linux.git/fs/ceph/caps.c, branch v5.13

Merge tag 'ceph-for-5.13-rc1' of git://github.com/ceph/ceph-client

2021-05-06T17:27:02+00:00

Pull ceph updates from Ilya Dryomov:
 "Notable items here are

   - a series to take advantage of David Howells' netfs helper library
     from Jeff

   - three new filesystem client metrics from Xiubo

   - ceph.dir.rsnaps vxattr from Yanhu

   - two auth-related fixes from myself, marked for stable.

  Interspersed is a smattering of assorted fixes and cleanups across the
  filesystem"

* tag 'ceph-for-5.13-rc1' of git://github.com/ceph/ceph-client: (24 commits)
  libceph: allow addrvecs with a single NONE/blank address
  libceph: don't set global_id until we get an auth ticket
  libceph: bump CephXAuthenticate encoding version
  ceph: don't allow access to MDS-private inodes
  ceph: fix up some bare fetches of i_size
  ceph: convert some PAGE_SIZE invocations to thp_size()
  ceph: support getting ceph.dir.rsnaps vxattr
  ceph: drop pinned_page parameter from ceph_get_caps
  ceph: fix inode leak on getattr error in __fh_to_dentry
  ceph: only check pool permissions for regular files
  ceph: send opened files/pinned caps/opened inodes metrics to MDS daemon
  ceph: avoid counting the same request twice or more
  ceph: rename the metric helpers
  ceph: fix kerneldoc copypasta over ceph_start_io_direct
  ceph: use attach/detach_page_private for tracking snap context
  ceph: don't use d_add in ceph_handle_snapdir
  ceph: don't clobber i_snap_caps on non-I_NEW inode
  ceph: fix fall-through warnings for Clang
  ceph: convert ceph_readpages to ceph_readahead
  ceph: convert ceph_write_begin to netfs_write_begin
  ...

ceph: fix up some bare fetches of i_size

2021-04-27T21:52:23+00:00

We need to use i_size_read(), which properly handles the torn read
case on 32-bit arches.

Signed-off-by: Jeff Layton 
Signed-off-by: Ilya Dryomov

ceph: drop pinned_page parameter from ceph_get_caps

2021-04-27T21:52:23+00:00

All of the existing callers that don't set this to NULL just drop the
page reference at some arbitrary point later in processing. There's no
point in keeping a page reference that we don't use, so just drop the
reference immediately after checking the Uptodate flag.

Signed-off-by: Jeff Layton 
Signed-off-by: Ilya Dryomov

ceph: fix fscache invalidation

2021-04-27T21:52:22+00:00

Ensure that we invalidate the fscache whenever we invalidate the
pagecache.

Signed-off-by: Jeff Layton 
Signed-off-by: Ilya Dryomov

ceph: rip out old fscache readpage handling

2021-04-27T21:52:21+00:00

With the new netfs read helper functions, we won't need a lot of this
infrastructure as it handles the pagecache pages itself. Rip out the
read handling for now, and much of the old infrastructure that deals in
individual pages.

The cookie handling is mostly unchanged, however.

Signed-off-by: Jeff Layton 
Signed-off-by: Ilya Dryomov

ceph: don't allow type or device number to change on non-I_NEW inodes

2021-03-08T15:19:37+00:00

Al pointed out that a malicious or broken MDS could change the type or
device number of a given inode number. It may also be possible for the
MDS to reuse an old inode number.

Ensure that we never allow fill_inode to change the type part of the
i_mode or the i_rdev unless I_NEW is set. Throw warnings if the MDS ever
changes these on us mid-stream, and return an error.

Don't set i_rdev directly, and rely on init_special_inode to do it.
Also, fix up error handling in the callers of ceph_get_inode.

In handle_cap_grant, check for and warn if the inode type changes, and
only overwrite the mode if it didn't.

Reported-by: Al Viro 
Signed-off-by: Jeff Layton 
Signed-off-by: Al Viro

ceph: defer flushing the capsnap if the Fb is used

2021-02-16T11:09:52+00:00

If the Fb cap is used it means the current inode is flushing the
dirty data to OSD, just defer flushing the capsnap.

URL: https://tracker.ceph.com/issues/48640
Signed-off-by: Xiubo Li 
Reviewed-by: Jeff Layton 
Signed-off-by: Ilya Dryomov

ceph: allow queueing cap/snap handling after putting cap references

2021-02-16T11:09:51+00:00

Testing with the fscache overhaul has triggered some lockdep warnings
about circular lock dependencies involving page_mkwrite and the
mmap_lock. It'd be better to do the "real work" without the mmap lock
being held.

Change the skip_checking_caps parameter in __ceph_put_cap_refs to an
enum, and use that to determine whether to queue check_caps, do it
synchronously or not at all. Change ceph_page_mkwrite to do a
ceph_put_cap_refs_async().

Signed-off-by: Jeff Layton 
Reviewed-by: Ilya Dryomov 
Signed-off-by: Ilya Dryomov

ceph: fix flush_snap logic after putting caps

2021-02-16T11:09:51+00:00

A primary reason for skipping ceph_check_caps after putting the
references was to avoid the locking in ceph_check_caps during a
reconnect. __ceph_put_cap_refs can still call ceph_flush_snaps in that
case though, and that takes many of the same inconvenient locks.

Fix the logic in __ceph_put_cap_refs to skip flushing snaps when the
skip_checking_caps flag is set.

Fixes: e64f44a88465 ("ceph: skip checking caps when session reconnecting and releasing reqs")
Signed-off-by: Jeff Layton 
Reviewed-by: Ilya Dryomov 
Signed-off-by: Ilya Dryomov

ceph: fix race in concurrent __ceph_remove_cap invocations

2020-12-14T22:21:47+00:00

A NULL pointer dereference may occur in __ceph_remove_cap with some of the
callbacks used in ceph_iterate_session_caps, namely trim_caps_cb and
remove_session_caps_cb. Those callers hold the session->s_mutex, so they
are prevented from concurrent execution, but ceph_evict_inode does not.

Since the callers of this function hold the i_ceph_lock, the fix is simply
a matter of returning immediately if caps->ci is NULL.

Cc: stable@vger.kernel.org
URL: https://tracker.ceph.com/issues/43272
Suggested-by: Jeff Layton 
Signed-off-by: Luis Henriques 
Reviewed-by: Jeff Layton 
Signed-off-by: Ilya Dryomov