projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 1d96dca) | patch
AMDGPU: Refine rcp/rsq intrinsic folding for modern FP rules
authorMatt Arsenault <Matthew.Arsenault@amd.com>
Fri, 22 May 2020 01:04:06 +0000 (21:04 -0400)
committerMatt Arsenault <Matthew.Arsenault@amd.com>
Sat, 23 May 2020 17:28:36 +0000 (13:28 -0400)
commit27fe841aa650a24fd98da2fb6c6eb2fca806a63f
tree9a44f6de2a4dde6b48fd60a54393f019a565f579tree | snapshot
parent1d96dca9491e3d75c11c3cd1acff5fcda8c2f613commit | diff
AMDGPU: Refine rcp/rsq intrinsic folding for modern FP rules

We have to assume undef could be an snan, which would need quieting so
returning qnan is safer than undef. Also consider strictfp, and don't
care if the result rounded.
llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp diff | blob | history
llvm/test/Transforms/InstCombine/AMDGPU/amdgcn-intrinsics.ll diff | blob | history
Domain: System / Toolchain;