From: Cyril Hrubis <chrubis@suse.cz>
To: Wei Gao <wegao@suse.com>
Cc: ltp@lists.linux.it
Subject: Re: [LTP] [PATCH v3] clone3: Add clone3's clone_args cgroup
Date: Wed, 26 Apr 2023 15:44:01 +0200 [thread overview]
Message-ID: <ZEkqoZKbORCnRlT5@rei> (raw)
In-Reply-To: <20230422014216.26294-1-wegao@suse.com>
Hi!
> diff --git a/include/tst_cgroup.h b/include/tst_cgroup.h
> index 2826ddad1..d8e621962 100644
> --- a/include/tst_cgroup.h
> +++ b/include/tst_cgroup.h
> @@ -157,6 +157,9 @@ const char *
> tst_cg_group_name(const struct tst_cg_group *const cg)
> __attribute__ ((nonnull, warn_unused_result));
>
> +int tst_cg_group_dir_fd(const struct tst_cg_group *const cg)
> + __attribute__ ((nonnull, warn_unused_result));
> +
> /* Remove a descendant CGroup */
> struct tst_cg_group *
> tst_cg_group_rm(struct tst_cg_group *const cg)
> diff --git a/include/tst_clone.h b/include/tst_clone.h
> index 9ffdc68d1..7b278dfa7 100644
> --- a/include/tst_clone.h
> +++ b/include/tst_clone.h
> @@ -11,6 +11,7 @@
> struct tst_clone_args {
> uint64_t flags;
> uint64_t exit_signal;
> + uint64_t cgroup;
> };
>
> /* clone3 with fallbacks to clone when possible. Be aware that it
> diff --git a/lib/tst_cgroup.c b/lib/tst_cgroup.c
> index 50699bc63..c63d04c67 100644
> --- a/lib/tst_cgroup.c
> +++ b/lib/tst_cgroup.c
> @@ -166,6 +166,7 @@ static const struct cgroup_file cgroup_ctrl_files[] = {
> { "cgroup.controllers", NULL, 0 },
> { "cgroup.subtree_control", NULL, 0 },
> { "cgroup.clone_children", "cgroup.clone_children", 0 },
> + { "cgroup.kill", NULL, 0 },
> { }
> };
This seems to be unrelated change.
> @@ -1093,6 +1094,11 @@ const char *tst_cg_group_name(const struct tst_cg_group *const cg)
> return cg->group_name;
> }
>
> +int tst_cg_group_dir_fd(const struct tst_cg_group *const cg)
> +{
> + return cg->dirs[0]->dir_fd;
Does this actually work in the case of mixed cgroup v1 and v2 setup?
Even if so I would just make sure that this call returns a fd pointing
to a v2 directory. Also the function should include "unified" in the
name so that it's clear that it returns fd to the v2 "unified"
hierarchy. I guess that tst_cg_group_unified_dir_fd() would be fitting
name.
To implement that we have to loop over dirs[] array and check which one
of them is v2, which should be reachable via the root pointer in the dir.
> +}
> +
> struct tst_cg_group *tst_cg_group_rm(struct tst_cg_group *const cg)
> {
> struct cgroup_dir **dir;
> diff --git a/lib/tst_clone.c b/lib/tst_clone.c
> index ecc84408c..2aa00beb1 100644
> --- a/lib/tst_clone.c
> +++ b/lib/tst_clone.c
> @@ -15,6 +15,7 @@ pid_t tst_clone(const struct tst_clone_args *tst_args)
> struct clone_args args = {
> .flags = tst_args->flags,
> .exit_signal = tst_args->exit_signal,
> + .cgroup = tst_args->cgroup,
> };
> int flags;
> pid_t pid = -1;
> diff --git a/runtest/syscalls b/runtest/syscalls
> index 9c23a4248..0b6adfd7f 100644
> --- a/runtest/syscalls
> +++ b/runtest/syscalls
> @@ -117,6 +117,7 @@ clone09 clone09
>
> clone301 clone301
> clone302 clone302
> +clone303 clone303
>
> close01 close01
> close02 close02
> diff --git a/testcases/kernel/syscalls/clone3/.gitignore b/testcases/kernel/syscalls/clone3/.gitignore
> index 604cb903e..10369954b 100644
> --- a/testcases/kernel/syscalls/clone3/.gitignore
> +++ b/testcases/kernel/syscalls/clone3/.gitignore
> @@ -1,2 +1,3 @@
> clone301
> clone302
> +clone303
> diff --git a/testcases/kernel/syscalls/clone3/clone303.c b/testcases/kernel/syscalls/clone3/clone303.c
> new file mode 100644
> index 000000000..e35cc8def
> --- /dev/null
> +++ b/testcases/kernel/syscalls/clone3/clone303.c
> @@ -0,0 +1,94 @@
> +// SPDX-License-Identifier: GPL-2.0-or-later
> +/*
> + * Copyright (c) 2023 SUSE LLC <wegao@suse.com>
> + */
> +
> +/*\
> + * [Description]
> + *
> + * This test case check clone3 CLONE_INTO_CGROUP flag
> + *
> + */
> +
> +#define _GNU_SOURCE
> +#include <stdio.h>
> +#include <stdlib.h>
> +#include <sys/wait.h>
> +
> +#include "tst_test.h"
> +#include "lapi/sched.h"
> +#include "lapi/pidfd.h"
> +
> +#define BUF_LEN 20
> +
> +static struct tst_cg_group *cg_child_test_simple;
> +static struct clone_args *args;
> +
> +static pid_t clone_into_cgroup(int cgroup_fd)
> +{
> + pid_t pid;
> +
> + args->flags = CLONE_INTO_CGROUP;
> + args->exit_signal = SIGCHLD;
> + args->cgroup = cgroup_fd;
> +
> + pid = clone3(args, sizeof(*args));
> +
> + if (pid < 0)
> + tst_res(TFAIL | TTERRNO, "clone3() failed !");
> +
> + return pid;
> +}
> +
> +static void run(void)
> +{
> + int fd;
> + pid_t pid;
> +
> + cg_child_test_simple = tst_cg_group_mk(tst_cg, "cg_test_simple");
> + fd = tst_cg_group_dir_fd(cg_child_test_simple);
I suppoose that this can be done once in the test setup.
> + pid = clone_into_cgroup(fd);
> +
> + if (!pid) {
> + TST_CHECKPOINT_WAIT(0);
> + return;
> + }
> +
> + char buf[BUF_LEN];
> +
> + SAFE_CG_READ(cg_child_test_simple, "cgroup.procs", buf, BUF_LEN);
> +
> + int x = atoi(buf);
> +
> + if (x == pid)
> + tst_res(TPASS, "clone3 case pass!");
> + else
> + tst_brk(TFAIL | TTERRNO, "clone3() failed !");
> +
> + TST_CHECKPOINT_WAKE(0);
> +
> + SAFE_WAITPID(pid, NULL, 0);
> +
> + cg_child_test_simple = tst_cg_group_rm(cg_child_test_simple);
And this could be done in the test cleanup.
> +}
> +
> +static void setup(void)
> +{
> + clone3_supported_by_kernel();
> +}
> +
> +static struct tst_test test = {
> + .test_all = run,
> + .setup = setup,
> + .forks_child = 1,
> + .max_runtime = 20,
> + .needs_cgroup_ctrls = (const char *const []){ "memory", NULL },
> + .needs_cgroup_ver = TST_CG_V2,
> + .needs_checkpoints = 1,
> + .min_kver = "5.7",
> + .bufs = (struct tst_buffers []) {
> + {&args, .size = sizeof(*args)},
> + {},
> + }
> +};
> --
> 2.35.3
>
>
> --
> Mailing list info: https://lists.linux.it/listinfo/ltp
--
Cyril Hrubis
chrubis@suse.cz
--
Mailing list info: https://lists.linux.it/listinfo/ltp
next prev parent reply other threads:[~2023-04-26 13:46 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-26 0:35 [LTP] [PATCH v1] clone3: Add clone3's clone_args cgroup Wei Gao via ltp
2023-03-23 9:26 ` Petr Vorel
2023-03-23 12:17 ` Wei Gao via ltp
2023-03-24 6:32 ` Petr Vorel
2023-03-24 6:54 ` Wei Gao via ltp
2023-03-24 8:26 ` Petr Vorel
2023-04-21 12:38 ` [LTP] [PATCH v2] " Wei Gao via ltp
2023-04-21 13:29 ` Cyril Hrubis
2023-04-22 1:42 ` [LTP] [PATCH v3] " Wei Gao via ltp
2023-04-26 13:44 ` Cyril Hrubis [this message]
2023-05-09 0:31 ` [LTP] [PATCH v4] " Wei Gao via ltp
2023-05-17 9:28 ` Petr Vorel
2023-05-17 12:08 ` [LTP] [PATCH v5] " Wei Gao via ltp
2023-08-25 10:36 ` Richard Palethorpe
2023-08-29 23:26 ` Wei Gao via ltp
2023-08-30 8:02 ` Richard Palethorpe
2023-08-29 23:18 ` [LTP] [PATCH v6] " Wei Gao via ltp
2023-08-31 6:47 ` [LTP] [PATCH v7] " Wei Gao via ltp
2023-08-31 9:19 ` Richard Palethorpe
2023-09-01 10:19 ` Petr Vorel
2023-09-01 10:22 ` Petr Vorel
2023-09-02 5:05 ` Wei Gao via ltp
2024-04-04 20:42 ` Petr Vorel
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=ZEkqoZKbORCnRlT5@rei \
--to=chrubis@suse.cz \
--cc=ltp@lists.linux.it \
--cc=wegao@suse.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