[Bug testsuite/101169] New: [12 regression] test case gcc.target/powerpc/fold-vec-extract-char.p7.c fails after r10-9880

[Bug testsuite/101169] New: [12 regression] test case gcc.target/powerpc/fold-vec-extract-char.p7.c fails after r10-9880

[seurer at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101169

            Bug ID: 101169
           Summary: [12 regression] test case
                    gcc.target/powerpc/fold-vec-extract-char.p7.c fails
                    after r10-9880
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: testsuite
          Assignee: unassigned at gcc dot gnu.org
          Reporter: seurer at gcc dot gnu.org
  Target Milestone: ---

g:eb13f3f81d56910626529af52e03e74770ffca98, r10-9880
make  -k check-gcc
RUNTESTFLAGS="powerpc.exp=gcc.target/powerpc/fold-vec-extract-char.p7.c"
FAIL: gcc.target/powerpc/fold-vec-extract-char.p7.c scan-assembler-times
\\maddi\\M 6
# of expected passes            7
# of unexpected failures        1


Previously there were 6 addi instructions and now there are 9.  This may have a
performance impact.


commit eb13f3f81d56910626529af52e03e74770ffca98
Author: Vladimir N. Makarov <vmakarov@redhat.com>
Date:   Tue Jan 12 11:26:15 2021 -0500

    [PR97969] LRA: Transform pattern `plus (plus (hard reg, const),    
pseudo)` after elimination

[Bug testsuite/101169] [10 regression] test case gcc.target/powerpc/fold-vec-extract-char.p7.c fails after r10-9880

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101169

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |10.4
           Keywords|                            |missed-optimization
      Known to work|                            |10.3.0
      Known to fail|                            |10.3.1, 11.1.0, 12.0
                 CC|                            |dje at gcc dot gnu.org

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
Can you check why/whether this does not happen with GCC 11+?  What fixed it
there?

Ah, I see

commit 994fb69ac1b7d52348e84a021c07b24e285294d0
Author: David Edelsohn <dje.gcc@gmail.com>
Date:   Sun Jan 17 19:33:04 2021 -0500

    testsuite: powerpc fold-vec and sse updates.

    Recent code generation changes have affected the count of some
instructions.
    This patch updates the instruction count for fold-vec-extract on P7 and P8.

    Also, some of SSE emulation intrinsics only work on LE systems.

    gcc/testsuite/ChangeLog:

            * gcc.target/powerpc/fold-vec-extract-char.p7.c: Adjust addi count.

which adjusted it to 9 ... so the same thing happens there.  So it must be
fine for you.

[Bug testsuite/101169] [10 regression] test case gcc.target/powerpc/fold-vec-extract-char.p7.c fails after r10-9880

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101169

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

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

--- Comment #2 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
GCC 10.4 is being released, retargeting bugs to GCC 10.5.

[Bug rtl-optimization/101169] [10 regression] test case gcc.target/powerpc/fold-vec-extract-char.p7.c fails after r10-9880

[linkw at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101169

Kewen Lin <linkw at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |linkw at gcc dot gnu.org
          Component|testsuite                   |rtl-optimization

--- Comment #4 from Kewen Lin <linkw at gcc dot gnu.org> ---
The culprit change did cause some code gen regression, take a function
testbc_var from fold-vec-extract-char.p7.c as example:

// compiled with -O2 -mcpu=power7

#include <altivec.h>

unsigned char
testbc_var (vector bool char vbc2, signed int si)
{
  return vec_extract (vbc2, si);
}

on both Linux and aix 64-bit, or aix 32-bit, it generates one more addi than
before:

sub-optimal:

.L.testbc_var:
        li 9,48
        addi 10,1,-64
        stxvw4x 34,10,9
        rldicl 5,5,0,60
        addi 9,5,-64    // these two insns
        add 5,9,1       // can be optimized (see below)
        lbz 3,48(5)
        blr

vs. the previous:

.L.testbc_var:
        addi 10,1,-64
        li 9,48
        stxvw4x 34,10,9
        rldicl 5,5,0,60
        add 5,10,5
        lbz 3,48(5)
        blr

[Bug rtl-optimization/101169] [10 regression] test case gcc.target/powerpc/fold-vec-extract-char.p7.c fails after r10-9880

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101169

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|10.5                        |11.0
         Resolution|---                         |FIXED
             Status|NEW                         |RESOLVED

--- Comment #5 from Richard Biener <rguenth at gcc dot gnu.org> ---
Unclear status, closing anyway as fixed in 11+.

[Bug rtl-optimization/101169] [10 regression] test case gcc.target/powerpc/fold-vec-extract-char.p7.c fails after r10-9880

[linkw at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101169

--- Comment #6 from Kewen Lin <linkw at gcc dot gnu.org> ---
PR111850 reminded me this bug, the sub-optimal issue described in #comment 4
has been fixed on latest trunk, I think it's r14-4664-g04c9cf5c786b94.

[Bug rtl-optimization/101169] [10 regression] test case gcc.target/powerpc/fold-vec-extract-char.p7.c fails after r10-9880

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101169

--- Comment #7 from GCC Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Alexandre Oliva <aoliva@gcc.gnu.org>:

https://gcc.gnu.org/g:ac5c6c90a7f1f432fe19bc058d70b08967c25e3e

commit r15-923-gac5c6c90a7f1f432fe19bc058d70b08967c25e3e
Author: Alexandre Oliva <oliva@adacore.com>
Date:   Wed May 29 02:52:18 2024 -0300

    [testsuite] [powerpc] adjust -m32 counts for fold-vec-extract*

    Codegen changes caused add instruction count mismatches on
    ppc-*-linux-gnu and other 32-bit ppc targets.  At some point the
    expected counts were adjusted for lp64, but ilp32 differences
    remained, and published test results confirm it.


    for  gcc/testsuite/ChangeLog

            PR testsuite/101169
            * gcc.target/powerpc/fold-vec-extract-double.p7.c: Adjust addi
            counts for ilp32.
            * gcc.target/powerpc/fold-vec-extract-float.p7.c: Likewise.
            * gcc.target/powerpc/fold-vec-extract-float.p8.c: Likewise.
            * gcc.target/powerpc/fold-vec-extract-int.p7.c: Likewise.
            * gcc.target/powerpc/fold-vec-extract-int.p8.c: Likewise.
            * gcc.target/powerpc/fold-vec-extract-short.p7.c: Likewise.
            * gcc.target/powerpc/fold-vec-extract-short.p8.c: Likewise.