https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122604
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Severity|normal |enhancement
Status|UNCONFIRMED |NEW
Last reconfirmed| |2025-11-07
Ever confirmed|0 |1
CC| |pinskia at gcc dot gnu.org
--- Comment #2 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Confirmed.
Well actually there might be other issues as we inline common_part early and
don't have a chance to do what I was thinking of.
Here we have:
```
if (found_3(D) != 0)
goto <bb 3>; [50.00%]
else
goto <bb 4>; [50.00%]
<bb 3> [local count: 536870912]:
k_5 = b_4(D) & 31;
_17 = k_5 + _27;
_18 = _17 ^ -1640531527;
# .MEM_25 = VDEF <.MEM_6(D)>
e_7(D)->_v[k_5] = _18;
_20 = (signed char) k_5;
_21 = _20 ^ _28;
_22 = (unsigned char) _21;
# .MEM_26 = VDEF <.MEM_25>
e_7(D)->_s[k_5] = _22;
goto <bb 5>; [100.00%]
<bb 4> [local count: 536870912]:
_1 = __builtin_ctz (m_8(D));
_10 = _1 + _27;
_11 = _10 ^ -1640531527;
# .MEM_23 = VDEF <.MEM_6(D)>
e_7(D)->_v[_1] = _11;
_13 = (signed char) _1;
_14 = _13 ^ _28;
_15 = (unsigned char) _14;
# .MEM_24 = VDEF <.MEM_23>
e_7(D)->_s[_1] = _15;
<bb 5> [local count: 1073741824]:
# .MEM_2 = PHI <.MEM_24(4), .MEM_26(3)>
# VUSE <.MEM_2>
```
We should be able pull the everything out from that phi but don't currently.
