Age | Commit message (Collapse) | Author | Files | Lines |
|
We add the new state to the nfsi->open_states list, making it
potentially visible to other threads, before we've finished initializing
it.
That wasn't a problem when all the readers were also taking the i_lock
(as we do here), but since we switched to RCU, there's now a possibility
that a reader could see the partially initialized state.
Symptoms observed were a crash when another thread called
nfs4_get_valid_delegation() on a NULL inode, resulting in an oops like:
BUG: unable to handle page fault for address: ffffffffffffffb0 ...
RIP: 0010:nfs4_get_valid_delegation+0x6/0x30 [nfsv4] ...
Call Trace:
nfs4_open_prepare+0x80/0x1c0 [nfsv4]
__rpc_execute+0x75/0x390 [sunrpc]
? finish_task_switch+0x75/0x260
rpc_async_schedule+0x29/0x40 [sunrpc]
process_one_work+0x1ad/0x370
worker_thread+0x30/0x390
? create_worker+0x1a0/0x1a0
kthread+0x10c/0x130
? kthread_park+0x80/0x80
ret_from_fork+0x22/0x30
Fixes: 9ae075fdd190 "NFSv4: Convert open state lookup to use RCU"
Reviewed-by: Seiichi Ikarashi <s.ikarashi@fujitsu.com>
Tested-by: Daisuke Matsuda <matsuda-daisuke@fujitsu.com>
Tested-by: Masayoshi Mizuma <m.mizuma@jp.fujitsu.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Cc: stable@vger.kernel.org # v4.20+
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
|
|
rxrpc currently uses a fixed 4s retransmission timeout until the RTT is
sufficiently sampled. This can cause problems with some fileservers with
calls to the cache manager in the afs filesystem being dropped from the
fileserver because a packet goes missing and the retransmission timeout is
greater than the call expiry timeout.
Fix this by:
(1) Copying the RTT/RTO calculation code from Linux's TCP implementation
and altering it to fit rxrpc.
(2) Altering the various users of the RTT to make use of the new SRTT
value.
(3) Replacing the use of rxrpc_resend_timeout to use the calculated RTO
value instead (which is needed in jiffies), along with a backoff.
Notes:
(1) rxrpc provides RTT samples by matching the serial numbers on outgoing
DATA packets that have the RXRPC_REQUEST_ACK set and PING ACK packets
against the reference serial number in incoming REQUESTED ACK and
PING-RESPONSE ACK packets.
(2) Each packet that is transmitted on an rxrpc connection gets a new
per-connection serial number, even for retransmissions, so an ACK can
be cross-referenced to a specific trigger packet. This allows RTT
information to be drawn from retransmitted DATA packets also.
(3) rxrpc maintains the RTT/RTO state on the rxrpc_peer record rather than
on an rxrpc_call because many RPC calls won't live long enough to
generate more than one sample.
(4) The calculated SRTT value is in units of 8ths of a microsecond rather
than nanoseconds.
The (S)RTT and RTO values are displayed in /proc/net/rxrpc/peers.
Fixes: 17926a79320a ([AF_RXRPC]: Provide secure RxRPC sockets for use by userspace and kernel both"")
Signed-off-by: David Howells <dhowells@redhat.com>
|
|
Pull block fixes from Jens Axboe:
- a small series fixing a use-after-free of bdi name (Christoph,Yufen)
- NVMe fix for a regression with the smaller CQ update (Alexey)
- NVMe fix for a hang at namespace scanning error recovery (Sagi)
- fix race with blk-iocost iocg->abs_vdebt updates (Tejun)
* tag 'block-5.7-2020-05-09' of git://git.kernel.dk/linux-block:
nvme: fix possible hang when ns scanning fails during error recovery
nvme-pci: fix "slimmer CQ head update"
bdi: add a ->dev_name field to struct backing_dev_info
bdi: use bdi_dev_name() to get device name
bdi: move bdi_dev_name out of line
vboxsf: don't use the source name in the bdi name
iocost: protect iocg->abs_vdebt with iocg->waitq.lock
|
|
Use the common interface bdi_dev_name() to get device name.
Signed-off-by: Yufen Yu <yuyufen@huawei.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Bart Van Assche <bvanassche@acm.org>
Add missing <linux/backing-dev.h> include BFQ
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Pull io_uring fixes from Jens Axboe:
- Fix finish_wait() balancing in file cancelation (Xiaoguang)
- Ensure early cleanup of resources in ring map failure (Xiaoguang)
- Ensure IORING_OP_SLICE does the right file mode checks (Pavel)
- Remove file opening from openat/openat2/statx, it's not needed and
messes with O_PATH
* tag 'io_uring-5.7-2020-05-08' of git://git.kernel.dk/linux-block:
io_uring: don't use 'fd' for openat/openat2/statx
splice: move f_mode checks to do_{splice,tee}()
io_uring: handle -EFAULT properly in io_uring_setup()
io_uring: fix mismatched finish_wait() calls in io_uring_cancel_files()
|
|
do_splice() doesn't expect len to be 0. Just always return 0 in this
case as splice(2) does.
Fixes: 7d67af2c0134 ("io_uring: add splice(2) support")
Reported-by: Jann Horn <jannh@google.com>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
There is a potential race in fscache operation enqueuing for reading and
copying multiple pages from cachefiles to netfs. The problem can be seen
easily on a heavy loaded system (for example many processes reading files
continually on an NFS share covered by fscache triggered this problem within
a few minutes).
The race is due to cachefiles_read_waiter() adding the op to the monitor
to_do list and then then drop the object->work_lock spinlock before
completing fscache_enqueue_operation(). Once the lock is dropped,
cachefiles_read_copier() grabs the op, completes processing it, and
makes it through fscache_retrieval_complete() which sets the op->state to
the final state of FSCACHE_OP_ST_COMPLETE(4). When cachefiles_read_waiter()
finally gets through the remainder of fscache_enqueue_operation()
it sees the invalid state, and hits the ASSERTCMP and the following
oops is seen:
[ 2259.612361] FS-Cache:
[ 2259.614785] FS-Cache: Assertion failed
[ 2259.618639] FS-Cache: 4 == 5 is false
[ 2259.622456] ------------[ cut here ]------------
[ 2259.627190] kernel BUG at fs/fscache/operation.c:70!
...
[ 2259.791675] RIP: 0010:[<ffffffffc061b4cf>] [<ffffffffc061b4cf>] fscache_enqueue_operation+0xff/0x170 [fscache]
[ 2259.802059] RSP: 0000:ffffa0263d543be0 EFLAGS: 00010046
[ 2259.807521] RAX: 0000000000000019 RBX: ffffa01a4d390480 RCX: 0000000000000006
[ 2259.814847] RDX: 0000000000000000 RSI: 0000000000000046 RDI: ffffa0263d553890
[ 2259.822176] RBP: ffffa0263d543be8 R08: 0000000000000000 R09: ffffa0263c2d8708
[ 2259.829502] R10: 0000000000001e7f R11: 0000000000000000 R12: ffffa01a4d390480
[ 2259.844483] R13: ffff9fa9546c5920 R14: ffffa0263d543c80 R15: ffffa0293ff9bf10
[ 2259.859554] FS: 00007f4b6efbd700(0000) GS:ffffa0263d540000(0000) knlGS:0000000000000000
[ 2259.875571] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 2259.889117] CR2: 00007f49e1624ff0 CR3: 0000012b38b38000 CR4: 00000000007607e0
[ 2259.904015] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 2259.918764] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 2259.933449] PKRU: 55555554
[ 2259.943654] Call Trace:
[ 2259.953592] <IRQ>
[ 2259.955577] [<ffffffffc03a7c12>] cachefiles_read_waiter+0x92/0xf0 [cachefiles]
[ 2259.978039] [<ffffffffa34d3942>] __wake_up_common+0x82/0x120
[ 2259.991392] [<ffffffffa34d3a63>] __wake_up_common_lock+0x83/0xc0
[ 2260.004930] [<ffffffffa34d3510>] ? task_rq_unlock+0x20/0x20
[ 2260.017863] [<ffffffffa34d3ab3>] __wake_up+0x13/0x20
[ 2260.030230] [<ffffffffa34c72a0>] __wake_up_bit+0x50/0x70
[ 2260.042535] [<ffffffffa35bdcdb>] unlock_page+0x2b/0x30
[ 2260.054495] [<ffffffffa35bdd09>] page_endio+0x29/0x90
[ 2260.066184] [<ffffffffa368fc81>] mpage_end_io+0x51/0x80
CPU1
cachefiles_read_waiter()
20 static int cachefiles_read_waiter(wait_queue_entry_t *wait, unsigned mode,
21 int sync, void *_key)
22 {
...
61 spin_lock(&object->work_lock);
62 list_add_tail(&monitor->op_link, &op->to_do);
63 spin_unlock(&object->work_lock);
<begin race window>
64
65 fscache_enqueue_retrieval(op);
182 static inline void fscache_enqueue_retrieval(struct fscache_retrieval *op)
183 {
184 fscache_enqueue_operation(&op->op);
185 }
58 void fscache_enqueue_operation(struct fscache_operation *op)
59 {
60 struct fscache_cookie *cookie = op->object->cookie;
61
62 _enter("{OBJ%x OP%x,%u}",
63 op->object->debug_id, op->debug_id, atomic_read(&op->usage));
64
65 ASSERT(list_empty(&op->pend_link));
66 ASSERT(op->processor != NULL);
67 ASSERT(fscache_object_is_available(op->object));
68 ASSERTCMP(atomic_read(&op->usage), >, 0);
<end race window>
CPU2
cachefiles_read_copier()
168 while (!list_empty(&op->to_do)) {
...
202 fscache_end_io(op, monitor->netfs_page, error);
203 put_page(monitor->netfs_page);
204 fscache_retrieval_complete(op, 1);
CPU1
58 void fscache_enqueue_operation(struct fscache_operation *op)
59 {
...
69 ASSERTIFCMP(op->state != FSCACHE_OP_ST_IN_PROGRESS,
70 op->state, ==, FSCACHE_OP_ST_CANCELLED);
Signed-off-by: Lei Xue <carmark.dlut@gmail.com>
Signed-off-by: Dave Wysochanski <dwysocha@redhat.com>
Signed-off-by: David Howells <dhowells@redhat.com>
|
|
Commit 402cb8dda949 ("fscache: Attach the index key and aux data to
the cookie") added the aux_data and aux_data_len to parameters to
fscache_acquire_cookie(), and updated the callers in the NFS client.
In the process it modified the aux_data to include the change_attr,
but missed adding change_attr to a couple places where aux_data was
used. Specifically, when opening a file and the change_attr is not
added, the following attempt to lookup an object will fail inside
cachefiles_check_object_xattr() = -116 due to
nfs_fscache_inode_check_aux() failing memcmp on auxdata and returning
FSCACHE_CHECKAUX_OBSOLETE.
Fix this by adding nfs_fscache_update_auxdata() to set the auxdata
from all relevant fields in the inode, including the change_attr.
Fixes: 402cb8dda949 ("fscache: Attach the index key and aux data to the cookie")
Signed-off-by: Dave Wysochanski <dwysocha@redhat.com>
Signed-off-by: David Howells <dhowells@redhat.com>
|
|
Commit f2aedb713c28 ("NFS: Add fs_context support.") reworked
NFS mount code paths for fs_context support which included
super_block initialization. In the process there was an extra
return left in the code and so we never call
nfs_fscache_get_super_cookie even if 'fsc' is given on as mount
option. In addition, there is an extra check inside
nfs_fscache_get_super_cookie for the NFS_OPTION_FSCACHE which
is unnecessary since the only caller nfs_get_cache_cookie
checks this flag.
Fixes: f2aedb713c28 ("NFS: Add fs_context support.")
Signed-off-by: Dave Wysochanski <dwysocha@redhat.com>
Signed-off-by: David Howells <dhowells@redhat.com>
|
|
Commit 402cb8dda949 ("fscache: Attach the index key and aux data to
the cookie") added the index_key and index_key_len parameters to
fscache_acquire_cookie(), and updated the callers in the NFS client.
One of the callers was inside nfs_fscache_get_super_cookie()
and was changed to use the full struct nfs_fscache_key as the
index_key. However, a couple members of this structure contain
pointers and thus will change each time the same NFS share is
remounted. Since index_key is used for fscache_cookie->key_hash
and this subsequently is used to compare cookies, the effectiveness
of fscache with NFS is reduced to the point at which a umount
occurs. Any subsequent remount of the same share will cause a
unique NFS super_block index_key and key_hash to be generated for
the same data, rendering any prior fscache data unable to be
found. A simple reproducer demonstrates the problem.
1. Mount share with 'fsc', create a file, drop page cache
systemctl start cachefilesd
mount -o vers=3,fsc 127.0.0.1:/export /mnt
dd if=/dev/zero of=/mnt/file1.bin bs=4096 count=1
echo 3 > /proc/sys/vm/drop_caches
2. Read file into page cache and fscache, then unmount
dd if=/mnt/file1.bin of=/dev/null bs=4096 count=1
umount /mnt
3. Remount and re-read which should come from fscache
mount -o vers=3,fsc 127.0.0.1:/export /mnt
echo 3 > /proc/sys/vm/drop_caches
dd if=/mnt/file1.bin of=/dev/null bs=4096 count=1
4. Check for READ ops in mountstats - there should be none
grep READ: /proc/self/mountstats
Looking at the history and the removed function, nfs_super_get_key(),
we should only use nfs_fscache_key.key plus any uniquifier, for
the fscache index_key.
Fixes: 402cb8dda949 ("fscache: Attach the index key and aux data to the cookie")
Signed-off-by: Dave Wysochanski <dwysocha@redhat.com>
Signed-off-by: David Howells <dhowells@redhat.com>
|
|
This reverts commit df5db5f9ee112e76b5202fbc331f990a0fc316d6.
This patch fixes a regression: patch df5db5f9ee112 allowed function
run_queue() to bypass its call to do_xmote() if revokes were queued for
the glock. That's wrong because its call to do_xmote() is what is
responsible for calling the go_sync() glops functions to sync both
the ail list and any revokes queued for it. By bypassing the call,
gfs2 could get into a stand-off where the glock could not be demoted
until its revokes are written back, but the revokes would not be
written back because do_xmote() was never called.
It "sort of" works, however, because there are other mechanisms like
the log flush daemon (logd) that can sync the ail items and revokes,
if it deems it necessary. The problem is: without file system pressure,
it might never deem it necessary.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
|
|
Before this patch, if the go_sync operation returned an error during
the do_xmote process (such as unable to sync metadata to the journal)
the code did goto out. That kept the glock locked, so it could not be
given away, which correctly avoids file system corruption. However,
it never set the withdraw bit or requeueing the glock work. So it would
hang forever, unable to ever demote the glock.
This patch changes to goto to a new label, skip_inval, so that errors
from go_sync are treated the same way as errors from go_inval:
The delayed withdraw bit is set and the work is requeued. That way,
the logd should eventually figure out there's a problem and withdraw
properly there.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
|
|
Pull ceph fixes from Ilya Dryomov:
"Fixes for an endianness handling bug that prevented mounts on
big-endian arches, a spammy log message and a couple error paths.
Also included a MAINTAINERS update"
* tag 'ceph-for-5.7-rc5' of git://github.com/ceph/ceph-client:
ceph: demote quotarealm lookup warning to a debug message
MAINTAINERS: remove myself as ceph co-maintainer
ceph: fix double unlock in handle_cap_export()
ceph: fix special error code in ceph_try_get_caps()
ceph: fix endianness bug when handling MDS session feature bits
|
|
This patch rearranges gfs2_add_revoke so that the extra glock
reference is added earlier on in the function to avoid races in which
the glock is freed before the new reference is taken.
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
|
|
Before this patch, function gfs2_quota_unlock checked if quotas are
turned off, and if so, it branched to label out, which called
gfs2_quota_unhold. With the new system of gfs2_qa_get and put, we
no longer want to call gfs2_quota_unhold or we won't balance our
gets and puts.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
|
|
Before this patch, function gfs2_quota_lock checked if it was called
from a privileged user, and if so, it bypassed the quota check:
superuser can operate outside the quotas.
That's the wrong place for the check because the lock/unlock functions
are separate from the lock_check function, and you can do lock and
unlock without actually checking the quotas.
This patch moves the check to gfs2_quota_lock_check.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
|
|
This patch removes a check from gfs2_quota_check for whether quotas
are enabled by the superblock. There is a test just prior for the
GIF_QD_LOCKED bit in the inode, and that can only be set by functions
that already check that quotas are enabled in the superblock.
Therefore, the check is redundant.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
|
|
Before this patch, gfs2_quota_change() would BUG_ON if the
qa_ref counter was not a positive number. This patch changes it to
be a withdraw instead. That way we can debug things more easily.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
|
|
This patch fixes a couple of places in which gfs2_qa_get and gfs2_qa_put are
not balanced: we now keep references around whenever a file is open for writing
(see gfs2_open_common and gfs2_release), so we need to put all references we
grab in function gfs2_create_inode. This was broken in the successful case and
on one error path.
This also means that we don't have a reference to put in gfs2_evict_inode.
In addition, gfs2_qa_put was called for the wrong inode in gfs2_link.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
|
|
A misconfigured cephx can easily result in having the kernel client
flooding the logs with:
ceph: Can't lookup inode 1 (err: -13)
Change this message to debug level.
Cc: stable@vger.kernel.org
URL: https://tracker.ceph.com/issues/44546
Signed-off-by: Luis Henriques <lhenriques@suse.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core
Pull driver core fixes from Greg KH:
"Here are a number of small driver core fixes for 5.7-rc5 to resolve a
bunch of reported issues with the current tree.
Biggest here are the reverts and patches from John Stultz to resolve a
bunch of deferred probe regressions we have been seeing in 5.7-rc
right now.
Along with those are some other smaller fixes:
- coredump crash fix
- devlink fix for when permissive mode was enabled
- amba and platform device dma_parms fixes
- component error silenced for when deferred probe happens
All of these have been in linux-next for a while with no reported
issues"
* tag 'driver-core-5.7-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core:
regulator: Revert "Use driver_deferred_probe_timeout for regulator_init_complete_work"
driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires
driver core: Use dev_warn() instead of dev_WARN() for deferred_probe_timeout warnings
driver core: Revert default driver_deferred_probe_timeout value to 0
component: Silence bind error on -EPROBE_DEFER
driver core: Fix handling of fw_devlink=permissive
coredump: fix crash when umh is disabled
amba: Initialize dma_parms for amba devices
driver core: platform: Initialize dma_parms for platform devices
|
|
Merge misc fixes from Andrew Morton:
"14 fixes and one selftest to verify the ipc fixes herein"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
mm: limit boost_watermark on small zones
ubsan: disable UBSAN_ALIGNMENT under COMPILE_TEST
mm/vmscan: remove unnecessary argument description of isolate_lru_pages()
epoll: atomically remove wait entry on wake up
kselftests: introduce new epoll60 testcase for catching lost wakeups
percpu: make pcpu_alloc() aware of current gfp context
mm/slub: fix incorrect interpretation of s->offset
scripts/gdb: repair rb_first() and rb_last()
eventpoll: fix missing wakeup for ovflist in ep_poll_callback
arch/x86/kvm/svm/sev.c: change flag passed to GUP fast in sev_pin_memory()
scripts/decodecode: fix trapping instruction formatting
kernel/kcov.c: fix typos in kcov_remote_start documentation
mm/page_alloc: fix watchdog soft lockups during set_zone_contiguous()
mm, memcg: fix error return value of mem_cgroup_css_alloc()
ipc/mqueue.c: change __do_notify() to bypass check_kill_permission()
|
|
It turns out that when extending an existing bio, gfs2_find_jhead fails to
check if the block number is consecutive, which leads to incorrect reads for
fragmented journals.
In addition, limit the maximum bio size to an arbitrary value of 2 megabytes:
since commit 07173c3ec276 ("block: enable multipage bvecs"), if we just keep
adding pages until bio_add_page fails, bios will grow much larger than useful,
which pins more memory than necessary with barely any additional performance
gains.
Fixes: f4686c26ecc3 ("gfs2: read journal in large chunks")
Cc: stable@vger.kernel.org # v5.2+
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
|
|
Make sure we don't walk past the end of the metadata in gfs2_walk_metadata: the
inode holds fewer pointers than indirect blocks.
Slightly clean up gfs2_iomap_get.
Fixes: a27a0c9b6a20 ("gfs2: gfs2_walk_metadata fix")
Cc: stable@vger.kernel.org # v5.3+
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
|
|
When the gfs2_logd daemon withdrew, the withdraw sequence called
into make_fs_ro() to make the file system read-only. That caused the
journal descriptors to be freed. However, those journal descriptors
were used by gfs2_logd's call to gfs2_ail_flush_reqd(). This caused
a use-after free and NULL pointer dereference.
This patch changes function gfs2_logd() so that it stops all logd
work until the thread is told to stop. Once a withdraw is done,
it only does an interruptible sleep.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
|
|
Before this patch, when the logd daemon was forced to withdraw, it
would try to request its journal be recovered by another cluster node.
However, in single-user cases with lock_nolock, there are no other
nodes to recover the journal. Function signal_our_withdraw() was
recognizing the lock_nolock situation, but not until after it had
evicted its journal inode. Since the journal descriptor that points
to the inode was never removed from the master list, when the unmount
occurred, it did another iput on the evicted inode, which resulted in
a BUG_ON(inode->i_state & I_CLEAR).
This patch moves the check for this situation earlier in function
signal_our_withdraw(), which avoids the extra iput, so the unmount
may happen normally.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
|
|
Before this patch, if an error was detected from glock function go_sync
by function do_xmote, it would return. But the function had temporarily
unlocked the gl_lockref spin_lock, and it never re-locked it. When the
caller of do_xmote tried to unlock it again, it was already unlocked,
which resulted in a corrupted spin_lock value.
This patch makes sure the gl_lockref spin_lock is re-locked after it is
unlocked.
Thanks to Wu Bo <wubo40@huawei.com> for reporting this problem.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
|
|
This patch does two things:
- fixes a lost wakeup introduced by commit 339ddb53d373 ("fs/epoll:
remove unnecessary wakeups of nested epoll")
- improves performance for events delivery.
The description of the problem is the following: if N (>1) threads are
waiting on ep->wq for new events and M (>1) events come, it is quite
likely that >1 wakeups hit the same wait queue entry, because there is
quite a big window between __add_wait_queue_exclusive() and the
following __remove_wait_queue() calls in ep_poll() function.
This can lead to lost wakeups, because thread, which was woken up, can
handle not all the events in ->rdllist. (in better words the problem is
described here: https://lkml.org/lkml/2019/10/7/905)
The idea of the current patch is to use init_wait() instead of
init_waitqueue_entry().
Internally init_wait() sets autoremove_wake_function as a callback,
which removes the wait entry atomically (under the wq locks) from the
list, thus the next coming wakeup hits the next wait entry in the wait
queue, thus preventing lost wakeups.
Problem is very well reproduced by the epoll60 test case [1].
Wait entry removal on wakeup has also performance benefits, because
there is no need to take a ep->lock and remove wait entry from the queue
after the successful wakeup. Here is the timing output of the epoll60
test case:
With explicit wakeup from ep_scan_ready_list() (the state of the
code prior 339ddb53d373):
real 0m6.970s
user 0m49.786s
sys 0m0.113s
After this patch:
real 0m5.220s
user 0m36.879s
sys 0m0.019s
The other testcase is the stress-epoll [2], where one thread consumes
all the events and other threads produce many events:
With explicit wakeup from ep_scan_ready_list() (the state of the
code prior 339ddb53d373):
threads events/ms run-time ms
8 5427 1474
16 6163 2596
32 6824 4689
64 7060 9064
128 6991 18309
After this patch:
threads events/ms run-time ms
8 5598 1429
16 7073 2262
32 7502 4265
64 7640 8376
128 7634 16767
(number of "events/ms" represents event bandwidth, thus higher is
better; number of "run-time ms" represents overall time spent
doing the benchmark, thus lower is better)
[1] tools/testing/selftests/filesystems/epoll/epoll_wakeup_test.c
[2] https://github.com/rouming/test-tools/blob/master/stress-epoll.c
Signed-off-by: Roman Penyaev <rpenyaev@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jason Baron <jbaron@akamai.com>
Cc: Khazhismel Kumykov <khazhy@google.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Heiher <r@hev.cc>
Cc: <stable@vger.kernel.org>
Link: http://lkml.kernel.org/r/20200430130326.1368509-2-rpenyaev@suse.de
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
In the event that we add to ovflist, before commit 339ddb53d373
("fs/epoll: remove unnecessary wakeups of nested epoll") we would be
woken up by ep_scan_ready_list, and did no wakeup in ep_poll_callback.
With that wakeup removed, if we add to ovflist here, we may never wake
up. Rather than adding back the ep_scan_ready_list wakeup - which was
resulting in unnecessary wakeups, trigger a wake-up in ep_poll_callback.
We noticed that one of our workloads was missing wakeups starting with
339ddb53d373 and upon manual inspection, this wakeup seemed missing to me.
With this patch added, we no longer see missing wakeups. I haven't yet
tried to make a small reproducer, but the existing kselftests in
filesystem/epoll passed for me with this patch.
[khazhy@google.com: use if/elif instead of goto + cleanup suggested by Roman]
Link: http://lkml.kernel.org/r/20200424190039.192373-1-khazhy@google.com
Fixes: 339ddb53d373 ("fs/epoll: remove unnecessary wakeups of nested epoll")
Signed-off-by: Khazhismel Kumykov <khazhy@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Roman Penyaev <rpenyaev@suse.de>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Roman Penyaev <rpenyaev@suse.de>
Cc: Heiher <r@hev.cc>
Cc: Jason Baron <jbaron@akamai.com>
Cc: <stable@vger.kernel.org>
Link: http://lkml.kernel.org/r/20200424025057.118641-1-khazhy@google.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
|
|
We currently make some guesses as when to open this fd, but in reality
we have no business (or need) to do so at all. In fact, it makes certain
things fail, like O_PATH.
Remove the fd lookup from these opcodes, we're just passing the 'fd' to
generic helpers anyway. With that, we can also remove the special casing
of fd values in io_req_needs_file(), and the 'fd_non_neg' check that
we have. And we can ensure that we only read sqe->fd once.
This fixes O_PATH usage with openat/openat2, and ditto statx path side
oddities.
Cc: stable@vger.kernel.org: # v5.6
Reported-by: Max Kellermann <mk@cm4all.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Pull configfs fix from Christoph Hellwig:
"Fix a refcount leak in configfs_rmdir (Xiyu Yang)"
* tag 'configfs-for-5.7' of git://git.infradead.org/users/hch/configfs:
configfs: fix config_item refcnt leak in configfs_rmdir()
|
|
do_splice() is used by io_uring, as will be do_tee(). Move f_mode
checks from sys_{splice,tee}() to do_{splice,tee}(), so they're
enforced for io_uring as well.
Fixes: 7d67af2c0134 ("io_uring: add splice(2) support")
Reported-by: Jann Horn <jannh@google.com>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Simplify the bdi name to mirror what we are doing elsewhere, and
drop them name in favor of just using a number. This avoids a
potentially very long bdi name.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
After a gfs2 file system withdraw, any attempt to read metadata is
automatically rejected by function gfs2_meta_read() except for reads
of the journal inode. This turns out to be a problem because function
signal_our_withdraw() repeatedly calls check_journal_clean() which reads
the metadata (both its dinode and indirect blocks) to see if the entire
journal is mapped. The dinode read works, but reading the indirect blocks
returns -EIO which gets sent back up and causes a consistency error.
This results in withdraw-from-withdraw, which becomes a deadlock.
This patch changes the test in gfs2_meta_read() to allow all metadata
reads for the journal. Instead of checking the journal block, it now
checks for the journal inode glock which is the same for all blocks in
the journal. This allows check_journal_clean() to properly check the
journal without trying to withdraw recursively.
Signed-off-by: Bob Peterson <rpeterso@redhat.com>
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
|
|
As per POSIX, the correct spelling is EACCES:
include/uapi/asm-generic/errno-base.h:#define EACCES 13 /* Permission denied */
Fixes: b8f7442bc46e48fb ("CIFS: refactor cifs_get_inode_info()")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Steve French <stfrench@microsoft.com>
|
|
If copy_to_user() in io_uring_setup() failed, we'll leak many kernel
resources, which will be recycled until process terminates. This bug
can be reproduced by using mprotect to set params to PROT_READ. To fix
this issue, refactor io_uring_create() a bit to add a new 'struct
io_uring_params __user *params' parameter and move the copy_to_user()
in io_uring_setup() to io_uring_setup(), if copy_to_user() failed,
we can free kernel resource properly.
Suggested-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Xiaoguang Wang <xiaoguang.wang@linux.alibaba.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
If the ceph_mdsc_open_export_target_session() return fails, it will
do a "goto retry", but the session mutex has already been unlocked.
Re-lock the mutex in that case to ensure that we don't unlock it
twice.
Signed-off-by: Wu Bo <wubo40@huawei.com>
Reviewed-by: "Yan, Zheng" <zyan@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
There are 3 speical error codes: -EAGAIN/-EFBIG/-ESTALE.
After calling try_get_cap_refs, ceph_try_get_caps test for the
-EAGAIN twice. Ensure that it tests for -ESTALE instead.
Signed-off-by: Wu Bo <wubo40@huawei.com>
Reviewed-by: Jeff Layton <jlayton@kernel.org>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
Eduard reported a problem mounting cephfs on s390 arch. The feature
mask sent by the MDS is little-endian, so we need to convert it
before storing and testing against it.
Cc: stable@vger.kernel.org
Reported-and-Tested-by: Eduard Shishkin <edward6@linux.ibm.com>
Signed-off-by: Jeff Layton <jlayton@kernel.org>
Reviewed-by: "Yan, Zheng" <zyan@redhat.com>
Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
|
|
cachefiles_read_or_alloc_pages()
The patch which changed cachefiles from calling ->bmap() to using the
bmap() wrapper overwrote the running return value with the result of
calling bmap(). This causes an assertion failure elsewhere in the code.
Fix this by using ret2 rather than ret to hold the return value.
The oops looks like:
kernel BUG at fs/nfs/fscache.c:468!
...
RIP: 0010:__nfs_readpages_from_fscache+0x18b/0x190 [nfs]
...
Call Trace:
nfs_readpages+0xbf/0x1c0 [nfs]
? __alloc_pages_nodemask+0x16c/0x320
read_pages+0x67/0x1a0
__do_page_cache_readahead+0x1cf/0x1f0
ondemand_readahead+0x172/0x2b0
page_cache_async_readahead+0xaa/0xe0
generic_file_buffered_read+0x852/0xd50
? mem_cgroup_commit_charge+0x6e/0x140
? nfs4_have_delegation+0x19/0x30 [nfsv4]
generic_file_read_iter+0x100/0x140
? nfs_revalidate_mapping+0x176/0x2b0 [nfs]
nfs_file_read+0x6d/0xc0 [nfs]
new_sync_read+0x11a/0x1c0
__vfs_read+0x29/0x40
vfs_read+0x8e/0x140
ksys_read+0x61/0xd0
__x64_sys_read+0x1a/0x20
do_syscall_64+0x60/0x1e0
entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x7f5d148267e0
Fixes: 10d83e11a582 ("cachefiles: drop direct usage of ->bmap method.")
Reported-by: David Wysochanski <dwysocha@redhat.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Tested-by: David Wysochanski <dwysocha@redhat.com>
cc: Carlos Maiolino <cmaiolino@redhat.com>
|
|
The prepare_to_wait() and finish_wait() calls in io_uring_cancel_files()
are mismatched. Currently I don't see any issues related this bug, just
find it by learning codes.
Signed-off-by: Xiaoguang Wang <xiaoguang.wang@linux.alibaba.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux
Pull more btrfs fixes from David Sterba:
"A few more stability fixes, minor build warning fixes and git url
fixup:
- fix partial loss of prealloc extent past i_size after fsync
- fix potential deadlock due to wrong transaction handle passing via
journal_info
- fix gcc 4.8 struct intialization warning
- update git URL in MAINTAINERS entry"
* tag 'for-5.7-rc3-tag' of git://git.kernel.org/pub/scm/linux/kernel/git/kdave/linux:
MAINTAINERS: btrfs: fix git repo URL
btrfs: fix gcc-4.8 build warning for struct initializer
btrfs: transaction: Avoid deadlock due to bad initialization timing of fs_info::journal_info
btrfs: fix partial loss of prealloc extent past i_size after fsync
|
|
Pull iomap fix from Darrick Wong:
"Hoist the check for an unrepresentable FIBMAP return value into
ioctl_fibmap.
The internal kernel function can handle 64-bit values (and is needed
to fix a regression on ext4 + jbd2). It is only the userspace ioctl
that is so old that it cannot deal"
* tag 'iomap-5.7-fixes-1' of git://git.kernel.org/pub/scm/fs/xfs/xfs-linux:
fibmap: Warn and return an error in case of block > INT_MAX
|
|
Pull NFS client bugfixes from Trond Myklebust:
"Highlights include:
Stable fixes:
- fix handling of backchannel binding in BIND_CONN_TO_SESSION
Bugfixes:
- Fix a credential use-after-free issue in pnfs_roc()
- Fix potential posix_acl refcnt leak in nfs3_set_acl
- defer slow parts of rpc_free_client() to a workqueue
- Fix an Oopsable race in __nfs_list_for_each_server()
- Fix trace point use-after-free race
- Regression: the RDMA client no longer responds to server disconnect
requests
- Fix return values of xdr_stream_encode_item_{present, absent}
- _pnfs_return_layout() must always wait for layoutreturn completion
Cleanups:
- Remove unreachable error conditions"
* tag 'nfs-for-5.7-4' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
NFS: Fix a race in __nfs_list_for_each_server()
NFSv4.1: fix handling of backchannel binding in BIND_CONN_TO_SESSION
SUNRPC: defer slow parts of rpc_free_client() to a workqueue.
NFSv4: Remove unreachable error condition due to rpc_run_task()
SUNRPC: Remove unreachable error condition
xprtrdma: Fix use of xdr_stream_encode_item_{present, absent}
xprtrdma: Fix trace point use-after-free race
xprtrdma: Restore wake-up-all to rpcrdma_cm_event_handler()
nfs: Fix potential posix_acl refcnt leak in nfs3_set_acl
NFS/pnfs: Fix a credential use-after-free issue in pnfs_roc()
NFS/pnfs: Ensure that _pnfs_return_layout() waits for layoutreturn completion
|
|
Pull io_uring fixes from Jens Axboe:
- Fix for statx not grabbing the file table, making AT_EMPTY_PATH fail
- Cover a few cases where async poll can handle retry, eliminating the
need for an async thread
- fallback request busy/free fix (Bijan)
- syzbot reported SQPOLL thread exit fix for non-preempt (Xiaoguang)
- Fix extra put of req for sync_file_range (Pavel)
- Always punt splice async. We'll improve this for 5.8, but wanted to
eliminate the inode mutex lock from the non-blocking path for 5.7
(Pavel)
* tag 'io_uring-5.7-2020-05-01' of git://git.kernel.dk/linux-block:
io_uring: punt splice async because of inode mutex
io_uring: check non-sync defer_list carefully
io_uring: fix extra put in sync_file_range()
io_uring: use cond_resched() in io_ring_ctx_wait_and_kill()
io_uring: use proper references for fallback_req locking
io_uring: only force async punt if poll based retry can't handle it
io_uring: enable poll retry for any file with ->read_iter / ->write_iter
io_uring: statx must grab the file table for valid fd
|
|
Nonblocking do_splice() still may wait for some time on an inode mutex.
Let's play safe and always punt it async.
Reported-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
io_req_defer() do double-checked locking. Use proper helpers for that,
i.e. list_empty_careful().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
[ 40.179474] refcount_t: underflow; use-after-free.
[ 40.179499] WARNING: CPU: 6 PID: 1848 at lib/refcount.c:28 refcount_warn_saturate+0xae/0xf0
...
[ 40.179612] RIP: 0010:refcount_warn_saturate+0xae/0xf0
[ 40.179617] Code: 28 44 0a 01 01 e8 d7 01 c2 ff 0f 0b 5d c3 80 3d 15 44 0a 01 00 75 91 48 c7 c7 b8 f5 75 be c6 05 05 44 0a 01 01 e8 b7 01 c2 ff <0f> 0b 5d c3 80 3d f3 43 0a 01 00 0f 85 6d ff ff ff 48 c7 c7 10 f6
[ 40.179619] RSP: 0018:ffffb252423ebe18 EFLAGS: 00010286
[ 40.179623] RAX: 0000000000000000 RBX: ffff98d65e929400 RCX: 0000000000000000
[ 40.179625] RDX: 0000000000000001 RSI: 0000000000000086 RDI: 00000000ffffffff
[ 40.179627] RBP: ffffb252423ebe18 R08: 0000000000000001 R09: 000000000000055d
[ 40.179629] R10: 0000000000000c8c R11: 0000000000000001 R12: 0000000000000000
[ 40.179631] R13: ffff98d68c434400 R14: ffff98d6a9cbaa20 R15: ffff98d6a609ccb8
[ 40.179634] FS: 0000000000000000(0000) GS:ffff98d6af580000(0000) knlGS:0000000000000000
[ 40.179636] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 40.179638] CR2: 00000000033e3194 CR3: 000000006480a003 CR4: 00000000003606e0
[ 40.179641] Call Trace:
[ 40.179652] io_put_req+0x36/0x40
[ 40.179657] io_free_work+0x15/0x20
[ 40.179661] io_worker_handle_work+0x2f5/0x480
[ 40.179667] io_wqe_worker+0x2a9/0x360
[ 40.179674] ? _raw_spin_unlock_irqrestore+0x24/0x40
[ 40.179681] kthread+0x12c/0x170
[ 40.179685] ? io_worker_handle_work+0x480/0x480
[ 40.179690] ? kthread_park+0x90/0x90
[ 40.179695] ret_from_fork+0x35/0x40
[ 40.179702] ---[ end trace 85027405f00110aa ]---
Opcode handler must never put submission ref, but that's what
io_sync_file_range_finish() do. use io_steal_work() there.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
While working on to make io_uring sqpoll mode support syscalls that need
struct files_struct, I got cpu soft lockup in io_ring_ctx_wait_and_kill(),
while (ctx->sqo_thread && !wq_has_sleeper(&ctx->sqo_wait))
cpu_relax();
above loop never has an chance to exit, it's because preempt isn't enabled
in the kernel, and the context calling io_ring_ctx_wait_and_kill() and
io_sq_thread() run in the same cpu, if io_sq_thread calls a cond_resched()
yield cpu and another context enters above loop, then io_sq_thread() will
always in runqueue and never exit.
Use cond_resched() can fix this issue.
Reported-by: syzbot+66243bb7126c410cefe6@syzkaller.appspotmail.com
Signed-off-by: Xiaoguang Wang <xiaoguang.wang@linux.alibaba.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|
|
Use ctx->fallback_req address for test_and_set_bit_lock() and
clear_bit_unlock().
Signed-off-by: Bijan Mottahedeh <bijan.mottahedeh@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
|