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 ***