diff options
| author | Matt Arsenault <Matthew.Arsenault@amd.com> | 2023-06-11 12:41:43 -0400 |
|---|---|---|
| committer | Matt Arsenault <Matthew.Arsenault@amd.com> | 2023-06-23 08:37:37 -0400 |
| commit | 89ccfa1b3978fa2c2945dec743401f090a23abec (patch) | |
| tree | e34664d61d446d33a6581c55dab55639073b889a /llvm/lib/Object/ObjectFile.cpp | |
| parent | 813f6a495bb695c46d5d18a35dd9b6db3e83af22 (diff) | |
| download | llvm-89ccfa1b3978fa2c2945dec743401f090a23abec.zip llvm-89ccfa1b3978fa2c2945dec743401f090a23abec.tar.gz llvm-89ccfa1b3978fa2c2945dec743401f090a23abec.tar.bz2 | |
AMDGPU: Use correct lowering for llvm.log2.f32
We previously directly codegened to v_log_f32, which is broken for
denormals. The lowering isn't complicated, you simply need to scale
denormal inputs and adjust the result. Note log and log10 are still
not accurate enough, and will be fixed separately.
Diffstat (limited to 'llvm/lib/Object/ObjectFile.cpp')
0 files changed, 0 insertions, 0 deletions