[Bug target/106577] New: [13 Regression] during RTL pass: subreg3 ICE: in extract_insn, at recog.cc:2791 (unrecognizable insn) with -O -mavx

[Bug target/106577] New: [13 Regression] during RTL pass: subreg3 ICE: in extract_insn, at recog.cc:2791 (unrecognizable insn) with -O -mavx

[zsojka at seznam dot cz]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106577

            Bug ID: 106577
           Summary: [13 Regression] during RTL pass: subreg3 ICE: in
                    extract_insn, at recog.cc:2791 (unrecognizable insn)
                    with -O -mavx
           Product: gcc
           Version: 13.0
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: zsojka at seznam dot cz
  Target Milestone: ---
              Host: x86_64-pc-linux-gnu
            Target: x86_64-pc-linux-gnu

Created attachment 53429
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=53429&action=edit
reduced testcase

Compiler output:
$ x86_64-pc-linux-gnu-gcc -O -mavx testcase.c
testcase.c: In function 'foo':
testcase.c:4:1: error: unrecognizable insn:
    4 | }
      | ^
(insn 28 27 30 2 (parallel [
            (set (reg:TI 98)
                (and:TI (reg:V4SI 100)
                    (reg:TI 99)))
            (clobber (reg:CC 17 flags))
        ]) "testcase.c":3:8 -1
     (nil))
during RTL pass: subreg3
testcase.c:4:1: internal compiler error: in extract_insn, at recog.cc:2791
0x77b2c8 _fatal_insn(char const*, rtx_def const*, char const*, int, char
const*)
        /repo/gcc-trunk/gcc/rtl-error.cc:108
0x77b34b _fatal_insn_not_found(rtx_def const*, char const*, int, char const*)
        /repo/gcc-trunk/gcc/rtl-error.cc:116
0x769bfd extract_insn(rtx_insn*)
        /repo/gcc-trunk/gcc/recog.cc:2791
0x24ab7b7 decompose_multiword_subregs
        /repo/gcc-trunk/gcc/lower-subreg.cc:1555
0x24ac8dd execute
        /repo/gcc-trunk/gcc/lower-subreg.cc:1820
Please submit a full bug report, with preprocessed source (by using
-freport-bug).
Please include the complete backtrace with any bug report.
See <https://gcc.gnu.org/bugs/> for instructions.

$ x86_64-pc-linux-gnu-gcc -v
Using built-in specs.
COLLECT_GCC=/repo/gcc-trunk/binary-latest-amd64/bin/x86_64-pc-linux-gnu-gcc
COLLECT_LTO_WRAPPER=/repo/gcc-trunk/binary-trunk-r13-2009-20220810070724-gc16d9f78dc8-checking-yes-rtl-df-extra-nobootstrap-amd64/bin/../libexec/gcc/x86_64-pc-linux-gnu/13.0.0/lto-wrapper
Target: x86_64-pc-linux-gnu
Configured with: /repo/gcc-trunk//configure --enable-languages=c,c++
--enable-valgrind-annotations --disable-nls --enable-checking=yes,rtl,df,extra
--disable-bootstrap --with-cloog --with-ppl --with-isl
--build=x86_64-pc-linux-gnu --host=x86_64-pc-linux-gnu
--target=x86_64-pc-linux-gnu --with-ld=/usr/bin/x86_64-pc-linux-gnu-ld
--with-as=/usr/bin/x86_64-pc-linux-gnu-as --disable-libstdcxx-pch
--prefix=/repo/gcc-trunk//binary-trunk-r13-2009-20220810070724-gc16d9f78dc8-checking-yes-rtl-df-extra-nobootstrap-amd64
Thread model: posix
Supported LTO compression algorithms: zlib zstd
gcc version 13.0.0 20220810 (experimental) (GCC)

[Bug target/106577] [13 Regression] during RTL pass: subreg3 ICE: in extract_insn, at recog.cc:2791 (unrecognizable insn) with -O -mavx since r13-2006-ga56c1641e9d25e46

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106577

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
            Summary|[13 Regression] during RTL  |[13 Regression] during RTL
                   |pass: subreg3 ICE: in       |pass: subreg3 ICE: in
                   |extract_insn, at            |extract_insn, at
                   |recog.cc:2791               |recog.cc:2791
                   |(unrecognizable insn) with  |(unrecognizable insn) with
                   |-O -mavx                    |-O -mavx since
                   |                            |r13-2006-ga56c1641e9d25e46
   Last reconfirmed|                            |2022-08-10
     Ever confirmed|0                           |1
                 CC|                            |marxin at gcc dot gnu.org,
                   |                            |sayle at gcc dot gnu.org

--- Comment #1 from Martin Liška <marxin at gcc dot gnu.org> ---
Started with r13-2006-ga56c1641e9d25e46.

[Bug target/106577] [13 Regression] during RTL pass: subreg3 ICE: in extract_insn, at recog.cc:2791 (unrecognizable insn) with -O -mavx since r13-2006-ga56c1641e9d25e46

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106577

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |13.0

[Bug target/106577] [13 Regression] during RTL pass: subreg3 ICE: in extract_insn, at recog.cc:2791 (unrecognizable insn) with -O -mavx since r13-2006-ga56c1641e9d25e46

[roger at nextmovesoftware dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106577

Roger Sayle <roger at nextmovesoftware dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |roger at nextmovesoftware dot com

--- Comment #2 from Roger Sayle <roger at nextmovesoftware dot com> ---
I'm bootstrapping and regression testing a fix/workaround, but I'll not assign
this bug to myself (yet) as the maintainers might not like my (hacky) solution.
 The problem is that calling "force_reg" may sometimes (on some targets)
clobber recog_data.operands, so it's not universally safe to call it from
within a pre-reload splitter(!?).

[Bug target/106577] [13 Regression] during RTL pass: subreg3 ICE: in extract_insn, at recog.cc:2791 (unrecognizable insn) with -O -mavx since r13-2006-ga56c1641e9d25e46

[roger at nextmovesoftware dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106577

--- Comment #3 from Roger Sayle <roger at nextmovesoftware dot com> ---
Patch proposed:
https://gcc.gnu.org/pipermail/gcc-patches/2022-August/599663.html

[Bug target/106577] [13 Regression] during RTL pass: subreg3 ICE: in extract_insn, at recog.cc:2791 (unrecognizable insn) with -O -mavx since r13-2006-ga56c1641e9d25e46

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106577

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P1

[Bug target/106577] [13 Regression] during RTL pass: subreg3 ICE: in extract_insn, at recog.cc:2791 (unrecognizable insn) with -O -mavx since r13-2006-ga56c1641e9d25e46

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106577

--- Comment #4 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Jakub Jelinek <jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:b3237a2c6847993f92218b65f96ece9831a8bfb0

commit r13-4462-gb3237a2c6847993f92218b65f96ece9831a8bfb0
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Fri Dec 2 11:08:45 2022 +0100

    i386: Save/restore recog_data in ix86_vector_duplicate_value [PR106577]

    On Tue, Aug 16, 2022 at 09:14:06AM +0100, Richard Sandiford via Gcc-patches
wrote:
    > IMO the correct low-effort fix is to save and restore recog_data
    > in ix86_vector_duplicate_value.  It's a relatively big copy,
    > but the current code is pretty wasteful anyway (allocating at
    > least a new SET and INSN for every query).  Compared to the
    > overhead of doing that, a copy to and from the stack shouldn't
    > be too bad.

    The following patch does that.
    It isn't the first spot in the compiler that does that, not even the first
    spot in the i386 backend.
    In i386-expand.cc beyond these 2 recog_memoized calls there is one in
    expand_vselect, but I think it is unlikely we'd run into these issues
trying
    to expand new permutations from splitters.

    2022-12-02  Jakub Jelinek  <jakub@redhat.com>

            PR target/106577
            * config/i386/i386-expand.cc (ix86_vector_duplicate_value):
Save/restore
            recog_data around recog_memoized calls.

            * gcc.target/i386/pr106577.c: New test.

[Bug target/106577] [13 Regression] during RTL pass: subreg3 ICE: in extract_insn, at recog.cc:2791 (unrecognizable insn) with -O -mavx since r13-2006-ga56c1641e9d25e46

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=106577

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |jakub at gcc dot gnu.org
             Status|NEW                         |RESOLVED
         Resolution|---                         |FIXED

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Fixed.