From: Shigeru Yoshida <syoshida-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
To: konishi.ryusuke-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
Cc: linux-nilfs-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
syzkaller-bugs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org,
Shigeru Yoshida
<syoshida-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>,
syzbot+f816fa82f8783f7a02bb-Pl5Pbv+GP7P466ipTTIvnc23WoclnBCfAL8bYrjMMd8@public.gmane.org
Subject: [PATCH] nilfs: Avoid use-after-free caused by nilfs->ns_writer
Date: Thu, 3 Nov 2022 23:17:59 +0900 [thread overview]
Message-ID: <20221103141759.1836312-1-syoshida@redhat.com> (raw)
syzbot reported use-after-free in nilfs_segctor_sync() [1].
The use-after-free occurs with nilfs->ns_writer. The scenario for the
issue is as follows:
Task1 Task2
----------------------------------------------------------------------
nilfs_construct_segment
nilfs_segctor_sync
init_wait
init_waitqueue_entry
add_wait_queue
schedule
nilfs_detach_log_writer
nilfs_segctor_destroy
kfree
finish_wait
_raw_spin_lock_irqsave
__raw_spin_lock_irqsave
do_raw_spin_lock
debug_spin_lock_before <-- use-after-free
While Task1 is sleeping, nilfs->ns_writer is freed by Task2. After
Task1 waked up, Task1 accesses nilfs->ns_writer which is already
freed.
This patch fixes the issue by taking nilfs->ns_segctor_sem in
nilfs_construct_segment() so that nilfs->ns_segctor_sem cannot be
freed while nilfs_segctor_sync() is sleeping.
Link: https://syzkaller.appspot.com/bug?id=79a4c002e960419ca173d55e863bd09e8112df8b [1]
Reported-by: syzbot+f816fa82f8783f7a02bb-Pl5Pbv+GP7P466ipTTIvnc23WoclnBCfAL8bYrjMMd8@public.gmane.org
Signed-off-by: Shigeru Yoshida <syoshida-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
---
fs/nilfs2/segment.c | 14 +++++++++++---
1 file changed, 11 insertions(+), 3 deletions(-)
diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c
index b4cebad21b48..d4f10d82664d 100644
--- a/fs/nilfs2/segment.c
+++ b/fs/nilfs2/segment.c
@@ -2239,16 +2239,24 @@ static void nilfs_segctor_wakeup(struct nilfs_sc_info *sci, int err)
int nilfs_construct_segment(struct super_block *sb)
{
struct the_nilfs *nilfs = sb->s_fs_info;
- struct nilfs_sc_info *sci = nilfs->ns_writer;
+ struct nilfs_sc_info *sci;
struct nilfs_transaction_info *ti;
+ int ret;
- if (!sci)
+ down_write(&nilfs->ns_segctor_sem);
+ sci = nilfs->ns_writer;
+ if (!sci) {
+ up_write(&nilfs->ns_segctor_sem);
return -EROFS;
+ }
/* A call inside transactions causes a deadlock. */
BUG_ON((ti = current->journal_info) && ti->ti_magic == NILFS_TI_MAGIC);
- return nilfs_segctor_sync(sci);
+ ret = nilfs_segctor_sync(sci);
+ up_write(&nilfs->ns_segctor_sem);
+
+ return ret;
}
/**
--
2.37.3
next reply other threads:[~2022-11-03 14:17 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-11-03 14:17 Shigeru Yoshida [this message]
[not found] ` <20221103141759.1836312-1-syoshida-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
2022-11-03 18:56 ` [PATCH] nilfs: Avoid use-after-free caused by nilfs->ns_writer Ryusuke Konishi
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=20221103141759.1836312-1-syoshida@redhat.com \
--to=syoshida-h+wxahxf7alqt0dzr+alfa@public.gmane.org \
--cc=konishi.ryusuke-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org \
--cc=linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=linux-nilfs-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=syzbot+f816fa82f8783f7a02bb-Pl5Pbv+GP7P466ipTTIvnc23WoclnBCfAL8bYrjMMd8@public.gmane.org \
--cc=syzkaller-bugs-/JYPxA39Uh5TLH3MbocFFw@public.gmane.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 read-only IMAP folder(s) and NNTP newsgroup(s).