diff options
| author | Matt Arsenault <Matthew.Arsenault@amd.com> | 2020-05-21 21:04:06 -0400 |
|---|---|---|
| committer | Matt Arsenault <Matthew.Arsenault@amd.com> | 2020-05-23 13:28:36 -0400 |
| commit | 27fe841aa650a24fd98da2fb6c6eb2fca806a63f (patch) | |
| tree | 9a44f6de2a4dde6b48fd60a54393f019a565f579 /llvm/lib/LTO/UpdateCompilerUsed.cpp | |
| parent | 1d96dca9491e3d75c11c3cd1acff5fcda8c2f613 (diff) | |
| download | llvm-27fe841aa650a24fd98da2fb6c6eb2fca806a63f.tar.gz llvm-27fe841aa650a24fd98da2fb6c6eb2fca806a63f.tar.bz2 llvm-27fe841aa650a24fd98da2fb6c6eb2fca806a63f.zip | |
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.
Diffstat (limited to 'llvm/lib/LTO/UpdateCompilerUsed.cpp')
0 files changed, 0 insertions, 0 deletions