[Bug tree-optimization/123038] [16 Regression] ICE at O3 during GIMPLE pass: vect

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123038

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
           Assignee|unassigned at gcc dot gnu.org      |rguenth at gcc dot gnu.org

--- Comment #4 from Richard Biener <rguenth at gcc dot gnu.org> ---
t.c:2:21: note:   vec_recog_ctz_ffs_pattern: detected: b_6 = __builtin_ffs
(b_10);
t.c:2:21: note:   created pattern stmt: patt_13 = patt_14 ? patt_9 : 0;
t.c:2:21: note:   ctz_ffs pattern recognized: patt_13 = patt_14 ? patt_9 : 0;
t.c:2:21: note:   extra pattern stmt: patt_8 = .CTZ (b_10, 32);
t.c:2:21: note:   extra pattern stmt: patt_9 = patt_8 + 1;
t.c:2:21: note:   extra pattern stmt: patt_14 = b_10 != 0;

this has b_10 used twice, so it's a pattern that needs to be guarded to not
apply to reductions.

I'll deal with this.