[Bug tree-optimization/125461] [17 Regression] znver5: 14.2 vs trunk differ on SRA store shape/LIM and loop vector width (ZMM vs XMM)

[rguenth at gcc dot gnu.org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=125461

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |17.0
            Summary|znver5: 14.2 vs trunk       |[17 Regression] znver5:
                   |differ on SRA store         |14.2 vs trunk differ on SRA
                   |shape/LIM and loop vector   |store shape/LIM and loop
                   |width (ZMM vs XMM)          |vector width (ZMM vs XMM)
             Target|                            |x86_64-*-*

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
I'll note that starting with GCC 16 we vectorize the loop while GCC 14 and 15
only vectorize on a basic-block level.

I have not looked in more detail yet.  Marking as 17 regression only for now,
does GCC 15 behave OK or not?