https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111801
Andrew Pinski <pinskia at gcc dot gnu.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Last reconfirmed| |2023-10-13 Status|UNCONFIRMED |NEW Ever confirmed|0 |1 --- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> --- Confirmed. What happens after the patch is `if (j == &g == 0)` still is kept around and we decide to jump thread into that block and duplicate the `if (f)` there (I think) and then there is a missing PRE (I think) where we could see that f is stored and we load from f.