[Bug target/99531] New: Performance regression since gcc 9 (argument passing / register allocation)

[Bug target/99531] New: Performance regression since gcc 9 (argument passing / register allocation)

[stefanrin at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

            Bug ID: 99531
           Summary: Performance regression since gcc 9 (argument passing /
                    register allocation)
           Product: gcc
           Version: 10.2.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: stefanrin at gmail dot com
  Target Milestone: ---
            Target: x86_64

For this source:

int func(int, int, int, int, int, int);
int caller(int a, int b, int c, int d, int e) { return func(0, a, b, c, d, e);
}

the code generated for caller is:

pushq %r12
movl %r8d, %r9d
popq %r12
movl %ecx, %r8d
movl %edx, %ecx
movl %esi, %edx
movl %edi, %esi
xorl %edi, %edi
jmp func

gcc 9 started producing the useless push/pop pair.

Mailing list link:
https://gcc.gnu.org/pipermail/gcc-help/2021-February/139885.html

[Bug target/99531] Performance regression since gcc 9 (argument passing / register allocation)

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2021-03-11
             Status|UNCONFIRMED                 |NEW
      Known to work|                            |8.4.0
      Known to fail|                            |11.0, 9.3.1

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Confirmed.

[Bug target/99531] [9/10/11 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|Performance regression      |[9/10/11 Regression]
                   |since gcc 9 (argument       |Performance regression
                   |passing / register          |since gcc 9 (argument
                   |allocation)                 |passing / register
                   |                            |allocation)
   Target Milestone|---                         |9.4

[Bug target/99531] [9/10/11 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

Martin Liška <marxin at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |marxin at gcc dot gnu.org,
                   |                            |vmakarov at gcc dot gnu.org

--- Comment #2 from Martin Liška <marxin at gcc dot gnu.org> ---
Started with r9-4310-geeae9314d4449b76:

caller:
.LFB0:
        .cfi_startproc
        pushq   %r12
        .cfi_def_cfa_offset 16
        .cfi_offset 12, -16
        movl    %r8d, %r9d
        popq    %r12
        .cfi_def_cfa_offset 8
        movl    %ecx, %r8d
        movl    %edx, %ecx
        movl    %esi, %edx
        movl    %edi, %esi
        xorl    %edi, %edi
        jmp     func
        .cfi_endproc

before the revision:

caller:
.LFB0:
        .cfi_startproc
        movl    %r8d, %r9d
        movl    %ecx, %r8d
        movl    %edx, %ecx
        movl    %esi, %edx
        movl    %edi, %esi
        xorl    %edi, %edi
        jmp     func
        .cfi_endproc

[Bug target/99531] [9/10/11 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P2

[Bug target/99531] [9/10/11/12 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|9.4                         |9.5

--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
GCC 9.4 is being released, retargeting bugs to GCC 9.5.

[Bug target/99531] [9/10/11/12 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[vmakarov at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

--- Comment #4 from Vladimir Makarov <vmakarov at gcc dot gnu.org> ---
  Thank you for reporting this.  It is true my patch caused this.

  I've reproduced the bug on master too.  I will be working on this PR.  I
think a fix will be ready on the next week the best as the fix will touch cost
calculations and will require a lot of testing on different targets.

[Bug target/99531] [9/10/11/12 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

--- Comment #5 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Vladimir Makarov <vmakarov@gcc.gnu.org>:

https://gcc.gnu.org/g:a7acb6dca941db2b1c135107dac3a34a20650d5c

commit r12-5944-ga7acb6dca941db2b1c135107dac3a34a20650d5c
Author: Vladimir N. Makarov <vmakarov@redhat.com>
Date:   Mon Dec 13 13:48:12 2021 -0500

    [PR99531] Modify pseudo class cost calculation when processing move
involving the pseudo and a hard register

    Pseudo class calculated on the 1st iteration should not have a
    special treatment in cost calculation when processing move involving
    the pseudo and a hard register.

    gcc/ChangeLog:

            PR target/99531
            * ira-costs.c (record_operand_costs): Do not take pseudo class
            calculated on the 1st iteration into account when processing move
            involving the pseudo and a hard register.

    gcc/testsuite/ChangeLog:

            PR target/99531
            * gcc.target/i386/pr99531.c: New test.

[Bug target/99531] [9/10/11/12 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

--- Comment #6 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Vladimir Makarov <vmakarov@gcc.gnu.org>:

https://gcc.gnu.org/g:973f6aedeb6489a9fdc7aeceaed0514348ee1e1a

commit r12-5957-g973f6aedeb6489a9fdc7aeceaed0514348ee1e1a
Author: Vladimir N. Makarov <vmakarov@redhat.com>
Date:   Tue Dec 14 08:57:30 2021 -0500

    [PR99531] Do not scan push insn for ia32 in the test

    The patch prohibits scanning push insn for ia32 as push are expected not to
be generated only for x86_64 Linux ABI.

    gcc/testsuite/ChangeLog:

            PR target/99531
            * gcc.target/i386/pr99531.c: Do not scan for ia32.

[Bug target/99531] [10/11/12/13 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|9.5                         |10.4

--- Comment #7 from Richard Biener <rguenth at gcc dot gnu.org> ---
GCC 9 branch is being closed

[Bug target/99531] [10/11/12/13 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|10.4                        |10.5

--- Comment #8 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
GCC 10.4 is being released, retargeting bugs to GCC 10.5.

[Bug target/99531] [10/11 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[roger at nextmovesoftware dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

Roger Sayle <roger at nextmovesoftware dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[10/11/12/13 Regression]    |[10/11 Regression]
                   |Performance regression      |Performance regression
                   |since gcc 9 (argument       |since gcc 9 (argument
                   |passing / register          |passing / register
                   |allocation)                 |allocation)
                 CC|                            |roger at nextmovesoftware dot com

--- Comment #9 from Roger Sayle <roger at nextmovesoftware dot com> ---
This has been fixed for GCC 12 and GCC 13.  Thanks to Vladimir.

[Bug target/99531] [11 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|10.5                        |11.5

--- Comment #10 from Richard Biener <rguenth at gcc dot gnu.org> ---
GCC 10 branch is being closed.

[Bug target/99531] [11 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[olegendo at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

Oleg Endo <olegendo at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |olegendo at gcc dot gnu.org

--- Comment #11 from Oleg Endo <olegendo at gcc dot gnu.org> ---
(In reply to GCC Commits from comment #5)
> The master branch has been updated by Vladimir Makarov
> <vmakarov@gcc.gnu.org>:
> 
> https://gcc.gnu.org/g:a7acb6dca941db2b1c135107dac3a34a20650d5c
> 
> commit r12-5944-ga7acb6dca941db2b1c135107dac3a34a20650d5c
> Author: Vladimir N. Makarov <vmakarov@redhat.com>
> Date:   Mon Dec 13 13:48:12 2021 -0500
> 
>     [PR99531] Modify pseudo class cost calculation when processing move
> involving the pseudo and a hard register
>     
>     Pseudo class calculated on the 1st iteration should not have a
>     special treatment in cost calculation when processing move involving
>     the pseudo and a hard register.
>     
>     gcc/ChangeLog:
>     
>             PR target/99531
>             * ira-costs.c (record_operand_costs): Do not take pseudo class
>             calculated on the 1st iteration into account when processing move
>             involving the pseudo and a hard register.
>     
>     gcc/testsuite/ChangeLog:
>     
>             PR target/99531
>             * gcc.target/i386/pr99531.c: New test.


This caused PR 115148

[Bug target/99531] [11 Regression] Performance regression since gcc 9 (argument passing / register allocation)

[olegendo at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=99531

--- Comment #12 from Oleg Endo <olegendo at gcc dot gnu.org> ---
(In reply to Oleg Endo from comment #11)
> 
> This caused PR 115148

I absolutely lack the imagination to see the connection of the change in #c6
and PR 115148.  This is the change in the output code that results in the
problem on SH:

.LVL108:
        bt/s    .L178           !
        mov     #-1,r0  !, <retval>
@@ -1832,36 +1830,39 @@
        .byte   .L215-.L190
        .byte   .L181-.L190
 .LVL109:
-       .align 1
-.L192:
 .LBE111:
 .LBE110:
        .loc 1 234 9
        .loc 1 234 14
+       .align 1
+.L287:
+       .align 1
+.L288:

Any ideas or suggestions are appreciated.