[PATCH] D49561: AMDGPU: Try to make isKnownNeverSNan more accurate

Steve Canon via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Thu Jul 19 14:42:38 PDT 2018


scanon added inline comments.


================
Comment at: lib/Target/AMDGPU/SIISelLowering.cpp:6729
+  case ISD::FSUB:
+  case ISD::FMUL:
+  case ISD::FMAD:
----------------
scanon wrote:
> arsenm wrote:
> > rampitec wrote:
> > > Do these quiet incoming sNaNs?
> > That's my understanding of how the basic operations work
> Yes, all computational operations quiet sNaNs. The only things that produce sNaN are fcopysign, fabs, (fneg would if we had it), and things like loads and bitcasts.
(and to be clear, fcopysign and fabs can only produce sNaN if their input is sNaN.)


https://reviews.llvm.org/D49561





More information about the llvm-commits mailing list