[PATCH] D133305: [AMDGPU] Simplify mad/mac patterns. NFC.

[Jay Foad via Phabricator via llvm-commits]

foad added inline comments.


================
Comment at: llvm/lib/Target/AMDGPU/SIInstructions.td:987
 // better to select mad instead of mac.
-class FMADPat <ValueType vt, Instruction inst, SDPatternOperator node>
-  : GCNPat <(vt (node (vt (VOP3NoMods vt:$src0)),
----------------
arsenm wrote:
> If we keep the node parameter, can we reuse this for v_fmac_*?
Not sure - it looks like V_FMAC_* lack the clamp and omod operands.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D133305/new/

https://reviews.llvm.org/D133305