linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Amir Goldstein <amir73il@gmail.com>
To: Christian Brauner <brauner@kernel.org>
Cc: Jan Kara <jack@suse.cz>,
	Alex Williamson <alex.williamson@redhat.com>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	linux-fsdevel@vger.kernel.org
Subject: [PATCH 2/3] fsnotify: disable notification by default for all pseudo files
Date: Mon,  3 Feb 2025 23:32:04 +0100	[thread overview]
Message-ID: <20250203223205.861346-3-amir73il@gmail.com> (raw)
In-Reply-To: <20250203223205.861346-1-amir73il@gmail.com>

Most pseudo files are not applicable for fsnotify events at all,
let alone to the new pre-content events.

Disable notifications to all files allocated with alloc_file_pseudo()
and enable legacy inotify events for the specific cases of pipe and
socket, which have known users of inotify events.

Pre-content events are also kept disabled for sockets and pipes.

Fixes: 20bf82a898b6 ("mm: don't allow huge faults for files with pre content watches")
Reported-by: Alex Williamson <alex.williamson@redhat.com>
Closes: https://lore.kernel.org/linux-fsdevel/20250131121703.1e4d00a7.alex.williamson@redhat.com/
Suggested-by: Linus Torvalds <torvalds@linux-foundation.org>
Link: https://lore.kernel.org/linux-fsdevel/CAHk-=wi2pThSVY=zhO=ZKxViBj5QCRX-=AS2+rVknQgJnHXDFg@mail.gmail.com/
Tested-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
---
 fs/file_table.c | 11 +++++++++++
 fs/open.c       |  4 ++--
 fs/pipe.c       |  6 ++++++
 net/socket.c    |  5 +++++
 4 files changed, 24 insertions(+), 2 deletions(-)

diff --git a/fs/file_table.c b/fs/file_table.c
index f0291a66f9db4..35b93da6c5cb1 100644
--- a/fs/file_table.c
+++ b/fs/file_table.c
@@ -375,7 +375,13 @@ struct file *alloc_file_pseudo(struct inode *inode, struct vfsmount *mnt,
 	if (IS_ERR(file)) {
 		ihold(inode);
 		path_put(&path);
+		return file;
 	}
+	/*
+	 * Disable all fsnotify events for pseudo files by default.
+	 * They may be enabled by caller with file_set_fsnotify_mode().
+	 */
+	file_set_fsnotify_mode(file, FMODE_NONOTIFY);
 	return file;
 }
 EXPORT_SYMBOL(alloc_file_pseudo);
@@ -400,6 +406,11 @@ struct file *alloc_file_pseudo_noaccount(struct inode *inode,
 		return file;
 	}
 	file_init_path(file, &path, fops);
+	/*
+	 * Disable all fsnotify events for pseudo files by default.
+	 * They may be enabled by caller with file_set_fsnotify_mode().
+	 */
+	file_set_fsnotify_mode(file, FMODE_NONOTIFY);
 	return file;
 }
 EXPORT_SYMBOL_GPL(alloc_file_pseudo_noaccount);
diff --git a/fs/open.c b/fs/open.c
index 3fcbfff8aede8..1be20de9f283a 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -936,8 +936,8 @@ static int do_dentry_open(struct file *f,
 
 	/*
 	 * Set FMODE_NONOTIFY_* bits according to existing permission watches.
-	 * If FMODE_NONOTIFY was already set for an fanotify fd, this doesn't
-	 * change anything.
+	 * If FMODE_NONOTIFY mode was already set for an fanotify fd or for a
+	 * pseudo file, this call will not change the mode.
 	 */
 	file_set_fsnotify_mode_from_watchers(f);
 	error = fsnotify_open_perm(f);
diff --git a/fs/pipe.c b/fs/pipe.c
index 94b59045ab44b..ce1af7592780d 100644
--- a/fs/pipe.c
+++ b/fs/pipe.c
@@ -960,6 +960,12 @@ int create_pipe_files(struct file **res, int flags)
 	res[1] = f;
 	stream_open(inode, res[0]);
 	stream_open(inode, res[1]);
+	/*
+	 * Disable permission and pre-content events, but enable legacy
+	 * inotify events for legacy users.
+	 */
+	file_set_fsnotify_mode(res[0], FMODE_NONOTIFY_PERM);
+	file_set_fsnotify_mode(res[1], FMODE_NONOTIFY_PERM);
 	return 0;
 }
 
diff --git a/net/socket.c b/net/socket.c
index 262a28b59c7f0..28bae5a942341 100644
--- a/net/socket.c
+++ b/net/socket.c
@@ -479,6 +479,11 @@ struct file *sock_alloc_file(struct socket *sock, int flags, const char *dname)
 	sock->file = file;
 	file->private_data = sock;
 	stream_open(SOCK_INODE(sock), file);
+	/*
+	 * Disable permission and pre-content events, but enable legacy
+	 * inotify events for legacy users.
+	 */
+	file_set_fsnotify_mode(file, FMODE_NONOTIFY_PERM);
 	return file;
 }
 EXPORT_SYMBOL(sock_alloc_file);
-- 
2.34.1


  parent reply	other threads:[~2025-02-03 22:32 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-02-03 22:32 [PATCH 0/3] Fix for huge faults regression Amir Goldstein
2025-02-03 22:32 ` [PATCH 1/3] fsnotify: use accessor to set FMODE_NONOTIFY_* Amir Goldstein
2025-02-04 10:43   ` Christian Brauner
2025-02-04 10:58     ` Amir Goldstein
2025-02-05 16:52     ` Jan Kara
2025-02-03 22:32 ` Amir Goldstein [this message]
2025-02-05 16:52   ` [PATCH 2/3] fsnotify: disable notification by default for all pseudo files Jan Kara
2025-02-03 22:32 ` [PATCH 3/3] fsnotify: disable pre-content and permission events by default Amir Goldstein
2025-02-05 16:59   ` Jan Kara
2025-02-05 22:20     ` Amir Goldstein
2025-02-05 13:12 ` [PATCH 0/3] Fix for huge faults regression 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=20250203223205.861346-3-amir73il@gmail.com \
    --to=amir73il@gmail.com \
    --cc=alex.williamson@redhat.com \
    --cc=brauner@kernel.org \
    --cc=jack@suse.cz \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=torvalds@linux-foundation.org \
    /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).