https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123374
--- Comment #12 from Siddhesh Poyarekar <siddhesh at gcc dot gnu.org> --- > So maybe waccess2 should be moved after vrp2 or have better control of the > warnings that are conditionalized by __builtin_constant_p . Maybe waccess2 should limit itself to dangling pointer analysis (since that's what it appears to have been introduced for) and skip all other access checks? So basically instead of a bool early_checks, have 3 levels, early_checks, dangling_ptr, late_checks? waccess3 should be able to catch real issues waccess2 identifies and maybe even reduce false positives.
