https://gcc.gnu.org/bugzilla/show_bug.cgi?id=79349
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Depends on| |23384
--- Comment #7 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
I am 99% sure this is actually due to the escape analysis is not fully flow
sensitive.
We think s can be modified by the call by foo but s does not escape until after
the call to foo (via the exception basic block).
So we don't optimize:
```
MEM[(struct _Alloc_hider *)&s]._M_p = &s.D.25271._M_local_buf;
__builtin_memcpy (&s.D.25271._M_local_buf, "abc", 3);
s._M_string_length = 3;
MEM[(char_type &)&s + 19] = 0;
foo ();
<bb 3> [local count: 1073741824]:
_24 = s._M_dataplus._M_p;
if (&s.D.25271._M_local_buf == _24)
```
Which means this depends on PR 23384.
Referenced Bugs:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=23384
[Bug 23384] escaped set should be flow sensitive
