https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95656
Jakub Jelinek <jakub at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |jakub at gcc dot gnu.org
--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
I don't see anything wrong on that.
You end up with essentially a != nullptr test, which is in some cases optimized
away, but during sanitization tests whether some address is non-NULL are
intentionally not optimized in case they would end up being NULL
(-fsanitize=undefined implies -fno-delete-null-pointer-checks).
You get the same result if you use -fno-delete-null-pointer-checks without
sanitization.
