From: libaokun@huaweicloud.com
To: netfs@lists.linux.dev, dhowells@redhat.com, jlayton@kernel.org
Cc: hsiangkao@linux.alibaba.com, jefflexu@linux.alibaba.com,
zhujia.zj@bytedance.com, linux-erofs@lists.ozlabs.org,
brauner@kernel.org, linux-fsdevel@vger.kernel.org,
linux-kernel@vger.kernel.org, libaokun@huaweicloud.com,
yangerkun@huawei.com, houtao1@huawei.com, yukuai3@huawei.com,
wozizhi@huawei.com, Baokun Li <libaokun1@huawei.com>
Subject: [PATCH v3 7/9] cachefiles: wait for ondemand_object_worker to finish when dropping object
Date: Fri, 28 Jun 2024 14:29:28 +0800 [thread overview]
Message-ID: <20240628062930.2467993-8-libaokun@huaweicloud.com> (raw)
In-Reply-To: <20240628062930.2467993-1-libaokun@huaweicloud.com>
From: Hou Tao <houtao1@huawei.com>
When queuing ondemand_object_worker() to re-open the object,
cachefiles_object is not pinned. The cachefiles_object may be freed when
the pending read request is completed intentionally and the related
erofs is umounted. If ondemand_object_worker() runs after the object is
freed, it will incur use-after-free problem as shown below.
process A processs B process C process D
cachefiles_ondemand_send_req()
// send a read req X
// wait for its completion
// close ondemand fd
cachefiles_ondemand_fd_release()
// set object as CLOSE
cachefiles_ondemand_daemon_read()
// set object as REOPENING
queue_work(fscache_wq, &info->ondemand_work)
// close /dev/cachefiles
cachefiles_daemon_release
cachefiles_flush_reqs
complete(&req->done)
// read req X is completed
// umount the erofs fs
cachefiles_put_object()
// object will be freed
cachefiles_ondemand_deinit_obj_info()
kmem_cache_free(object)
// both info and object are freed
ondemand_object_worker()
When dropping an object, it is no longer necessary to reopen the object,
so use cancel_work_sync() to cancel or wait for ondemand_object_worker()
to finish.
Fixes: 0a7e54c1959c ("cachefiles: resend an open request if the read request's object is closed")
Signed-off-by: Hou Tao <houtao1@huawei.com>
Signed-off-by: Baokun Li <libaokun1@huawei.com>
Reviewed-by: Jia Zhu <zhujia.zj@bytedance.com>
Acked-by: Jeff Layton <jlayton@kernel.org>
---
fs/cachefiles/ondemand.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/fs/cachefiles/ondemand.c b/fs/cachefiles/ondemand.c
index 36b97ded16b4..1d5b970206d0 100644
--- a/fs/cachefiles/ondemand.c
+++ b/fs/cachefiles/ondemand.c
@@ -694,6 +694,9 @@ void cachefiles_ondemand_clean_object(struct cachefiles_object *object)
}
}
xa_unlock(&cache->reqs);
+
+ /* Wait for ondemand_object_worker() to finish to avoid UAF. */
+ cancel_work_sync(&object->ondemand->ondemand_work);
}
int cachefiles_ondemand_init_obj_info(struct cachefiles_object *object,
--
2.39.2
next prev parent reply other threads:[~2024-06-28 6:31 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-28 6:29 [PATCH v3 0/9] cachefiles: random bugfixes libaokun
2024-06-28 6:29 ` [PATCH v3 1/9] netfs, fscache: export fscache_put_volume() and add fscache_try_get_volume() libaokun
2024-06-28 6:29 ` [PATCH v3 2/9] cachefiles: fix slab-use-after-free in fscache_withdraw_volume() libaokun
2024-06-28 6:29 ` [PATCH v3 3/9] cachefiles: fix slab-use-after-free in cachefiles_withdraw_cookie() libaokun
2024-06-28 6:29 ` [PATCH v3 4/9] cachefiles: propagate errors from vfs_getxattr() to avoid infinite loop libaokun
2024-06-28 7:30 ` Gao Xiang
2024-06-28 6:29 ` [PATCH v3 5/9] cachefiles: stop sending new request when dropping object libaokun
2024-06-28 6:51 ` Gao Xiang
2024-07-02 12:29 ` [External] " Jia Zhu
2024-06-28 6:29 ` [PATCH v3 6/9] cachefiles: cancel all requests for the object that is being dropped libaokun
2024-06-28 7:21 ` Gao Xiang
2024-07-02 12:31 ` [External] " Jia Zhu
2024-06-28 6:29 ` libaokun [this message]
2024-06-28 7:22 ` [PATCH v3 7/9] cachefiles: wait for ondemand_object_worker to finish when dropping object Gao Xiang
2024-06-28 6:29 ` [PATCH v3 8/9] cachefiles: cyclic allocation of msg_id to avoid reuse libaokun
2024-07-02 12:34 ` [External] " Jia Zhu
2024-06-28 6:29 ` [PATCH v3 9/9] cachefiles: add missing lock protection when polling libaokun
2024-06-28 7:39 ` [PATCH v3 0/9] cachefiles: random bugfixes Gao Xiang
2024-06-28 11:37 ` Baokun Li
2024-07-02 12:25 ` Baokun Li
2024-07-03 8:30 ` (subset) " Christian Brauner
2024-07-03 8:37 ` 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=20240628062930.2467993-8-libaokun@huaweicloud.com \
--to=libaokun@huaweicloud.com \
--cc=brauner@kernel.org \
--cc=dhowells@redhat.com \
--cc=houtao1@huawei.com \
--cc=hsiangkao@linux.alibaba.com \
--cc=jefflexu@linux.alibaba.com \
--cc=jlayton@kernel.org \
--cc=libaokun1@huawei.com \
--cc=linux-erofs@lists.ozlabs.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=netfs@lists.linux.dev \
--cc=wozizhi@huawei.com \
--cc=yangerkun@huawei.com \
--cc=yukuai3@huawei.com \
--cc=zhujia.zj@bytedance.com \
/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).