[Bug tree-optimization/101758] Inconsistent optimizations with UBSan

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101758

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
          Component|sanitizer                   |tree-optimization
             Status|UNCONFIRMED                 |NEW
           Keywords|                            |missed-optimization
           Severity|normal                      |enhancement
   Last reconfirmed|                            |2021-08-03

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
This is ipa-icf going interesting.  We mark c and d as clones of b.  When we go
and do inlining, we fix up that a is no longer addressable because the check
for UBSAN_NULL is removed already.

So the real issue is after UBSAN_NULL is removed, there is no redoing
addressability of a.