From: Petr Vorel <pvorel@suse.cz>
To: Stephen Bertram <sbertram@redhat.com>
Cc: ltp@lists.linux.it
Subject: Re: [LTP] [PATCH v3] Add new tests for clone and clone3
Date: Wed, 12 Nov 2025 21:11:24 +0100 [thread overview]
Message-ID: <20251112201124.GA157071@pevik> (raw)
In-Reply-To: <20251112180319.630584-1-sbertram@redhat.com>
Hi Ben,
> Confirming EPERM is returned when CAP_SYS_ADMIN is
> removed from clone and clone3.
> And for clone3 the set_tid_size is greater than 0.
> runtest/syscalls | 2 +
> testcases/kernel/syscalls/clone/.gitignore | 1 +
> testcases/kernel/syscalls/clone/clone11.c | 72 +++++++++++++++++++++
> testcases/kernel/syscalls/clone3/.gitignore | 1 +
> testcases/kernel/syscalls/clone3/clone304.c | 63 ++++++++++++++++++
> 5 files changed, 139 insertions(+)
This looks better then v2. Generally LGTM, few notes below.
Reviewed-by: Petr Vorel <pvorel@suse.cz>
I see adding the test for both clone() and clone3() is because Li asked
[1] for it.
[1] https://lore.kernel.org/ltp/CAEemH2cQam9tOeQj_4SdvoYG5ZNmar6Quc5M6mtr==2-HQR9yQ@mail.gmail.com/
nit: I would phrase the subject as "Add EPERM test for clone and clone3" to
have at least a bit of information in the subject.
...
> +/*\
> + * This test verifies that clone() fals with EPERM when CAP_SYS_ADMIN
s/fals/fails/
If new version is needed, could you please use :man2:`clone` ?
This will link man clone(2) [2] in our test catalog [3].
[2] https://man7.org/linux/man-pages/man2/clone.2.html
[3] https://linux-test-project.readthedocs.io/en/latest/users/test_catalog.html
> + * has been dropped.
> + */
> +
> +#define _GNU_SOURCE
> +
> +#include "tst_test.h"
> +#include "clone_platform.h"
NOTE: we need to add
#include "lapi/sched.h"
otherwise old systems fail to compile due missing CLONE_NEWCGROUP.
This should really be fixed in a new version or before merge.
FYI we support kernel from 4.4 and glibc 2.22 [4].
[4] https://linux-test-project.readthedocs.io/en/latest/users/supported_systems.html
> +
> +static void *child_stack;
> +static int *child_pid;
> +
> +static struct tcase {
> + uint64_t flags;
> + char *sflags;
> +} tcases[] = {
> + {CLONE_NEWPID, "CLONE_NEWPID"},
nit: If new version needed using macro is more elegant than copy pasting
definition as a string:
#define DESC(x) .flags = x, .sflags = #x
{ DESC(CLONE_NEWPID) },
{ DESC(CLONE_NEWCGROUP) },
...
> + {CLONE_NEWCGROUP, "CLONE_NEWCGROUP"},
> + {CLONE_NEWIPC, "CLONE_NEWIPC"},
> + {CLONE_NEWNET, "CLONE_NEWNET"},
> + {CLONE_NEWNS, "CLONE_NEWNS"},
> + {CLONE_NEWUTS, "CLONE_NEWUTS"},
> +};
> +
> +static int child_fn(void *arg LTP_ATTRIBUTE_UNUSED)
> +{
> + *child_pid = getpid();
> + exit(0);
nit: there are 2 tabs.
> +}
> +
> +static void run(unsigned int n)
> +{
> + struct tcase *tc = &tcases[n];
> +
> + TST_EXP_FAIL(ltp_clone(tc->flags, child_fn, NULL, CHILD_STACK_SIZE,
> + child_stack), EPERM, "clone(%s) should fail with EPERM"
> + , tc->sflags);
very nit: ',' is usually on a previous line.
> +}
> +
> +static void setup(void)
> +{
> + child_pid = SAFE_MMAP(NULL, sizeof(*child_pid), PROT_READ | PROT_WRITE,
> + MAP_SHARED | MAP_ANONYMOUS, -1, 0);
> +}
> +
> +static void cleanup(void)
> +{
> + if (child_pid)
> + SAFE_MUNMAP(child_pid, sizeof(*child_pid));
> +}
> +
> +static struct tst_test test = {
> + .tcnt = ARRAY_SIZE(tcases),
> + .setup = setup,
> + .test = run,
> + .cleanup = cleanup,
> + .needs_root = 1,
> + .caps = (struct tst_cap []) {
> + TST_CAP(TST_CAP_DROP, CAP_SYS_ADMIN),
> + {},
> + },
> + .bufs = (struct tst_buffers []) {
> + {&child_stack, .size = CHILD_STACK_SIZE},
> + {},
> + },
> +};
> diff --git a/testcases/kernel/syscalls/clone3/.gitignore b/testcases/kernel/syscalls/clone3/.gitignore
> index 10369954b..e9b5312f4 100644
> --- a/testcases/kernel/syscalls/clone3/.gitignore
> +++ b/testcases/kernel/syscalls/clone3/.gitignore
> @@ -1,3 +1,4 @@
> clone301
> clone302
> clone303
> +clone304
> diff --git a/testcases/kernel/syscalls/clone3/clone304.c b/testcases/kernel/syscalls/clone3/clone304.c
> new file mode 100644
> index 000000000..e47ab313e
> --- /dev/null
> +++ b/testcases/kernel/syscalls/clone3/clone304.c
> @@ -0,0 +1,63 @@
> +// SPDX-License-Identifier: GPL-2.0-or-later
> +/*
> + * Copyright (c) 2025 Stephen Bertram <sbertram@redhat.com>
> + */
> +
> +/*\
> + * This test verifies that clone3() fals with EPERM when CAP_SYS_ADMIN
> + * has been dropped and ``clone_args.set_tid_size`` is greater than zero.
> + */
> +
> +#define _GNU_SOURCE
> +#include "tst_test.h"
> +#include "lapi/sched.h"
> +
> +static struct clone_args args = {0};
> +static pid_t tid_array[4] = {0, 0, 0, 0};
> +
> +static struct tcase {
> + uint64_t flags;
> + char *sflags;
> +} tcases[] = {
> + {CLONE_NEWPID, "CLONE_NEWPID"},
> + {CLONE_NEWCGROUP, "CLONE_NEWCGROUP"},
> + {CLONE_NEWIPC, "CLONE_NEWIPC"},
> + {CLONE_NEWNET, "CLONE_NEWNET"},
> + {CLONE_NEWNS, "CLONE_NEWNS"},
> + {CLONE_NEWUTS, "CLONE_NEWUTS"},
> +};
> +
> +static void run(unsigned int n)
> +{
> + struct tcase *tc = &tcases[n];
> +
> + args.flags = tc->flags;
> +
> + TST_EXP_FAIL(clone3(&args, sizeof(args)), EPERM, "clone3(%s) should fail with EPERM"
> + , tc->sflags);
> +}
> +
> +static void setup(void)
> +{
> + clone3_supported_by_kernel();
> +
> + memset(&args, 0, sizeof(args));
> + args.set_tid = (uintptr_t)tid_array;
> + /* set_tid_size greater than zero - requires CAP_SYS_ADMIN */
> + args.set_tid_size = 4;
> +}
> +
> +static struct tst_test test = {
NOTE: man clone(2) claims set_tid_size is "since Linux 5.5".
But fortunately test TCONF correctly:
../../../../include/lapi/sched.h:88: TCONF: syscall(435) __NR_clone3 not supported on your arch
Therefore we don't need to add .min_kver = "5.5".
> + .tcnt = ARRAY_SIZE(tcases),
> + .setup = setup,
> + .test = run,
> + .needs_root = 1,
> + .caps = (struct tst_cap []) {
> + TST_CAP(TST_CAP_DROP, CAP_SYS_ADMIN),
> + {},
> + },
> + .bufs = (struct tst_buffers []) {
> + {&args, .size = sizeof(struct clone_args)},
> + {},
> + },
> +};
--
Mailing list info: https://lists.linux.it/listinfo/ltp
next prev parent reply other threads:[~2025-11-12 20:11 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-11-12 18:03 [LTP] [PATCH v3] Add new tests for clone and clone3 Stephen Bertram via ltp
2025-11-12 20:11 ` Petr Vorel [this message]
2025-11-12 21:05 ` Stephen Bertram via ltp
2025-11-18 14:54 ` 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=20251112201124.GA157071@pevik \
--to=pvorel@suse.cz \
--cc=ltp@lists.linux.it \
--cc=sbertram@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.