[Bug c/97880] New: [9/10/11 Regression] ICE in emit_library_call_value_1, at calls.c:5298

[Bug c/97880] New: [9/10/11 Regression] ICE in emit_library_call_value_1, at calls.c:5298

[gscfq@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97880

            Bug ID: 97880
           Summary: [9/10/11 Regression] ICE in emit_library_call_value_1,
                    at calls.c:5298
           Product: gcc
           Version: 11.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
          Assignee: unassigned at gcc dot gnu.org
          Reporter: gscfq@t-online.de
  Target Milestone: ---

With option -ftrapv at -O0, down to version r7 :


$ cat z1.c
void f ()
{
  #pragma acc parallel loop tile(2, 3)
  for (int i = 0; i < 8; i++)
    for (long j = 0; j < 8; j++);
}


$ gcc-6 -c z1.c -fopenacc -ftrapv -O0
$
$ gcc-11-20201115 -c z1.c -fopenacc
$ gcc-11-20201115 -c z1.c -fopenacc -ftrapv -O1
$
$ gcc-11-20201115 -c z1.c -fopenacc -ftrapv -O0
during RTL pass: expand
z1.c: In function 'f._omp_fn.0':
z1.c:3:11: internal compiler error: in emit_library_call_value_1, at
calls.c:5298
    3 |   #pragma acc parallel loop tile(2, 3)
      |           ^~~
0x7304a0 emit_library_call_value_1(int, rtx_def*, rtx_def*, libcall_type,
machine_mode, int, std::pair<rtx_def*, machine_mode>*)
        ../../gcc/calls.c:5296
0xa3b4c8 emit_library_call_value(rtx_def*, rtx_def*, libcall_type,
machine_mode, rtx_def*, machine_mode, rtx_def*, machine_mode)
        ../../gcc/rtl.h:4258
0xa3b4c8 expand_binop(machine_mode, optab_tag, rtx_def*, rtx_def*, rtx_def*,
int, optab_methods)
        ../../gcc/optabs.c:1837
0x83bd01 expand_mult(machine_mode, rtx_def*, rtx_def*, rtx_def*, int, bool)
        ../../gcc/expmed.c:3571
0x858a04 expand_expr_real_2(separate_ops*, rtx_def*, machine_mode,
expand_modifier)
        ../../gcc/expr.c:9175
0x74af02 expand_gimple_stmt_1
        ../../gcc/cfgexpand.c:3816
0x74af02 expand_gimple_stmt
        ../../gcc/cfgexpand.c:3877
0x74fb37 expand_gimple_basic_block
        ../../gcc/cfgexpand.c:5918
0x75219e execute
        ../../gcc/cfgexpand.c:6602

[Bug c/97880] [8/9/10/11 Regression] ICE in emit_library_call_value_1, at calls.c:5298

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97880

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |openacc
   Target Milestone|---                         |8.5
            Summary|[9/10/11 Regression] ICE in |[8/9/10/11 Regression] ICE
                   |emit_library_call_value_1,  |in
                   |at calls.c:5298             |emit_library_call_value_1,
                   |                            |at calls.c:5298

[Bug c/97880] [8/9/10/11 Regression] ICE in emit_library_call_value_1, at calls.c:5298

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97880

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever confirmed|0                           |1
   Last reconfirmed|                            |2020-11-18
                 CC|                            |cesar at codesourcery dot com,
                   |                            |marxin at gcc dot gnu.org,
                   |                            |nathan at gcc dot gnu.org

--- Comment #1 from Martin Liška <marxin at gcc dot gnu.org> ---
Confirmed, started with r7-6598-g02889d23ee3b0285.

[Bug c/97880] [8/9/10/11 Regression] ICE in emit_library_call_value_1, at calls.c:5298

[burnus at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97880

Tobias Burnus <burnus at gcc dot gnu.org> changed:

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

--- Comment #2 from Tobias Burnus <burnus at gcc dot gnu.org> ---
With trunk, I get for 
  for (int i = 0; i < 8; i++)
    for (long j = 0; j < 8; j++);

foocc.c: In function ‘f._omp_fn.0’:
foocc.c:6:1: error: type mismatch in binary expression
    6 | }
      | ^
signed long

long int

int

D.2125 = .tile.5 * .tile.7;
during IPA pass: *free_lang_data


Draft patch:

--- a/gcc/omp-expand.c
+++ b/gcc/omp-expand.c
@@ -7561 +7561,2 @@ expand_oacc_for (struct omp_region *region, struct
omp_for_data *fd)
-       expr = fold_build2 (MULT_EXPR, diff_type, counts[ix].tile, expr);
+       expr = fold_build2 (MULT_EXPR, diff_type,
+                           fold_convert (diff_type, counts[ix].tile), expr);

[Bug c/97880] [8/9/10/11 Regression] [OpenACC] ICE in emit_library_call_value_1, at calls.c:5298

[burnus at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97880

Tobias Burnus <burnus at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[8/9/10/11 Regression] ICE  |[8/9/10/11 Regression]
                   |in                          |[OpenACC] ICE in
                   |emit_library_call_value_1,  |emit_library_call_value_1,
                   |at calls.c:5298             |at calls.c:5298

--- Comment #3 from Tobias Burnus <burnus at gcc dot gnu.org> ---
(In reply to Tobias Burnus from comment #2)
> +++ b/gcc/omp-expand.c
> @@ -7561 +7561,2 @@ expand_oacc_for (struct omp_region *region, struct

I think my draft patch is not quite right:

  expand_oacc_for (struct omp_region *region, struct omp_for_data *fd)

has:
  tree iter_type = TREE_TYPE (v);
  tree diff_type = iter_type;
and
  if (fd->collapse > 1 || fd->tiling)
      tree total = expand_oacc_collapse_init (fd, &gsi, counts,
                                              TREE_TYPE (fd->loop.n2), loc);
The called  expand_oacc_collapse_init  also extracts the types.
as does   expand_oacc_collapse_vars  .

At least for some of those loop *_type, the largest  TYPE_PRECISION has to be
used.

[Bug c/97880] [8/9/10/11 Regression] [OpenACC] ICE in emit_library_call_value_1, at calls.c:5298

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97880

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

https://gcc.gnu.org/g:f324479caf0ac326534f4fcf72cb12991ccddb3d

commit r11-5489-gf324479caf0ac326534f4fcf72cb12991ccddb3d
Author: Tobias Burnus <tobias@codesourcery.com>
Date:   Fri Nov 27 11:17:50 2020 +0100

    OpenACC: Fix integer-type issue with collapse/tile [PR97880]

    gcc/ChangeLog:

            PR c/97880
            * omp-expand.c (expand_oacc_collapse_init,
expand_oacc_collapse_vars):
            Use now passed diff_type.
            (expand_oacc_for): Take largest type for diff_type, taking tiling
            and collapsing into account.

    gcc/testsuite/ChangeLog:

            PR c/97880
            * gcc.dg/goacc/tile-1.c: New test.

[Bug c/97880] [8/9/10 Regression] [OpenACC] ICE in emit_library_call_value_1, at calls.c:5298

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97880

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to work|                            |11.0
            Summary|[8/9/10/11 Regression]      |[8/9/10 Regression]
                   |[OpenACC] ICE in            |[OpenACC] ICE in
                   |emit_library_call_value_1,  |emit_library_call_value_1,
                   |at calls.c:5298             |at calls.c:5298
           Priority|P3                          |P2

--- Comment #5 from Richard Biener <rguenth at gcc dot gnu.org> ---
Fixed on trunk I assume.

[Bug c/97880] [8/9/10 Regression] [OpenACC] ICE in emit_library_call_value_1, at calls.c:5298

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97880

--- Comment #6 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-10 branch has been updated by Tobias Burnus
<burnus@gcc.gnu.org>:

https://gcc.gnu.org/g:af408874e3d94492ac08ba17462b3ff8ecb4d791

commit r10-9716-gaf408874e3d94492ac08ba17462b3ff8ecb4d791
Author: Tobias Burnus <tobias@codesourcery.com>
Date:   Fri Nov 27 11:17:50 2020 +0100

    OpenACC: Fix integer-type issue with collapse/tile [PR97880]

    gcc/ChangeLog:

            PR c/97880
            * omp-expand.c (expand_oacc_collapse_init,
expand_oacc_collapse_vars):
            Use now passed diff_type.
            (expand_oacc_for): Take largest type for diff_type, taking tiling
            and collapsing into account.

    gcc/testsuite/ChangeLog:

            PR c/97880
            * gcc.dg/goacc/tile-1.c: New test.

    (cherry picked from commit f324479caf0ac326534f4fcf72cb12991ccddb3d)

[Bug c/97880] [8/9/10 Regression] [OpenACC] ICE in emit_library_call_value_1, at calls.c:5298

[burnus at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=97880

Tobias Burnus <burnus at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|NEW                         |RESOLVED

--- Comment #7 from Tobias Burnus <burnus at gcc dot gnu.org> ---
FIXED on GCC 11 (mainline) and GCC 10. (Cf. related also PR98088, fixed for the
same branches.)

I do not intent to backport it to GCC 8 or 9. Hence, close as FIXED

Thanks for this and the other reports!