From: Jann Horn <jannh@google.com>
To: David Howells <dhowells@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>,
raven@themaw.net, Miklos Szeredi <mszeredi@redhat.com>,
Christian Brauner <christian@brauner.io>,
Linux API <linux-api@vger.kernel.org>,
linux-fsdevel <linux-fsdevel@vger.kernel.org>,
kernel list <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH 15/19] vfs: Add superblock notifications [ver #16]
Date: Thu, 20 Feb 2020 00:08:11 +0100 [thread overview]
Message-ID: <CAG48ez2B2J_3-+EjR20ukRu3noPnAccZsOTaea0jtKK4=+bkhQ@mail.gmail.com> (raw)
In-Reply-To: <158204561120.3299825.5242636508455859327.stgit@warthog.procyon.org.uk>
On Tue, Feb 18, 2020 at 6:07 PM David Howells <dhowells@redhat.com> wrote:
> Add a superblock event notification facility whereby notifications about
> superblock events, such as I/O errors (EIO), quota limits being hit
> (EDQUOT) and running out of space (ENOSPC) can be reported to a monitoring
> process asynchronously. Note that this does not cover vfsmount topology
> changes. watch_mount() is used for that.
[...]
> @@ -354,6 +356,10 @@ void deactivate_locked_super(struct super_block *s)
> {
> struct file_system_type *fs = s->s_type;
> if (atomic_dec_and_test(&s->s_active)) {
> +#ifdef CONFIG_SB_NOTIFICATIONS
> + if (s->s_watchers)
> + remove_watch_list(s->s_watchers, s->s_unique_id);
> +#endif
> cleancache_invalidate_fs(s);
> unregister_shrinker(&s->s_shrink);
> fs->kill_sb(s);
[...]
> +/**
> + * sys_watch_sb - Watch for superblock events.
> + * @dfd: Base directory to pathwalk from or fd referring to superblock.
> + * @filename: Path to superblock to place the watch upon
> + * @at_flags: Pathwalk control flags
> + * @watch_fd: The watch queue to send notifications to.
> + * @watch_id: The watch ID to be placed in the notification (-1 to remove watch)
> + */
> +SYSCALL_DEFINE5(watch_sb,
> + int, dfd,
> + const char __user *, filename,
> + unsigned int, at_flags,
> + int, watch_fd,
> + int, watch_id)
> +{
> + struct watch_queue *wqueue;
> + struct super_block *s;
> + struct watch_list *wlist = NULL;
> + struct watch *watch = NULL;
> + struct path path;
> + unsigned int lookup_flags =
> + LOOKUP_DIRECTORY | LOOKUP_FOLLOW | LOOKUP_AUTOMOUNT;
> + int ret;
[...]
> + wqueue = get_watch_queue(watch_fd);
> + if (IS_ERR(wqueue))
> + goto err_path;
> +
> + s = path.dentry->d_sb;
> + if (watch_id >= 0) {
> + ret = -ENOMEM;
> + if (!s->s_watchers) {
READ_ONCE() ?
> + wlist = kzalloc(sizeof(*wlist), GFP_KERNEL);
> + if (!wlist)
> + goto err_wqueue;
> + init_watch_list(wlist, NULL);
> + }
> +
> + watch = kzalloc(sizeof(*watch), GFP_KERNEL);
> + if (!watch)
> + goto err_wlist;
> +
> + init_watch(watch, wqueue);
> + watch->id = s->s_unique_id;
> + watch->private = s;
> + watch->info_id = (u32)watch_id << 24;
> +
> + ret = security_watch_sb(watch, s);
> + if (ret < 0)
> + goto err_watch;
> +
> + down_write(&s->s_umount);
> + ret = -EIO;
> + if (atomic_read(&s->s_active)) {
> + if (!s->s_watchers) {
> + s->s_watchers = wlist;
> + wlist = NULL;
> + }
> +
> + ret = add_watch_to_object(watch, s->s_watchers);
> + if (ret == 0) {
> + spin_lock(&sb_lock);
> + s->s_count++;
> + spin_unlock(&sb_lock);
Where is the corresponding decrement of s->s_count? I'm guessing that
it should be in the ->release_watch() handler, except that there isn't
one...
> + watch = NULL;
> + }
> + }
> + up_write(&s->s_umount);
> + } else {
> + ret = -EBADSLT;
> + if (READ_ONCE(s->s_watchers)) {
(Nit: I don't get why you do a lockless check here before taking the
lock - it'd be more straightforward to take the lock first, and it's
not like you want to optimize for the case where someone calls
sys_watch_sb() with invalid arguments...)
> + down_write(&s->s_umount);
> + ret = remove_watch_from_object(s->s_watchers, wqueue,
> + s->s_unique_id, false);
> + up_write(&s->s_umount);
> + }
> + }
> +
> +err_watch:
> + kfree(watch);
> +err_wlist:
> + kfree(wlist);
> +err_wqueue:
> + put_watch_queue(wqueue);
> +err_path:
> + path_put(&path);
> + return ret;
> +}
> +#endif
[...]
> +/**
> + * notify_sb: Post simple superblock notification.
> + * @s: The superblock the notification is about.
> + * @subtype: The type of notification.
> + * @info: WATCH_INFO_FLAG_* flags to be set in the record.
> + */
> +static inline void notify_sb(struct super_block *s,
> + enum superblock_notification_type subtype,
> + u32 info)
> +{
> +#ifdef CONFIG_SB_NOTIFICATIONS
> + if (unlikely(s->s_watchers)) {
READ_ONCE() ?
> + struct superblock_notification n = {
> + .watch.type = WATCH_TYPE_SB_NOTIFY,
> + .watch.subtype = subtype,
> + .watch.info = watch_sizeof(n) | info,
> + .sb_id = s->s_unique_id,
> + };
> +
> + post_sb_notification(s, &n);
> + }
> +
> +#endif
> +}
> +
> +/**
> + * notify_sb_error: Post superblock error notification.
> + * @s: The superblock the notification is about.
> + * @error: The error number to be recorded.
> + */
> +static inline int notify_sb_error(struct super_block *s, int error)
> +{
> +#ifdef CONFIG_SB_NOTIFICATIONS
> + if (unlikely(s->s_watchers)) {
READ_ONCE() ?
> + struct superblock_error_notification n = {
> + .s.watch.type = WATCH_TYPE_SB_NOTIFY,
> + .s.watch.subtype = NOTIFY_SUPERBLOCK_ERROR,
> + .s.watch.info = watch_sizeof(n),
> + .s.sb_id = s->s_unique_id,
> + .error_number = error,
> + .error_cookie = 0,
> + };
> +
> + post_sb_notification(s, &n.s);
> + }
> +#endif
> + return error;
> +}
next prev parent reply other threads:[~2020-02-19 23:08 UTC|newest]
Thread overview: 67+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-02-18 17:04 [PATCH 00/19] VFS: Filesystem information and notifications [ver #16] David Howells
2020-02-18 17:05 ` [PATCH 01/19] vfs: syscall: Add fsinfo() to query filesystem information " David Howells
2020-02-19 16:31 ` Darrick J. Wong
2020-02-19 20:07 ` Jann Horn
2020-02-20 10:34 ` David Howells
2020-02-20 15:48 ` Darrick J. Wong
2020-02-20 11:03 ` David Howells
2020-02-20 14:54 ` Jann Horn
2020-02-20 15:31 ` Darrick J. Wong
2020-02-18 17:05 ` [PATCH 02/19] fsinfo: Add syscalls to other arches " David Howells
2020-02-21 14:51 ` Christian Brauner
2020-02-21 18:10 ` Geert Uytterhoeven
2020-02-18 17:05 ` [PATCH 03/19] fsinfo: Provide a bitmap of supported features " David Howells
2020-02-19 16:37 ` Darrick J. Wong
2020-02-20 12:22 ` David Howells
2020-02-18 17:05 ` [PATCH 04/19] vfs: Add mount change counter " David Howells
2020-02-21 14:48 ` Christian Brauner
2020-02-18 17:05 ` [PATCH 05/19] vfs: Introduce a non-repeating system-unique superblock ID " David Howells
2020-02-19 16:53 ` Darrick J. Wong
2020-02-20 12:45 ` David Howells
2020-02-18 17:05 ` [PATCH 06/19] vfs: Allow fsinfo() to look up a mount object by " David Howells
2020-02-21 15:09 ` Christian Brauner
2020-02-18 17:05 ` [PATCH 07/19] vfs: Allow mount information to be queried by fsinfo() " David Howells
2020-02-18 17:05 ` [PATCH 08/19] vfs: fsinfo sample: Mount listing program " David Howells
2020-02-18 17:06 ` [PATCH 09/19] fsinfo: Allow the mount topology propogation flags to be retrieved " David Howells
2020-02-18 17:06 ` [PATCH 10/19] fsinfo: Add API documentation " David Howells
2020-02-18 17:06 ` [PATCH 11/19] afs: Support fsinfo() " David Howells
2020-02-19 21:01 ` Jann Horn
2020-02-20 12:58 ` David Howells
2020-02-20 14:58 ` Jann Horn
2020-02-21 13:26 ` David Howells
2020-02-18 17:06 ` [PATCH 12/19] security: Add hooks to rule on setting a superblock or mount watch " David Howells
2020-02-18 17:06 ` [PATCH 13/19] vfs: Add a mount-notification facility " David Howells
2020-02-19 22:40 ` Jann Horn
2020-02-19 22:55 ` Jann Horn
2020-02-21 12:24 ` David Howells
2020-02-21 15:49 ` Jann Horn
2020-02-21 17:06 ` David Howells
2020-02-21 17:36 ` seq_lock and lockdep_is_held() assertions Jann Horn
2020-02-21 18:02 ` John Stultz
2020-02-18 17:06 ` [PATCH 14/19] notifications: sample: Display mount tree change notifications [ver #16] David Howells
2020-02-18 17:06 ` [PATCH 15/19] vfs: Add superblock " David Howells
2020-02-19 23:08 ` Jann Horn [this message]
2020-02-21 14:23 ` David Howells
2020-02-21 15:44 ` Jann Horn
2020-02-21 16:33 ` David Howells
2020-02-21 16:41 ` Jann Horn
2020-02-21 17:11 ` David Howells
2020-02-18 17:06 ` [PATCH 16/19] fsinfo: Provide superblock notification counter " David Howells
2020-02-18 17:07 ` [PATCH 17/19] notifications: sample: Display superblock notifications " David Howells
2020-02-18 17:07 ` [PATCH 18/19] ext4: Add example fsinfo information " David Howells
2020-02-19 17:04 ` Darrick J. Wong
2020-02-20 0:53 ` kbuild test robot
2020-02-21 14:43 ` David Howells
2020-02-21 16:26 ` Darrick J. Wong
2020-02-18 17:07 ` [PATCH 19/19] nfs: Add example filesystem " David Howells
2020-02-20 2:13 ` kbuild test robot
2020-02-20 2:20 ` kbuild test robot
2020-02-18 18:12 ` David Howells
2020-02-19 10:23 ` [PATCH 00/19] VFS: Filesystem information and notifications " Stefan Metzmacher
2020-02-19 14:46 ` Christian Brauner
2020-02-19 15:50 ` Darrick J. Wong
2020-02-20 4:42 ` Ian Kent
2020-02-20 9:09 ` Christian Brauner
2020-02-20 11:30 ` Ian Kent
2020-02-19 16:16 ` David Howells
2020-02-21 12:57 ` David Howells
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='CAG48ez2B2J_3-+EjR20ukRu3noPnAccZsOTaea0jtKK4=+bkhQ@mail.gmail.com' \
--to=jannh@google.com \
--cc=christian@brauner.io \
--cc=dhowells@redhat.com \
--cc=linux-api@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mszeredi@redhat.com \
--cc=raven@themaw.net \
--cc=viro@zeniv.linux.org.uk \
/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;
as well as URLs for NNTP newsgroup(s).