From: Tom de Vries <tdevries@suse.de>
To: Guinevere Larsen <guinevere@redhat.com>, gdb-patches@sourceware.org
Subject: Re: [PATCH] gdb/testsuite: add test for memory requirements of gcore
Date: Mon, 24 Feb 2025 16:24:58 +0100 [thread overview]
Message-ID: <5f6b3d35-bb54-45d3-9256-4c8714b9fc22@suse.de> (raw)
In-Reply-To: <20250224140407.252987-1-guinevere@redhat.com>
On 2/24/25 15:04, Guinevere Larsen wrote:
> For a long time, Fedora has been carrying an out-of-tree patch with a
> similar test to the one proposed in this patch, that ensures that the
> memory requirements don't grow with the inferior's memory. It's been
> so long that the context for why this test exists has been lost, but
> it looked like it could be interesting for upstream.
>
> The test runs twice, once with the inferior allocating 4Mb of memory,
> and the other allocating 64Mb. My plan was to find the rate at which
> things increase based on inferior size, and have that tested to ensure
> we're not growing that requirement accidentally, but my testing
> actually showed memory requirements going down as the inferior increases,
> so instead I just hardcoded that we need less than 2Mb for the command,
> and it can be tweaked later if necessary.
I tried out the test-case, and it passed for me, but after doing:
...
$ echo "1" | sudo tee /proc/sys/kernel/yama/ptrace_scope
...
I get:
...
(gdb) attach 29408^M
Attaching to program:
/data/vries/gdb/leap-15-6/build/gdb/testsuite/outputs/gdb.base/gcore-memory-usage/gcore-memory-usage,
process 29408^M
ptrace: Operation not permitted.^M
(gdb) PASS: gdb.base/gcore-memory-usage.exp: 4 Mb: attach 29408
PASS: gdb.base/gcore-memory-usage.exp: 4 Mb: before: Managed to read the
memory usage
gcore
/data/vries/gdb/leap-15-6/build/gdb/testsuite/outputs/gdb.base/gcore-memory-usage/gcore-memory-usage.core^M
You can't do that without a process to debug.^M
(gdb) FAIL: gdb.base/gcore-memory-usage.exp: 4 Mb: create the corefile
...
So, this is missing "require can_spawn_for_attach".
Thanks,
- Tom
> ---
> gdb/testsuite/gdb.base/gcore-memory-usage.c | 37 +++++++
> gdb/testsuite/gdb.base/gcore-memory-usage.exp | 97 +++++++++++++++++++
> 2 files changed, 134 insertions(+)
> create mode 100644 gdb/testsuite/gdb.base/gcore-memory-usage.c
> create mode 100644 gdb/testsuite/gdb.base/gcore-memory-usage.exp
>
> diff --git a/gdb/testsuite/gdb.base/gcore-memory-usage.c b/gdb/testsuite/gdb.base/gcore-memory-usage.c
> new file mode 100644
> index 00000000000..514fdb905cf
> --- /dev/null
> +++ b/gdb/testsuite/gdb.base/gcore-memory-usage.c
> @@ -0,0 +1,37 @@
> +/* This testcase is part of GDB, the GNU debugger.
> +
> + Copyright 2025 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/>. */
> +
> +
> +#include <stdio.h>
> +#include <stdlib.h>
> +#include <unistd.h>
> +
> +int main(int argc, char **argv)
> +{
> + /* Use argv to calculate how many megabytes to allocate.
> + Do this to see how much gcore memory usage increases
> + based on inferior dynamic memory. */
> + int megs = atoi (argv[1]);
> + char *p = malloc (megs * 1024 * 1024);
> +
> + /* Wait a long time so GDB can attach to this. */
> + sleep (10);
> +
> + /* Let's be nice citizens :). */
> + free (p);
> + return 0;
> +}
> diff --git a/gdb/testsuite/gdb.base/gcore-memory-usage.exp b/gdb/testsuite/gdb.base/gcore-memory-usage.exp
> new file mode 100644
> index 00000000000..cdb6d73c11b
> --- /dev/null
> +++ b/gdb/testsuite/gdb.base/gcore-memory-usage.exp
> @@ -0,0 +1,97 @@
> +# Copyright (C) 2025 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 gcore doesn't end up using excessive memory.
> +
> +require {istarget "*-*-linux*"}
> +
> +standard_testfile
> +
> +if {[build_executable "failed to prepare" $testfile $srcfile {debug}] == -1} {
> + return -1
> +}
> +
> +# Read the proc_pid_status page, to find how much memory the given
> +# PID is using. This is meant to be used to find the
> +# memory usage for the GDB in this test.
> +# Returns memory usage in Kb, or -1 if it couldn't be found.
> +proc get_mem_usage {pid prefix} {
> + set fd [open "/proc/$pid/status"]
> + set memory -1
> + while {[gets $fd line] != -1} {
> + if {[regexp {VmSize:\s*([0-9]+) kB} $line full mem]} {
> + set memory $mem
> + break
> + }
> + }
> + close $fd
> +
> + gdb_assert {$memory != -1} "$prefix: Managed to read the memory usage"
> +
> + return $memory
> +}
> +
> +# This proc restarts GDB, runs the inferior with the desired
> +# amount of memory, then checks how much memory is necessary
> +# to run the gcore command. It will return -1 if the gcore
> +# command fails, 0 otherwise.
> +proc run_test {megs} {
> + with_test_prefix "$megs Mb" {
> + clean_restart $::testfile
> +
> + set bin [standard_output_file $::testfile]
> + set corefile [standard_output_file "${::testfile}.core"]
> + set inferior_pid [eval exec $bin $megs &]
> + set gdb_pid [exp_pid -i [board_info host fileid]]
> +
> + # wait for memory allocation to finish.
> + sleep 1
> +
> + # Get the important info.
> + gdb_test "attach $inferior_pid" "Attaching to.*"
> + set mem_before [get_mem_usage $gdb_pid before]
> + if {![gdb_gcore_cmd $corefile "create the corefile"]} {
> + return -1
> + }
> + set mem_after [get_mem_usage $gdb_pid after]
> +
> + # Do the main part of the test: How much is the memory
> + # usage of GDB going to grow after using the gcore command.
> + set diff_k [expr $mem_after - $mem_before]
> + set diff [expr $diff_k/1024]
> + verbose -log "The gcore command used $diff Mb ($diff_k Kb)"
> + # The original plan was to compare to a multiple of MEGS
> + # but since the requirements don't seem to go up as the
> + # inferior allocated more memory, we instead just hardcode
> + # 2 megs, since sometimes 1 is used.
> + gdb_assert {$diff < 2} \
> + "gdb did not use as much memory as the inferior"
> +
> + # Kill the inferior so we don't have to wait until the process
> + # finishes.
> + gdb_test "signal SIGKILL" ".*The program no longer exists."
> + }
> + return 0
> +}
> +
> +# If we couldn't create the first corefile, there's no point
> +# in running the second part of the test.
> +if {[run_test 4] != 0} {
> + return
> +}
> +# Surprisingly enough, the larger inferior doesn't seem to use
> +# any extra memory, it usually uses less memory. Which is good,
> +# it means our memory requirements aren't growing with the inferior.
> +run_test 64
next prev parent reply other threads:[~2025-02-24 15:25 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-02-24 14:04 Guinevere Larsen
2025-02-24 15:24 ` Tom de Vries [this message]
2025-02-24 19:47 ` Guinevere Larsen
2025-02-24 18:21 ` Andrew Burgess
2025-02-24 20:00 ` Guinevere Larsen
2025-02-24 20:18 ` Guinevere Larsen
2025-02-25 13:45 ` [PATCH v2] " Guinevere Larsen
2025-02-26 17:50 ` [PATCH v3] " Guinevere Larsen
2025-03-04 18:19 ` Tom Tromey
2025-03-05 20:03 ` Guinevere Larsen
2025-03-05 20:21 ` Tom Tromey
2025-03-05 20:39 ` Guinevere Larsen
2025-03-11 16:14 ` Simon Marchi
2025-03-11 16:43 ` 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=5f6b3d35-bb54-45d3-9256-4c8714b9fc22@suse.de \
--to=tdevries@suse.de \
--cc=gdb-patches@sourceware.org \
--cc=guinevere@redhat.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