[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20240426200941.GP2118490@ZenIV>
Date: Fri, 26 Apr 2024 21:09:41 +0100
From: Al Viro <viro@...iv.linux.org.uk>
To: Lucas Karpinski <lkarpins@...hat.com>
Cc: brauner@...nel.org, jack@...e.cz, linux-fsdevel@...r.kernel.org,
linux-kernel@...r.kernel.org, alexl@...hat.com, echanude@...hat.com,
ikent@...hat.com
Subject: Re: [RFC v2 1/1] fs/namespace: defer RCU sync for MNT_DETACH umount
On Fri, Apr 26, 2024 at 03:53:48PM -0400, Lucas Karpinski wrote:
> -static void namespace_unlock(void)
> +static void free_mounts(struct hlist_head *mount_list)
> {
> - struct hlist_head head;
> struct hlist_node *p;
> struct mount *m;
> +
> + hlist_for_each_entry_safe(m, p, mount_list, mnt_umount) {
> + hlist_del(&m->mnt_umount);
> + mntput(&m->mnt);
.. which may block in quite a few ways.
> + }
> +}
> +
> +static void delayed_mount_release(struct rcu_head *head)
> +{
> + struct mount_delayed_release *drelease =
> + container_of(head, struct mount_delayed_release, rcu);
> +
> + free_mounts(&drelease->release_list);
.. and therefore so can this.
> + kfree(drelease);
> +}
> + call_rcu(&drelease->rcu, delayed_mount_release);
.. which is a bad idea, since call_rcu() callbacks are run
from interrupt context. Which makes blocking in them a problem.
Powered by blists - more mailing lists