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
+++ b/mlir/lib/Dialect/Vector/Transforms/VectorEmulateNarrowType.cpp
@@ -331,6 +331,9 @@ static void atomicStore(OpBuilder &builder, Location loc,
                         MemRefValue linearizedMemref, Value linearizedIndex,
                         VectorValue valueToStore, Value mask) {
   assert(valueToStore.getType().getRank() == 1 && "expected 1-D vector");
+
+  // Create an atomic load-modify-write region using
+  // `memref.generic_atomic_rmw`.
   auto atomicOp = builder.create<memref::GenericAtomicRMWOp>(
       loc, linearizedMemref, ValueRange{linearizedIndex});
   Value origValue = atomicOp.getCurrentValue();
@@ -340,16 +343,18 @@ static void atomicStore(OpBuilder &builder, Location loc,
 
   // Construct the vector value from the scalar.
   auto oneVectorType = VectorType::get({1}, origValue.getType());
-  Value origVecValue = builder.create<vector::FromElementsOp>(loc, oneVectorType,
-                                                         ValueRange{origValue});
-  origVecValue =
-      builder.create<vector::BitCastOp>(loc, valueToStore.getType(), origVecValue);
+  Value origVecValue = builder.create<vector::FromElementsOp>(
+      loc, oneVectorType, ValueRange{origValue});
+  origVecValue = builder.create<vector::BitCastOp>(loc, valueToStore.getType(),
+                                                   origVecValue);
 
   // Construct the final masked value and yield it.
   Value maskedValue =
       builder.create<arith::SelectOp>(loc, mask, valueToStore, origVecValue);
-  maskedValue = builder.create<vector::BitCastOp>(loc, oneVectorType, maskedValue);
-  auto scalarMaskedValue = builder.create<vector::ExtractOp>(loc, maskedValue, 0);
+  maskedValue =
+      builder.create<vector::BitCastOp>(loc, oneVectorType, maskedValue);
+  auto scalarMaskedValue =
+      builder.create<vector::ExtractOp>(loc, maskedValue, 0);
   builder.create<memref::AtomicYieldOp>(loc, scalarMaskedValue);
 }
 
@@ -359,17 +364,21 @@ static void rmwStore(OpBuilder &builder, Location loc,
                      MemRefValue linearizedMemref, Value linearizedIndex,
                      VectorValue valueToStore, Value mask) {
   assert(valueToStore.getType().getRank() == 1 && "expected 1-D vector");
+
+  // Construct the vector value from the scalar.
   auto oneElemVecType =
       VectorType::get({1}, linearizedMemref.getType().getElementType());
-  auto origValue = builder.create<vector::LoadOp>(
+  Value origVecValue = builder.create<vector::LoadOp>(
       loc, oneElemVecType, linearizedMemref, ValueRange{linearizedIndex});
-  auto castedValue =
-      builder.create<vector::BitCastOp>(loc, valueToStore.getType(), origValue);
-  auto result =
-      builder.create<arith::SelectOp>(loc, mask, castedValue, valueToStore);
-  auto resultBitcast =
-      builder.create<vector::BitCastOp>(loc, oneElemVecType, result);
-  builder.create<vector::StoreOp>(loc, resultBitcast, linearizedMemref,
+  origVecValue = builder.create<vector::BitCastOp>(loc, valueToStore.getType(),
+                                                   origVecValue);
+
+  // Construct the final masked value and yield it.
+  Value maskedValue =
+      builder.create<arith::SelectOp>(loc, mask, origVecValue, valueToStore);
+  maskedValue =
+      builder.create<vector::BitCastOp>(loc, oneElemVecType, maskedValue);
+  builder.create<vector::StoreOp>(loc, maskedValue, linearizedMemref,
                                   linearizedIndex);
 }
 



More information about the Mlir-commits mailing list