[Bug c/103587] New: [10/11/12 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850

[Bug c/103587] New: [10/11/12 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850

[gscfq@t-online.de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

            Bug ID: 103587
           Summary: [10/11/12 Regression] ICE in c_parser_consume_token,
                    at c/c-parser.c:850
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
          Assignee: unassigned at gcc dot gnu.org
          Reporter: gscfq@t-online.de
  Target Milestone: ---

Started between 20191110 and 20191117 :


$ cat z1.c
[[abc(
#pragma omp simd


$ gcc-10-20191110 -c z1.c -fopenmp
z1.c:1:1: error: expected identifier or '(' before '[' token
    1 | [[abc(
      | ^


$ gcc-12-20211205 -c z1.c -fopenmp
z1.c:2:9: internal compiler error: in c_parser_consume_token, at
c/c-parser.c:850
    2 | #pragma omp simd
      |         ^~~
0x78afa7 c_parser_consume_token(c_parser*)
        ../../gcc/c/c-parser.c:850
0x78c55f c_parser_balanced_token_sequence
        ../../gcc/c/c-parser.c:4850
0x79e41f c_parser_std_attribute
        ../../gcc/c/c-parser.c:4956
0x79e41f c_parser_std_attribute_specifier
        ../../gcc/c/c-parser.c:4998
0x79e529 c_parser_std_attribute_specifier_sequence
        ../../gcc/c/c-parser.c:5114
0x7955ce c_parser_declspecs(c_parser*, c_declspecs*, bool, bool, bool, bool,
bool, bool, bool, c_lookahead_kind)
        ../../gcc/c/c-parser.c:2826
0x7b0669 c_parser_declaration_or_fndef
        ../../gcc/c/c-parser.c:1997
0x7b930f c_parser_external_declaration
        ../../gcc/c/c-parser.c:1779
0x7b9c3b c_parser_translation_unit
        ../../gcc/c/c-parser.c:1652
0x7b9c3b c_parse_file()
        ../../gcc/c/c-parser.c:23279
0x80b0e2 c_common_parse_file()
        ../../gcc/c-family/c-opts.c:1238

[Bug c/103587] [10/11/12 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |10.0
   Last reconfirmed|                            |2021-12-06
             Status|UNCONFIRMED                 |NEW
     Ever confirmed|0                           |1

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Confirmed.

[Bug c/103587] [10/11/12 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850 since r10-4633-g4e03c3a7c1149a8e

[marxin at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[10/11/12 Regression] ICE   |[10/11/12 Regression] ICE
                   |in c_parser_consume_token,  |in c_parser_consume_token,
                   |at c/c-parser.c:850         |at c/c-parser.c:850 since
                   |                            |r10-4633-g4e03c3a7c1149a8e
                 CC|                            |jsm28 at gcc dot gnu.org,
                   |                            |marxin at gcc dot gnu.org

--- Comment #2 from Martin Liška <marxin at gcc dot gnu.org> ---
Started with r10-4633-g4e03c3a7c1149a8e.

[Bug c/103587] [10/11/12 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850 since r10-4633-g4e03c3a7c1149a8e

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Priority|P3                          |P2

[Bug c/103587] [10/11/12 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850 since r10-4633-g4e03c3a7c1149a8e

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|10.0                        |10.4

[Bug c/103587] [10/11/12 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850 since r10-4633-g4e03c3a7c1149a8e

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

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

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

--- Comment #3 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Not really specific to OpenMP, e.g.
[[foo::bar(
#pragma GCC ivdep
)]];
ICEs too.
cp_parser_skip_balanced_tokens ignores CPP_PRAGMA_EOL tokens if the function
wasn't called from inside of pragma parsing (otherwise treats it like CPP_EOF
and
determines it wasn't balanced).
Though, maybe it should also look at CPP_PRAGMA and ensure the tokens are
balanced with it too,
(
#pragma GCC ivdep )
isn't really balanced but C++ FE doesn't really take the balancing as pedantic
as C FE does - for our C++ parser ( [ { ) } ] is balanced, for C it isn't.

[Bug c/103587] [10/11/12 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850 since r10-4633-g4e03c3a7c1149a8e

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Assignee|unassigned at gcc dot gnu.org      |jakub at gcc dot gnu.org
             Status|NEW                         |ASSIGNED

--- Comment #4 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Created attachment 51969
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=51969&action=edit
gcc12-pr103587.patch

We don't really know what the pragma can mean in that particular context, so
I think it is best to skip over it like the C++ FE does.
This patch doesn't care if inside of the pragma the tokens are balanced or not.

[Bug c/103587] [10/11/12 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850 since r10-4633-g4e03c3a7c1149a8e

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

--- Comment #5 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:e163dbbc4433e598cad7e6011b255d1d6ad93a3b

commit r12-5947-ge163dbbc4433e598cad7e6011b255d1d6ad93a3b
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Tue Dec 14 12:02:55 2021 +0100

    c: Fix ICE on deferred pragma in unknown attribute arguments [PR103587]

    We ICE on the following testcase, because c_parser_balanced_token_sequence
    when encountering a deferred pragma will just use c_parser_consume_token
    which the FE doesn't allow for CPP_PRAGMA tokens (and if that wasn't
    the case, it could ICE on CPP_PRAGMA_EOL similarly).
    We don't know in what exact context the pragma appears when we don't
    know what those arguments semantically mean, so I think we should just
    skip over them, like e.g. the C++ FE does.  And, I think (/[/{ vs. )/]/}
    from outside of the pragma shouldn't be paired with those inside of
    the pragma and it doesn't seem to be necessary to check that inside of
    the pragma line itself all the paren kinds are balanced.

    2021-12-14  Jakub Jelinek  <jakub@redhat.com>

            PR c/103587
            * c-parser.c (c_parser_balanced_token_sequence): For CPP_PRAGMA,
            consume the pragma and silently skip to the pragma eol.

            * gcc.dg/pr103587.c: New test.

[Bug c/103587] [10/11 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850 since r10-4633-g4e03c3a7c1149a8e

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[10/11/12 Regression] ICE   |[10/11 Regression] ICE in
                   |in c_parser_consume_token,  |c_parser_consume_token, at
                   |at c/c-parser.c:850 since   |c/c-parser.c:850 since
                   |r10-4633-g4e03c3a7c1149a8e  |r10-4633-g4e03c3a7c1149a8e

--- Comment #6 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Fixed on the trunk so far.

[Bug c/103587] [10/11 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850 since r10-4633-g4e03c3a7c1149a8e

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

--- Comment #7 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-11 branch has been updated by Jakub Jelinek
<jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:c239267759be2cf3e1afadea86aff5ba3517d934

commit r11-9490-gc239267759be2cf3e1afadea86aff5ba3517d934
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Tue Dec 14 12:02:55 2021 +0100

    c: Fix ICE on deferred pragma in unknown attribute arguments [PR103587]

    We ICE on the following testcase, because c_parser_balanced_token_sequence
    when encountering a deferred pragma will just use c_parser_consume_token
    which the FE doesn't allow for CPP_PRAGMA tokens (and if that wasn't
    the case, it could ICE on CPP_PRAGMA_EOL similarly).
    We don't know in what exact context the pragma appears when we don't
    know what those arguments semantically mean, so I think we should just
    skip over them, like e.g. the C++ FE does.  And, I think (/[/{ vs. )/]/}
    from outside of the pragma shouldn't be paired with those inside of
    the pragma and it doesn't seem to be necessary to check that inside of
    the pragma line itself all the paren kinds are balanced.

    2021-12-14  Jakub Jelinek  <jakub@redhat.com>

            PR c/103587
            * c-parser.c (c_parser_balanced_token_sequence): For CPP_PRAGMA,
            consume the pragma and silently skip to the pragma eol.

            * gcc.dg/pr103587.c: New test.

    (cherry picked from commit e163dbbc4433e598cad7e6011b255d1d6ad93a3b)

[Bug c/103587] [10 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850 since r10-4633-g4e03c3a7c1149a8e

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[10/11 Regression] ICE in   |[10 Regression] ICE in
                   |c_parser_consume_token, at  |c_parser_consume_token, at
                   |c/c-parser.c:850 since      |c/c-parser.c:850 since
                   |r10-4633-g4e03c3a7c1149a8e  |r10-4633-g4e03c3a7c1149a8e

--- Comment #8 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Fixed for 11.3 too.

[Bug c/103587] [10 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850 since r10-4633-g4e03c3a7c1149a8e

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

--- Comment #9 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The releases/gcc-10 branch has been updated by Jakub Jelinek
<jakub@gcc.gnu.org>:

https://gcc.gnu.org/g:58e8e03d18ad950e8ca84054ddf23f70f8b0f055

commit r10-10660-g58e8e03d18ad950e8ca84054ddf23f70f8b0f055
Author: Jakub Jelinek <jakub@redhat.com>
Date:   Tue Dec 14 12:02:55 2021 +0100

    c: Fix ICE on deferred pragma in unknown attribute arguments [PR103587]

    We ICE on the following testcase, because c_parser_balanced_token_sequence
    when encountering a deferred pragma will just use c_parser_consume_token
    which the FE doesn't allow for CPP_PRAGMA tokens (and if that wasn't
    the case, it could ICE on CPP_PRAGMA_EOL similarly).
    We don't know in what exact context the pragma appears when we don't
    know what those arguments semantically mean, so I think we should just
    skip over them, like e.g. the C++ FE does.  And, I think (/[/{ vs. )/]/}
    from outside of the pragma shouldn't be paired with those inside of
    the pragma and it doesn't seem to be necessary to check that inside of
    the pragma line itself all the paren kinds are balanced.

    2021-12-14  Jakub Jelinek  <jakub@redhat.com>

            PR c/103587
            * c-parser.c (c_parser_balanced_token_sequence): For CPP_PRAGMA,
            consume the pragma and silently skip to the pragma eol.

            * gcc.dg/pr103587.c: New test.

    (cherry picked from commit e163dbbc4433e598cad7e6011b255d1d6ad93a3b)

[Bug c/103587] [10 Regression] ICE in c_parser_consume_token, at c/c-parser.c:850 since r10-4633-g4e03c3a7c1149a8e

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103587

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

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

--- Comment #10 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Fixed for 10.4 too.