[PATCH] D142166: [X86] Make `shouldExpandLogicAtomicRMWInIR` able to match both operands.

[Noah Goldstein via Phabricator via llvm-commits]

goldstein.w.n added inline comments.


================
Comment at: llvm/test/CodeGen/X86/atomic-rm-bit-test.ll:159
   %1 = atomicrmw or ptr %v, i8 %shl monotonic, align 1
-  %conv7 = and i8 %1, %shl
+  %conv7 = and i8 %shl, %1
   ret i8 %conv7
----------------
This seems to change RA (or maybe something else). Whatever it is there is some logic that is a bit faulty and make assumption about operands.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D142166/new/

https://reviews.llvm.org/D142166