[PATCH] D57281: [X86] combineCarryThroughADD - add support for X86::COND_A commutations (PR24545)

Sat Jan 26 13:11:12 PST 2019

RKSimon added inline comments.

================
Comment at: llvm/trunk/lib/Target/X86/X86ISelLowering.cpp:34904
+          //
+          if (CarryOp1.getOpcode() == X86ISD::SUB && CarryOp1.hasOneUse() &&
+              CarryOp1.getValueType().isInteger() &&
----------------
craig.topper wrote:
> Do we need to make sure the data output of the subtract doesn't have any users? CarryOp1.hasOneUse() will only check the flag result has one use. CarryOp1->hasOneUse() will check that the entire node only has one use.
I can do that - interestingly the code in combineAddOrSubToADCOrSBB (where I got this from) doesn't do this either

Repository:
  rL LLVM

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D57281/new/

https://reviews.llvm.org/D57281