On June 6, 2021 12:54:03 AM GMT+02:00, Andrew Pinski via Gcc <gcc@gcc.gnu.org> wrote: >While debugging PR 100925, I found that after my match.pd (for >A?CST0:CST1) and phiopt patch to use match-and-simplify, gcc will >produce an negate assignment gimple which has a pointer type (and >offset type). Before we would produce "cond ? -1B : 0B" in this case >but now we produce: > _4 = a_2(D) == 0; > _5 = (void *) _4; > _6 = -_5; >Which I think would be invalid gimple but the verifiers don't say it >is. >Should the verifier reject this gimple and similar ones? I am going >to do some simple testing to see what code hits it too.
I think this is invalid GIMPLE. We only allow BIT_xxx on pointer types. But digging in history might reveal sth I don't remember... Richard. >Thanks, >Andrew Pinski
