[PATCH] D133300: [InstCombine] Matrix multiplication negation optimisation

[Sanjay Patel via Phabricator via llvm-commits]

spatel added inline comments.


================
Comment at: llvm/lib/Transforms/InstCombine/InstCombineCalls.cpp:1850-1858
+    bool MatchOp0 = match(Op0, m_FNeg(m_Value(X)));
+    bool MatchOp1 = match(Op1, m_FNeg(m_Value(X)));
+    if (MatchOp0 && MatchOp1) {
+      Instruction *FNegOp0 = cast<Instruction>(Op0);
+      Instruction *FNegOp1 = cast<Instruction>(Op1);
+      replaceInstUsesWith(*FNegOp0, FNegOp0->getOperand(0));
+      replaceInstUsesWith(*FNegOp1, FNegOp1->getOperand(0));
----------------
I don't think this is correct if an fneg has multiple uses (similar to the bug noted earlier, and I repeat my suggestion to create new instructions rather than modifying existing ones).

Please split this change and tests to its own review ahead of the original transforms in this patch. 


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D133300/new/

https://reviews.llvm.org/D133300