From: Guinevere Larsen <guinevere@redhat.com>
To: "Gerlicher, Klaus" <klaus.gerlicher@intel.com>,
"gdb-patches@sourceware.org" <gdb-patches@sourceware.org>
Cc: Guinevere Larsen <blarsen@redhat.com>,
Jan Kratochvil <jan@jankratochvil.net>
Subject: Re: [PATCH v5 5/5] gdb/testsuite: Test for a backtrace through object without debuginfo
Date: Wed, 9 Oct 2024 11:56:33 -0300 [thread overview]
Message-ID: <f8666ccf-19f3-4938-b0b2-745387ba3661@redhat.com> (raw)
In-Reply-To: <SN7PR11MB70914F7AF260BD8EC601673FE8712@SN7PR11MB7091.namprd11.prod.outlook.com>
On 10/3/24 3:58 AM, Gerlicher, Klaus wrote:
> Hi Guienvere,
>
> Just one suggestion inline.
>
> Thanks
> Klaus
>
>> -----Original Message-----
>> From: Guinevere Larsen <guinevere@redhat.com>
>> Sent: Tuesday, October 1, 2024 8:43 PM
>> To: gdb-patches@sourceware.org
>> Cc: Guinevere Larsen <blarsen@redhat.com>; Jan Kratochvil
>> <jan@jankratochvil.net>
>> Subject: [PATCH v5 5/5] gdb/testsuite: Test for a backtrace through object
>> without debuginfo
>>
>> From: Guinevere Larsen <blarsen@redhat.com>
>>
>> Fedora has been carrying this test since back in the Project Archer
>> days. A change back then caused GDB to stop being able to backtrace when
>> only some of the object files had debug information. Even though the
>> changed code never seems to have made its way into the main GDB project,
>> I think it makes sense to bring the test along to ensure something like
>> this doesn't pass unnoticed.
>>
>> Co-Authored-By: Jan Kratochvil <jan@jankratochvil.net>
>> ---
>> .../backtrace-through-cu-nodebug-caller.c | 28 ++++++
>> .../backtrace-through-cu-nodebug-main.c | 32 +++++++
>> .../gdb.base/backtrace-through-cu-nodebug.exp | 90
>> +++++++++++++++++++
>> 3 files changed, 150 insertions(+)
>> create mode 100644 gdb/testsuite/gdb.base/backtrace-through-cu-
>> nodebug-caller.c
>> create mode 100644 gdb/testsuite/gdb.base/backtrace-through-cu-
>> nodebug-main.c
>> create mode 100644 gdb/testsuite/gdb.base/backtrace-through-cu-
>> nodebug.exp
>>
>> diff --git a/gdb/testsuite/gdb.base/backtrace-through-cu-nodebug-caller.c
>> b/gdb/testsuite/gdb.base/backtrace-through-cu-nodebug-caller.c
>> new file mode 100644
>> index 00000000000..3a63d72a468
>> --- /dev/null
>> +++ b/gdb/testsuite/gdb.base/backtrace-through-cu-nodebug-caller.c
>> @@ -0,0 +1,28 @@
>> +/* This testcase is part of GDB, the GNU debugger.
>> +
>> + Copyright 2005-2024 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/>. */
>> +
>> +typedef int (*callback_t) (void);
>> +
>> +int
>> +caller (callback_t callback)
>> +{
>> + /* Ensure some frame content to push away the return address. */
>> + volatile const long one = 1;
>> +
>> + /* Modify the return value to prevent any tail-call optimization. */
>> + return (*callback) () - one;
>> +}
>> diff --git a/gdb/testsuite/gdb.base/backtrace-through-cu-nodebug-main.c
>> b/gdb/testsuite/gdb.base/backtrace-through-cu-nodebug-main.c
>> new file mode 100644
>> index 00000000000..3e7ac57a166
>> --- /dev/null
>> +++ b/gdb/testsuite/gdb.base/backtrace-through-cu-nodebug-main.c
>> @@ -0,0 +1,32 @@
>> +/* This testcase is part of GDB, the GNU debugger.
>> +
>> + Copyright 2005-2024 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/>. */
>> +
>> +typedef int (*callback_t) (void);
>> +
>> +extern int caller (callback_t callback);
>> +
>> +int
>> +callback (void)
>> +{
>> + return 1;
>> +}
>> +
>> +int
>> +main (void)
>> +{
>> + return caller (callback);
>> +}
>> diff --git a/gdb/testsuite/gdb.base/backtrace-through-cu-nodebug.exp
>> b/gdb/testsuite/gdb.base/backtrace-through-cu-nodebug.exp
>> new file mode 100644
>> index 00000000000..7951ab41141
>> --- /dev/null
>> +++ b/gdb/testsuite/gdb.base/backtrace-through-cu-nodebug.exp
>> @@ -0,0 +1,90 @@
>> +# Copyright 2010-2024 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/>.
>> +
>> +# Test that GDB can generate accurate backtraces even if some of the stack
>> +# trace goes through a function with no debug information.
>> +
>> +standard_testfile -caller.c -main.c
>> +set objmainfile ${testfile}-main.o
>> +set objcallerfile ${testfile}-caller.o
>> +
>> +# recompile the inferior with or without CFI information, then run the
>> +# inferior until the point where the important test starts
>> +# returns TRUE on an ERROR.
>> +proc prepare_test {has_cfi} {
>> + global srcdir subdir srcfile srcfile2 objmainfile objcallerfile binfile
>> + if {$has_cfi} {
>> + set extension "cfi"
>> + if {[gdb_compile "${srcdir}/${subdir}/${srcfile}" ${objcallerfile} \
>> + object [list {additional_flags=-fomit-frame-pointer \
>> + -funwind-tables -fasynchronous-unwind-tables}]] != "" } {
>> + untested "couldn't compile with cfi"
>> + return true
>> + }
>> + } else {
>> + set extension "no-cfi"
>> + if {[gdb_compile "${srcdir}/${subdir}/${srcfile}" ${objcallerfile} \
>> + object [list {additional_flags=-fomit-frame-pointer \
>> + -fno-unwind-tables \
>> + -fno-asynchronous-unwind-tables}]] != "" } {
>> + untested "couldn't compile without cfi"
>> + return true
>> + }
>> + }
>> + if {[gdb_compile "${objmainfile} ${objcallerfile}" \
>> + "${binfile}-${extension}" binfile {}] != ""} {
>> + untested "couldn't link object files"
>> + return true
>> + }
>> +
>> + clean_restart "$binfile-${extension}"
>> +
>> + with_test_prefix "${extension}" {
>> +
>> + if ![runto callback] then {
>> + fail "has_cfi=$has_cfi: Can't run to callback"
>> + return true
>> + }
>> + gdb_test_no_output "maint frame-unwinder disable ARCH"
>> + return false
>> + }
>> +}
>> +
>> +if {[gdb_compile "${srcdir}/${subdir}/${srcfile2}" ${objmainfile} \
>> + object {debug}] != "" } {
>> + untested "couldn't compile main file"
>> + return
>> +}
>> +
>> +if { [prepare_test false] } {
>> + untested ${testfile}.exp
>> +} else {
>> + gdb_test "bt" "frame_unwind_find_by_frame failed.*" \
>> + "verify unwind fail without CFI"
>> +}
>> +
>> +if { [prepare_test true] } {
>> + untested ${testfile}.exp
>> +} else {
>> + if { [istarget "arm*-*-*"] } {
>> + setup_kfail backtrace/31950 *-*-*
>> + }
>> + # #0 callback () at ...
>> + # #1 0x00000000004004e9 in caller ()
>> + # #2 0x00000000004004cd in main () at ...
>> + gdb_test "bt" \
>> + "#0 +callback \[^\r\n\]+\r\n#1 \[^\r\n\]+ in caller \[^\r\n\]+\r\n#2
>> \[^\r\n\]+ in main \[^\r\n\]+" \
>> + "verify unwinding works for CFI without DIEs"
>> +}
> You could do below to reduce the line to under 80 cols, I wonder why we don't have this defined somewhere
> globally like "hex" or "decimal", though it's not really used that often.
>
> set r {[^\r\n]+}
> gdb_test "bt" "#0 +callback $r+\r\n#1 $r+ in caller $r+\r\n#2> $r+ in main $r+"
Thanks for the suggestion! I think calling it "r" makes things pretty
hard to read, especially with how often it is followed by \r.
I called it "text" instead, which makes the line 84 lines, but much more
readable IMO.
>
>> --
>> 2.46.2
>>
> Intel Deutschland GmbH
> Registered Address: Am Campeon 10, 85579 Neubiberg, Germany
> Tel: +49 89 99 8853-0, www.intel.de
> Managing Directors: Sean Fennelly, Jeffrey Schneiderman, Tiffany Doon Silva
> Chairperson of the Supervisory Board: Nicole Lau
> Registered Office: Munich
> Commercial Register: Amtsgericht Muenchen HRB 186928
>
--
Cheers,
Guinevere Larsen
She/Her/Hers
prev parent reply other threads:[~2024-10-09 14:57 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-10-01 18:42 [PATCH v5 0/5] Modernize frame unwinders and add disable feature Guinevere Larsen
2024-10-01 18:42 ` [PATCH v5 1/5] gdb: make gdbarch store a vector of frame unwinders Guinevere Larsen
2024-10-02 21:49 ` Thiago Jung Bauermann
2024-10-08 17:01 ` Guinevere Larsen
2024-10-03 18:33 ` Simon Marchi
2024-10-04 18:37 ` Tom Tromey
2024-10-12 1:34 ` Thiago Jung Bauermann
2024-10-14 18:18 ` Guinevere Larsen
2024-10-17 22:53 ` Tom Tromey
2024-10-18 17:40 ` Guinevere Larsen
2024-10-17 23:41 ` Tom Tromey
2024-10-01 18:42 ` [PATCH v5 2/5] gdb: add "unwinder class" to " Guinevere Larsen
2024-10-02 22:08 ` Thiago Jung Bauermann
2024-10-03 18:46 ` Simon Marchi
2024-10-08 18:22 ` Guinevere Larsen
2024-10-08 18:37 ` Simon Marchi
2024-10-01 18:42 ` [PATCH v5 3/5] gdb: Migrate frame unwinders to use C++ classes Guinevere Larsen
2024-10-03 0:23 ` Thiago Jung Bauermann
2024-10-09 18:16 ` Guinevere Larsen
2024-10-03 20:06 ` Simon Marchi
2024-10-04 5:21 ` Simon Marchi
2024-10-10 14:10 ` Guinevere Larsen
2024-10-10 16:28 ` Simon Marchi
2024-10-09 20:00 ` Guinevere Larsen
2024-10-01 18:42 ` [PATCH v5 4/5] gdb: introduce ability to disable frame unwinders Guinevere Larsen
2024-10-02 6:10 ` Eli Zaretskii
2024-10-04 17:57 ` Guinevere Larsen
2024-10-03 2:45 ` Thiago Jung Bauermann
2024-10-08 19:23 ` Guinevere Larsen
2024-10-06 2:51 ` Simon Marchi
2024-10-09 13:32 ` Guinevere Larsen
2024-10-09 15:38 ` Simon Marchi
2024-10-01 18:42 ` [PATCH v5 5/5] gdb/testsuite: Test for a backtrace through object without debuginfo Guinevere Larsen
2024-10-03 2:47 ` Thiago Jung Bauermann
2024-10-03 6:58 ` Gerlicher, Klaus
2024-10-09 14:56 ` Guinevere Larsen [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=f8666ccf-19f3-4938-b0b2-745387ba3661@redhat.com \
--to=guinevere@redhat.com \
--cc=blarsen@redhat.com \
--cc=gdb-patches@sourceware.org \
--cc=jan@jankratochvil.net \
--cc=klaus.gerlicher@intel.com \
/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