public inbox for linux-fsdevel@vger.kernel.org
 help / color / mirror / Atom feed
From: Christian Brauner <brauner@kernel.org>
To: linux-fsdevel@vger.kernel.org
Cc: Linus Torvalds <torvalds@linux-foundation.org>,
	 linux-kernel@vger.kernel.org,
	Alexander Viro <viro@zeniv.linux.org.uk>,
	 Jens Axboe <axboe@kernel.dk>, Jan Kara <jack@suse.cz>,
	 Tejun Heo <tj@kernel.org>, Jann Horn <jannh@google.com>,
	 Christian Brauner <brauner@kernel.org>
Subject: [PATCH RFC v3 24/26] fs: start all kthreads in nullfs
Date: Wed, 11 Mar 2026 22:44:07 +0100	[thread overview]
Message-ID: <20260311-work-kthread-nullfs-v3-24-3dd2cbe92ad0@kernel.org> (raw)
In-Reply-To: <20260311-work-kthread-nullfs-v3-0-3dd2cbe92ad0@kernel.org>

Point init_task's fs_struct (root and pwd) at a private nullfs instance
instead of the mutable rootfs. All kthreads now start isolated in nullfs
and must use scoped_with_init_fs() for any path resolution.

PID 1 is moved from nullfs into the initramfs by init_userspace_fs().
Usermodehelper threads use userspace_init_fs via the umh flag in
copy_fs(). All subsystems that need init's filesystem state for path
resolution already use scoped_with_init_fs() from earlier commits in
this series.

This isolates kthreads from userspace filesystem state and makes it
hard to perform filesystem operations from kthread context.

Signed-off-by: Christian Brauner <brauner@kernel.org>
---
 fs/namespace.c | 16 ++++++++++------
 1 file changed, 10 insertions(+), 6 deletions(-)

diff --git a/fs/namespace.c b/fs/namespace.c
index e23d2fa7e255..5d318e2e1e4a 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -6143,12 +6143,14 @@ static void __init init_mount_tree(void)
 	struct path root;
 
 	/*
-	 * We create two mounts:
+	 * We create three mounts:
 	 *
 	 * (1) nullfs with mount id 1
 	 * (2) mutable rootfs with mount id 2
+	 * (3) private nullfs for kthreads (SB_KERNMOUNT)
 	 *
-	 * with (2) mounted on top of (1).
+	 * with (2) mounted on top of (1). The init_task's root and pwd
+	 * are pointed at (3) so all kthreads start isolated in nullfs.
 	 */
 	nullfs_mnt = vfs_kern_mount(&nullfs_fs_type, 0, "nullfs", NULL);
 	if (IS_ERR(nullfs_mnt))
@@ -6188,12 +6190,14 @@ static void __init init_mount_tree(void)
 		init_mnt_ns.nr_mounts++;
 	}
 
+	nullfs_mnt = kern_mount(&nullfs_fs_type);
+	if (IS_ERR(nullfs_mnt))
+		panic("VFS: Failed to create private nullfs instance");
+	root.mnt	= nullfs_mnt;
+	root.dentry	= nullfs_mnt->mnt_root;
+
 	init_task.nsproxy->mnt_ns = &init_mnt_ns;
 	get_mnt_ns(&init_mnt_ns);
-
-	/* The root and pwd always point to the mutable rootfs. */
-	root.mnt	= mnt;
-	root.dentry	= mnt->mnt_root;
 	set_fs_pwd(current->fs, &root);
 	set_fs_root(current->fs, &root);
 

-- 
2.47.3


  parent reply	other threads:[~2026-03-11 21:57 UTC|newest]

Thread overview: 28+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-03-11 21:43 [PATCH RFC v3 00/26] fs,kthread: start all kthreads in nullfs Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 01/26] fs: add switch_fs_struct() Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 02/26] fs: notice when init abandons fs sharing Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 03/26] fs: add scoped_with_init_fs() Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 04/26] fs: add real_fs to track task's actual fs_struct Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 05/26] fs: make userspace_init_fs a dynamically-initialized pointer Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 06/26] rnbd: use scoped_with_init_fs() for block device open Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 07/26] crypto: ccp: use scoped_with_init_fs() for SEV file access Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 08/26] scsi: target: use scoped_with_init_fs() for ALUA metadata Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 09/26] scsi: target: use scoped_with_init_fs() for APTPL metadata Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 10/26] btrfs: use scoped_with_init_fs() for update_dev_time() Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 11/26] coredump: use scoped_with_init_fs() for coredump path resolution Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 12/26] fs: use scoped_with_init_fs() for kernel_read_file_from_path_initns() Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 13/26] ksmbd: use scoped_with_init_fs() for share path resolution Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 14/26] ksmbd: use scoped_with_init_fs() for filesystem info path lookup Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 15/26] ksmbd: use scoped_with_init_fs() for VFS path operations Christian Brauner
2026-03-11 21:43 ` [PATCH RFC v3 16/26] pnfs/blocklayout: use scoped_with_init_fs() for SCSI device lookup Christian Brauner
2026-03-11 21:44 ` [PATCH RFC v3 17/26] initramfs: use scoped_with_init_fs() for rootfs unpacking Christian Brauner
2026-03-11 21:44 ` [PATCH RFC v3 18/26] af_unix: use scoped_with_init_fs() for coredump socket lookup Christian Brauner
2026-03-11 21:44 ` [PATCH RFC v3 19/26] fs: stop sharing fs_struct between init_task and pid 1 Christian Brauner
2026-03-11 21:44 ` [PATCH RFC v3 20/26] fs: add umh argument to struct kernel_clone_args Christian Brauner
2026-03-11 21:44 ` [PATCH RFC v3 21/26] fs: add kthread_mntns() Christian Brauner
2026-03-11 22:13   ` Thomas Weißschuh
2026-03-11 21:44 ` [PATCH RFC v3 22/26] devtmpfs: create private mount namespace Christian Brauner
2026-03-11 21:44 ` [PATCH RFC v3 23/26] nullfs: make nullfs multi-instance Christian Brauner
2026-03-11 21:44 ` Christian Brauner [this message]
2026-03-11 21:44 ` [PATCH RFC v3 25/26] fs: stop rewriting kthread fs structs Christian Brauner
2026-03-11 21:44 ` [PATCH RFC v3 26/26] fs: stop rewriting paths for PF_EXITING | PF_DUMPCORE Christian Brauner

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=20260311-work-kthread-nullfs-v3-24-3dd2cbe92ad0@kernel.org \
    --to=brauner@kernel.org \
    --cc=axboe@kernel.dk \
    --cc=jack@suse.cz \
    --cc=jannh@google.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=tj@kernel.org \
    --cc=torvalds@linux-foundation.org \
    --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