Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: Guinevere Larsen <guinevere@redhat.com>
To: gdb-patches@sourceware.org, Guinevere Larsen <guinevere@redhat.com>
Subject: [PING][PATCH v7] gdb: Print linker namespace when showing a frame
Date: Thu, 5 Mar 2026 09:18:11 -0300	[thread overview]
Message-ID: <b6ca0af5-61da-42a6-8465-d972e50a1d48@redhat.com> (raw)
In-Reply-To: <20260113174341.3532973-1-guinevere@redhat.com>

Ping :)

On 1/13/26 2:43 PM, Guinevere Larsen wrote:
> When a user is stopped in a private linker namespace, the only way for
> them to realize that is using the _linker_namespace convenience
> variable. While serviceable, this is a sub-optimal solution, as most
> users are unaware of convenience variables.
>
> This commit introduces a new way for users to be informed of the linker
> namespace of a function, by printing it along with the function name.
> This is done by using the proposed syntax for symbols and locations,
> like so:
>
>    #0  [[0]]::main ()
>
> This is done by introducing a new method to general_symbol_info, which
> uses part of the functionality behind the _linker_namespace variable
> , namely, find the linker namespace that contains the given address on
> the given program space.  The reason print_name was not changed to do
> this is that the old method returns a "const char *", but the new one
> needs to return a string, and the refactor was very complicated, so the
> new method is introduced.
>
> The namespace ID is only printed if multiple namespaces are active,
> otherwise no change in behavior is expected. This commit also updates
> the test gdb.base/dlmopen-ns-ids.exp to test this functionality.
> ---
>   gdb/NEWS                                      |  4 +++
>   gdb/doc/gdb.texinfo                           |  9 ++++++
>   gdb/solib.c                                   | 29 ++++++++++-------
>   gdb/solib.h                                   |  8 +++++
>   gdb/stack.c                                   |  9 ++++--
>   gdb/symtab.c                                  | 15 +++++++++
>   gdb/symtab.h                                  | 13 ++++++--
>   .../gdb.base/dlmopen-ns-ids-framefilter.py    | 31 +++++++++++++++++++
>   gdb/testsuite/gdb.base/dlmopen-ns-ids.exp     | 25 +++++++++++++--
>   gdb/testsuite/gdb.mi/mi-dlmopen.exp           |  2 +-
>   10 files changed, 125 insertions(+), 20 deletions(-)
>   create mode 100644 gdb/testsuite/gdb.base/dlmopen-ns-ids-framefilter.py
>
> diff --git a/gdb/NEWS b/gdb/NEWS
> index cd0303fa199..80dfc25f824 100644
> --- a/gdb/NEWS
> +++ b/gdb/NEWS
> @@ -54,6 +54,10 @@
>     on the index.  Any existing indexes should be regenerated.
>   * Support for Floating Point Mode Register (FPMR) in AArch64.
>   
> +* When debugging an inferior with multiple linker namespaces, functions
> +  will be printed like so [[N]]::foo, where N is the namespace that
> +  contains foo.  For example [[0]]::main.
> +
>   * New targets
>   
>   GNU/Linux/MicroBlaze (gdbserver) microblazeel-*linux*
> diff --git a/gdb/doc/gdb.texinfo b/gdb/doc/gdb.texinfo
> index 7059f73935c..d18710ca273 100644
> --- a/gdb/doc/gdb.texinfo
> +++ b/gdb/doc/gdb.texinfo
> @@ -22406,6 +22406,15 @@ possibility that those objects may cross-talk.  Each set of isolated
>   shared objects is said to belong to a ``namespace'', and linker related
>   actions such as relocations do not cross namespace boundaries.
>   
> +When multiple linker namespaces are active, @value{GDBN} will print the
> +identifier of the namespace that contains the function, when a function
> +is printed.  It will look like this:
> +
> +@smallexample
> +#0  [[1]]::inc (n=0) at dlmopen-ns-ids-lib.c
> +#1  0x0000000000400591 in [[0]]::main () at dlmopen-ns-ids-main.c
> +@end smallexample
> +
>   @kindex info dll
>   @item info dll @var{regex}
>   This is an alias of @code{info sharedlibrary}.
> diff --git a/gdb/solib.c b/gdb/solib.c
> index 492ff5580c9..a3e40940ddd 100644
> --- a/gdb/solib.c
> +++ b/gdb/solib.c
> @@ -1825,6 +1825,22 @@ solib_linker_namespace_count (program_space *pspace)
>     return 0;
>   }
>   
> +/* See solib.h.  */
> +
> +int
> +linker_namespace_for_addr (CORE_ADDR addr, program_space *pspace)
> +{
> +  for (const solib &so : pspace->solibs ())
> +    if (solib_contains_address_p (so, addr))
> +      {
> +	if (so.ops().supports_namespaces ())
> +	  return so.ops ().find_solib_ns (so);
> +	break;
> +      }
> +
> +  return 0;
> +}
> +
>   /* Implementation of the linker_namespace convenience variable.
>   
>      This returns the GDB internal identifier of the linker namespace,
> @@ -1833,19 +1849,10 @@ solib_linker_namespace_count (program_space *pspace)
>   
>   static value *
>   linker_namespace_make_value (gdbarch *gdbarch, internalvar *var,
> -				     void *ignore)
> +			     void *ignore)
>   {
> -  int nsid = 0;
>     CORE_ADDR curr_pc = get_frame_pc (get_selected_frame ());
> -
> -  for (const solib &so : current_program_space->solibs ())
> -    if (solib_contains_address_p (so, curr_pc))
> -      {
> -	if (so.ops ().supports_namespaces ())
> -	  nsid = so.ops ().find_solib_ns (so);
> -
> -	break;
> -      }
> +  int nsid = linker_namespace_for_addr (curr_pc, current_program_space);
>   
>     /* If the PC is not in an SO, or the solib_ops doesn't support
>        linker namespaces, the inferior is in the default namespace.  */
> diff --git a/gdb/solib.h b/gdb/solib.h
> index 4709669c54e..48cc709f2de 100644
> --- a/gdb/solib.h
> +++ b/gdb/solib.h
> @@ -335,6 +335,14 @@ extern const char *solib_name_from_address (struct program_space *, CORE_ADDR);
>   
>   extern bool solib_contains_address_p (const solib &, CORE_ADDR);
>   
> +/* Given the address ADDR, return which linker namespace contains
> +   this address in PSPACE.  If ADDR is present in multiple namespaces,
> +   an arbitrary one is returned.  If ADDR is not present in any
> +   namespace, or if the target doesn't support linker namespaces,
> +   returns 0.  */
> +
> +extern int linker_namespace_for_addr (CORE_ADDR addr, program_space *pspace);
> +
>   /* Return whether the data starting at VADDR, size SIZE, must be kept
>      in a core file for shared libraries loaded before "gcore" is used
>      to be handled correctly when the core file is loaded.  This only
> diff --git a/gdb/stack.c b/gdb/stack.c
> index 9d8e9da6aa8..bdfd41b7827 100644
> --- a/gdb/stack.c
> +++ b/gdb/stack.c
> @@ -1276,7 +1276,9 @@ find_frame_funname (const frame_info_ptr &frame, enum language *funlang,
>     func = get_frame_function (frame);
>     if (func)
>       {
> -      const char *print_name = func->print_name ();
> +      std::string print_name
> +	= func->print_name_with_namespace (get_frame_pc_if_available (frame),
> +					   get_frame_program_space (frame));
>   
>         *funlang = func->language ();
>         if (funcp)
> @@ -1288,13 +1290,13 @@ find_frame_funname (const frame_info_ptr &frame, enum language *funlang,
>   	     stored in the symbol table, but we stored a version
>   	     with DMGL_PARAMS turned on, and here we don't want to
>   	     display parameters.  So remove the parameters.  */
> -	  funname = cp_remove_params (print_name);
> +	  funname = cp_remove_params (print_name.c_str ());
>   	}
>   
>         /* If we didn't hit the C++ case above, set *funname
>   	 here.  */
>         if (funname == NULL)
> -	funname.reset (xstrdup (print_name));
> +	funname.reset (xstrdup (print_name.c_str ()));
>       }
>     else
>       {
> @@ -1362,6 +1364,7 @@ print_frame (struct ui_out *uiout,
>       annotate_frame_function_name ();
>   
>       string_file stb;
> +
>       gdb_puts (funname ? funname.get () : "??", &stb);
>       uiout->field_stream ("func", stb, function_name_style.style ());
>       uiout->wrap_hint (3);
> diff --git a/gdb/symtab.c b/gdb/symtab.c
> index 63505a94d41..bf4c3c7183f 100644
> --- a/gdb/symtab.c
> +++ b/gdb/symtab.c
> @@ -756,6 +756,21 @@ gdb_mangle_name (struct type *type, int method_id, int signature_id)
>   
>   /* See symtab.h.  */
>   
> +std::string
> +general_symbol_info::print_name_with_namespace (std::optional <CORE_ADDR> pc,
> +						program_space *pspace)
> +{
> +  std::string name = print_name ();
> +  if (pc.has_value ()
> +      && solib_linker_namespace_count (pspace) > 1)
> +    return string_printf
> +      ("[[%d]]::%s", linker_namespace_for_addr (*pc, pspace),
> +       name.c_str ());
> +  return name;
> +}
> +
> +/* See symtab.h.  */
> +
>   void
>   general_symbol_info::set_demangled_name (const char *name,
>   					 struct obstack *obstack)
> diff --git a/gdb/symtab.h b/gdb/symtab.h
> index b771851f03a..4b0e21c4cee 100644
> --- a/gdb/symtab.h
> +++ b/gdb/symtab.h
> @@ -448,9 +448,10 @@ struct general_symbol_info
>     /* Short version as to when to use which name accessor:
>        Use natural_name () to refer to the name of the symbol in the original
>        source code.  Use linkage_name () if you want to know what the linker
> -     thinks the symbol's name is.  Use print_name () for output.  Use
> -     demangled_name () if you specifically need to know whether natural_name ()
> -     and linkage_name () are different.  */
> +     thinks the symbol's name is.  Use print_name_with_namespace () for most
> +     output.  Use print_name () if you have a reason to always omit the
> +     namespace identifier.  Use demangled_name () if you specifically need
> +     to know whether natural_name () and linkage_name () are different.  */
>   
>     const char *linkage_name () const
>     { return m_name; }
> @@ -470,6 +471,12 @@ struct general_symbol_info
>     const char *print_name () const
>     { return demangle ? natural_name () : linkage_name (); }
>   
> +  /* Similar to print_name, but may prefix the name of the symbol with the
> +     linker namespace identifier, if the target supports it and the inferior
> +     has more than one identifier loaded.  */
> +  std::string print_name_with_namespace
> +    (std::optional <CORE_ADDR> pc, program_space *pspace);
> +
>     /* Return the demangled name for a symbol based on the language for
>        that symbol.  If no demangled name exists, return NULL.  */
>     const char *demangled_name () const;
> diff --git a/gdb/testsuite/gdb.base/dlmopen-ns-ids-framefilter.py b/gdb/testsuite/gdb.base/dlmopen-ns-ids-framefilter.py
> new file mode 100644
> index 00000000000..11d96fcfbc4
> --- /dev/null
> +++ b/gdb/testsuite/gdb.base/dlmopen-ns-ids-framefilter.py
> @@ -0,0 +1,31 @@
> +# Copyright (C) 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/>.
> +
> +
> +# This file is part of the GDB testsuite.  It tests if printing the
> +# linker namespace id is affected by a null frame filter
> +
> +import gdb
> +
> +
> +class NullFilter:
> +    def __init__(self):
> +        self.name = "null filter"
> +        self.priority = 100
> +        self.enabled = True
> +        gdb.frame_filters[self.name] = self
> +
> +    def filter(self, iterator):
> +        return iterator
> diff --git a/gdb/testsuite/gdb.base/dlmopen-ns-ids.exp b/gdb/testsuite/gdb.base/dlmopen-ns-ids.exp
> index e40bc213454..e9832664566 100644
> --- a/gdb/testsuite/gdb.base/dlmopen-ns-ids.exp
> +++ b/gdb/testsuite/gdb.base/dlmopen-ns-ids.exp
> @@ -21,7 +21,7 @@
>   
>   require allow_dlmopen_tests
>   
> -standard_testfile -main.c -lib.c
> +standard_testfile -main.c -lib.c -framefilter.py
>   
>   set srcfile_lib $srcfile2
>   set so_name dlmopen-lib.so
> @@ -192,6 +192,11 @@ proc_with_prefix test_conv_vars {} {
>       gdb_test "print \$_linker_namespace" ".* = 0" \
>   	"Still in the default namespace"
>   
> +    # There should be no namespace ID visible, since there's
> +    # only one namespace loaded.
> +    gdb_test "backtrace" "\#0\\s+main .*" \
> +	"No namespace ID in backtrace"
> +
>       gdb_breakpoint "inc" allow-pending
>       gdb_breakpoint [gdb_get_line_number "TAG: first dlclose"]
>   
> @@ -200,14 +205,30 @@ proc_with_prefix test_conv_vars {} {
>   
>   	gdb_test "print \$_linker_namespace" ".* = $dl" \
>   	    "Verify we're in namespace $dl"
> +
> +	gdb_test "frame" "\#0\\s+\\\[\\\[$dl\\\]\\\]::inc.*" \
> +	    "Namespace ID in the frame"
>       }
>   
>       # Check that we display the namespace of the selected
>       # frame, not the lowermost one.
> -    gdb_test "up" "\#1.*in main.*"
> +    gdb_test "up" "\#1.*in \\\[\\\[0\\\]\\\]::main.*"
>       gdb_test "print \$_linker_namespace" ".* = 0" \
>   	"print namespace of selected frame"
>   
> +    gdb_test "backtrace" [multi_line \
> +	    "\#0\\s+\\\[\\\[1\\\]\\\]::inc \[^\r\n\]*" \
> +	    "\#1\\s+$::hex in \\\[\\\[0\\\]\\\]::main \[^\r\n\]*" ] \
> +	"raw backtrace"
> +
> +    set remote_python_file [gdb_remote_download host \
> +				$::srcdir/$::subdir/$::srcfile3]
> +    gdb_test_no_output "source $remote_python_file" "load frame filter"
> +    gdb_test "backtrace" [multi_line \
> +	    "\#0\\s+\\\[\\\[1\\\]\\\]::inc \[^\r\n\]*" \
> +	    "\#1\\s+$::hex in \\\[\\\[0\\\]\\\]::main \[^\r\n\]*" ] \
> +	"With null frame filter"
> +
>       gdb_continue_to_breakpoint "first dlclose"
>       gdb_test "print \$_linker_namespace_count" "4" "all SOs loaded"
>   
> diff --git a/gdb/testsuite/gdb.mi/mi-dlmopen.exp b/gdb/testsuite/gdb.mi/mi-dlmopen.exp
> index ff854ac7dd4..832c6ea6fe6 100644
> --- a/gdb/testsuite/gdb.mi/mi-dlmopen.exp
> +++ b/gdb/testsuite/gdb.mi/mi-dlmopen.exp
> @@ -156,7 +156,7 @@ proc check_solib_unload_events {} {
>   	-disp keep -func main -file ".*$::srcfile" -line $::bp_main
>   
>       # Run past all the dlopen and dlmopen calls.
> -    mi_execute_to "exec-continue" "breakpoint-hit" main "" ".*" $::bp_loaded \
> +    mi_execute_to "exec-continue" "breakpoint-hit" {\[\[0\]\]::main} "" ".*" $::bp_loaded \
>   	{"" "disp=\"keep\""} "continue until all libraries are loaded"
>   
>       # Check that the dynamic linker has now been loaded multiple times.
>
> base-commit: 6b8fb74a9403092c4e4813c728e20ed10a84676f


-- 
Cheers,
Guinevere Larsen
It/she


  parent reply	other threads:[~2026-03-05 12:18 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-01-13 17:43 [PATCH " Guinevere Larsen
2026-01-13 20:08 ` Eli Zaretskii
2026-03-05 12:18 ` Guinevere Larsen [this message]
2026-04-20 16:28 ` Andrew Burgess
2026-05-06 20:07   ` Guinevere Larsen

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=b6ca0af5-61da-42a6-8465-d972e50a1d48@redhat.com \
    --to=guinevere@redhat.com \
    --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