I have observed that the current size of 8M gives us pretty poor
fragmentation on multi-threaded workloads which do lots of writes.
Generally, I can increase the size of local alloc windows and observe a
marked decrease in fragmentation, even up and beyond window sizes of 512
megabytes. This makes sense for a couple reasons - larger local alloc means
more room for reservation windows. On multi-node workloads the larger local
alloc helps as well because we don't have to do window slides as often.
Also, I removed the OCFS2_DEFAULT_LOCAL_ALLOC_SIZE constant as it is no
longer used and the comment above it was out of date.
To test fragmentation, I used a workload which launched 4 threads that did
4k writes into a series of about 140 alternating files.
With resv_level=2, and a 4k/4k file system I observed the following average
fragmentation for various localalloc= parameters:
localalloc= avg. fragmentation
8 48
32 16
64 10
120 7
On larger cluster sizes, the difference is more dramatic.
The new default size top out at 256M, which we'll only get for cluster
sizes of 32K and above.
Signed-off-by: Mark Fasheh <mfasheh@suse.com>
Signed-off-by: Joel Becker <joel.becker@oracle.com>
This patch pulls the local alloc sizing code into localalloc.c and provides
a callout to it from ocfs2_fill_super(). Behavior is essentially unchanged
except that I correctly calculate the maximum local alloc size. The old code
in ocfs2_parse_options() calculated the max size as:
ocfs2_local_alloc_size(sb) * 8
which is correct, in bits. Unfortunately though the option passed in is in
megabytes. Ultimately, this bug made no real difference - the shrink code
would catch a too-large size and bring it down to something reasonable.
Still, it's less than efficient as-is.
Signed-off-by: Mark Fasheh <mfasheh@suse.com>
Signed-off-by: Joel Becker <joel.becker@oracle.com>
Inodes are always allocated from the global bitmap now so we don't need this
any more. Also, the existing implementation bounces reservations around
needlessly.
Signed-off-by: Mark Fasheh <mfasheh@suse.com>
Otherwise, the need for a very large contiguous allocation tends to
wreak havoc on many inode allocation reservations on the local alloc, thus
ruining any chances for contiguousness.
Signed-off-by: Mark Fasheh <mfasheh@suse.com>
Use the reservations system for unindexed dir tree allocations. We don't
bother with the indexed tree as reads from it are mostly random anyway.
Directory reservations are marked seperately, to allow the reservations code
a chance to optimize their window sizes. This patch allocates only 8 bits
for directory windows as they generally are not expected to grow as quickly
as file data. Future improvements to dir window sizing can trivially be
made.
Signed-off-by: Mark Fasheh <mfasheh@suse.com>
This patch improves Ocfs2 allocation policy by allowing an inode to
reserve a portion of the local alloc bitmap for itself. The reserved
portion (allocation window) is advisory in that other allocation
windows might steal it if the local alloc bitmap becomes
full. Otherwise, the reservations are honored and guaranteed to be
free. When the local alloc window is moved to a different portion of
the bitmap, existing reservations are discarded.
Reservation windows are represented internally by a red-black
tree. Within that tree, each node represents the reservation window of
one inode. An LRU of active reservations is also maintained. When new
data is written, we allocate it from the inodes window. When all bits
in a window are exhausted, we allocate a new one as close to the
previous one as possible. Should we not find free space, an existing
reservation is pulled off the LRU and cannibalized.
Signed-off-by: Mark Fasheh <mfasheh@suse.com>
jbd[2]_journal_dirty_metadata() only returns 0. It's been returning 0
since before the kernel moved to git. There is no point in checking
this error.
ocfs2_journal_dirty() has been faithfully returning the status since the
beginning. All over ocfs2, we have blocks of code checking this can't
fail status. In the past few years, we've tried to avoid adding these
checks, because they are pointless. But anyone who looks at our code
assumes they are needed.
Finally, ocfs2_journal_dirty() is made a void function. All error
checking is removed from other files. We'll BUG_ON() the status of
jbd2_journal_dirty_metadata() just in case they change it someday. They
won't.
Signed-off-by: Joel Becker <joel.becker@oracle.com>
...rather than the secType. This allows us to get rid of the MSKerberos
securityEnum. The client just makes a decision at upcall time.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
So that we can reasonably set up the secType based on both the
NegotiateProtocol response and the parsed mount options.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
When CONFIG_BLOCK is not enabled:
fs/logfs/super.c:142: error: implicit declaration of function 'bdev_get_queue'
fs/logfs/super.c:142: error: invalid type argument of '->' (have 'int')
Found by Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Joern Engel <joern@logfs.org>
Convert udf_ioctl to an unlocked_ioctl and push the BKL down into it.
Signed-off-by: John Kacur <jkacur@redhat.com
Signed-off-by: Jan Kara <jack@suse.cz>
We need to report both the accounting and enforcing flags if we are
in enforcing mode.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
This patch contains various tweaks to how log flushes and active item writeback
work. gfs2_logd is now managed by a waitqueue, and gfs2_log_reseve now waits
for gfs2_logd to do the log flushing. Multiple functions were rewritten to
remove the need to call gfs2_log_lock(). Instead of using one test to see if
gfs2_logd had work to do, there are now seperate tests to check if there
are two many buffers in the incore log or if there are two many items on the
active items list.
This patch is a port of a patch Steve Whitehouse wrote about a year ago, with
some minor changes. Since gfs2_ail1_start always submits all the active items,
it no longer needs to keep track of the first ai submitted, so this has been
removed. In gfs2_log_reserve(), the order of the calls to
prepare_to_wait_exclusive() and wake_up() when firing off the logd thread has
been switched. If it called wake_up first there was a small window for a race,
where logd could run and return before gfs2_log_reserve was ready to get woken
up. If gfs2_logd ran, but did not free up enough blocks, gfs2_log_reserve()
would be left waiting for gfs2_logd to eventualy run because it timed out.
Finally, gt_logd_secs, which controls how long to wait before gfs2_logd times
out, and flushes the log, can now be set on mount with ar_commit.
Signed-off-by: Benjamin Marzinski <bmarzins@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jlbec/ocfs2:
ocfs2: Avoid a gcc warning in ocfs2_wipe_inode().
ocfs2: Avoid direct write if we fall back to buffered I/O
ocfs2_dlmfs: Fix math error when reading LVB.
ocfs2: Update VFS inode's id info after reflink.
ocfs2: potential ERR_PTR dereference on error paths
ocfs2: Add directory entry later in ocfs2_symlink() and ocfs2_mknod()
ocfs2: use OCFS2_INODE_SKIP_ORPHAN_DIR in ocfs2_mknod error path
ocfs2: use OCFS2_INODE_SKIP_ORPHAN_DIR in ocfs2_symlink error path
ocfs2: add OCFS2_INODE_SKIP_ORPHAN_DIR flag and honor it in the inode wipe code
ocfs2: Reset status if we want to restart file extension.
ocfs2: Compute metaecc for superblocks during online resize.
ocfs2: Check the owner of a lockres inside the spinlock
ocfs2: one more warning fix in ocfs2_file_aio_write(), v2
ocfs2_dlmfs: User DLM_* when decoding file open flags.
Unregister and destroy the bdi in put_super, after mount is r/o, but before
put_anon_super releases the device name.
For symmetry, bdi_destroy in destroy_client (we bdi_init in create_client).
Only set s_bdi if bdi_register succeeds, since we use it to decide whether
to bdi_unregister.
Signed-off-by: Sage Weil <sage@newdream.net>
li_refcount was not re-initialized in function logfs_init_inode(), small
patch that will fix the problem
Signed-off-by: Prasad Joshi <prasadjoshi124@gmail.com>
Signed-off-by: Joern Engel <joern@logfs.org>
gcc warns that a variable is uninitialized. It's actually handled, but
an early return fools gcc. Let's just initialize the variable to a
garbage value that will crash if the usage is ever broken.
Signed-off-by: Joel Becker <joel.becker@oracle.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client:
ceph: remove bad auth_x kmem_cache
ceph: fix lockless caps check
ceph: clear dir complete, invalidate dentry on replayed rename
ceph: fix direct io truncate offset
ceph: discard incoming messages with bad seq #
ceph: fix seq counting for skipped messages
ceph: add missing #includes
ceph: fix leaked spinlock during mds reconnect
ceph: print more useful version info on module load
ceph: fix snap realm splits
ceph: clear dir complete on d_move
It's useless, since our allocations are already a power of 2. And it was
allocated per-instance (not globally), which caused a name collision when
we tried to mount a second file system with auth_x enabled.
Signed-off-by: Sage Weil <sage@newdream.net>
If a rename operation is resent to the MDS following an MDS restart, the
client does not get a full reply (containing the resulting metadata) back.
In that case, a ceph_rename() needs to compensate by doing anything useful
that fill_inode() would have, like d_move().
It also needs to invalidate the dentry (to workaround the vfs_rename_dir()
bug) and clear the dir complete flag, just like fill_trace().
Signed-off-by: Sage Weil <sage@newdream.net>
We can get old message seq #'s after a tcp reconnect for stateful sessions
(i.e., the MDS). If we get a higher seq #, that is an error, and we
shouldn't see any bad seq #'s for stateless (mon, osd) connections.
Signed-off-by: Sage Weil <sage@newdream.net>
The snap realm split was checking i_snap_realm, not the list_head, to
determine if an inode belonged in the new realm. The check always failed,
which meant we always moved the inode, corrupting the old realm's list and
causing various crashes.
Also wait to release old realm reference to avoid possibility of use after
free.
Signed-off-by: Sage Weil <sage@newdream.net>
d_move() reorders the d_subdirs list, breaking the readdir result caching.
Unless/until d_move preserves that ordering, clear CEPH_I_COMPLETE on
rename.
Signed-off-by: Sage Weil <sage@newdream.net>
As of 32a88aa1, __sync_filesystem() will return 0 if s_bdi is not set.
And nilfs does not set s_bdi anywhere. I noticed this problem by the
warning introduced by the recent commit 5129a469 ("Catch filesystem
lacking s_bdi").
WARNING: at fs/super.c:959 vfs_kern_mount+0xc5/0x14e()
Hardware name: PowerEdge 2850
Modules linked in: nilfs2 loop tpm_tis tpm tpm_bios video shpchp pci_hotplug output dcdbas
Pid: 3773, comm: mount.nilfs2 Not tainted 2.6.34-rc6-debug #38
Call Trace:
[<c1028422>] warn_slowpath_common+0x60/0x90
[<c102845f>] warn_slowpath_null+0xd/0x10
[<c1095936>] vfs_kern_mount+0xc5/0x14e
[<c1095a03>] do_kern_mount+0x32/0xbd
[<c10a811e>] do_mount+0x671/0x6d0
[<c1073794>] ? __get_free_pages+0x1f/0x21
[<c10a684f>] ? copy_mount_options+0x2b/0xe2
[<c107b634>] ? strndup_user+0x48/0x67
[<c10a81de>] sys_mount+0x61/0x8f
[<c100280c>] sysenter_do_call+0x12/0x32
This ensures to set s_bdi for nilfs and fixes the sync silent failure.
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Acked-by: Jens Axboe <jens.axboe@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In the replay case, the
renew_client(session->se_client);
happens after we've droppped the sessionid_lock, and without holding a
reference on the session; so there's nothing preventing the session
being freed before we get here.
Thanks to Benny Halevy for catching a bug in an earlier version of this
patch.
Signed-off-by: J. Bruce Fields <bfields@citi.umich.edu>
Acked-by: Benny Halevy <bhalevy@panasas.com>
Fix a number of RCU issues in the NFSv4 delegation code.
(1) delegation->cred doesn't need to be RCU protected as it's essentially an
invariant refcounted structure.
By the time we get to nfs_free_delegation(), the delegation is being
released, so no one else should be attempting to use the saved
credentials, and they can be cleared.
However, since the list of delegations could still be under traversal at
this point by such as nfs_client_return_marked_delegations(), the cred
should be released in nfs_do_free_delegation() rather than in
nfs_free_delegation(). Simply using rcu_assign_pointer() to clear it is
insufficient as that doesn't stop the cred from being destroyed, and nor
does calling put_rpccred() after call_rcu(), given that the latter is
asynchronous.
(2) nfs_detach_delegation_locked() and nfs_inode_set_delegation() should use
rcu_derefence_protected() because they can only be called if
nfs_client::cl_lock is held, and that guards against anyone changing
nfsi->delegation under it. Furthermore, the barrier imposed by
rcu_dereference() is superfluous, given that the spin_lock() is also a
barrier.
(3) nfs_detach_delegation_locked() is now passed a pointer to the nfs_client
struct so that it can issue lockdep advice based on clp->cl_lock for (2).
(4) nfs_inode_return_delegation_noreclaim() and nfs_inode_return_delegation()
should use rcu_access_pointer() outside the spinlocked region as they
merely examine the pointer and don't follow it, thus rendering unnecessary
the need to impose a partial ordering over the one item of interest.
These result in an RCU warning like the following:
[ INFO: suspicious rcu_dereference_check() usage. ]
---------------------------------------------------
fs/nfs/delegation.c:332 invoked rcu_dereference_check() without protection!
other info that might help us debug this:
rcu_scheduler_active = 1, debug_locks = 0
2 locks held by mount.nfs4/2281:
#0: (&type->s_umount_key#34){+.+...}, at: [<ffffffff810b25b4>] deactivate_super+0x60/0x80
#1: (iprune_sem){+.+...}, at: [<ffffffff810c332a>] invalidate_inodes+0x39/0x13a
stack backtrace:
Pid: 2281, comm: mount.nfs4 Not tainted 2.6.34-rc1-cachefs #110
Call Trace:
[<ffffffff8105149f>] lockdep_rcu_dereference+0xaa/0xb2
[<ffffffffa00b4591>] nfs_inode_return_delegation_noreclaim+0x5b/0xa0 [nfs]
[<ffffffffa0095d63>] nfs4_clear_inode+0x11/0x1e [nfs]
[<ffffffff810c2d92>] clear_inode+0x9e/0xf8
[<ffffffff810c3028>] dispose_list+0x67/0x10e
[<ffffffff810c340d>] invalidate_inodes+0x11c/0x13a
[<ffffffff810b1dc1>] generic_shutdown_super+0x42/0xf4
[<ffffffff810b1ebe>] kill_anon_super+0x11/0x4f
[<ffffffffa009893c>] nfs4_kill_super+0x3f/0x72 [nfs]
[<ffffffff810b25bc>] deactivate_super+0x68/0x80
[<ffffffff810c6744>] mntput_no_expire+0xbb/0xf8
[<ffffffff810c681b>] release_mounts+0x9a/0xb0
[<ffffffff810c689b>] put_mnt_ns+0x6a/0x79
[<ffffffffa00983a1>] nfs_follow_remote_path+0x5a/0x146 [nfs]
[<ffffffffa0098334>] ? nfs_do_root_mount+0x82/0x95 [nfs]
[<ffffffffa00985a9>] nfs4_try_mount+0x75/0xaf [nfs]
[<ffffffffa0098874>] nfs4_get_sb+0x291/0x31a [nfs]
[<ffffffff810b2059>] vfs_kern_mount+0xb8/0x177
[<ffffffff810b2176>] do_kern_mount+0x48/0xe8
[<ffffffff810c810b>] do_mount+0x782/0x7f9
[<ffffffff810c8205>] sys_mount+0x83/0xbe
[<ffffffff81001eeb>] system_call_fastpath+0x16/0x1b
Also on:
fs/nfs/delegation.c:215 invoked rcu_dereference_check() without protection!
[<ffffffff8105149f>] lockdep_rcu_dereference+0xaa/0xb2
[<ffffffffa00b4223>] nfs_inode_set_delegation+0xfe/0x219 [nfs]
[<ffffffffa00a9c6f>] nfs4_opendata_to_nfs4_state+0x2c2/0x30d [nfs]
[<ffffffffa00aa15d>] nfs4_do_open+0x2a6/0x3a6 [nfs]
...
And:
fs/nfs/delegation.c:40 invoked rcu_dereference_check() without protection!
[<ffffffff8105149f>] lockdep_rcu_dereference+0xaa/0xb2
[<ffffffffa00b3bef>] nfs_free_delegation+0x3d/0x6e [nfs]
[<ffffffffa00b3e71>] nfs_do_return_delegation+0x26/0x30 [nfs]
[<ffffffffa00b406a>] __nfs_inode_return_delegation+0x1ef/0x1fe [nfs]
[<ffffffffa00b448a>] nfs_client_return_marked_delegations+0xc9/0x124 [nfs]
...
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Ensure that we correctly rcu-dereference the delegation itself, and that we
protect against removal while we're changing the contents.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
logfs_seek_hole() may return the same offset it is passed as argument.
Found by Prasad Joshi <prasadjoshi124@gmail.com>
Signed-off-by: Joern Engel <joern@logfs.org>
when we fall back to buffered write from direct write, we call
__generic_file_aio_write() but that will end up doing direct write
even we are only prepared to do buffered write because the file
has the O_DIRECT flag set. This is a fix for
https://bugzilla.novell.com/show_bug.cgi?id=591039
revised with Joel's comments.
Signed-off-by: Li Dongyang <lidongyang@novell.com>
Acked-by: Mark Fasheh <mfasheh@suse.com>
Signed-off-by: Joel Becker <joel.becker@oracle.com>
Commit 7fe2b3190b fixed possible
misordering of completion asts (casts) and blocking asts (basts)
for kernel locks. This patch does the same for locks taken by
user space applications.
Signed-off-by: David Teigland <teigland@redhat.com>
Smatch complains because "lkb" is never NULL. Looking at it, the original
code actually adds the new element to the end of the list fine, so we can
just get rid of the if condition. This code is four years old and no one
has complained so it must work.
Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: David Teigland <teigland@redhat.com>
CONFIG_INOTIFY_USER defined but CONFIG_ANON_INODES undefined will result
in the following build failure:
LD vmlinux
fs/built-in.o: In function 'sys_inotify_init1':
(.text.sys_inotify_init1+0x22c): undefined reference to 'anon_inode_getfd'
fs/built-in.o: In function `sys_inotify_init1':
(.text.sys_inotify_init1+0x22c): relocation truncated to fit: R_MIPS_26 against 'anon_inode_getfd'
make[2]: *** [vmlinux] Error 1
make[1]: *** [sub-make] Error 2
make: *** [all] Error 2
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
