[mlir][LLVM] Fix unsupported FP lowering in VectorConvertToLLVMPattern

[matthias-springer]

Fixes a bug in VectorConvertToLLVMPattern, which converted operations with unsupported FP types. E.g., arith.addf ... : f4E2M1FN was lowered to llvm.fadd ... : i4, which does not verify. There are a few more patterns that have the same bug. Those will be fixed in follow-up PRs.

This commit is in preparation of adding an APFloat-based lowering for arith operations with unsupported floating-point types.

[llvmbot]

@llvm/pr-subscribers-mlir

@llvm/pr-subscribers-mlir-llvm

Author: Matthias Springer (matthias-springer)

Changes

Fixes a bug in VectorConvertToLLVMPattern, which converted operations with unsupported FP types. E.g., arith.addf ... : f4E2M1FN was lowered to llvm.fadd ... : i4, which does not verify.

---

Full diff: https://github.com/llvm/llvm-project/pull/166513.diff

2 Files Affected:

• (modified) mlir/include/mlir/Conversion/LLVMCommon/VectorPattern.h (+25)
• (modified) mlir/test/Conversion/ArithToLLVM/arith-to-llvm.mlir (+26)

diff --git a/mlir/include/mlir/Conversion/LLVMCommon/VectorPattern.h b/mlir/include/mlir/Conversion/LLVMCommon/VectorPattern.h
index 964281592cc65..d8483114f1137 100644
--- a/mlir/include/mlir/Conversion/LLVMCommon/VectorPattern.h
+++ b/mlir/include/mlir/Conversion/LLVMCommon/VectorPattern.h
@@ -92,12 +92,37 @@ class VectorConvertToLLVMPattern : public ConvertOpToLLVMPattern<SourceOp> {
   using ConvertOpToLLVMPattern<SourceOp>::ConvertOpToLLVMPattern;
   using Super = VectorConvertToLLVMPattern<SourceOp, TargetOp>;
 
+  /// Return "true" if the given type (or its element type) is a floating point
+  /// type.
+  static FloatType getFloatingPointType(Type type) {
+    if (auto floatType = dyn_cast<FloatType>(type))
+      return floatType;
+    if (auto vecType = dyn_cast<VectorType>(type))
+      return dyn_cast<FloatType>(vecType.getElementType());
+    return nullptr;
+  }
+
   LogicalResult
   matchAndRewrite(SourceOp op, typename SourceOp::Adaptor adaptor,
                   ConversionPatternRewriter &rewriter) const override {
     static_assert(
         std::is_base_of<OpTrait::OneResult<SourceOp>, SourceOp>::value,
         "expected single result op");
+
+    // The pattern should not apply if a floating-point operand is converted to
+    // a non-floating-point type. This indicates that the floating point type
+    // is not supported by the LLVM lowering. (Such types are converted to
+    // integers.)
+    for (Value operand : op->getOperands()) {
+      FloatType floatType = getFloatingPointType(operand.getType());
+      if (!floatType)
+        continue;
+      Type convertedType = this->getTypeConverter()->convertType(floatType);
+      if (!isa<FloatType>(convertedType))
+        return rewriter.notifyMatchFailure(op,
+                                           "unsupported floating point type");
+    }
+
     // Determine attributes for the target op
     AttrConvert<SourceOp, TargetOp> attrConvert(op);
 
diff --git a/mlir/test/Conversion/ArithToLLVM/arith-to-llvm.mlir b/mlir/test/Conversion/ArithToLLVM/arith-to-llvm.mlir
index ba12ff29ebef9..b5dcb01d3dc6b 100644
--- a/mlir/test/Conversion/ArithToLLVM/arith-to-llvm.mlir
+++ b/mlir/test/Conversion/ArithToLLVM/arith-to-llvm.mlir
@@ -747,3 +747,29 @@ func.func @memref_bitcast(%1: memref<?xi16>) -> memref<?xbf16> {
   %2 = arith.bitcast %1 : memref<?xi16> to memref<?xbf16>
   func.return %2 : memref<?xbf16>
 }
+
+// -----
+
+// CHECK-LABEL: func @unsupported_fp_type
+//       CHECK:   arith.addf {{.*}} : f4E2M1FN
+//       CHECK:   arith.addf {{.*}} : vector<4xf4E2M1FN>
+//       CHECK:   arith.addf {{.*}} : vector<8x4xf4E2M1FN>
+func.func @unsupported_fp_type(%arg0: f4E2M1FN, %arg1: vector<4xf4E2M1FN>, %arg2: vector<8x4xf4E2M1FN>) -> (f4E2M1FN, vector<4xf4E2M1FN>, vector<8x4xf4E2M1FN>) {
+  %0 = arith.addf %arg0, %arg0 : f4E2M1FN
+  %1 = arith.addf %arg1, %arg1 : vector<4xf4E2M1FN>
+  %2 = arith.addf %arg2, %arg2 : vector<8x4xf4E2M1FN>
+  return %0, %1, %2 : f4E2M1FN, vector<4xf4E2M1FN>, vector<8x4xf4E2M1FN>
+}
+
+// -----
+
+//   CHECK-LABEL: func @supported_fp_type
+//         CHECK:   llvm.fadd {{.*}} : f32
+//         CHECK:   llvm.fadd {{.*}} : vector<4xf32>
+// CHECK-COUNT-4:   llvm.fadd {{.*}} : vector<8xf32>
+func.func @supported_fp_type(%arg0: f32, %arg1: vector<4xf32>, %arg2: vector<4x8xf32>) -> (f32, vector<4xf32>, vector<4x8xf32>) {
+  %0 = arith.addf %arg0, %arg0 : f32
+  %1 = arith.addf %arg1, %arg1 : vector<4xf32>
+  %2 = arith.addf %arg2, %arg2 : vector<4x8xf32>
+  return %0, %1, %2 : f32, vector<4xf32>, vector<4x8xf32>
+}

[gysit]

LGTM