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
next prev 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