[llvm-branch-commits] [llvm] [AMDGPU][InstCombine] Fold constant add/sub into the dot accumulator (PR #225002)

[Harrison Hao via llvm-branch-commits]

================
@@ -1968,6 +1968,44 @@ GCNTTIImpl::instCombineIntrinsic(InstCombiner &IC, IntrinsicInst &II) const {
     Result = scalbn(Result, Scale, RoundingMode::NearestTiesToEven);
     return IC.replaceInstUsesWith(II, ConstantFP::get(Src->getType(), Result));
   }
+  case Intrinsic::amdgcn_sdot2:
+  case Intrinsic::amdgcn_udot2:
+  case Intrinsic::amdgcn_sdot4:
+  case Intrinsic::amdgcn_udot4:
+  case Intrinsic::amdgcn_sdot8:
+  case Intrinsic::amdgcn_udot8: {
+    if (!match(II.getArgOperand(3), m_Zero()) || !II.hasOneUse())
----------------
harrisonGPU wrote:

I initially wanted to use `m_False()` as well, but it does not exist in LLVM IR's PatternMatch API. :-)

https://github.com/llvm/llvm-project/pull/225002