kernel/fs/cachefiles/ondemand.c, branch 0x221E-v0.0.1-v6.19

cachefiles: Fix the incorrect return value in __cachefiles_write()

2025-07-10T07:40:17Z

In __cachefiles_write(), if the return value of the write operation > 0, it is set to 0. This makes it impossible to distinguish scenarios where a partial write has occurred, and will affect the outer calling functions: 1) cachefiles_write_complete() will call "term_func" such as netfs_write_subrequest_terminated(). When "ret" in __cachefiles_write() is used as the "transferred_or_error" of this function, it can not distinguish the amount of data written, makes the WARN meaningless. 2) cachefiles_ondemand_fd_write_iter() can only assume all writes were successful by default when "ret" is 0, and unconditionally return the full length specified by user space. Fix it by modifying "ret" to reflect the actual number of bytes written. Furthermore, returning a value greater than 0 from __cachefiles_write() does not affect other call paths, such as cachefiles_issue_write() and fscache_write(). Fixes: 047487c947e8 ("cachefiles: Implement the I/O routines") Signed-off-by: Zizhi Wo Link: https://lore.kernel.org/20250703024418.2809353-1-wozizhi@huaweicloud.com Signed-off-by: Christian Brauner

make use of anon_inode_getfile_fmode()

2025-02-21T09:25:31Z

["fallen through the cracks" misc stuff] A bunch of anon_inode_getfile() callers follow it with adjusting ->f_mode; we have a helper doing that now, so let's make use of it. Signed-off-by: Al Viro Link: https://lore.kernel.org/r/20250118014434.GT1977892@ZenIV Reviewed-by: Jan Kara Signed-off-by: Christian Brauner

cachefiles: Fix NULL pointer dereference in object->file

2024-11-11T13:39:38Z

At present, the object->file has the NULL pointer dereference problem in ondemand-mode. The root cause is that the allocated fd and object->file lifetime are inconsistent, and the user-space invocation to anon_fd uses object->file. Following is the process that triggers the issue: [write fd] [umount] cachefiles_ondemand_fd_write_iter fscache_cookie_state_machine cachefiles_withdraw_cookie if (!file) return -ENOBUFS cachefiles_clean_up_object cachefiles_unmark_inode_in_use fput(object->file) object->file = NULL // file NULL pointer dereference! __cachefiles_write(..., file, ...) Fix this issue by add an additional reference count to the object->file before write/llseek, and decrement after it finished. Fixes: c8383054506c ("cachefiles: notify the user daemon when looking up cookie") Signed-off-by: Zizhi Wo Link: https://lore.kernel.org/r/20241107110649.3980193-5-wozizhi@huawei.com Reviewed-by: David Howells Signed-off-by: Christian Brauner

cachefiles: Fix missing pos updates in cachefiles_ondemand_fd_write_iter()

2024-11-11T13:39:38Z

In the erofs on-demand loading scenario, read and write operations are usually delivered through "off" and "len" contained in read req in user mode. Naturally, pwrite is used to specify a specific offset to complete write operations. However, if the write(not pwrite) syscall is called multiple times in the read-ahead scenario, we need to manually update ki_pos after each write operation to update file->f_pos. This step is currently missing from the cachefiles_ondemand_fd_write_iter function, added to address this issue. Fixes: c8383054506c ("cachefiles: notify the user daemon when looking up cookie") Signed-off-by: Zizhi Wo Link: https://lore.kernel.org/r/20241107110649.3980193-3-wozizhi@huawei.com Acked-by: David Howells Signed-off-by: Christian Brauner

cachefiles: Fix incorrect length return value in cachefiles_ondemand_fd_write_iter()

2024-11-11T13:39:37Z

cachefiles_ondemand_fd_write_iter() function first aligns "pos" and "len" to block boundaries. When calling __cachefiles_write(), the aligned "pos" is passed in, but "len" is the original unaligned value(iter->count). Additionally, the returned length of the write operation is the modified "len" aligned by block size, which is unreasonable. The alignment of "pos" and "len" is intended only to check whether the cache has enough space. But the modified len should not be used as the return value of cachefiles_ondemand_fd_write_iter() because the length we passed to __cachefiles_write() is the previous "len". Doing so would result in a mismatch in the data written on-demand. For example, if the length of the user state passed in is not aligned to the block size (the preread scene/DIO writes only need 512 alignment/Fault injection), the length of the write will differ from the actual length of the return. To solve this issue, since the __cachefiles_prepare_write() modifies the size of "len", we pass "aligned_len" to __cachefiles_prepare_write() to calculate the free blocks and use the original "len" as the return value of cachefiles_ondemand_fd_write_iter(). Fixes: c8383054506c ("cachefiles: notify the user daemon when looking up cookie") Signed-off-by: Zizhi Wo Link: https://lore.kernel.org/r/20241107110649.3980193-2-wozizhi@huawei.com Reviewed-by: David Howells Signed-off-by: Christian Brauner

Merge patch series "cachefiles: random bugfixes"

2024-07-05T16:40:40Z

libaokun@huaweicloud.com says: This is the third version of this patch series, in which another patch set is subsumed into this one to avoid confusing the two patch sets. (https://patchwork.kernel.org/project/linux-fsdevel/list/?series=854914) We've been testing ondemand mode for cachefiles since January, and we're almost done. We hit a lot of issues during the testing period, and this patch series fixes some of the issues. The patches have passed internal testing without regression. The following is a brief overview of the patches, see the patches for more details. Patch 1-2: Add fscache_try_get_volume() helper function to avoid fscache_volume use-after-free on cache withdrawal. Patch 3: Fix cachefiles_lookup_cookie() and cachefiles_withdraw_cache() concurrency causing cachefiles_volume use-after-free. Patch 4: Propagate error codes returned by vfs_getxattr() to avoid endless loops. Patch 5-7: A read request waiting for reopen could be closed maliciously before the reopen worker is executing or waiting to be scheduled. So ondemand_object_worker() may be called after the info and object and even the cache have been freed and trigger use-after-free. So use cancel_work_sync() in cachefiles_ondemand_clean_object() to cancel the reopen worker or wait for it to finish. Since it makes no sense to wait for the daemon to complete the reopen request, to avoid this pointless operation blocking cancel_work_sync(), Patch 1 avoids request generation by the DROPPING state when the request has not been sent, and Patch 2 flushes the requests of the current object before cancel_work_sync(). Patch 8: Cyclic allocation of msg_id to avoid msg_id reuse misleading the daemon to cause hung. Patch 9: Hold xas_lock during polling to avoid dereferencing reqs causing use-after-free. This issue was triggered frequently in our tests, and we found that anolis 5.10 had fixed it. So to avoid failing the test, this patch is pushed upstream as well. Baokun Li (7): netfs, fscache: export fscache_put_volume() and add fscache_try_get_volume() cachefiles: fix slab-use-after-free in fscache_withdraw_volume() cachefiles: fix slab-use-after-free in cachefiles_withdraw_cookie() cachefiles: propagate errors from vfs_getxattr() to avoid infinite loop cachefiles: stop sending new request when dropping object cachefiles: cancel all requests for the object that is being dropped cachefiles: cyclic allocation of msg_id to avoid reuse Hou Tao (1): cachefiles: wait for ondemand_object_worker to finish when dropping object Jingbo Xu (1): cachefiles: add missing lock protection when polling fs/cachefiles/cache.c | 45 ++++++++++++++++++++++++++++- fs/cachefiles/daemon.c | 4 +-- fs/cachefiles/internal.h | 3 ++ fs/cachefiles/ondemand.c | 52 ++++++++++++++++++++++++++++++---- fs/cachefiles/volume.c | 1 - fs/cachefiles/xattr.c | 5 +++- fs/netfs/fscache_volume.c | 14 +++++++++ fs/netfs/internal.h | 2 -- include/linux/fscache-cache.h | 6 ++++ include/trace/events/fscache.h | 4 +++ 10 files changed, 123 insertions(+), 13 deletions(-) Link: https://lore.kernel.org/r/20240628062930.2467993-1-libaokun@huaweicloud.com Signed-off-by: Christian Brauner

cachefiles: cyclic allocation of msg_id to avoid reuse

2024-07-03T08:36:16Z

Reusing the msg_id after a maliciously completed reopen request may cause a read request to remain unprocessed and result in a hung, as shown below: t1 | t2 | t3 ------------------------------------------------- cachefiles_ondemand_select_req cachefiles_ondemand_object_is_close(A) cachefiles_ondemand_set_object_reopening(A) queue_work(fscache_object_wq, &info->work) ondemand_object_worker cachefiles_ondemand_init_object(A) cachefiles_ondemand_send_req(OPEN) // get msg_id 6 wait_for_completion(&req_A->done) cachefiles_ondemand_daemon_read // read msg_id 6 req_A cachefiles_ondemand_get_fd copy_to_user // Malicious completion msg_id 6 copen 6,-1 cachefiles_ondemand_copen complete(&req_A->done) // will not set the object to close // because ondemand_id && fd is valid. // ondemand_object_worker() is done // but the object is still reopening. // new open req_B cachefiles_ondemand_init_object(B) cachefiles_ondemand_send_req(OPEN) // reuse msg_id 6 process_open_req copen 6,A.size // The expected failed copen was executed successfully Expect copen to fail, and when it does, it closes fd, which sets the object to close, and then close triggers reopen again. However, due to msg_id reuse resulting in a successful copen, the anonymous fd is not closed until the daemon exits. Therefore read requests waiting for reopen to complete may trigger hung task. To avoid this issue, allocate the msg_id cyclically to avoid reusing the msg_id for a very short duration of time. Fixes: c8383054506c ("cachefiles: notify the user daemon when looking up cookie") Signed-off-by: Baokun Li Link: https://lore.kernel.org/r/20240628062930.2467993-9-libaokun@huaweicloud.com Acked-by: Jeff Layton Reviewed-by: Gao Xiang Reviewed-by: Jia Zhu Signed-off-by: Christian Brauner

cachefiles: wait for ondemand_object_worker to finish when dropping object

2024-07-03T08:36:16Z

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 Signed-off-by: Baokun Li Link: https://lore.kernel.org/r/20240628062930.2467993-8-libaokun@huaweicloud.com Acked-by: Jeff Layton Reviewed-by: Jia Zhu Reviewed-by: Gao Xiang Signed-off-by: Christian Brauner

cachefiles: cancel all requests for the object that is being dropped

2024-07-03T08:36:15Z

Because after an object is dropped, requests for that object are useless, cancel them to avoid causing other problems. This prepares for the later addition of cancel_work_sync(). After the reopen requests is generated, cancel it to avoid cancel_work_sync() blocking by waiting for daemon to complete the reopen requests. Signed-off-by: Baokun Li Link: https://lore.kernel.org/r/20240628062930.2467993-7-libaokun@huaweicloud.com Acked-by: Jeff Layton Reviewed-by: Gao Xiang Reviewed-by: Jia Zhu Signed-off-by: Christian Brauner

cachefiles: stop sending new request when dropping object

2024-07-03T08:36:15Z

Added CACHEFILES_ONDEMAND_OBJSTATE_DROPPING indicates that the cachefiles object is being dropped, and is set after the close request for the dropped object completes, and no new requests are allowed to be sent after this state. This prepares for the later addition of cancel_work_sync(). It prevents leftover reopen requests from being sent, to avoid processing unnecessary requests and to avoid cancel_work_sync() blocking by waiting for daemon to complete the reopen requests. Signed-off-by: Baokun Li Link: https://lore.kernel.org/r/20240628062930.2467993-6-libaokun@huaweicloud.com Acked-by: Jeff Layton Reviewed-by: Gao Xiang Reviewed-by: Jia Zhu Signed-off-by: Christian Brauner