https://gcc.gnu.org/bugzilla/show_bug.cgi?id=121295
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Last reconfirmed| |2025-07-29
Ever confirmed|0 |1
Status|UNCONFIRMED |ASSIGNED
Assignee|unassigned at gcc dot gnu.org |pinskia at gcc dot
gnu.org
--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
So we have:
```
if (_2 == 0)
goto <bb 5>; [50.00%]
else
goto <bb 6>; [50.00%]
bb5:
a_lsm.6_20 = _2;
goto <bb 7>; [100.00%]
bb6:
a_lsm.6_21 = c.1_3;
bb7:
# RANGE [irange] int [0, 1] MASK 0x1 VALUE 0x0
# iftmp.0_5 = PHI <1(5), _13(6)>
# a_lsm.6_8 = PHI <a_lsm.6_20(5), a_lsm.6_21(6)>
```
We factor out a_lsm.6_21 assignment which is fine but we don't leave behind a
new phi for it.
And things get confused by that.
so I have a simple patch which reverts my previous patch (since it is overly
complicated) and changes the phi a_lsm.6_8 to be with the new arguments/result
since we know this is a 2 argument only phi it is ok.
