[Bug target/107271] New: [13 Regression] ICE: in expand_vec_perm_shufps_shufps, at config/i386/i386-expand.cc:19676 with __builtin_shuffle()

[Bug target/107271] New: [13 Regression] ICE: in expand_vec_perm_shufps_shufps, at config/i386/i386-expand.cc:19676 with __builtin_shuffle()

[zsojka at seznam dot cz]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107271

            Bug ID: 107271
           Summary: [13 Regression] ICE: in expand_vec_perm_shufps_shufps,
                    at config/i386/i386-expand.cc:19676 with
                    __builtin_shuffle()
           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 53707
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=53707&action=edit
reduced testcase

Compiler output:
$ x86_64-pc-linux-gnu-gcc testcase.c
during RTL pass: expand
In function 'bar',
    inlined from 'foo' at testcase.c:12:10:
testcase.c:6:10: internal compiler error: in expand_vec_perm_shufps_shufps, at
config/i386/i386-expand.cc:19676
    6 |   return __builtin_shuffle ((V){}, v128u32_0, v128u32_0);
      |          ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
0x8505c7 expand_vec_perm_shufps_shufps
        /repo/gcc-trunk/gcc/config/i386/i386-expand.cc:19676
0x8505c7 ix86_expand_vec_perm_const_1
        /repo/gcc-trunk/gcc/config/i386/i386-expand.cc:22268
0x185bebb ix86_vectorize_vec_perm_const(machine_mode, machine_mode, rtx_def*,
rtx_def*, rtx_def*, vec_perm_indices const&)
        /repo/gcc-trunk/gcc/config/i386/i386-expand.cc:22598
0x1286429 expand_vec_perm_const(machine_mode, rtx_def*, rtx_def*,
int_vector_builder<poly_int<1u, long> > const&, machine_mode, rtx_def*)
        /repo/gcc-trunk/gcc/optabs.cc:6254
0xfd7e72 expand_expr_real_2(separate_ops*, rtx_def*, machine_mode,
expand_modifier)
        /repo/gcc-trunk/gcc/expr.cc:10275
0xea9268 expand_gimple_stmt_1
        /repo/gcc-trunk/gcc/cfgexpand.cc:3983
0xea9268 expand_gimple_stmt
        /repo/gcc-trunk/gcc/cfgexpand.cc:4044
0xeaf286 expand_gimple_basic_block
        /repo/gcc-trunk/gcc/cfgexpand.cc:6096
0xeb0cb7 execute
        /repo/gcc-trunk/gcc/cfgexpand.cc:6822
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-r13-3311-20221015001738-gbaeec7cc83b-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-r13-3311-20221015001738-gbaeec7cc83b-checking-yes-rtl-df-extra-nobootstrap-amd64
Thread model: posix
Supported LTO compression algorithms: zlib zstd
gcc version 13.0.0 20221015 (experimental) (GCC)

[Bug target/107271] [13 Regression] ICE: in expand_vec_perm_shufps_shufps, at config/i386/i386-expand.cc:19676 with __builtin_shuffle()

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107271

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
   Target Milestone|---                         |13.0
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2022-10-17

--- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> ---
19675         /* When lone_idx is not 0, it must from second op(count == 1). 
*/
19676         gcc_assert (count == (lone_idx ? 1 : 3));

(gdb) p count
$5 = 3
(gdb) p lone_idx
$6 = 3

and d->perm[] is { 7, 4, 4, 0 }

maybe the function expects some canonicalization that isn't done?

[Bug target/107271] [13 Regression] ICE: in expand_vec_perm_shufps_shufps, at config/i386/i386-expand.cc:19676 with __builtin_shuffle() since r13-2843-g3db8e9c2422d924a

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107271

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[13 Regression] ICE: in     |[13 Regression] ICE: in
                   |expand_vec_perm_shufps_shuf |expand_vec_perm_shufps_shuf
                   |ps, at                      |ps, at
                   |config/i386/i386-expand.cc: |config/i386/i386-expand.cc:
                   |19676 with                  |19676 with
                   |__builtin_shuffle()         |__builtin_shuffle() since
                   |                            |r13-2843-g3db8e9c2422d924a
                 CC|                            |marxin at gcc dot gnu.org

--- Comment #2 from Martin Liška <marxin at gcc dot gnu.org> ---
Started with r13-2843-g3db8e9c2422d924a.

[Bug target/107271] [13 Regression] ICE: in expand_vec_perm_shufps_shufps, at config/i386/i386-expand.cc:19676 with __builtin_shuffle() since r13-2843-g3db8e9c2422d924a

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107271

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

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

[Bug target/107271] [13 Regression] ICE: in expand_vec_perm_shufps_shufps, at config/i386/i386-expand.cc:19676 with __builtin_shuffle() since r13-2843-g3db8e9c2422d924a

[crazylht at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107271

--- Comment #3 from Hongtao.liu <crazylht at gmail dot com> ---
I remember GIMPLE will canonicalize the selector index to make the first index
always comes the frist vector, means

It excepts canonicalized form like { 3, 0, 0, 4 } with op0 and op1 swapped.

[Bug target/107271] [13 Regression] ICE: in expand_vec_perm_shufps_shufps, at config/i386/i386-expand.cc:19676 with __builtin_shuffle() since r13-2843-g3db8e9c2422d924a

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107271

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

https://gcc.gnu.org/g:1442e2031e0bc2d0a5bf88ef3c92c5410e044bab

commit r13-3368-g1442e2031e0bc2d0a5bf88ef3c92c5410e044bab
Author: liuhongt <hongtao.liu@intel.com>
Date:   Tue Oct 18 16:58:52 2022 +0800

    Canonicalize vec_perm index to make the first index come from the first
vector.

    Fix unexpected non-canon form from gimple vector selector.

    gcc/ChangeLog:

            PR target/107271
            * config/i386/i386-expand.cc (ix86_vec_perm_index_canon): New.
            (expand_vec_perm_shufps_shufps): Call
            ix86_vec_perm_index_canon

    gcc/testsuite/ChangeLog:

            * gcc.target/i386/pr107271.c: New test.

[Bug target/107271] [13 Regression] ICE: in expand_vec_perm_shufps_shufps, at config/i386/i386-expand.cc:19676 with __builtin_shuffle() since r13-2843-g3db8e9c2422d924a

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107271

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

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

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