| Age | Commit message (Collapse) | Author |
|---|
|
Pull ceph updates from Ilya Dryomov:
"A wide variety of mostly CephFS fixes and cleanups, split between
changes that address edge cases (Sam, Xiubo, Matthew), efficiency
improvements (Max) and AI-assisted hardening (Michael, Jeremy).
One thing that stands out is Alex's change to how CephFS behaves in
NEARFULL scenarios: the long-standing "make all writes synchronous"
behavior has become opt-in. It was always somewhat controversial and
doesn't make much sense for modern deployments; the new default is to
continue normal operation (i.e. buffer writes as MDS allows, etc). The
behavior in case the cluster reaches any FULL state remains the same
as before"
* tag 'ceph-for-7.3-rc1' of https://github.com/ceph/ceph-client: (32 commits)
ceph: force a cap message when a deferred revoke can't be acked immediately
libceph: reject buckets with mismatched CRUSH ids
ceph: reject export_targets ranks >= CEPH_MAX_MDS in mdsmap decode
ceph: fix leaked inode reference on writeback abort at umount
libceph: remove ceph_put_page_vector()
libceph: validate banner payload length
ceph: make nearfull sync writes opt-in
ceph: do not repeat ceph_trim_dentries() if no progress possible
ceph: drop mdsc->mutex before decoding the MDS reply
ceph: fix UAF in check_new_map() on session freed during unlock
ceph: fix UAF in __kick_flushing_caps() on cf entry freed during unlock
ceph: pass inode pointer around instead of reloading it
ceph: mark cap remove with RB_CLEAR_NODE() instead of setting ci=NULL
ceph: add helper function ceph_cap_is_removed()
ceph: make __ceph_remove_cap() static
ceph: cap delegated inode count in ceph_parse_deleg_inos()
ceph: bound num_export_targets array for mds info v2/v3
ceph: bound MDSCapAuth path and fs_name decode in handle_session()
ceph: bound xattr value length in __build_xattrs()
ceph: bound copied dentry name length in NFS export get_name
...
|
|
When the MDS revokes capabilities, handle_cap_grant() normally
guarantees a response by setting `CHECK_CAPS_FLUSH_FORCE` (see
commit 31634d7597d8 ("ceph: force sending a cap update msg back to MDS
for revoke op")), so ceph_check_caps() sends a cap message even if the
client would otherwise decide it has nothing to do. That guarantee is
skipped whenever the revoke has to be deferred (via revoke_wait):
revoking Fb while dirty data is still buffered (writeback is queued
first) or revoking Fc while pages are cached (async invalidation is
queued first).
In those cases, the ack is left to the deferred completion
(ceph_put_wrbuffer_cap_refs() after writeback, or the invalidate
worker after invalidation); both of which call ceph_check_caps(ci,0)
i.e. without `CHECK_CAPS_FLUSH_FORCE`. Nothing gets sent under one
of the following conditions:
- the inode is retaining caps because the file was used recently
(file_wanted != 0; retain |= CEPH_CAP_ANY)
- the revoked cap is still used because the page was re-cached (e.g. a
file being re-read)
- the MDS has meanwhile re-granted, so `issued==implemented` and the
client sees nothing being revoked
The client then never emits the cap message which the MDS is waiting
for. The MDS blocks on the revoke indefinitely and logs, for minutes
or hours:
client.NNN isn't responding to mclientcaps(revoke), ino 0x... pending
pAsxLsXsxFsxcrwb issued pAsxLsXsxFsxcrwb, sent 964.899182 seconds ago
The client-side state at that point shows the full cap set still
issued, nothing in the revoking/flushing sets. Thus nothing gets
sent.
This patch fixes it by remembering that a forced response is expected.
When a revoke is deferred, set `CEPH_I_FLUSH_FORCE` on the inode.
ceph_check_caps() replays it as `CHECK_CAPS_FLUSH_FORCE`, so whichever
path re-checks the inode next (the writeback/invalidate completion,
the delayed worker, or any other caller) is guaranteed to send a cap
message to the MDS. __prep_cap() clears the flag once a message is
actually built.
This is the deferred-path counterpart of the existing
`CHECK_CAPS_FLUSH_FORCE` handling; a normal (non-deferred) revoke
still forces the response inline as before.
Cc: stable@vger.kernel.org
Fixes: 31634d7597d8 ("ceph: force sending a cap update msg back to MDS for revoke op")
Fixes: 257e6172ab36 ("ceph: don't let check_caps skip sending responses for revoke msgs")
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
MDSMap export_targets entries are monitor controlled. check_new_map()
uses each entry as a bit number in a fixed stack bitmap, so a rank
outside the protocol namespace can make set_bit() write past the end of
the array.
Reject ranks outside CEPH_MAX_MDS while decoding the map. Do not
validate against possible_max_rank here because maps may legitimately
reference ranks beyond a temporarily reduced max_mds.
Cc: stable@vger.kernel.org
Fixes: d517b3983dd3 ("ceph: reconnect to the export targets on new mdsmaps")
Signed-off-by: Jérémy Jean <Jeremy.Jean@oss.cyber.gouv.fr>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
ceph_dirty_folio() takes a wrbuffer claim on each newly dirtied folio: it
bumps i_wrbuffer_ref (taking an ihold() on the 0->1 transition) and
attaches the snap_context to folio->private. That claim is released only
by ceph_put_wrbuffer_cap_refs(), which for a submitted write runs from
writepages_finish().
In ceph_submit_write(), if ceph_inc_osd_stopping_blocker() fails -- which
happens during umount -- the request is aborted before submission: the
already-collected folios are only redirtied and unlocked, so
writepages_finish() never runs and the claim is leaked.
redirty_page_for_writepage() -> folio_redirty_for_writepage() ->
filemap_dirty_folio() sets PG_dirty directly and does not go through
->dirty_folio, so ceph_dirty_folio() is not re-entered to rebalance it.
Because every subsequent writeback also fails the osd_stopping_blocker,
i_wrbuffer_ref never returns to 0, the ihold() is never dropped, and the
inode cannot be evicted:
VFS: Busy inodes after unmount of ceph
kernel BUG at fs/super.c:650!
Release the orphaned claim in the abort path before redirtying, via
ceph_undo_wrbuffer_claim(): detach the snap_context, drop the wrbuffer
reference (letting i_wrbuffer_ref reach 0 and iput() the inode), and drop
the snap_context reference -- i.e. do what writepages_finish() would have
done for these never-submitted folios.
Only the locked_pages entries are undone; folios still in the fbatch were
never dirty-cleared by this call (folio_clear_dirty_for_io() is the
ownership-transfer point, and a successful move NULLs the fbatch slot), so
they hold no claim this call owns.
Cc: stable@vger.kernel.org
Fixes: fd7449d937e7 ("ceph: fix generic/421 test failure")
Signed-off-by: Matthew Brown <matthew@bargrove.com>
Reviewed-by: Xiubo Li <xiubo.li@clyso.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
ceph_put_page_vector() was paired with ceph_get_direct_page_vector(),
which was removed in commit 97a385e55829 ("libceph: remove
ceph_get_direct_page_vector()"). Its only remaining caller,
finish_netfs_read(), uses it to put a page vector allocated with
iov_iter_get_pages_alloc2(), which is confusing. Open-code the
put_page() loop and kvfree() there instead.
The caller passed dirty = false, so this also removes the dead dirty
branch and with it a call to the deprecated set_page_dirty_lock().
Signed-off-by: Tal Zussman <tz2294@columbia.edu>
Reviewed-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
The kernel CephFS client has historically treated a cluster or pool
NEARFULL condition as a request to force successful writes through
generic_write_sync(). That effectively turns otherwise buffered writes
into synchronous writes and can cause a severe throughput drop as soon
as a single OSD or the file data pool crosses the nearfull threshold.
On modern large clusters, NEARFULL is primarily an operator health
signal rather than an immediate client-side capacity failure. Operators
can still have substantial usable capacity while a cluster is
rebalancing, splitting PGs, or expanding onto new devices. RBD, RGW and
the userspace CephFS client do not impose this extra client-side
sync-write throttle, so the kernel client behavior is surprising and
operationally painful.
Change the default behavior so NEARFULL no longer changes normal
write-sync semantics. FULL and pool FULL still fail with -ENOSPC, and
explicitly synchronous writes continue to be synced by
generic_write_sync().
Add a nearfull_sync mount option for deployments that want the legacy
backpressure behavior. When this option is set, successful writes are
promoted to IOCB_DSYNC if the cluster or file data pool is marked
NEARFULL, preserving the old behavior for conservative deployments.
Link: https://tracker.ceph.com/issues/74849
Signed-off-by: Alex Markuze <amarkuze@redhat.com>
Reviewed-by: Xiubo Li <xiubo.li@clyso.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
ceph_cap_reclaim_work() re-queues itself for as long as
ceph_trim_dentries() returns -EAGAIN, which happens whenever a lease
walk exhausts its `nr_to_scan` budget. This creates a busy loop that
consumes CPU without making any progress when there is nothing to
reclaim: with no cap pressure (`count==0`) and every scanned lease
still valid, each pass runs the full scan budget down to zero and
returns `-EAGAIN`, only to be queued again immediately.
The dir-lease walk made this worse. When `expire_dir_lease` is
`false` (i.e. we have no intention of reclaiming dir leases),
__dir_lease_check() returned `TOUCH` for every valid lease. `TOUCH`
moves the dentry to the tail of the list and resets `di->time` via
__dentry_dir_lease_touch(), so a walk over N valid leases pointlessly
rewrote the list, refreshed the timestamps (preventing them from ever
aging out) and always drained `nr_to_scan`, guaranteeing the `-EAGAIN`
requeue.
Fix this in three steps:
- Return `KEEP` instead of `TOUCH` when `expire_dir_lease` is
`false`. If we are not going to reclaim the lease, leave it in
place instead of churning the list and resetting its timestamp; the
walk then terminates naturally (or via `STOP` at the first fresh
lease).
- Only return `-EAGAIN` from the first (dentry-lease) walk when something
was actually freed. A full batch that frees nothing means retrying
the same list immediately is futile; fall through to the dir-lease
walk instead.
- After both walks, bail out with success (0) when nothing was freed
and there is no cap pressure (`count==0`). There is no reason to
keep retrying when we are not over the cap limit and made no
progress.
Under real cap pressure (`count>0`) the reclaim path is unchanged and
still retries via `-EAGAIN`.
Without this patch, I saw 500 ceph_trim_dentries() calls per second on
our web servers. This is very visible in `/proc/lock_stat` (5 minute
capture):
class name con-bounces contentions waittime-min waittime-max waittime-total waittime-avg acq-bounces acquisitions holdtime-min holdtime-max holdtime-total holdtime-avg
&mdsc->dentry_list_lock: 126180 128218 0.04 8063.44 15986965.20 124.69 1573354 5296812 0.04 8291.28 74164526.48 14.00
-----------------------
&mdsc->dentry_list_lock 111736 [<000000007b11e319>] __ceph_dentry_dir_lease_touch+0x7c/0xa8
&mdsc->dentry_list_lock 2631 [<0000000050597999>] __dentry_leases_walk+0x64/0x2c8
&mdsc->dentry_list_lock 3878 [<00000000c0022f62>] __ceph_dentry_lease_touch+0x5c/0xa8
&mdsc->dentry_list_lock 9973 [<000000002f27cb6f>] __dentry_lease_unlist+0x50/0xa0
-----------------------
&mdsc->dentry_list_lock 123621 [<0000000050597999>] __dentry_leases_walk+0x64/0x2c8
&mdsc->dentry_list_lock 1822 [<000000007b11e319>] __ceph_dentry_dir_lease_touch+0x7c/0xa8
&mdsc->dentry_list_lock 2720 [<000000002f27cb6f>] __dentry_lease_unlist+0x50/0xa0
&mdsc->dentry_list_lock 55 [<00000000c0022f62>] __ceph_dentry_lease_touch+0x5c/0xa8
With this patch:
class name con-bounces contentions waittime-min waittime-max waittime-total waittime-avg acq-bounces acquisitions holdtime-min holdtime-max holdtime-total holdtime-avg
&mdsc->dentry_list_lock: 1203 1215 0.16 408.88 33082.88 27.23 4320501 7357389 0.04 500.64 1961578.00 0.27
-----------------------
&mdsc->dentry_list_lock 1029 [<000000003c9aea8a>] __ceph_dentry_dir_lease_touch+0x7c/0xa8
&mdsc->dentry_list_lock 169 [<000000002038c577>] __dentry_lease_unlist+0x50/0xa0
&mdsc->dentry_list_lock 16 [<00000000c991106d>] __ceph_dentry_lease_touch+0x5c/0xa8
&mdsc->dentry_list_lock 1 [<00000000612fe15f>] __dentry_leases_walk+0x64/0x2c8
-----------------------
&mdsc->dentry_list_lock 158 [<000000002038c577>] __dentry_lease_unlist+0x50/0xa0
&mdsc->dentry_list_lock 858 [<000000003c9aea8a>] __ceph_dentry_dir_lease_touch+0x7c/0xa8
&mdsc->dentry_list_lock 182 [<00000000612fe15f>] __dentry_leases_walk+0x64/0x2c8
&mdsc->dentry_list_lock 17 [<00000000c991106d>] __ceph_dentry_lease_touch+0x5c/0xa8
__dentry_leases_walk() is almost gone. The total wait time is reduced
by a factor of 483. That will give some latency gains to
ceph_readdir().
Cc: stable@vger.kernel.org
Fixes: 37c4efc1ddf9 ("ceph: periodically trim stale dentries")
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
handle_reply() held `mdsc->mutex` across parse_reply_info(),
i.e. across the full decode of the reply message. For large replies
(a big readdir allocates and parses many dir_entries), this can take a
while and blocks ceph_mdsc_submit_request() calls meanwhile.
The decode does not need `mdsc->mutex`: parse_reply_info() mostly
fills the request's `r_reply_info`. Create replies may also add
delegated inode numbers to the session xarray, but that xarray is
protected by its own lock and is not serialized by `mdsc->mutex`
today. By the time we reach parse_reply_info(), all
`mdsc->mutex`-protected state has already been updated under the lock
(the request has either been unregistered (safe reply) or added to the
session's unsafe list (unsafe reply)) and the request is pinned by the
reference taken in lookup_get_request().
Drop `mdsc->mutex` before calling parse_reply_info() so reply decoding
no longer blocks request submission. This only widens the existing
unlocked window that already covers the heavier ceph_fill_trace() /
ceph_readdir_prepopulate() processing, so no new races are introduced.
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Xiubo Li <xiubo.li@clyso.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
check_new_map() iterates mdsc->sessions[] and for each active session
drops mdsc->mutex to perform per-session operations. The forced-close
path (rank removed from map) correctly takes a reference on s via
ceph_get_mds_session() before releasing mdsc->mutex, but three other
paths do not:
Path A (address changed): mutex_unlock → mutex_lock(&s->s_mutex)
Path B (reconnect): mutex_unlock → send_mds_reconnect(mdsc, s)
Path C (active transition): mutex_unlock → mutex_lock(&s->s_mutex)
Without the extra reference, another thread can acquire mdsc->mutex
during the unlock window, call __unregister_session() which drops the
last reference on s, and free it. The original thread then accesses
freed memory via s->s_mutex.
Fix by adding ceph_get_mds_session(s) before each mutex_unlock and
ceph_put_mds_session(s) after the corresponding mutex_lock, matching
the pattern already used in the forced-close path.
Race timeline (Path A):
Thread A (check_new_map) Thread B (another map update
holds mdsc->mutex or session teardown)
-------------------------- --------------------------
s = mdsc->sessions[i]
(refcount == 1, held only by
sessions[] array)
mutex_unlock(&mdsc->mutex)
---> acquires mdsc->mutex
__unregister_session(mdsc, s)
sessions[i] = NULL
ceph_put_mds_session(s)
refcount: 1 -> 0
kfree(s) <--- freed!
mutex_lock(&s->s_mutex)
UAF on freed s->s_mutex
Cc: stable@vger.kernel.org
Signed-off-by: Xiubo Li <xiubo.li@clyso.com>
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
list_for_each_entry() iterates ci->i_cap_flush_list but drops
i_ceph_lock to send cap messages. During the unlock window,
handle_cap_flush_ack() can acquire i_ceph_lock, detach cf entries
with tid <= flush_tid from the list, release i_ceph_lock, and free
them via ceph_free_cap_flush() outside any lock. When the original
thread reacquires i_ceph_lock and the for-loop macro advances via
cf = list_next_entry(cf, i_list), it dereferences cf->i_list.next
on freed memory.
The race timeline:
__kick_flushing_caps() handle_cap_flush_ack()
----------------------- -----------------------
holds i_ceph_lock <---
iterates to cf (tid=10)
prepares FLUSH message
drops i_ceph_lock <---
__send_cap() ── FLUSH(tid=10)
MDS sends FLUSH_ACK(tid=10)
---> acquires i_ceph_lock
cf->tid(10) <= flush_tid(10),
detaches cf from i_cap_flush_list
drops i_ceph_lock
ceph_free_cap_flush(cf) <- frees it!
acquires i_ceph_lock <---
for-loop advances:
cf = list_next_entry(cf, i_list)
-- UAF on freed cf->i_list.next
The cf was just sent by __kick_flushing_caps itself via __send_cap().
The MDS may respond with FLUSH_ACK quickly enough that
handle_cap_flush_ack() frees cf before __kick_flushing_caps can
finish the iteration.
Fix by converting to a manual while loop: save the next pointer
under i_ceph_lock before dropping it, then use the saved pointer
after reacquiring, so the potentially-freed cf is never accessed again.
Cc: stable@vger.kernel.org
Signed-off-by: Xiubo Li <xiubo.li@clyso.com>
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
All these functions already have a ceph_inode_info pointer, so let's
use that instead of letting every function reload it from RAM
(i.e. `ceph_cap.ci`). This eliminates several memory accesses.
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
__ceph_remove_cap() erases the ceph_cap object from the RB tree, thus
it seems natural to use RB_CLEAR_NODE() / RB_EMPTY_NODE() for the
removal check.
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
Having it as a wrapper allows replacing the implementation, which the
next patch will do.
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
It's only used from within caps.c.
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
ceph_parse_deleg_inos() decodes interval sets of delegated inode numbers
from an MDS create-with-delegation reply. For each set it reads a 64-bit
start and a 64-bit len with ceph_decode_64_safe(), which only validates
that the eight bytes are present in the message, not the value, and then
loops over len while inserting entries into s_delegated_inos.
len is fully attacker controlled. A malicious or compromised MDS can send
one huge interval, many intervals in one reply, duplicate intervals, or
repeated replies that accumulate delegated inodes on the same session.
The original code bounded none of these and could spin the insert loop or
grow the xarray without limit.
Bound both dimensions with a single enforcement point. Track the number
of delegated inodes held by each MDS session in an atomic counter and
grow it only in ceph_insert_deleg_ino(), which uses atomic_add_unless()
to refuse to push the count past CEPH_MAX_DELEG_INOS. Because that helper
is the only place the counter grows, the per-session population can never
exceed the cap, so no separate per-session pre-check is needed. The
counter is decremented when async create consumes a delegated inode or
when an insert fails, incremented when a delegated inode is restored,
initialized with the session xarray, and reset when reconnect destroys
the xarray.
A per-session cap alone still lets one reply spin the insert loop on
duplicate ranges without growing the counter, so also cap the aggregate
interval length accepted from a single reply. Together these bound both
the loop trip count per reply and the xarray population across replies.
The cap is a fixed, client-chosen constant rather than a value derived
from the MDS. mds_client_prealloc_inos is a userspace MDS configuration
option; it is never sent to the kernel client on the wire, and a
server-supplied bound could not be trusted for a defensive limit in any
case. The constant is set well above that option's documented default of
1000 (a generous multiple), so legitimate refill behavior is unaffected
while the CPU and xarray memory a malformed delegation stream can consume
stays bounded.
Impact: a malicious or compromised Ceph MDS can no longer make a client
spin through an unbounded delegated-inode interval or grow one session's
delegated-inode xarray without limit.
Cc: stable@vger.kernel.org
Fixes: d48464878708 ("ceph: decode interval_sets for delegated inos")
Suggested-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
ceph_mdsmap_decode() in fs/ceph/mdsmap.c reads num_export_targets from
each per-mds info record and advances the decode cursor by
num_export_targets * sizeof(u32) without first checking that many bytes
remain. The only upper-bound check that catches a runaway cursor
(*p > info_end) is gated on info_v >= 4, because info_end is left NULL
for info_v 2 and 3. When the monitor sends an MDS map whose per-mds
info version is 2 or 3 with an oversized num_export_targets, the cursor
moves past the message front buffer and the later export-targets loop
calls the unchecked ceph_decode_32() on out-of-bounds memory.
A kernel client processes CEPH_MSG_MDS_MAP from its monitor session
(net/ceph/mon_client.c dispatches it; fs/ceph/super.c routes it to
ceph_mdsc_handle_mdsmap(), which sets end to the front buffer bound and
calls ceph_mdsmap_decode()). A malicious or compromised monitor, or an
on-path attacker on an unsigned/unencrypted messenger session, can
therefore drive an out-of-bounds read in the client kernel; on x86_64
with KASAN it is reported as a slab-out-of-bounds read in
ceph_mdsmap_decode(). The decoded values land in the internal
info->export_targets[] array, so the consequence is a kernel
out-of-bounds read, not an information leak to the attacker.
Impact: a malicious or compromised Ceph monitor sending an MDS map with
a per-mds info version of 2 or 3 and an oversized num_export_targets
field triggers an out-of-bounds read in the CephFS client kernel.
Add a ceph_decode_need() for the export-targets array before advancing
the cursor, so the bound is enforced for every info_v >= 2, not only
info_v >= 4. This mirrors the count-then-need idiom already used for
m_data_pg_pools later in the same function.
Compute the export-targets byte count with size_mul() and reuse that
checked length when advancing the cursor, so the attacker-controlled
num_export_targets multiplication fails closed on overflow rather than
relying on the later kcalloc() guard.
Cc: stable@vger.kernel.org
Fixes: d463a43d69f4 ("ceph: CEPH_FEATURE_MDSENC support")
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
handle_session() decodes the MDSCapAuth records carried by a
CEPH_SESSION_OPEN message (msg_version >= 6). For each record the
match.path and match.fs_name byte strings are read by first decoding a
32-bit length and then copying that many bytes with the bare
ceph_decode_copy(). Unlike the surrounding fields, which all use the
_safe decode variants, these two copies are not preceded by a
ceph_decode_need() bounds check, and the enclosing MDSCapAuth and
MDSCapMatch struct_len fields are skipped rather than enforced as an
upper bound. A length larger than the bytes remaining in the message
front makes ceph_decode_copy() read past the end of the front buffer.
The message front is a dedicated allocation (ceph_msg_new2() ->
kvmalloc), so the over-read runs off that object. A malicious or
compromised MDS can trigger this with the first post-connect message on
mount, with no client-side user interaction; under KASAN it is reported
as a slab-out-of-bounds read in handle_session().
Impact: a malicious MDS can force the kernel client to read up to 4 GiB
past the message front allocation during session setup, crashing the
client (out-of-bounds read).
Switch both copies to ceph_decode_copy_safe(), which performs the
ceph_decode_need() bounds check before the copy and branches to the
existing bad label, matching the rest of the decoder and the error path
that frees the partially decoded cap_auths array.
Cc: stable@vger.kernel.org
Fixes: 1d17de9534cb ("ceph: save cap_auths in MDS client when session is opened")
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
__build_xattrs() decodes the MDS-supplied xattr blob one attribute at a
time. For each attribute it reads a 32-bit name length, advances past the
name bytes, reads a 32-bit value length, records the value pointer, and
advances past the value bytes. The two length fields are read with
ceph_decode_32_safe(), but the value bytes themselves are advanced over
with a bare "p += len" and no ceph_decode_need() check that "len" bytes
remain in the blob.
For every attribute except the last, the next iteration's
ceph_decode_32_safe() on the following name length implicitly verifies
that the previous value did not run past the blob end. The final
attribute has no successor, so its decoded value length is never checked
against the blob bounds. A malicious or compromised metadata server can
set the last attribute's value length larger than the bytes actually
present in the blob.
The blob is a dedicated kvmalloc() allocation sized to the wire length
(ceph_buffer_new() in ceph_fill_inode()). __set_xattr() records the
oversized length in xattr->val_len verbatim, and a later getxattr(2) runs
memcpy(value, xattr->val, xattr->val_len) into a user-supplied buffer,
copying bytes past the end of the allocation back to user space.
Impact: a malicious metadata server discloses adjacent kernel heap bytes
to a local user via getxattr(2) on a CephFS file. Add the missing
ceph_decode_need() so an out-of-bounds value length on the final
attribute fails the decode and returns -EIO instead of being stored.
Cc: stable@vger.kernel.org
Fixes: 355da1eb7a1f ("ceph: inode operations")
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
ceph_get_name() copies the MDS-supplied name into the caller's
NAME_MAX-sized buffer with memcpy(name, rinfo->dname, rinfo->dname_len)
and then writes name[rinfo->dname_len] = 0, without checking dname_len
against NAME_MAX. A malicious or buggy MDS that returns a LOOKUPNAME reply
with dname_len > NAME_MAX overflows the buffer. __get_snap_name() copies
rde->name / rde->name_len the same unchecked way.
Impact: a malicious or compromised Ceph MDS overflows the NAME_MAX name
buffer in a client's NFS-export get_name path, a slab out-of-bounds write
reported by KASAN. Reachable when a CephFS mount is re-exported over NFS.
Add ceph_export_copy_name(), which rejects lengths above NAME_MAX with
-ENAMETOOLONG before the copy, and use it in both ceph_get_name() and
__get_snap_name().
Cc: stable@vger.kernel.org
Fixes: 19913b4eac4a ("ceph: add get_name() NFS export callback")
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Michael Bommarito <michael.bommarito@gmail.com>
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
For O_APPEND writes, ki_pos is set to the current EOF via
generic_write_checks() after fetching i_size from the MDS. However,
ceph_get_caps() may need to wait for Fwx exclusive caps if the write
extends the file (endoff > i_max_size). While waiting for Fwx, the
previous Fwx holder (another client) may have already extended the
file. When the MDS grants us Fwx, the cap grant message updates the
local i_size, but ki_pos remains at the old EOF, causing the append
write to land at a stale offset and overwrite data from the other
client.
Fix by re-reading i_size_read(inode) after ceph_get_caps() returns.
At this point we hold Fwx exclusive caps, no other client can modify
the file, and i_size reflects the true EOF from the MDS cap grant.
No extra MDS round-trip is needed. Only adjust ki_pos when the EOF
has actually changed.
After adjusting ki_pos forward, the write range [pos, pos+count) may
now exceed the i_max_size that was validated by ceph_get_caps() for
the old range. Re-check against i_max_size and truncate the write
if necessary to stay within the MDS-granted limit.
Link: https://tracker.ceph.com/issues/7333
Fixes: 8e4473bb50a1 ("ceph: do not execute direct write in parallel if O_APPEND is specified")
Signed-off-by: Xiubo Li <xiubo.li@clyso.com>
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
The NULL check for "ci" in __ceph_remove_cap() was dead code because
ci was dereferenced via &ci->netfs.inode before the check, and
cap->session was dereferenced via session->s_mdsc->fsc->client even
earlier. On a double-remove, both cap->ci and cap->session are set
to NULL by the first call, so the second call would crash before
ever reaching the guard.
Move ci, session, cl, and inode initializations after the NULL check
so that the early-return actually works.
Signed-off-by: Xiubo Li <xiubo.li@clyso.com>
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
When a LOOKUP/LOOKUPSNAP in a snapped directory returns ENOENT
without a trace, ceph_finish_lookup() creates a negative dentry
via d_add(dentry, NULL). For live directories this is fine — the
dentry naturally expires. But for snapped directories,
ceph_d_revalidate() unconditionally trusts all cached dentries
(valid = 1), so a negative dentry created by a transient error
persists forever, hiding entries that genuinely exist in the
snapshot.
Only cache negative dentries for live (non-snapshotted) parent
directories. For snapped parents, skip the negative dentry so
that VFS retries the lookup on the next access. Since the
conditions that trigger a negative dentry (MDS transient error,
local ENOENT shortcut, or MDS null dentry lease) are all rare in
snapped directories, the performance impact of this change is
negligible.
Link: https://tracker.ceph.com/issues/78529
Reported-by: Andras Pataki <apataki@flatironinstitute.org>
Signed-off-by: Xiubo Li <xiubo.li@clyso.com>
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
__ceph_pool_perm_get() has six allocations for building OSD STAT
requests, five of which used GFP_NOFS and one (the page vector
allocation) used GFP_KERNEL, making them inconsistent.
The function is only called from ceph_try_get_caps() and
__ceph_get_caps(), both of which are in the user I/O path (read,
write, fallocate, mmap fault), not in the writeback path. There is
no risk of recursive writeback, so GFP_NOFS is unnecessarily
restrictive. Use GFP_KERNEL consistently for all six allocations.
Signed-off-by: Xiubo Li <xiubo.li@clyso.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
ceph_alloc_cap_flush() is called from ceph_writepages_start() inside
the writeback layer, where other allocations in the same path
(ceph_osdc_alloc_request, ceph_osdc_alloc_messages) already use
GFP_NOFS. A GFP_KERNEL allocation here can trigger direct reclaim
that recursively enters the filesystem writeback path:
ceph_writepages_start() // inode A writeback
ceph_alloc_cap_flush()
kmem_cache_alloc(..., GFP_KERNEL)
[direct reclaim]
try_to_free_pages()
shrink_slab()
super_cache_scan()
prune_icache_sb()
inode_lru_isolate()
iput() -> evict(inode_B)
[inode_B has dirty pages]
filemap_flush()
ceph_writepages_start() // re-enters writeback
ceph_alloc_cap_flush()
-> RECURSION / STACK OVERFLOW
All 11 callers of ceph_alloc_cap_flush() are in write or writeback
contexts: writepages (x2), write_iter, fallocate, copy_file_range,
setxattr, setattr, and page_mkwrite.
Signed-off-by: Xiubo Li <xiubo.li@clyso.com>
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
Eliminate some redundant code.
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Xiubo Li <xiubo.li@clyso.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
__touch_cap() moves one capability to the end of the LRU list; this
list is sorted by access time for just one thing: ceph_trim_caps().
That function is supposed to discard the least-recently used
capabilities.
__touch_cap() is called extremely often - several times for every
system call, but ceph_trim_caps() is only called rarely.
__touch_cap() causes considerable lock contention on
`ceph_mds_session.s_cap_lock`; this is a /proc/lock_stat I captured on
one of our web servers for 5 minutes:
class name con-bounces contentions waittime-min waittime-max waittime-total waittime-avg acq-bounces acquisitions holdtime-min holdtime-max holdtime-total holdtime-avg
&s->s_cap_lock: 336304046 341686597 0.04 4905.76 418498578.76 1.22 892783632 1957814739 0.04 959.40 355752146.24 0.18
--------------
&s->s_cap_lock 339379730 [<00000000a2197200>] __ceph_caps_issued_mask+0x1bc/0x240
&s->s_cap_lock 1268054 [<00000000c96a24b7>] ceph_add_cap+0x234/0x3e0
&s->s_cap_lock 1021360 [<00000000aa76f996>] ceph_add_cap+0x108/0x3e0
&s->s_cap_lock 16042 [<0000000099463548>] __ceph_remove_cap+0x1f4/0x270
--------------
&s->s_cap_lock 338509619 [<00000000a2197200>] __ceph_caps_issued_mask+0x1bc/0x240
&s->s_cap_lock 1937864 [<00000000c96a24b7>] ceph_add_cap+0x234/0x3e0
&s->s_cap_lock 1203451 [<00000000aa76f996>] ceph_add_cap+0x108/0x3e0
&s->s_cap_lock 202 [<00000000888f212a>] __ceph_remove_cap+0x7c/0x270
In this /proc/lock_stat output, __touch_cap() is inlined in
__ceph_caps_issued_mask(). It is responsible for 99% of all
contentions.
Since __touch_cap() is called so often, it is acceptable to just skip
most calls. The most busy capabilities will still gravitate towards
the end of the linked list, and if not, it doesn't hurt as much as the
lock contention. This is still good enough for ceph_trim_caps().
This patch adds a static variable that gets incremented with each
call, and 255 out of 256 calls will just be skipped. I didn't bother
to make the increment atomic or use READ_ONCE because I don't think
that makes a practical difference for this use case.
Another /proc/lock_stat for 5 minutes with this patch (__touch_cap()
is no longer inlined probably because it contains a static variable):
class name con-bounces contentions waittime-min waittime-max waittime-total waittime-avg acq-bounces acquisitions holdtime-min holdtime-max holdtime-total holdtime-avg
&s->s_cap_lock: 1043711 1065182 0.04 502.72 737472.88 0.69 10522578 25069948 0.04 796.44 11053669.64 0.44
--------------
&s->s_cap_lock 1043074 [<00000000f4367d73>] __touch_cap.isra.0+0x50/0xa8
&s->s_cap_lock 12147 [<0000000096f45706>] ceph_add_cap+0x234/0x3e0
&s->s_cap_lock 9472 [<0000000038a23e0f>] ceph_add_cap+0x108/0x3e0
&s->s_cap_lock 471 [<00000000e2eba934>] __ceph_remove_cap+0x1f4/0x270
--------------
&s->s_cap_lock 978499 [<00000000f4367d73>] __touch_cap.isra.0+0x50/0xa8
&s->s_cap_lock 57794 [<0000000038a23e0f>] ceph_add_cap+0x108/0x3e0
&s->s_cap_lock 27226 [<0000000096f45706>] ceph_add_cap+0x234/0x3e0
&s->s_cap_lock 1581 [<00000000e2eba934>] __ceph_remove_cap+0x1f4/0x270
__touch_cap() is still responsible for 91% of all contentions, but the
number of contentions has been reduced by a factor of 320 and the
total wait time by a factor of 567.
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
system_wq (per-CPU) and system_unbound_wq (unbound) are the older
workqueue name, replaced by system_{percpu|dfl}_wq.
The new workqueues have been introduced by:
128ea9f6ccfb ("workqueue: Add system_percpu_wq and system_dfl_wq")
Usage of older workqueues will now trigger a pr_warn_once() because they are
marked as deprecated as per commit:
64d8eae3f895 ("workqueue: Add warnings and fallback if system_{unbound}_wq is used")
So change the used workqueue with the newer, keeping the same behavior.
Suggested-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Marco Crivellari <marco.crivellari@suse.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
The fscrypt subsystem uses the scatterlist crypto API, inheriting its
requirement that any buffers are in the linear mapping region. However,
the messenger client uses kvmalloc() to create buffers for messages,
which will occasionally place those buffers in the vmalloc() region when
physical memory fragmentation doesn't permit a large enough kmalloc().
The various callers of ceph_fname_to_usr() directly pass (slices of) raw
messages from the MDS without considering that the messages may be in
vmalloc() buffers, resulting in oopses especially on non-x86 platforms
(see 'Closes:' for more details and a reproducer).
Make ceph_fname_to_usr() explicitly tolerant of vmalloc()-allocated
fname->ctext, fname->name, and/or oname->name buffers, using `tname`
(which, when non-null, must be a linear address; when null, is briefly
allocated as necessary) as a bounce buffer to avoid passing any
inappropriate addresses to fscrypt_fname_disk_to_usr().
Additionally change parse_reply_info_readdir() -- the only function to
supply its own `tname` -- to follow the new "tname must never come from
vmalloc()" rule by passing NULL when the message is not in the linear
region. Though this causes a per-dentry kmalloc()+kfree(), this overhead
exists only when processing the minority of messages that spill into
vmalloc(). My (crude) testing puts this at only about 1 in 8,000 readdir
messages. Still, if the overhead proves unreasonable in the future, it
is easy enough to mitigate: a future change could allocate a bounce
buffer in parse_reply_info_readdir() and use that as `tname` instead.
Cc: stable@vger.kernel.org # 888d33b208bd: ceph: pass fscrypt `tname` buffers directly
Cc: stable@vger.kernel.org
Fixes: 457117f077c6 ("ceph: add helpers for converting names for userland presentation")
Closes: https://lore.kernel.org/ceph-devel/20260415034020.11530-1-CFSworks@gmail.com/
Signed-off-by: Sam Edwards <CFSworks@gmail.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
ceph_fname_to_usr() needs a temporary buffer for some operations
(currently only base64-decoding ciphertext) and it is convenient to
allow the caller to specify this buffer to avoid a heap allocation, so
it has a (nullable) `tname` argument. Until now, this argument was a
`struct fscrypt_str`; however, this is unnecessary for two reasons:
1. `tname->len` isn't used anywhere: ceph_fname_to_usr() assumes a
buffer large enough to hold the ciphertext, and
parse_reply_info_readdir() -- the only caller to use tname -- doesn't
set it.
2. While the `tname` parameter is documented "may be NULL,"
parse_reply_info_readdir() always passes it but with `tname->name`
sometimes NULL in violation of the contract, indicating that the
unnecessary container creates actual confusion.
Therefore, change the type to `unsigned char *` and pass the buffer
directly.
Signed-off-by: Sam Edwards <CFSworks@gmail.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
When mkdir succeeds, ceph_mkdir() sets ret to ERR_PTR(0) which is
incorrect. It should return NULL instead for success.
Fixes: 88d5baf69082 ("Change inode_operations.mkdir to return struct dentry *")
Signed-off-by: Hongling Zeng <zenghongling@kylinos.cn>
Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull misc vfs updates from Christian Brauner:
"Bigger cleanups:
- The lockref dead-count handling is tidied up.
The open-coded check for a count below zero as the dead marker
relies on information the caller should not have.
- make put_mnt_ns() leave mounts connected. Destroying a mount
namespace disconnected its mounts from their mount points. So a
file descriptor still open on the parent of a mount point could be
used to peek under it.
Locked mounts were already kept connected to prevent exactly that.
But a mount is only locked when its tree is copied across a user
namespace boundary. So a mount namespace set up by a privileged
component had no locked mounts and its mounts were disconnected.
Passing UMOUNT_CONNECTED keeps every mount connected and prevents
that bug.
- vfs_prepare_mode() passes S_IFDIR for directories. I meant to fix
that ago but didn't get to it. So now someone finally did it.
This kills the exception where the mode could be 0 when a directory
was created whereas every other creation operation passed it
explicitly already.
- move long delayed work for ufs, jffs2, hfsplus, hfs and affs from
the per-cpu system_long_wq to the new unbound system_dfl_long_wq.
None of that work relies on per-cpu state and the work item is
enqueued with queue_delayed_work() whose timer is global anyway. So
it may as well benefit from scheduler task placement.
Smaller fixes and cleanups:
- unlock_buffer() and journal_end_buffer_io_sync() use
clear_and_wake_up_bit()
- the pipe page pools are unified into a single per-pipe pool and the
extra wake_up(rd_wait) is limited to EPOLLET consumers
- eventpoll now computes its timer slack lazily in ep_poll()
- shrink_dcache_for_umount() keeps making progress on busy roots
- excess xarray nodes are freed in clear_inode()
- romfs detects hard link cycles
- the user path of nested backing files is fixed
- pidfd holds exec_update_lock around the namespace ioctl
- non-memcg-aware nr_cached_objects is skipped during memcg slab
shrink
- iomap_write_iter() always returns status
- mangle_path() is renamed to seq_mangle_path()
- inode timestamp accessors are annotated
- new regression test for pipe->poll_usage.
- a few documentation, kernel-doc and selftest fixes"
* tag 'vfs-7.3-rc1.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs: (67 commits)
selftests/namespaces: Fix racy pipe handshake in timens and pidns_separate
selftests/epoll: add a regression test for pipe->poll_usage
pipe: only enable the extra wake_up(rd_wait) for EPOLLET consumers
pidfd: hold exec_update_lock around namespace ioctl
fs: fix user path of nested backing files
fs: remove stale inode_insert5() kernel-doc parameter
fs: fix switch/case indentation in sysfs() syscall
fs: document semantics of kstat::{uid,gid} fields
dcache: keep shrink_dcache_for_umount() making progress on busy roots
seq_file: rename mangle_path to seq_mangle_path
nstree: add/fix struct ns_id_req kernel-doc member fields
dcache: use lockref routines for dead count checks
lockref: tidy up dead count handling
initramfs: fix typo in reserve_initrd_mem comment
fs/pipe: unify the page pools into a single per-pipe pool
fs: annotate inode timestamp accessors
eventpoll: compute timer slack lazily in ep_poll()
selftests/filesystems: add mntns cleanup test
put_mnt_ns(): leave mounts connected
affs: Move long delayed work on system_dfl_long_wq
...
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs
Pull vfs lookup updates from Christian Brauner:
"This refactors lookup_open() and adds vfs_lookup_open() for nfsd.
mnt_want_write() and parent locking are moved into lookup_open()
itself.
audit_inode_child() is also now called in lookup_open() on failure.
That is the calling convention in vfs_create() and vfs_mkdir(), but
lookup_open() made no such call when atomic_open() should have created
a file and did not. And neither did the regular ->create() path fwiw.
This also contains work to remove the unneeded excl argument from the
->create() inode op"
* tag 'vfs-7.3-rc1.lookup' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
fs/namei.c: fix coding style in atomic_open() and lookup_open()
fs/namei.c: fix kerneldoc of atomic_open() and vfs_lookup_open()
fs/namei.c: update stale comments in lookup_open()
Remove excl arg to ->create inode_operation
fs/namei.c: update kerneldoc of atomic_open()
vfs: call audit_inode_child() in lookup_open() on failure
vfs: move create error && negative dentry case in lookup_open() up
VFS: add vfs_lookup_open() for nfsd
VFS: move delegated_inode retry loop into lookup_open()
VFS: move mnt_want_write() and locking into lookup_open()
|
|
ceph_ioctl_set_layout() and ceph_ioctl_set_layout_policy() call
inode_owner_or_capable() with &nop_mnt_idmap instead of the idmap of the
mount the ioctl was issued on.
CephFS supports idmapped mounts (FS_ALLOW_IDMAP), so on such a mount this
compares the caller's fsuid against the unmapped on-disk owner rather than
the mapped owner: the actual owner can be wrongly denied with -EACCES and
an unrelated caller wrongly allowed. Both functions already have the
struct file, so use file_mnt_idmap(file) instead.
Cc: stable@vger.kernel.org
Fixes: cee38bbf5556 ("ceph: add owner/capability checks for CEPH_IOC_SET_LAYOUT*")
Signed-off-by: Zhan Xusheng <zhanxusheng@xiaomi.com>
Reviewed-by: Xiubo Li <xiubo.li@clyso.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
CEPH_MDS_IS_READY() is parsed so that the ternary expression can
return true for an MDS entry with state 0 when it is not laggy. This
allows the random selector to choose a down/DNE rank.
Group the ternary expression under the state check so zero-state ranks
are not treated as ready.
Cc: stable@vger.kernel.org
Fixes: b38c9eb4757d ("ceph: add possible_max_rank and make the code more readable")
Link: https://tracker.ceph.com/issues/78648
Signed-off-by: Yiming Zhu <zhuyiming@kuaishou.com>
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
The only time that 'false' is passed as the 'excl' arg to the ->create
inode_operation is in lookup_open() when ->atomic_open is not provided
by the parent directory.
*all* directory inode_operations which do not have ->atomic_open
completely ignore the 'excl' arg.
Therefore we don't need the 'excl' arg. Those few ->create operations
which pay attention to the arg are only ever called with a value of
'true'.
We remove that arg and change all ->create operations to behave as those
thhe arg were 'true'.
Signed-off-by: NeilBrown <neil@brown.name>
Link: https://patch.msgid.link/178290671516.27465.15984496764174914338@noble.neil.brown.name
Reviewed-by: Jori Koolstra <jkoolstra@xs4all.nl>
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org>
|
|
1. put the dead val into a macro so that it can be used in other places
2. __lockref_is_dead():
- drop the __ suffix, this is not an internal routine
- drop the spurious cast, the value is already a signed int
- use READ_ONCE to prevent any compile shenanigans
3. provide lockref_is_dead_or_zero()
Signed-off-by: Mateusz Guzik <mjguzik@gmail.com>
Link: https://patch.msgid.link/20260724171422.429284-2-mjguzik@gmail.com
Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org>
|
|
handle_reply() stores a `ceph_mds_request` pointer in
`current->journal_info` while filling the inode and dentry cache from
an MDS reply.
An allocation in this section can enter direct reclaim and prune
dentries from another filesystem. If this dirties an ext4 inode, ext4
starts a JBD2 transaction. JBD2 interprets the Ceph request in
`current->journal_info` as a journal handle and dereferences the
request's `r_tid` as `h_transaction`, causing a kernel crash, e.g.:
Unable to handle kernel paging request at virtual address 00000000077b4818
[...]
Internal error: Oops: 0000000096000004 [#1] SMP
Modules linked in:
CPU: 6 UID: 0 PID: 2699135 Comm: kworker/6:3 Tainted: G W 6.18.38-i3 #1113 NONE
[...]
Workqueue: ceph-msgr ceph_con_workfn
pstate: 80400009 (Nzcv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
pc : jbd2__journal_start+0x2c/0x208
lr : __ext4_journal_start_sb+0x100/0x178
[...]
Call trace:
jbd2__journal_start+0x2c/0x208 (P)
__ext4_journal_start_sb+0x100/0x178
ext4_dirty_inode+0x3c/0x90
__mark_inode_dirty+0x58/0x400
iput.part.0+0x2b0/0x370
iput+0x18/0x30
dentry_unlink_inode+0xc0/0x158
__dentry_kill+0x80/0x250
shrink_dentry_list+0x90/0x130
prune_dcache_sb+0x60/0x98
super_cache_scan+0xe8/0x190
do_shrink_slab+0x174/0x388
shrink_slab+0xd8/0x4c0
shrink_node+0x31c/0x908
do_try_to_free_pages+0xd0/0x508
try_to_free_pages+0x11c/0x238
__alloc_frozen_pages_noprof+0x4d0/0xdd0
__folio_alloc_noprof+0x18/0x70
__filemap_get_folio+0x248/0x440
ceph_readdir_prepopulate+0x570/0x9e8
mds_dispatch+0x1424/0x1ba0
ceph_con_process_message+0x74/0xa0
ceph_con_v1_try_read+0x3a0/0x1510
ceph_con_workfn+0x260/0x460
Enter a scoped NOFS allocation context and leave it after clearing
`journal_info`. This prevents filesystem reclaim from recursing into
another filesystem while the field contains Ceph-private data.
Cc: stable@vger.kernel.org
Fixes: 315f24088048 ("ceph: fix security xattr deadlock")
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Reviewed-by: Xiubo Li <xiubo.li@clyso.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
These permission checks were already missing in the initial
impementation of these ioctls. This Ceph allows any user who owns a
file descriptor to manipulate the layout of any file, even if they
don't have write permissions.
It might be a good idea to guard other ioctls with permission checks
as well or even disallow regular users (even if they own the file) to
manipulate layout settings completely, as this may be abused to DoS
the Ceph servers, but right now, I find it most urgent to have setter
checks at all.
Cc: stable@vger.kernel.org
Fixes: 8f4e91dee2a2 ("ceph: ioctls")
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Xiubo Li <xiubo.li@clyso.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
A reader can hang forever in __ceph_get_caps() when the client no
longer holds `FILE_RD`, but local cap state still says that the
capability is already wanted (via `mds_wanted`).
One way to trigger this is through MDS cap revocation. If another
client performs a conflicting operation, the MDS can revoke `FILE_RD`
from the reader; the next read then has to reacquire `FILE_RD`. If
the cap update that should request `FILE_RD` never reaches the MDS
after `cap->mds_wanted` was raised, the reader is left holding only
non-file caps while local `mds_wanted` still includes the file read
caps.
In that state, try_get_cap_refs() sees `need <= mds_wanted` and
returns 0, so __ceph_get_caps() just waits on `i_cap_wq`. If the cap
update that was supposed to request `FILE_RD never reaches the MDS
after `cap->mds_wanted was` raised, no further request is sent and the
waiter can sleep indefinitely until unrelated cap traffic happens to
wake it up.
The ordering issue is that `cap->mds_wanted` is updated in
__prep_cap() before the `CEPH_MSG_CLIENT_CAPS message` is actually
queued for send. That makes one field serve two different meanings at
once: what this client wants, and what the client believes the MDS
already knows it wants.
A proper fix would be to split those states and track whether a cap
update is actually in flight or has been observed by the MDS.
However, simply moving the `cap->mds_wanted assignment` later would
not be sufficient: queueing the message in the messenger does not
guarantee that the MDS processed that specific wanted set, and
reconnect or message loss can still invalidate that assumption.
Fixing that properly would require a larger rework of the cap state
machine.
To allow simpler backports to stable kernels, this patch implements a
simpler workaround:
- stop waiting forever in __ceph_get_caps(); after a bounded wait,
fall back to the renew path
- make ceph_renew_caps() issue a synchronous `OPEN` request whenever
the inode still does not actually hold the wanted caps, instead of
only calling ceph_check_caps()
The extra issued-vs-wanted check in ceph_renew_caps() is necessary
because the previous test only checked whether the inode still had any
real caps at all. That is not enough after revocation: the client can
still hold something like `pLs` and yet be missing `FILE_RD`
completely. In that case, falling back to ceph_check_caps() is not
sufficient, because it still trusts `cap->mds_wanted` and may resend
nothing. By requiring `(issued & wanted) == wanted` before taking the
asynchronous path, the code only uses ceph_check_caps() when the
`wanted caps` are already actually issued. Otherwise, it sends the
synchronous `OPEN` renew.
This preserves the existing asynchronous fast path when the wanted
caps are already issued, avoids changing cap-state semantics, and
fixes the hang by guaranteeing that a stalled waiter eventually
retries through a path that does not rely on the stale `mds_wanted`
state.
[ idryomov: move CEPH_GET_CAPS_WAIT_TIMEOUT from libceph.h to
mds_client.h, formatting ]
Cc: stable@vger.kernel.org
Fixes: 0a454bdd501a ("ceph: reorganize __send_cap for less spinlock abuse")
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
The ceph_readdir() function allocates a ceph_mds_request via
ceph_mdsc_create_request() and stores it in dfi->last_readdir. In
the directory entry processing loop, if the entry's offset is less
than ctx->pos or if the inode pointer is unexpectedly NULL, the
function returns -EIO without releasing the reference held by
dfi->last_readdir, causing a refcount leak.
Fix this by adding ceph_mdsc_put_request(dfi->last_readdir) before
returning on these error paths. Also set dfi->last_readdir to NULL
for safety, matching the cleanup done at the normal exit.
Cc: stable@vger.kernel.org
Fixes: af9ffa6df7e3 ("ceph: add support to readdir for encrypted names")
Signed-off-by: WenTao Liang <vulab@iscas.ac.cn>
Reviewed-by: Viacheslav Dubeyko <slava@dubeyko.com>
Reviewed-by: Alex Markuze <amarkuze@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
write_folio_nounlock() increments fsc->writeback_count to track
in-flight writeback operations. On several error paths where the
function returns early (folio lookup failure, snapshot context
allocation failure, and writepages submission failure), the function
returns without calling atomic_long_dec_return() to decrement the
counter.
Each leaked increment keeps the counter above zero, which can prevent
the filesystem from cleanly unmounting or suspending writes.
Add atomic_long_dec_return() calls on all error paths that currently
return without decrementing the counter.
Cc: stable@vger.kernel.org
Fixes: d55207717ded ("ceph: add encryption support to writepage and writepages")
Signed-off-by: Wentao Liang <vulab@iscas.ac.cn>
Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
ceph_handle_caps() reads snap_trace_len from the wire-format
ceph_mds_caps header and uses it unconditionally to build a fake
end pointer (snaptrace + snaptrace_len) that is later handed to
ceph_update_snap_trace() in the CEPH_CAP_OP_IMPORT case:
snaptrace = h + 1;
snaptrace_len = le32_to_cpu(h->snap_trace_len);
p = snaptrace + snaptrace_len;
...
case CEPH_CAP_OP_IMPORT:
if (snaptrace_len) {
...
if (ceph_update_snap_trace(mdsc, snaptrace,
snaptrace + snaptrace_len,
false, &realm)) { ... }
ceph_update_snap_trace() then decodes a struct ceph_mds_snap_realm
from snaptrace using ceph_decode_need(&p, e, sizeof(*ri), bad)
with the attacker-supplied fake end e == snaptrace + snaptrace_len.
With snaptrace_len == 0xFFFFFFFF the bound check is trivially
satisfied, ri = p reads sizeof(struct ceph_mds_snap_realm) past
the legitimate msg->front buffer, and ri->num_snaps /
ri->num_prior_parent_snaps then drive further out-of-bounds
reads of the encoded snap arrays.
The eleven msg_version >= 2 .. msg_version >= 12 decoder blocks
above the op switch each catch this OOB through their
ceph_decode_*_safe() / ceph_decode_need() helpers, but they sit
behind a hdr.version-gated if, so a malicious or compromised
MDS that sets msg->hdr.version = 1 reaches the IMPORT path with
no version-gated decoder having validated snap_trace_len. The
shape has been present since ceph_handle_caps() was introduced.
Validate snap_trace_len against the message front buffer before
consuming it, using the canonical ceph_decode_need() / ceph_has_room()
helper. The helper bounds the length with subtraction (n <= end - p,
guarded by end >= p) rather than pointer addition, so it is wrap-safe
for the attacker-controlled u32 length on 32-bit builds where
p + snap_trace_len could overflow the address space. This matches the
rest of the ceph decode path (e.g. the pool_ns_len check a few lines
below), and the existing goto bad cleanup already covers this exit
path.
Cc: stable@vger.kernel.org
Fixes: a8599bd821d0 ("ceph: capability management")
Signed-off-by: Bryam Vargas <hexlabsecurity@proton.me>
Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
vfs_mkdir() now sets the S_IFDIR type bit in the mode it passes to
->mkdir(), so OR-ing S_IFDIR into the mode again in ceph_mkdir() is
redundant. Drop it.
Assisted-by: LLM
Signed-off-by: Jori Koolstra <jkoolstra@xs4all.nl>
Link: https://patch.msgid.link/20260630105400.68459-8-jkoolstra@xs4all.nl
Reviewed-by: NeilBrown <neil@brown.name>
Signed-off-by: Christian Brauner (Amutable) <brauner@kernel.org>
|
|
Pull ceph updates from Ilya Dryomov:
"This adds support for manual client session reset in CephFS, allowing
operators to get out of tricky livelock situations involving caps and
file locks without evicting the problematic client instance on the MDS
side or rebooting the client node both of which can be disruptive"
* tag 'ceph-for-7.2-rc1' of https://github.com/ceph/ceph-client:
ceph: add manual reset debugfs control and tracepoints
ceph: add client reset state machine and session teardown
ceph: add diagnostic timeout loop to wait_caps_flush()
ceph: harden send_mds_reconnect and handle active-MDS peer reset
ceph: use proper endian conversion for flock_len in reconnect
ceph: convert inode flags to named bit positions and atomic bitops
rbd: switch to dynamic root device
|
|
Add the debugfs and trace plumbing used to trigger and observe
manual client reset.
The reset interface exposes a trigger file for operator-initiated
reset and a status file for tracking the most recent run. The
tracepoints record scheduling, completion, and blocked caller
behavior so reset progress can be diagnosed from the client side.
debugfs layout under /sys/kernel/debug/ceph/<client>/reset/:
trigger - write to initiate a manual reset
status - read to see the most recent reset result
The reset directory is cleaned up via debugfs_remove_recursive()
on the parent, so individual file dentries are not stored.
Tracepoints:
ceph_client_reset_schedule - reset queued
ceph_client_reset_complete - reset finished (success or failure)
ceph_client_reset_blocked - caller blocked waiting for reset
ceph_client_reset_unblocked - caller unblocked after reset
All tracepoints use a null-safe access for monc.auth->global_id
to guard against early-init or late-teardown edge cases.
Signed-off-by: Alex Markuze <amarkuze@redhat.com>
Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
Add the client-side reset state machine, request gating, and manual
session teardown implementation.
Manual reset is an operator-triggered escape hatch for client/MDS
stalemates in which caps, locks, or unsafe metadata state stop making
forward progress. The reset blocks new metadata work, attempts a
bounded best-effort drain of dirty client state while sessions are
still alive, and finally asks the MDS to close sessions before tearing
local session state down directly.
The reset state machine tracks four phases: IDLE -> QUIESCING ->
DRAINING -> TEARDOWN -> IDLE. QUIESCING is set synchronously by
schedule_reset() before the workqueue item is dispatched, so that new
metadata requests and file-lock acquisitions are gated immediately --
even before the work function begins running. All non-IDLE phases
block callers on blocked_wq, preventing races with session teardown.
The drain phase flushes mdlog state, dirty caps, and pending cap
releases for a bounded interval. State that still cannot make progress
within that interval is discarded during teardown, which is the point
of the reset: break the stalemate and allow fresh sessions to rebuild
clean state.
The session teardown follows the established check_new_map()
forced-close pattern: unregister sessions under mdsc->mutex, then clean
up caps and requests under s->s_mutex. Reconnect is not attempted
because the MDS only accepts reconnects during its own RECONNECT phase
after restart, not from an active client.
Blocked callers are released when reset completes and observe the final
result via -EAGAIN (reset failed) or 0 (success). Internal work-function
errors such as -ENOMEM are not propagated to unrelated callers like
open() or flock(); the detailed error remains in debugfs and
tracepoints.
The work function checks st->shutdown before each phase transition
(DRAINING, TEARDOWN) so that a concurrent ceph_mdsc_destroy() is not
overwritten. If destroy already took ownership, the work function
releases session references and returns without touching the state.
The timeout calculation for blocked-request waiters uses max_t() to
prevent jiffies underflow when the deadline has already passed.
The close-grace sleep before teardown is a best-effort nudge to let
queued REQUEST_CLOSE messages egress; it is not a correctness
requirement since the MDS still has session_autoclose as a fallback.
The destroy path marks reset as failed and wakes blocked waiters before
cancel_work_sync() so unmount does not stall.
Signed-off-by: Alex Markuze <amarkuze@redhat.com>
Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
Convert wait_caps_flush() from a silent indefinite wait into a diagnostic
wait loop that periodically dumps pending cap flush state.
The underlying wait semantics remain intact: callers still wait until the
requested cap flushes complete. The difference is that long stalls now
produce actionable diagnostics instead of looking like a silent hang.
CEPH_CAP_FLUSH_MAX_DUMP_ENTRIES limits the number of entries
emitted per diagnostic dump, and CEPH_CAP_FLUSH_MAX_DUMP_ITERS
limits the number of timed diagnostic dumps before the wait
continues silently. When more entries exist than the per-dump
limit, a truncation count is reported. When the dump iteration
limit is reached, a final suppression message is emitted so the
transition to silence is explicit.
The diagnostic dump collects flush entry data under cap_dirty_lock into
a bounded on-stack array, then prints after releasing the lock. This
avoids holding the spinlock across printk calls.
A null cf->ci on the global flush list indicates a bug since all
cap_flush entries are initialized with a valid ci before being added.
Signal this with WARN_ON_ONCE while still printing enough context for
debugging.
READ_ONCE is used for the i_last_cap_flush_ack field, which is read
outside the inode lock domain. Flush tids are monotonically increasing
and acks are processed in order under i_ceph_lock, so the latest ack
tid is always the most recently written value.
Add a ci pointer to struct ceph_cap_flush so that the diagnostic
dump can identify which inode each pending flush belongs to. The
new i_last_cap_flush_ack field tracks the latest acknowledged flush
tid per inode for diagnostic correlation.
This improves reset-drain observability and is also useful for
existing sync and writeback troubleshooting paths.
Signed-off-by: Alex Markuze <amarkuze@redhat.com>
Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
Change send_mds_reconnect() to return an error code so callers can detect
and report reconnect failures instead of silently ignoring them. Add early
bailout checks for sessions that are already closed, rejected, or
unregistered, which avoids sending reconnect messages for sessions that
can no longer be recovered.
The early -ESTALE and -ENOENT bailouts use a separate fail_return label
that skips the pr_err_client diagnostic, since these codes indicate
expected concurrent-teardown races rather than genuine reconnect build
failures.
Move the "reconnect start" log after the early-bailout checks so it
only appears for sessions that actually proceed with reconnect.
Save the prior session state before transitioning to RECONNECTING,
and restore it in the failure path. Without this, a transient
build or encoding failure (-ENOMEM, -ENOSPC) strands the session
in RECONNECTING indefinitely because check_new_map() only retries
sessions in RESTARTING state.
Rewrite mds_peer_reset() to handle the case where the MDS is past its
RECONNECT phase (i.e. active). An active MDS rejects CLIENT_RECONNECT
messages because it only accepts them during its own RECONNECT window
after restart. Previously, the client would send a doomed reconnect
that the MDS would reject or ignore. Now, the client tears the session
down locally and lets new requests re-open a fresh session, which is
the correct recovery for this scenario. The RECONNECTING state is
handled on the same teardown path, since the MDS will reject reconnect
attempts from an active client regardless of the session's local state.
Add explicit cases for CLOSED and REJECTED session states in
mds_peer_reset() since these are terminal states where a connection
drop is expected behavior.
The session teardown path in mds_peer_reset() follows the established
drop-and-reacquire locking pattern from check_new_map(): take
mdsc->mutex for session unregistration, release it, then take s->s_mutex
separately for cleanup. This avoids introducing a new simultaneous lock
nesting pattern.
Log reconnect failures from check_new_map() and mds_peer_reset() at
pr_warn level rather than pr_err, since return codes like -ESTALE
(closed/rejected session) and -ENOENT (unregistered session) are
expected during concurrent teardown. Log dropped messages for
unregistered sessions via doutc() (dynamic debug) rather than
pr_info, as post-reset message arrival is routine and does not
warrant unconditional logging.
Signed-off-by: Alex Markuze <amarkuze@redhat.com>
Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
Replace the __force __le32 cast with cpu_to_le32() for the flock_len field
in reconnect_caps_cb(). The old code used a type-system bypass to silence
sparse; the new form uses the proper endian conversion macro.
Also switch from a raw bitmask test against i_ceph_flags to test_bit() on
the named CEPH_I_ERROR_FILELOCK_BIT, which is the correct accessor for the
unsigned long flags field after the bit-position conversion.
Remove the now-unused CEPH_I_ERROR_FILELOCK mask define since all callers
use the _BIT form with test_bit/set_bit/clear_bit.
Signed-off-by: Alex Markuze <amarkuze@redhat.com>
Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
Define named bit-position constants for all CEPH_I_* inode flags and
derive the bitmask values from them. This gives every flag a named
_BIT constant usable with the test_bit/set_bit/clear_bit family.
The intentionally unused bit position 1 is documented inline.
Convert all flag modifications to use atomic bitops (set_bit,
clear_bit, test_and_clear_bit). The previous code mixed lockless
atomic ops on some flags (ERROR_WRITE, ODIRECT) with non-atomic
read-modify-write (|= / &= ~) on other flags sharing the same
unsigned long. A concurrent non-atomic RMW can clobber an
adjacent lockless atomic update -- for example, a lockless
clear_bit(ERROR_WRITE) could be silently resurrected by a
concurrent ci->i_ceph_flags |= CEPH_I_FLUSH under the spinlock.
Using atomic bitops for all modifications eliminates this class
of race entirely.
Flags whose only users are now the _BIT form (ERROR_WRITE,
ASYNC_CHECK_CAPS) have their old mask defines removed to document
that callers must use the _BIT constant with the set_bit/test_bit
family. ERROR_FILELOCK and SHUTDOWN retain their mask defines
because they are still used via bitmask tests in lockless readers
(ceph_inode_is_shutdown, reconnect_caps_cb).
The direct assignment in ceph_finish_async_create() is converted
from i_ceph_flags = CEPH_I_ASYNC_CREATE to set_bit(). This
inode is I_NEW at this point -- still invisible to other threads
and guaranteed to have zero flags from alloc_inode -- so either
form is safe, but set_bit() keeps the conversion uniform.
Signed-off-by: Alex Markuze <amarkuze@redhat.com>
Reviewed-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Viacheslav Dubeyko <Slava.Dubeyko@ibm.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
