[flang-commits] [flang] [mlir] [RFC][mlir] Conditional support for fast-math attributes. (PR #125620)

[Slava Zakharin via flang-commits]

================
@@ -589,10 +589,18 @@ struct CallOpConversion : public fir::FIROpConversion<fir::CallOp> {
     // Convert arith::FastMathFlagsAttr to LLVM::FastMathFlagsAttr.
     mlir::arith::AttrConvertFastMathToLLVM<fir::CallOp, mlir::LLVM::CallOp>
         attrConvert(call);
-    rewriter.replaceOpWithNewOp<mlir::LLVM::CallOp>(
-        call, resultTys, adaptor.getOperands(),
+    auto llvmCall = rewriter.create<mlir::LLVM::CallOp>(
+        call.getLoc(), resultTys, adaptor.getOperands(),
         addLLVMOpBundleAttrs(rewriter, attrConvert.getAttrs(),
                              adaptor.getOperands().size()));
+    auto fmi =
+        mlir::cast<mlir::LLVM::FastmathFlagsInterface>(llvmCall.getOperation());
+    if (!fmi.isFastmathApplicable())
+      llvmCall->setAttr(
----------------
vzakhari wrote:

Right. Will fix.

https://github.com/llvm/llvm-project/pull/125620