[flang-commits] [PATCH] D139351: Lowering and runtime support for F08 transformational intrinsics: BESSEL_JN and BESSEL_YN

[Slava Zakharin via Phabricator via flang-commits]

vzakhari added inline comments.


================
Comment at: flang/lib/Lower/IntrinsicCall.cpp:2682
+    mlir::Value cmpXEq0 = builder.create<mlir::arith::CmpFOp>(
+        loc, mlir::arith::CmpFPredicate::UEQ, x, zero);
+    mlir::Value cmpN1LtN2 = builder.create<mlir::arith::CmpIOp>(
----------------
tarunprabhu wrote:
> vzakhari wrote:
> > Hello @tarunprabhu, is unordered predicate here intentional?
> I think I intended this to replicate the behavior of gfortran in the case when x is NaN. I'll have to double-check that though.
Yes, please double-check it, because treating `NaN` argument as zero may not be the right thing to do.  We'd better propagate the `NaN` argument to the result.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D139351/new/

https://reviews.llvm.org/D139351