[PATCH] D69275: Add constrained int->FP intrinsics

[Ulrich Weigand via Phabricator via llvm-commits]

uweigand added a comment.

In D69275#1786456 <https://reviews.llvm.org/D69275#1786456>, @craig.topper wrote:

> General question for you and @uweigand that I realized today. Do we need to set the FPExcept bit in the flags for new nodes when we expand/promote operations?


Ah yes, we need to do that.

This reminds me of a general concern: for all other flag bits, omitting the flag is conservatively safe, it just may impact performance.  But for FPExcept, omitting the flag impacts correctness.  Maybe we should go ahead and invert the sense of the flag (i.e. use a FPNoExcept flag instead of FPExcept) after all ...


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D69275/new/

https://reviews.llvm.org/D69275