On Wed, 2022-09-07 at 17:51 +0200, Tim Lange wrote: > Hi Dave, > > while re-reading that patch, I noticed a small mistake. I accidently > did > not move the op == PLUS_EXPR or MULT_EXPR guard in > symbolic_greater_than > when implementing the "eliminate operands on both sides" feature, > which > lead to the old patch also eliminating operands on both sides if the > operator decreases the value, which is obviously wrong. > > I moved the guard outside and added test coverage for this in > symbolic-gt-1.c. The patch passed the regrtests with the fix > included. > > I assume it is still okay for trunk?
Yes it is - thanks for the updated patch. Dave
