Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Tom de Vries <tdevries@suse.de>
To: gdb-patches@sourceware.org
Subject: Re: [PING][PATCH] [gdb/exp] Fix ns var lookup when stopped at inlined fn call
Date: Mon, 13 Jul 2026 19:10:08 +0200	[thread overview]
Message-ID: <1e033cc7-06b6-4530-9b6b-1ba0a1df8d67@suse.de> (raw)
In-Reply-To: <155fd32b-ce9f-4c46-958b-b9ff89e749bf@suse.de>

On 6/18/26 9:54 AM, Tom de Vries wrote:
> On 6/4/26 1:19 PM, Tom de Vries wrote:
>> Consider test.c:
>> ...
>>       1    namespace mod_a {
>>       2      int xxx = 10;
>>       3    }
>>       4
>>       5    static inline int __attribute__((always_inline))
>>       6    inlined () {
>>       7      return 0;
>>       8    }
>>       9
>>      10    int main () {
>>      11      using namespace mod_a;
>>      12      int res = inlined ();
>>      13      return res + xxx;
>>      14    }
>> ...
>> compiled with "g++ test.c -g".
>>
>> Trying to print variable xxx at line 12 fails:
>> ...
>> $ gdb -q -batch a.out -ex start -ex "p xxx"
>>    ...
>> Temporary breakpoint 1, main () at test.c:12
>> 12      int res = inlined ();
>> No symbol "xxx" in current context.
>> ...
>>
>> The problem is here in function using_direct::valid_line:
>> ...
>>        CORE_ADDR curr_pc = get_frame_pc (get_selected_frame (nullptr));
>>        symtab_and_line curr_sal = find_sal_for_pc (curr_pc, 0);
>>        return (decl_line <= curr_sal.line)
>>          || (decl_line >= boundary);
>> ...
>> where we're trying to decide whether "using namespace mod_a" is 
>> applicable.
>>
>> The decl_line is 11, as expected.
>>
>> If curr_sal.line were 12, decl_line <= curr_sal.line would be true, and
>> using_direct::valid_line would return true.
>>
>> But instead, curr_sal.line is 7.
>>
>> This is sort of correct, the current PC maps to that line.  It's just 
>> that gdb
>> steps into inlined functions in two steps, each with identical PC:
>> - once stopping at the call site (line 12 in this case)
>> - once stopping at the PC line (line 7 in this case)
>>
>> The function using_direct::valid_line doesn't apply this logic, and
>> consequently line 7 is used for both cases.
>>
>> Fix this by using find_frame_sal instead.
>>
> 
> Ping.
> 

I've pushed this, with a trivial update to the test-case to add an xfail 
for GCC PR debug/108716.

Thanks,
- Tom

> Thanks,
> - Tom
> 
>> Tested on x86_64-linux.
>>
>> Bug: https://sourceware.org/bugzilla/show_bug.cgi?id=34201
>> ---
>>   gdb/namespace.c                     |  4 +--
>>   gdb/testsuite/gdb.cp/ns-inlined.c   | 41 +++++++++++++++++++++++
>>   gdb/testsuite/gdb.cp/ns-inlined.exp | 52 +++++++++++++++++++++++++++++
>>   3 files changed, 95 insertions(+), 2 deletions(-)
>>   create mode 100644 gdb/testsuite/gdb.cp/ns-inlined.c
>>   create mode 100644 gdb/testsuite/gdb.cp/ns-inlined.exp
>>
>> diff --git a/gdb/namespace.c b/gdb/namespace.c
>> index 5b5749369ee..fec5e5cf67d 100644
>> --- a/gdb/namespace.c
>> +++ b/gdb/namespace.c
>> @@ -111,8 +111,8 @@ using_direct::valid_line (unsigned int boundary) 
>> const
>>   {
>>     try
>>       {
>> -      CORE_ADDR curr_pc = get_frame_pc (get_selected_frame (nullptr));
>> -      symtab_and_line curr_sal = find_sal_for_pc (curr_pc, 0);
>> +      frame_info_ptr frame = get_selected_frame ();
>> +      symtab_and_line curr_sal = find_frame_sal (frame);
>>         return (decl_line <= curr_sal.line)
>>            || (decl_line >= boundary);
>>       }
>> diff --git a/gdb/testsuite/gdb.cp/ns-inlined.c b/gdb/testsuite/gdb.cp/ 
>> ns-inlined.c
>> new file mode 100644
>> index 00000000000..26e72f6ba3f
>> --- /dev/null
>> +++ b/gdb/testsuite/gdb.cp/ns-inlined.c
>> @@ -0,0 +1,41 @@
>> +/* This testcase is part of GDB, the GNU debugger.
>> +
>> +   Copyright 2026 Free Software Foundation, Inc.
>> +
>> +   This program is free software; you can redistribute it and/or modify
>> +   it under the terms of the GNU General Public License as published by
>> +   the Free Software Foundation; either version 3 of the License, or
>> +   (at your option) any later version.
>> +
>> +   This program is distributed in the hope that it will be useful,
>> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
>> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>> +   GNU General Public License for more details.
>> +
>> +   You should have received a copy of the GNU General Public License
>> +   along with this program.  If not, see <http://www.gnu.org/ 
>> licenses/>.  */
>> +
>> +namespace mod_a
>> +{
>> +  int xxx = 10;
>> +}
>> +
>> +static inline int __attribute__((always_inline))
>> +inlined ()
>> +{
>> +  return 0;
>> +}
>> +
>> +static void
>> +foo ()
>> +{
>> +}
>> +
>> +int
>> +main ()
>> +{
>> +  foo ();                   /* stop 1.  */
>> +  using namespace mod_a;
>> +  int res = inlined ();     /* stop 2.  */
>> +  return res + xxx;         /* stop 3.  */
>> +}
>> diff --git a/gdb/testsuite/gdb.cp/ns-inlined.exp b/gdb/testsuite/ 
>> gdb.cp/ns-inlined.exp
>> new file mode 100644
>> index 00000000000..90cfbc0ff23
>> --- /dev/null
>> +++ b/gdb/testsuite/gdb.cp/ns-inlined.exp
>> @@ -0,0 +1,52 @@
>> +# Copyright 2026 Free Software Foundation, Inc.
>> +
>> +# This program is free software; you can redistribute it and/or modify
>> +# it under the terms of the GNU General Public License as published by
>> +# the Free Software Foundation; either version 3 of the License, or
>> +# (at your option) any later version.
>> +#
>> +# This program is distributed in the hope that it will be useful,
>> +# but WITHOUT ANY WARRANTY; without even the implied warranty of
>> +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
>> +# GNU General Public License for more details.
>> +#
>> +# You should have received a copy of the GNU General Public License
>> +# along with this program.  If not, see <http://www.gnu.org/licenses/>.
>> +
>> +standard_testfile .c
>> +
>> +set stop2_line [gdb_get_line_number "stop 2"]
>> +set stop3_line [gdb_get_line_number "stop 3"]
>> +
>> +if {[prepare_for_testing "failed to prepare" $testfile $srcfile \
>> +     {debug c++}]} {
>> +    return
>> +}
>> +
>> +if {![runto_main]} {
>> +    return
>> +}
>> +
>> +with_test_prefix "stop 1" {
>> +    gdb_test "print xxx" \
>> +    [string_to_regexp {No symbol "xxx" in current context.}]
>> +}
>> +
>> +gdb_test "next" \
>> +    "\r\n$stop2_line\t\[^\r\n\]+" \
>> +    "next to stop 2"
>> +
>> +with_test_prefix "stop 2" {
>> +    # Regression test for PR exp/34201.
>> +    gdb_test "print xxx" \
>> +    "$valnum_re = 10"
>> +}
>> +
>> +gdb_test "next" \
>> +    "\r\n$stop3_line\t\[^\r\n\]+" \
>> +    "next to stop 3"
>> +
>> +with_test_prefix "stop 3" {
>> +    gdb_test "print xxx" \
>> +    "$valnum_re = 10"
>> +}
>>
>> base-commit: 48997323b0f929d3255a4845f8a3800c6c00a53a
> 


      reply	other threads:[~2026-07-13 17:10 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-06-04 11:19 [PATCH] " Tom de Vries
2026-06-18  7:54 ` [PING][PATCH] " Tom de Vries
2026-07-13 17:10   ` Tom de Vries [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1e033cc7-06b6-4530-9b6b-1ba0a1df8d67@suse.de \
    --to=tdevries@suse.de \
    --cc=gdb-patches@sourceware.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox