[Bug target/97192] [11 Regression] ICE: in gimple_expand_vec_cond_expr, at gimple-isel.cc:133 with -O -ftracer -mavx512vl

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97192

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |DUPLICATE
             Status|UNCONFIRMED                 |RESOLVED

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Looks like the same case as for PR97085 for aarch64:

-  _19 = VEC_COND_EXPR <_18, { -1, -1, -1, -1, -1, -1, -1, -1 }, { 0, 0, 0, 0,
0, 0, 0, 0 }>;
-  _20 = {a.2_2, a.2_2, a.2_2, a.2_2, a.2_2, a.2_2, a.2_2, a.2_2};
-  _21 = _19 == _20;
+  _21 = VEC_COND_EXPR <_18, { 0, 0, 0, 0, 0, 0, 0, 0 }, { -1, -1, -1, -1, -1,
-1, -1, -1 }>;
   _22 = VEC_COND_EXPR <_21, { -1, -1, -1, -1, -1, -1, -1, -1 }, { 0, 0, 0, 0,
0, 0, 0, 0 }>;

so we compute the VEC_COND_EXPR mask using a VEC_COND_EXPR.  And it's indeed
fixed by the fix for said PR.  I will add the testcase.

*** This bug has been marked as a duplicate of bug 97085 ***