linux/fs/ceph
Patrick Donnelly ccda9910d8 ceph: fix cap ref leak via netfs init_request
Log recovered from a user's cluster:

    <7>[ 5413.970692] ceph:  get_cap_refs 00000000958c114b ret 1 got Fr
    <7>[ 5413.970695] ceph:  start_read 00000000958c114b, no cache cap
    ...
    <7>[ 5473.934609] ceph:   my wanted = Fr, used = Fr, dirty -
    <7>[ 5473.934616] ceph:  revocation: pAsLsXsFr -> pAsLsXs (revoking Fr)
    <7>[ 5473.934632] ceph:  __ceph_caps_issued 00000000958c114b cap 00000000f7784259 issued pAsLsXs
    <7>[ 5473.934638] ceph:  check_caps 10000000e68.fffffffffffffffe file_want - used Fr dirty - flushing - issued pAsLsXs revoking Fr retain pAsLsXsFsr  AUTHONLY NOINVAL FLUSH_FORCE

The MDS subsequently complains that the kernel client is late releasing
caps.

Approximately, a series of changes to this code by commits 4987005600
("ceph: convert ceph_readpages to ceph_readahead"), 2de1604173
("netfs: Change ->init_request() to return an error code") and
a5c9dc4451 ("ceph: Make ceph_init_request() check caps on readahead")
resulted in subtle resource cleanup to be missed. The main culprit is
the change in error handling in 2de1604173 which meant that a failure
in init_request() would no longer cause cleanup to be called. That
would prevent the ceph_put_cap_refs() call which would cleanup the
leaked cap ref.

Cc: stable@vger.kernel.org
Fixes: a5c9dc4451 ("ceph: Make ceph_init_request() check caps on readahead")
Link: https://tracker.ceph.com/issues/67008
Signed-off-by: Patrick Donnelly <pdonnell@redhat.com>
Reviewed-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
2024-10-03 09:31:08 +02:00
..
acl.c ceph: allow idmapped set_acl inode op 2023-11-03 23:28:34 +01:00
addr.c ceph: fix cap ref leak via netfs init_request 2024-10-03 09:31:08 +02:00
cache.c ceph: rename _to_client() to _to_fs_client() 2023-11-03 23:28:33 +01:00
cache.h netfs: Provide invalidate_folio and release_folio calls 2023-12-24 15:08:51 +00:00
caps.c ceph: Fix typo in the comment 2024-09-24 22:51:33 +02:00
ceph_frag.c
crypto.c Two items: 2023-11-10 09:52:56 -08:00
crypto.h ceph: add support for encrypted snapshot names 2023-08-24 11:24:36 +02:00
debugfs.c ceph: print cluster fsid and client global_id in all debug logs 2023-11-03 23:28:33 +01:00
dir.c Three CephFS fixes from Xiubo and Luis and a bunch of assorted 2024-09-28 08:40:36 -07:00
export.c ceph: d_obtain_{alias,root}(ERR_PTR(...)) will do the right thing 2024-01-15 15:40:51 +01:00
file.c ceph: check the cephx mds auth access for async dirop 2024-05-23 10:35:47 +02:00
inode.c Three CephFS fixes from Xiubo and Luis and a bunch of assorted 2024-09-28 08:40:36 -07:00
io.c
io.h
ioctl.c ceph: print cluster fsid and client global_id in all debug logs 2023-11-03 23:28:33 +01:00
ioctl.h
Kconfig ceph: select FS_ENCRYPTION_ALGS if FS_ENCRYPTION 2024-01-15 15:40:50 +01:00
locks.c ceph: adapt to breakup of struct file_lock 2024-02-05 13:11:42 +01:00
Makefile ceph: fscrypt_auth handling for ceph 2023-08-22 09:01:48 +02:00
mds_client.c ceph: Fix typo in the comment 2024-09-24 22:51:33 +02:00
mds_client.h ceph: rename ceph_flush_cap_releases() to ceph_flush_session_cap_releases() 2024-09-24 22:51:19 +02:00
mdsmap.c ceph: switch to corrected encoding of max_xattr_size in mdsmap 2024-02-26 19:20:30 +01:00
mdsmap.h ceph: switch to corrected encoding of max_xattr_size in mdsmap 2024-02-26 19:20:30 +01:00
metric.c ceph: print cluster fsid and client global_id in all debug logs 2023-11-03 23:28:33 +01:00
metric.h
quota.c ceph: fix invalid pointer access if get_quota_realm return ERR_PTR 2024-01-15 15:40:51 +01:00
snap.c Two items: 2023-11-10 09:52:56 -08:00
strings.c
super.c ceph: flush all caps releases when syncing the whole filesystem 2024-09-24 22:51:28 +02:00
super.h ceph: Remove empty definition in header file 2024-09-24 22:51:33 +02:00
util.c
xattr.c Two items: 2023-11-10 09:52:56 -08:00