Mirror of the gdb-patches mailing list
 help / color / mirror / Atom feed
From: "Gerlicher, Klaus" <klaus.gerlicher@intel.com>
To: Guinevere Larsen <guinevere@redhat.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: Thu, 3 Oct 2024 06:58:31 +0000	[thread overview]
Message-ID: <SN7PR11MB70914F7AF260BD8EC601673FE8712@SN7PR11MB7091.namprd11.prod.outlook.com> (raw)
In-Reply-To: <20241001184235.3710608-6-guinevere@redhat.com>

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+"

> --
> 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


  parent reply	other threads:[~2024-10-03  6:59 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 [this message]
2024-10-09 14:56     ` 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=SN7PR11MB70914F7AF260BD8EC601673FE8712@SN7PR11MB7091.namprd11.prod.outlook.com \
    --to=klaus.gerlicher@intel.com \
    --cc=blarsen@redhat.com \
    --cc=gdb-patches@sourceware.org \
    --cc=guinevere@redhat.com \
    --cc=jan@jankratochvil.net \
    /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