[PATCH] D69978: Separately track input and output denormal mode

[Matt Arsenault via Phabricator via cfe-commits]

arsenm marked 3 inline comments as done.
arsenm added inline comments.


================
Comment at: llvm/docs/LangRef.rst:1829
+   operations. The second indicates the handling of denormal inputs to
+   floating point instructions.
+
----------------
andrew.w.kaylor wrote:
> Based on the changes below, if the second value is omitted the input mode will be assumed to be the same as the output mode. That should probably be documented. I guess you intend for that not to happen, but the documentation here leaves the result ambiguous if it does happen.
I've added a note for this


================
Comment at: llvm/docs/LangRef.rst:1848
+   should be converted to 0 as if by ``@llvm.canonicalize`` during
+   lowering.
+
----------------
andrew.w.kaylor wrote:
> Is this saying that if a backend generates an instruction that doesn't handle the hardware daz mode then it must insert instructions to check for normals and convert them to zero? If so, do you intend this to apply to all such instructions or only instructions that aren't able to accept denormal inputs?
Only in cases where denormal inputs are invalid or unhandled. The case I'm thinking of is the one user in DAGCombiner, where if denormals are not flushed the result ends up incorrect (see https://bugs.llvm.org/show_bug.cgi?id=34994)


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D69978/new/

https://reviews.llvm.org/D69978