public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug target/114960] New: [12/13/14/15 Regression] fails to clean up vector casts
@ 2024-05-06 14:09 amonakov at gcc dot gnu.org
  2024-05-06 16:29 ` [Bug rtl-optimization/114960] " pinskia at gcc dot gnu.org
  2024-05-07  8:48 ` rguenth at gcc dot gnu.org
  0 siblings, 2 replies; 3+ messages in thread
From: amonakov at gcc dot gnu.org @ 2024-05-06 14:09 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114960

            Bug ID: 114960
           Summary: [12/13/14/15 Regression] fails to clean up vector
                    casts
           Product: gcc
           Version: 12.3.1
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: amonakov at gcc dot gnu.org
  Target Milestone: ---
            Target: x86_64-*-*

#include <immintrin.h>

__m128i f(__m128i dummy, __m128i x)
{
    __v16qi v = { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 };
    return (__m128i)((__v16qi)x + v);
}

Starting from gcc-12, we get

        movdqa  .LC0(%rip), %xmm2
        paddb   %xmm1, %xmm2
        movdqa  %xmm2, %xmm0
        ret

while previously it was

        movdqa  .LC0(%rip), %xmm0
        paddb   %xmm1, %xmm0
        ret

Diverges in RTL land, starting in combine.

^ permalink raw reply	[flat|nested] 3+ messages in thread
* [Bug rtl-optimization/114960] [12/13/14/15 Regression] fails to clean up vector casts
  2024-05-06 14:09 [Bug target/114960] New: [12/13/14/15 Regression] fails to clean up vector casts amonakov at gcc dot gnu.org
@ 2024-05-06 16:29 ` pinskia at gcc dot gnu.org
  2024-05-07  8:48 ` rguenth at gcc dot gnu.org
  1 sibling, 0 replies; 3+ messages in thread
From: pinskia at gcc dot gnu.org @ 2024-05-06 16:29 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114960

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|target                      |rtl-optimization
           Keywords|                            |missed-optimization, ra

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
The difference between GCC 11 and 12 is just:
(subreg:V16QI (reg:V2DI 84 [ <retval> ]) 0)

vs
(reg:V16QI 88)

For the lhs of the plus.

This is a RA issue. There are a few bugs about subreg and RA somewhere ...

^ permalink raw reply	[flat|nested] 3+ messages in thread
* [Bug rtl-optimization/114960] [12/13/14/15 Regression] fails to clean up vector casts
  2024-05-06 14:09 [Bug target/114960] New: [12/13/14/15 Regression] fails to clean up vector casts amonakov at gcc dot gnu.org
  2024-05-06 16:29 ` [Bug rtl-optimization/114960] " pinskia at gcc dot gnu.org
@ 2024-05-07  8:48 ` rguenth at gcc dot gnu.org
  1 sibling, 0 replies; 3+ messages in thread
From: rguenth at gcc dot gnu.org @ 2024-05-07  8:48 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114960

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |12.4

^ permalink raw reply	[flat|nested] 3+ messages in thread
end of thread, other threads:[~2024-05-07  8:48 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-05-06 14:09 [Bug target/114960] New: [12/13/14/15 Regression] fails to clean up vector casts amonakov at gcc dot gnu.org
2024-05-06 16:29 ` [Bug rtl-optimization/114960] " pinskia at gcc dot gnu.org
2024-05-07  8:48 ` rguenth at gcc dot gnu.org

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).