[Mlir-commits] [mlir] Introduce `arith.scaling_extf` and `arith.scaling_truncf` (PR #141965)

Umang Yadav llvmlistbot at llvm.org
Fri Jun 6 16:41:48 PDT 2025


================
@@ -357,7 +356,8 @@ struct F8E8M0ExtFOpConverter : public OpRewritePattern<arith::ExtFOp> {
     f32Bits = b.create<arith::SelectOp>(isNan, cF32NaN, f32Bits);
     Value result = b.create<arith::BitcastOp>(f32Ty, f32Bits);
     if (resultETy.getIntOrFloatBitWidth() < 32) {
-      result = b.create<arith::TruncFOp>(resultTy, result);
+      result = b.create<arith::TruncFOp>(resultTy, result, nullptr,
+                                         op.getFastmathAttr());
     } else if (resultETy.getIntOrFloatBitWidth() > 32) {
       result = b.create<arith::ExtFOp>(resultTy, result);
----------------
umangyadav wrote:

Yes, Added flag and tests.

https://github.com/llvm/llvm-project/pull/141965


More information about the Mlir-commits mailing list