[Mlir-commits] [mlir] [MLIR][Arith] Add rounding mode attribute to `truncf` (PR #86152)

[Jakub Kuderski]

================
@@ -1398,8 +1420,15 @@ OpFoldResult arith::TruncFOp::fold(FoldAdaptor adaptor) {
   const llvm::fltSemantics &targetSemantics = resElemType.getFloatSemantics();
   return constFoldCastOp<FloatAttr, FloatAttr>(
       adaptor.getOperands(), getType(),
-      [&targetSemantics](const APFloat &a, bool &castStatus) {
-        FailureOr<APFloat> result = convertFloatValue(a, targetSemantics);
+      [this, &targetSemantics](const APFloat &a, bool &castStatus) {
+        FailureOr<APFloat> result;
+        if (std::optional<RoundingMode> roundingMode = getRoundingmode()) {
----------------
kuhar wrote:

nit: Could we use `getRoundingmode().value_or(theDefault)` to reduce branching?

https://github.com/llvm/llvm-project/pull/86152