This is related to the discussion in D142071. It can solve the regression in D142071. I tested llvm/tests/Transforms/CodeGenPrepare/X86/overflow-intrinsics.ll, no new regression. Usually MathUsed is Add->hasNUsesOrMore(2), for optbranch_64, the add just has two uses. MathUsed is true. If we could check these use must come from the same BB, perhaps MathUsed could be false.

This is related to the discussion in D142071. It can solve the regression in D142071. I copied llvm/tests/Transforms/CodeGenPrepare/X86/overflow-intrinsics.ll to test shouldFormOverflowOp. Usually MathUsed is Add->hasNUsesOrMore(2), for optbranch_64, the add just has two uses. MathUsed is true. If we could check these use must come from the same BB, perhaps MathUsed could be false.

This is related to the discussion in D142071. It can solve the regression in D142071. I testI copied llvm/tests/Transforms/CodeGenPrepare/X86/overflow-intrinsics.ll, no new regression to test shouldFormOverflowOp. Usually MathUsed is Add->hasNUsesOrMore(2), for optbranch_64, the add just has two uses. MathUsed is true. If we could check these use must come from the same BB, perhaps MathUsed could be false.