From mboxrd@z Thu Jan 1 00:00:00 1970 Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.remote-ip=100.103.45.18 ARC-Seal: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; cv=none; t=1795219795; b=1fbSNBFf9HKQDr20tOuHFver5tfjSwVBwM+SSOt/Dqd5DZWyErzaJ8uTJAdbKs+T+Yow PkULuypBjrQj5tSyjgMWULBIJsp+7CNq9EAsAtG3cCM7Hfoq/YAujV0HOsOM6V3qx0D71 OlQaGA0oOlc00RLcoiJQwX8HA6wQo8CBCLK7NiErUwYPLbas86g0X+RybwS93dcjOqrp5 g1Smh1IfhCgscM1D6jHINMbETVVX7Nm7SXoCwiP3VroEFZhroncFahBvK6aM9gAyHd9bi Ohs9ZadeWAPssAnpEMwUqyfTi+rFaQNHMz6e/+97vfxLGiOfVRjs8EhYJiLCukW607w== ARC-Message-Signature: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; c=relaxed/relaxed; t=1795219795; h=Received:DKIM-Signature:From:To:Cc:Subject:Date:Message-ID:X-Mailer: In-Reply-To:References:MIME-Version:Content-Transfer-Encoding; bh=hUPm8hcXsD5eOEjWcMgviJk+m28ETz0ZEKJMYNL1zxw=; b=BSR+cjIvZ8sSDN7heH3B2QiKoi0If5Bjs6XqyWcwrd0vcw84c7MLOypnfvlXVx+EsMnF Z4W83BxA8djl2BgBWGICLqPpIcFHdvo1JeFL91atcXpt2ZZ5epgrQus+II/s48bqvnVAY tNqXxqpt6CP9ato2tHBK5NVa1qWMltsMfkKy3aYH/Tu8z4llWreLbTgNYfbOm3kujNPYU E72WQUowJxKCwxnpOTFDEQSAUnDZQt7vNbecpU8TeOMl0DWL27Q8/qY8mWbMp2eY0OhST 9+p1y+z2Ta8BFWN0b1n3N62DvhdQFqPsiyktVjI3t3wXRHBj4VYuGGPqutc7mjUjHWA== ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass header.d=kernel.org header.i=@kernel.org header.a=rsa-sha256 header.s=k20260515 header.b=GZ7QEq8W; arc=none smtp.remote-ip=100.103.45.18 Received: from smtp.kernel.org (quasi.space.kernel.org [100.103.45.18]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (prime256v1)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with UTF8SMTPS id D5C1D60A2618; Sat, 21 Nov 2026 00:09:55 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.a=rsa-sha256 header.s=k20260515 header.b=GZ7QEq8W Received: by smtp.kernel.org (Postfix) with UTF8SMTPSA id 20BBB1F000EE; Sat, 21 Nov 2026 00:09:54 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1795219795; bh=hUPm8hcXsD5eOEjWcMgviJk+m28ETz0ZEKJMYNL1zxw=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=GZ7QEq8W4PyqW/qT6hinM44l4wrZpmfeOC9JOLKK2k6wKiN0e1qfZqkzP0UeGB91T FjWoocLZlUeY8+VY37AuOxfG4t7/QzYtAmFvcXZ6H2I7b1qkFxUq+87V6A1XX3h3KE fMsWln9AAQyWd2q0sZpQ0IsDZDZTyyWktN4kghO43q10fKlnRXGzJomJbPQg1Rbz7G DP9i0NkXw3jSZUIHsTs0dtNfsJoSgiLX4ERp86rnP5yosVsCesjlJEDyNBTK6xG5b5 xS+z19fkG0n0gvRxR2mZxvxqNjkdeF6J8JgbmTA4zN5twSFDiWPGatBmO2Yn4ArTzQ tnyqYo9ce+gvg== From: Jarkko Sakkinen To: linux-fsdevel@vger.kernel.org Cc: Jarkko Sakkinen , linux-api@vger.kernel.org, linux-doc@vger.kernel.org, linux-kernel@vger.kernel.org, linux-kselftest@vger.kernel.org, linux-mm@kvack.org, linux-security-module@vger.kernel.org, Alexander Mikhalitsyn , Alexander Viro , Andy Lutomirski , Christian Brauner , David Howells , Jan Kara , Jonathan Corbet , Josef Bacik , Kees Cook , Neal Gompa , Paul Moore , Serge Hallyn , Zygmunt Krynicki Subject: [RFC PATCH v5 14/25] rootns: Add ROOTNS_WAIT ioctl Date: Sat, 21 Nov 2026 02:08:33 +0200 Message-ID: <20261121000844.1863324-15-jarkko@kernel.org> X-Mailer: git-send-email 2.47.3 In-Reply-To: <20261121000844.1863324-1-jarkko@kernel.org> References: <20261121000844.1863324-1-jarkko@kernel.org> X-Mailing-List: linux-fsdevel@vger.kernel.org List-Id: List-Post: List-Help: List-Subscribe: List-Unsubscribe: List-Owner: Precedence: list MIME-Version: 1.0 Content-Transfer-Encoding: 8bit wait4() and waitid() only allow a parent or subreaper to collect exit status, so an external manager holding a rootns file descriptor cannot wait on a rootns init process forked by another process or observe exit status multiple times across observers. Add the ROOTNS_WAIT ioctl and struct rootns_wait_args to wait freezably and interruptibly on a rootns file descriptor until its init thread group exits and report the recorded wait status. Return the init TGID translated into the caller's active PID namespace, 0 for WNOHANG while running, or -ECHILD if init was never started or is not visible to the caller. Co-developed-by: David Howells Signed-off-by: David Howells Signed-off-by: Jarkko Sakkinen --- include/uapi/linux/rootns.h | 26 +++++++++++++++++++++ kernel/rootns.c | 46 +++++++++++++++++++++++++++++++++++++ 2 files changed, 72 insertions(+) diff --git a/include/uapi/linux/rootns.h b/include/uapi/linux/rootns.h index bd643bdfb43c..3fd215185398 100644 --- a/include/uapi/linux/rootns.h +++ b/include/uapi/linux/rootns.h @@ -8,6 +8,7 @@ #define _UAPI_LINUX_ROOTNS_H #include +#include /** * enum rootns_flags - Flags for rootns() @@ -24,6 +25,16 @@ enum rootns_flags { ROOTNS_ALL_FLAGS = 0x00000003, }; +/** + * struct rootns_wait_args - Arguments for %ROOTNS_WAIT + * @flags: Wait options (0 or %WNOHANG). + * @status: Output wait status of rootns init in waitpid(2) format. + */ +struct rootns_wait_args { + __u32 flags; + __s32 status; +}; + #define ROOTNS_IOCTL_MAGIC 0xB9 /** @@ -41,4 +52,19 @@ enum rootns_flags { */ #define ROOTNS_GET_NSFD _IO(ROOTNS_IOCTL_MAGIC, 0x01) +/** + * DOC: ROOTNS_WAIT + * + * ``ioctl(rootnsfd, ROOTNS_WAIT, &args)`` waits interruptibly (unless + * ``WNOHANG`` is set in ``args.flags``) until ``rootnsfd`` reaches + * ``ROOTNS_DEAD``, then writes the recorded exit status of ``rootns`` init + * into ``args.status``. + * + * Returns the caller-visible TGID of ``rootns`` init on success, ``0`` if + * ``WNOHANG`` is set and the ``rootns`` has not terminated, or ``-1`` with + * ``errno`` set to ``ECHILD`` if init was never started or is not visible in + * the caller's active PID namespace. + */ +#define ROOTNS_WAIT _IOWR(ROOTNS_IOCTL_MAGIC, 0x02, struct rootns_wait_args) + #endif /* _UAPI_LINUX_ROOTNS_H */ diff --git a/kernel/rootns.c b/kernel/rootns.c index 5d59fc9a591d..cb74ddc8363c 100644 --- a/kernel/rootns.c +++ b/kernel/rootns.c @@ -117,6 +117,47 @@ static long rootns_get_nsfd(struct rootns *rootns, unsigned long nstype) return nsproxy_open_namespace(rootns->ns, rootns->user_ns, nstype); } +static long rootns_wait(struct rootns *rootns, void __user *uarg) +{ + struct pid_namespace *pid_ns = task_active_pid_ns(current); + struct rootns_wait_args args; + long ret; + + if (!uarg) + return -EINVAL; + if (copy_from_user(&args, uarg, sizeof(args))) + return -EFAULT; + if (args.flags & ~WNOHANG) + return -EINVAL; + if (!pidns_is_ancestor(rootns->ns->pid_ns_for_children, pid_ns)) + return -ECHILD; + if (READ_ONCE(rootns->state) == ROOTNS_INIT) + return -ECHILD; + + if (!(args.flags & WNOHANG)) { + ret = wait_event_freezable(rootns->waitq, + READ_ONCE(rootns->state) == ROOTNS_DEAD || + READ_ONCE(rootns->state) == ROOTNS_INIT); + if (ret) + return ret; + } + + scoped_guard(spinlock, &rootns->lock) { + if (rootns->state == ROOTNS_INIT) + return -ECHILD; + if (rootns->state != ROOTNS_DEAD) + return 0; + ret = pid_nr_ns(rootns->init_pid, pid_ns); + if (ret <= 0) + return -ECHILD; + args.status = rootns->exit_code; + } + + if (copy_to_user(uarg, &args, sizeof(args))) + return -EFAULT; + return ret; +} + static long rootns_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { struct rootns *rootns = file->private_data; @@ -124,6 +165,8 @@ static long rootns_ioctl(struct file *file, unsigned int cmd, unsigned long arg) switch (cmd) { case ROOTNS_GET_NSFD: return rootns_get_nsfd(rootns, arg); + case ROOTNS_WAIT: + return rootns_wait(rootns, (void __user *)arg); default: return -ENOIOCTLCMD; } @@ -134,6 +177,8 @@ static long rootns_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { switch (cmd) { + case ROOTNS_WAIT: + return rootns_ioctl(file, cmd, (unsigned long)compat_ptr(arg)); case ROOTNS_GET_NSFD: return rootns_ioctl(file, cmd, arg); default: @@ -328,3 +373,4 @@ SYSCALL_DEFINE1(rootns, unsigned int, flags) return fd; } + -- 2.47.3