[PATCH] D30134: AMDGPU: Fold FP clamp as modifier bit

[Jay Foad via Phabricator via llvm-commits]

foad added inline comments.
Herald added subscribers: kosarev, jsilvanus, kerbowa, t-tye, jvesely.
Herald added a project: All.


================
Comment at: test/CodeGen/AMDGPU/clamp.ll:108
 
 ; FIXME: Better to fold neg into max
+; SI: v_cvt_f32_f16_e64 [[CVT:v[0-9]+]], -[[A]] clamp{{$}}
----------------
Does this FIXME still make sense? Not sure what it was trying to say in the first place.


================
Comment at: test/CodeGen/AMDGPU/clamp.ll:128
 
 ; FIXME: Better to fold neg/abs into max
 
----------------
Ditto.


CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D30134/new/

https://reviews.llvm.org/D30134