[Bug exp/34203] [gdb/exp] Namespace variable found before using statement at end of block

https://sourceware.org/bugzilla/show_bug.cgi?id=34203

--- Comment #3 from Sourceware Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Tom de Vries <[email protected]>:

https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;h=c20c23f2a738cfc641ae2ffd9574f6288c248b81

commit c20c23f2a738cfc641ae2ffd9574f6288c248b81
Author: Tom de Vries <[email protected]>
Date:   Mon Jul 13 18:13:24 2026 +0200

    [gdb/exp] Limit workaround in using_direct::valid_line to broken GCC
versions

    Consider test.c:
    ...
         1  namespace mod_a {
         2    int xxx = 10;
         3  }
         4
         5  static void
         6  foo ()
         7  {
         8  }
         9
        10  int
        11  main ()
        12  {
        13    {
        14      foo ();
        15      using namespace mod_a;
        16    }
        17
        18    return mod_a::xxx;
        19  }
    ...
    compiled with "g++ test.c -g".

    Attempting to print xxx at line 14 shouldn't find anything (because it's
    before the "using namespace mod_a"), but it does:
    ...
    $ gdb -q -batch a.out -ex start -ex "p xxx"
    ...
    Temporary breakpoint 1, main () at test.c:14
    14          foo ();
    $1 = 10
    ...

    This happens because using_direct::valid_line returns true here:
    ...
          return (decl_line <= curr_sal.line)
                 || (decl_line >= boundary);
    ...

    Since we have decl_line == 15 and curr_sal.line == 14,
    "(decl_line <= curr_sal.line)" evaluates to false.

    But boundary == 14, so "(decl_line >= boundary)" evaluates to true.

    The "(decl_line >= boundary)" bit was added as a workaround for
    GCC PR debug/108716.

    Since I'm using GCC 15, the workaround is not needed.

    Fix this by limiting the workaround to broken GCC versions.

    Tested on x86_64-linux, using GCC 15.2, and 7.5.

    Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=34203

-- 
You are receiving this mail because:
You are on the CC list for the bug.