[Bug debug/100515] New: [12 Regression] ICE in output_rnglists, at dwarf2out.c:12294

[Bug debug/100515] New: [12 Regression] ICE in output_rnglists, at dwarf2out.c:12294

[asolokha at gmx dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100515

            Bug ID: 100515
           Summary: [12 Regression] ICE in output_rnglists, at
                    dwarf2out.c:12294
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code, openmp
          Severity: normal
          Priority: P3
         Component: debug
          Assignee: unassigned at gcc dot gnu.org
          Reporter: asolokha at gmx dot com
  Target Milestone: ---

gcc-12.0.0-alpha20210509 snapshot (g:5e0236d3b0e0d7ad98bcee36128433fa755b5558)
ICEs when compiling the following testcase, reduced from
clang/testsuite/AST/ast-dump-openmp-taskloop-simd.c, w/ -O2 -fopenmp -g:

void
foo (int x)
{
#pragma omp taskloop
  for (int i = 0; i < x; i++)
    ;
}

void
bar (int x)
{
#pragma omp taskloop
  for (int i = 0; i < x; i++)
    ;
}

% gcc-12.0.0 -O2 -fopenmp -g -c wuhgfgai.c
wuhgfgai.c:15:1: internal compiler error: in output_rnglists, at
dwarf2out.c:12294
   15 | }
      | ^
0x64302c output_rnglists
       
/var/tmp/portage/sys-devel/gcc-12.0.0_alpha20210509/work/gcc-12-20210509/gcc/dwarf2out.c:12294
0xa4d492 dwarf2out_finish
       
/var/tmp/portage/sys-devel/gcc-12.0.0_alpha20210509/work/gcc-12-20210509/gcc/dwarf2out.c:32176

[Bug debug/100515] [12 Regression] ICE in output_rnglists, at dwarf2out.c:12294

[asolokha at gmx dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100515

--- Comment #1 from Arseny Solokha <asolokha at gmx dot com> ---
(In reply to Arseny Solokha from comment #0)
> reduced from
> clang/testsuite/AST/ast-dump-openmp-taskloop-simd.c

Reduced from test/AST/ast-dump-openmp-taskloop-simd.c from the clang 11.1.0
test suite.

[Bug debug/100515] [12 Regression] ICE in output_rnglists, at dwarf2out.c:12294 since r12-574-ge69ac0203725fb8d

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100515

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Last reconfirmed|                            |2021-05-11
            Summary|[12 Regression] ICE in      |[12 Regression] ICE in
                   |output_rnglists, at         |output_rnglists, at
                   |dwarf2out.c:12294           |dwarf2out.c:12294 since
                   |                            |r12-574-ge69ac0203725fb8d
                 CC|                            |edlinger at gcc dot gnu.org,
                   |                            |marxin at gcc dot gnu.org
      Known to work|                            |11.1.0
     Ever confirmed|0                           |1
      Known to fail|                            |12.0
             Status|UNCONFIRMED                 |NEW
   Target Milestone|---                         |12.0

--- Comment #2 from Martin Liška <marxin at gcc dot gnu.org> ---
Started with r12-574-ge69ac0203725fb8d.

[Bug debug/100515] [12 Regression] ICE in output_rnglists, at dwarf2out.c:12294 since r12-574-ge69ac0203725fb8d

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100515

--- Comment #3 from Martin Liška <marxin at gcc dot gnu.org> ---
I noticed the same problem with Ada bootstrap:

[  128s] +===========================GNAT BUG
DETECTED==============================+
[  128s] | 12.0.0 20210511 (experimental) [revision
6ab1176667734bd6de20833f8d263c03a418c452] (x86_64-suse-linux) GCC error:|
[  128s] | in output_rnglists, at dwarf2out.c:12294                            
    |
[  128s] | Error detected around
/usr/lib64/gcc/x86_64-suse-linux/12/adainclude/s-atacco.ads:39:16|
[  128s] | Please submit a bug report; see https://gcc.gnu.org/bugs/ .         
    |
[  128s] | Use a subject line meaningful to you and us to track the bug.       
    |
[  128s] | Include the entire contents of this bug box in the report.          
    |
[  128s] | Include the exact command that you entered.                         
    |
[  128s] | Also include sources listed below.                                  
    |

[Bug debug/100515] [12 Regression] ICE in output_rnglists, at dwarf2out.c:12294 since r12-574-ge69ac0203725fb8d

[edlinger at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100515

--- Comment #4 from Bernd Edlinger <edlinger at gcc dot gnu.org> ---
Created attachment 50795
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=50795&action=edit
Proposed patch

This is what I'm currently testing.

[Bug debug/100515] [12 Regression] ICE in output_rnglists, at dwarf2out.c:12294 since r12-574-ge69ac0203725fb8d

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100515

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

https://gcc.gnu.org/g:1ecd1e6c894fbdbc10fdcfee419922b24e1115ee

commit r12-740-g1ecd1e6c894fbdbc10fdcfee419922b24e1115ee
Author: Bernd Edlinger <bernd.edlinger@hotmail.de>
Date:   Tue May 11 17:55:18 2021 +0200

    Fix ICE in output_rnglists, at dwarf2out.c:12294

    In this testcase the compile unit consists of a single
    text section with a single embedded DECL_IGNORED_P function.
    So we have a kind of multi-range text section here.
    To avoid an ICE in output_rnglists we need to make sure
    that have_multiple_function_sections is set to true.
    This is a regression from
    e69ac020372 ("Add line debug info for virtual thunks")

    2021-05-12  Bernd Edlinger  <bernd.edlinger@hotmail.de>

            PR debug/100515
            * dwarf2out.c (dwarf2out_finish): Set
            have_multiple_function_sections with multi-range text_section.

            * gcc.dg/debug/dwarf2/pr100515.c: New testcase.

[Bug debug/100515] [12 Regression] ICE in output_rnglists, at dwarf2out.c:12294 since r12-574-ge69ac0203725fb8d

[edlinger at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100515

Bernd Edlinger <edlinger at gcc dot gnu.org> changed:

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

--- Comment #6 from Bernd Edlinger <edlinger at gcc dot gnu.org> ---
The ICE should be fixed now.

[Bug debug/100515] [12 Regression] ICE in output_rnglists, at dwarf2out.c:12294 since r12-574-ge69ac0203725fb8d

[edlinger at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100515

--- Comment #7 from Bernd Edlinger <edlinger at gcc dot gnu.org> ---
Okay, while the ICE is fixed, there is something odd in the test case:

foo._omp_fn.0:
.LVL0:
.LFB2:
        .cfi_startproc
        .file 1 "pr100515.c"
        .loc 1 10 5 view -0
        ret
        .cfi_endproc
.LFE2:
        .size   foo._omp_fn.0, .-foo._omp_fn.0
        .p2align 4
        .type   bar._omp_fn.0, @function
bar._omp_fn.0:
.LFB5:
        .cfi_startproc
        ret
        .cfi_endproc
.LFE5:
        .size   bar._omp_fn.0, .-bar._omp_fn.0
        .p2align 4
        .globl  foo
        .type   foo, @function

foo._omp_fn.0 does have a correct .loc statement,
while bar._omp_fn.0 does not have a .loc statement
and therefore no debugging in bar is possible.
This must be a pre-existing issue already there before
r12-574-ge69ac0203725fb8d.

[Bug debug/100515] [12 Regression] ICE in output_rnglists, at dwarf2out.c:12294 since r12-574-ge69ac0203725fb8d

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100515

--- Comment #8 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Christophe Lyon <clyon@gcc.gnu.org>:

https://gcc.gnu.org/g:b050cf6a4d9c305daff4a96e5a2489ece69dc287

commit r12-823-gb050cf6a4d9c305daff4a96e5a2489ece69dc287
Author: Christophe Lyon <christophe.lyon@linaro.org>
Date:   Mon May 17 09:25:43 2021 +0000

    testsuite: Require openmp effective-target for PR100515

    The related test uses -fopenmp, which is not supported by newlib-based
    targets such as arm-eabi or aarch64-elf.

    Requiring the openmp effective-target makes the test unsupported
    rather than failed.

    2021-05-17  Christophe Lyon  <christophe.lyon@linaro.org>

            PR debug/100515
            gcc/testsuite
            * gcc.dg/debug/dwarf2/pr100515.c: Require openmp effective-target.