Currently the xfs module init/exit code is a mess. It's farmed out over a
lot of function with very little error checking. This patch makes sure we
propagate all initialization failures properly and clean up after them.
Various runtime initializations are replaced with compile-time
initializations where possible to make this easier. The exit path is
similarly consolidated.
There's now split out function to create/destroy the kmem zones and
alloc/free the trace buffers. I've also changed the ktrace allocations to
KM_MAYFAIL and handled errors resulting from that.
And yes, we really should replace the XFS_*_TRACE ifdefs with a single
XFS_TRACE..
SGI-PV: 976035
SGI-Modid: xfs-linux-melb:xfs-kern:31354a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Niv Sardi <xaiki@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
xfs_mount is already pretty linux-specific so merge it into
xfs_fs_fill_super to allow for a more structured mount code in the next
patches. xfs_start_flags and xfs_finish_flags also move to xfs_super.c.
SGI-PV: 981951
SGI-Modid: xfs-linux-melb:xfs-kern:31189a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
xfs_unmount is small and already pretty Linux specific, so merge it into
the callers. The real unmount path is simplified a little by doing a
WARN_ON on the xfs_unmount_flush retval directly instead of propagating
the error back to the caller, and the mout failure case in simplified
significantly by removing the forced shutdown case and all the dmapi
events that shouldn't be sent because the dmapi mount event hasn't been
sent by that time either.
SGI-PV: 981951
SGI-Modid: xfs-linux-melb:xfs-kern:31188a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
xfs_mntupdate already is completely Linux specific due to the VFS flags
passed in, so it might aswell be merged into xfs_fs_remount.
SGI-PV: 981498
SGI-Modid: xfs-linux-melb:xfs-kern:31185a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
kmem_free() function takes (ptr, size) arguments but doesn't actually use
second one.
This patch removes size argument from all callsites.
SGI-PV: 981498
SGI-Modid: xfs-linux-melb:xfs-kern:31050a
Signed-off-by: Denys Vlasenko <vda.linux@googlemail.com>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
features2 fields.
Previously, mounting with noattr2 failed to achieve anything because
although it cleared the attr2 mount flag, it would set it again as soon as
it processed the superblock fields. The fix now has an explicit noattr2
flag and uses it later to fix up the versionnum and features2 fields.
SGI-PV: 980021
SGI-Modid: xfs-linux-melb:xfs-kern:31003a
Signed-off-by: Tim Shimmin <tes@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
xfssyncd triggers the logging of superblock counters every 30s if the
filesystem is made with lazy-count=1. This will prevent disks from idling
and spinning down as there will be a log write every 30s. With the way
counter recovery works for lazy-count=1, this code is unnecessary and
provides no real benefit, so just remove it.
SGI-PV: 980145
SGI-Modid: xfs-linux-melb:xfs-kern:30840a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Barry Naujok <bnaujok@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
xfs_trans_commit() can return errors when there are problems in the
transaction subsystem. They are indicative that the entire transaction may
be incomplete, and hence the error should be propagated as there is a good
possibility that there is something fatally wrong in the filesystem. Catch
and propagate or warn about commit errors in the places where they are
currently ignored.
SGI-PV: 980084
SGI-Modid: xfs-linux-melb:xfs-kern:30795a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Niv Sardi <xaiki@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
The ticket allocator is just a simple slab implementation internal to the
log. It requires the icloglock to be held when manipulating it and this
contributes to contention on that lock.
Just kill the entire allocator and use a memory zone instead. While there,
allow us to gracefully fail allocation with ENOMEM.
SGI-PV: 978729
SGI-Modid: xfs-linux-melb:xfs-kern:30771a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Most VN_RELE calls either directly contain a XFS_ITOV or have the
corresponding xfs_inode already in scope. Use the IRELE helper instead of
VN_RELE to clarify the code. With a little more work we can kill VN_RELE
altogether and define IRELE in terms of iput directly.
SGI-PV: 976035
SGI-Modid: xfs-linux-melb:xfs-kern:30710a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Remove the xfs_icluster structure and replace with a radix tree lookup.
We don't need to keep a list of inodes in each cluster around anymore as
we can look them up quickly when we need to. The only time we need to do
this now is during inode writeback.
Factor the inode cluster writeback code out of xfs_iflush and convert it
to use radix_tree_gang_lookup() instead of walking a list of inodes built
when we first read in the inodes.
This remove 3 pointers from each xfs_inode structure and the xfs_icluster
structure per inode cluster. Hence we reduce the cache footprint of the
xfs_inodes by between 5-10% depending on cluster sparseness.
To be truly efficient we need a radix_tree_gang_lookup_range() call to
stop searching once we are past the end of the cluster instead of trying
to find a full cluster's worth of inodes.
Before (ia64):
$ cat /sys/slab/xfs_inode/object_size 536
After:
$ cat /sys/slab/xfs_inode/object_size 512
SGI-PV: 977460
SGI-Modid: xfs-linux-melb:xfs-kern:30502a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Remove the xfs_refcache, it was only needed while we were still
building for 2.4 kernels.
SGI-PV: 971186
SGI-Modid: xfs-linux-melb:xfs-kern:30472a
Signed-off-by: Donald Douwsma <donaldd@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Remove macro-to-small-function indirection from xfs_sb.h, and remove some
which are completely unused.
SGI-PV: 976035
SGI-Modid: xfs-linux-melb:xfs-kern:30528a
Signed-off-by: Eric Sandeen <sandeen@sandeen.net>
Signed-off-by: Donald Douwsma <donaldd@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
the "ikeep" option is set rather than "noikeep".
This regression was introduced in 970451.
With no mount options specified, xfs_parseargs() does the following:
int ikeep = 0;
args->flags |= XFSMNT_BARRIER;
args->flags2 |= XFSMNT2_COMPAT_IOSIZE;
if (!options)
goto done;
It only sets the above two options by default and before, it also used to
set XFSMNT_IDELETE by default.
If options are specified, then
if (!(args->flags & XFSMNT_DMAPI) && !ikeep)
args->flags |= XFSMNT_IDELETE;
is executed later on which is skipped by the "goto done;" above.
The solution is to invert the logic.
SGI-PV: 977771
SGI-Modid: xfs-linux-melb:xfs-kern:30590a
Signed-off-by: Niv Sardi <xaiki@sgi.com>
Signed-off-by: Barry Naujok <bnaujok@sgi.com>
Signed-off-by: Josef 'Jeff' Sipek <jeffpc@josefsipek.net>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
The only caller (xfs_fs_fill_super) can simplify call igrab on the root
inode.
SGI-PV: 971186
SGI-Modid: xfs-linux-melb:xfs-kern:30393a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
- merge xfs_fid2 into it's only caller xfs_dm_inode_to_fh.
- remove xfs_vget and opencode it in the two callers, simplifying
both of them by avoiding the awkward calling convetion.
- assign directly to the dm_fid_t members in various places in the
dmapi code instead of casting them to xfs_fid_t first (which
is identical to dm_fid_t)
SGI-PV: 974747
SGI-Modid: xfs-linux-melb:xfs-kern:30258a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Vlad Apostolov <vapo@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Mount option parsing is platform specific. Move it out of core code into
the platform specific superblock operation file.
SGI-PV: 971186
SGI-Modid: xfs-linux-melb:xfs-kern:30012a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
There are several mount options that don't show up in /proc/mounts. Add
them in and clean up the showargs code at the same time.
SGI-PV: 971186
SGI-Modid: xfs-linux-melb:xfs-kern:30004a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
These are mostly locking annotations, marking things static, casts where
needed and declaring stuff in header files.
SGI-PV: 971186
SGI-Modid: xfs-linux-melb:xfs-kern:30002a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
No need to have a wrapper just two call two more functions.
SGI-PV: 971186
SGI-Modid: xfs-linux-melb:xfs-kern:29816a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Donald Douwsma <donaldd@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
We were already filling the Linux struct statfs anyway, and doing this
trivial task directly in xfs_fs_statfs makes the code quite a bit cleaner.
While I was at it I also moved copying attributes that don't change over
the lifetime of the filesystem outside the superblock lock.
xfs_fs_fill_super used to get the magic number and blocksize through
xfs_statvfs, but assigning them directly is a lot cleaner and will save
some stack space during mount.
SGI-PV: 971186
SGI-Modid: xfs-linux-melb:xfs-kern:29802a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Tim Shimmin <tes@sgi.com>
remove spinlock init abstraction macro in spin.h, remove the callers, and
remove the file. Move no-op spinlock_destroy to xfs_linux.h Cleanup
spinlock locals in xfs_mount.c
SGI-PV: 970382
SGI-Modid: xfs-linux-melb:xfs-kern:29751a
Signed-off-by: Eric Sandeen <sandeen@sandeen.net>
Signed-off-by: Donald Douwsma <donaldd@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Un-obfuscate XFS_SB_LOCK, remove XFS_SB_LOCK->mutex_lock->spin_lock
macros, call spin_lock directly, remove extraneous cookie holdover from
old xfs code, and change lock type to spinlock_t.
SGI-PV: 970382
SGI-Modid: xfs-linux-melb:xfs-kern:29746a
Signed-off-by: Eric Sandeen <sandeen@sandeen.net>
Signed-off-by: Donald Douwsma <donaldd@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Un-obfuscate dabuf_global_lock, remove mutex_lock->spin_lock macros, call
spin_lock directly, remove extraneous cookie holdover from old xfs code,
and change lock type to spinlock_t.
SGI-PV: 970382
SGI-Modid: xfs-linux-melb:xfs-kern:29744a
Signed-off-by: Eric Sandeen <sandeen@sandeen.net>
Signed-off-by: Donald Douwsma <donaldd@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Currently there is an indirection called ioops in the XFS data I/O path.
Various functions are called by functions pointers, but there is no
coherence in what this is for, and of course for XFS itself it's entirely
unused. This patch removes it instead and significantly reduces source and
binary size of XFS while making maintaince easier.
SGI-PV: 970841
SGI-Modid: xfs-linux-melb:xfs-kern:29737a
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Tim Shimmin <tes@sgi.com>
A large part of xfs_sync_inodes is conditional on the SYNC_BDFLUSH which
is never passed to it. This patch removes it and adds an assert that
triggers in case some new code tries to pass SYNC_BDFLUSH to it.
SGI-PV: 970242
SGI-Modid: xfs-linux-melb:xfs-kern:29630a
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Currently XFs has three different fid types: struct fid, struct xfs_fid
and struct xfs_fid2 with hte latter two beeing identicaly and the first
one beeing the same size but an unstructured array with the same size.
This patch consolidates all this to alway uuse struct xfs_fid.
This patch is required for an upcoming patch series from me that revamps
the nfs exporting code and introduces a Linux-wide struct fid.
SGI-PV: 970336
SGI-Modid: xfs-linux-melb:xfs-kern:29651a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Removes STATIC on xfs_freeze function which was not manually
applied for SGI-Modid: xfs-linux-melb:xfs-kern:29504a.
Reviewed-by: Lachlan McIlroy <lachlan@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Fix for a regression caused by a recent patch
that moved the DMAPI mount option processing inside xfs_parseargs(). The
DMAPI mount option used to be processed in the DMAPI module loaded before
xfs_parseargs() was invoked.
SGI-PV: 970451
SGI-Modid: xfs-linux-melb:xfs-kern:29683a
Signed-off-by: Vlad Apostolov <vapo@sgi.com>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Now that struct bhv_vfs doesn't have any members left we can kill it and
go directly from the super_block to the xfs_mount everywhere.
SGI-PV: 969608
SGI-Modid: xfs-linux-melb:xfs-kern:29509a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
All flags are added to xfs_mount's m_flag instead. Note that the 32bit
inode flag was duplicated in both of them, but only cleared in the mount
when it was not nessecary due to the filesystem beeing small enough. Two
flags are still required here - one to indicate the mount option setting,
and one to indicate if it applies or not.
SGI-PV: 969608
SGI-Modid: xfs-linux-melb:xfs-kern:29507a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Also remove the now dead behavior code.
SGI-PV: 969608
SGI-Modid: xfs-linux-melb:xfs-kern:29505a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
All vfs ops now take struct xfs_mount pointers and the behaviour related
glue is split out into methods of its own.
SGI-PV: 969608
SGI-Modid: xfs-linux-melb:xfs-kern:29504a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Mount options are now parsed by the main XFS module and rejected if quota
support is not available, and there are some new quota operation for the
quotactl syscall and calls to quote in the mount, unmount and sync
callchains.
SGI-PV: 969608
SGI-Modid: xfs-linux-melb:xfs-kern:29503a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Mount options are now parsed by the main XFS module and rejected if dmapi
support is not available, and there is a new dm operation to send the
mount event.
SGI-PV: 969608
SGI-Modid: xfs-linux-melb:xfs-kern:29502a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
In the next patch we need to look at the mount structure until just before
it's freed, so we need to be able to free it as the very last thing in
xfs_unmount.
SGI-PV: 969608
SGI-Modid: xfs-linux-melb:xfs-kern:29501a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
struct bhv_vnode is on it's way out, so move the I/O count to the XFS
inode.
SGI-PV: 969608
SGI-Modid: xfs-linux-melb:xfs-kern:29497a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
XFS inodes are dynamically allocated on demand, rather than being
allocated at mkfs time. Chunks of 64 inodes are allocated at once, but
they are never freed. Over time, this can lead to filesystem
fragmentation, clusters of inodes and the btrees which point at them can
be scattered around the system.
By freeing clusters as they are emptied, we will reduce fragmentation of
the free space after removing files. This in turn will allow us to make
better placement decisions when repopulating a filesystem. The
XFSMNT_IDELETE mount option enables freeing clusters when they get empty.
Unfortunately a side effect of freeing inode clusters is that the inode
generation numbers of such inodes would be reset to zero when the cluster
is reclaimed. This is a problem in particular for a DMAPI enabled
filesystem as the the DMAPI handles need to be unique and persistent in
time. An unique DMAPI handle is built with the help of the inode
generation number. When the last one is prematurely reset by an inode
cluster reclaim, there is a high probability of different generation
inodes to end up having identical DMAPI handles.
To avoid the problem with identical DMAPI handles, the XFSMNT_IDELETE
mount option should be set as default, only if the filesystem is not
mounted with XFSMNT_DMAPI.
SGI-PV: 969192
SGI-Modid: xfs-linux-melb:xfs-kern:29486a
Signed-off-by: Vlad Apostolov <vapo@sgi.com>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Mark Goodwin <markgw@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
One of the perpetual scaling problems XFS has is indexing it's incore
inodes. We currently uses hashes and the default hash sizes chosen can
only ever be a tradeoff between memory consumption and the maximum
realistic size of the cache.
As a result, anyone who has millions of inodes cached on a filesystem
needs to tunes the size of the cache via the ihashsize mount option to
allow decent scalability with inode cache operations.
A further problem is the separate inode cluster hash, whose size is based
on the ihashsize but is smaller, and so under certain conditions (sparse
cluster cache population) this can become a limitation long before the
inode hash is causing issues.
The following patchset removes the inode hash and cluster hash and
replaces them with radix trees to avoid the scalability limitations of the
hashes. It also reduces the size of the inodes by 3 pointers....
SGI-PV: 969561
SGI-Modid: xfs-linux-melb:xfs-kern:29481a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Tim Shimmin <tes@sgi.com>
If we fail to open the the log device buftarg, we can fall through to
error handling code that fails to check for a NULL log device buftarg
before calling xfs_free_buftarg().
This patch fixes the issue by checking mp->m_logdev_targp against NULL in
xfs_unmountfs_close() and doing the proper xfs_blkdev_put(logdev); and
xfs_blkdev_put(rtdev); on (!mp->m_rtdev_targp) in xfs_mount().
Discovered by the Coverity checker.
SGI-PV: 968563
SGI-Modid: xfs-linux-melb:xfs-kern:29328a
Signed-off-by: Jesper Juhl <jesper.juhl@gmail.com>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
xfs_start_flags can make use of is_power_of_2 to tidy up the test a little
bit.
SGI-PV: 968563
SGI-Modid: xfs-linux-melb:xfs-kern:29327a
Signed-off-by: Eric Sandeen <sandeen@sandeen.net>
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
In media spaces, video is often stored in a frame-per-file format. When
dealing with uncompressed realtime HD video streams in this format, it is
crucial that files do not get fragmented and that multiple files a placed
contiguously on disk.
When multiple streams are being ingested and played out at the same time,
it is critical that the filesystem does not cross the streams and
interleave them together as this creates seek and readahead cache miss
latency and prevents both ingest and playout from meeting frame rate
targets.
This patch set creates a "stream of files" concept into the allocator to
place all the data from a single stream contiguously on disk so that RAID
array readahead can be used effectively. Each additional stream gets
placed in different allocation groups within the filesystem, thereby
ensuring that we don't cross any streams. When an AG fills up, we select a
new AG for the stream that is not in use.
The core of the functionality is the stream tracking - each inode that we
create in a directory needs to be associated with the directories' stream.
Hence every time we create a file, we look up the directories' stream
object and associate the new file with that object.
Once we have a stream object for a file, we use the AG that the stream
object point to for allocations. If we can't allocate in that AG (e.g. it
is full) we move the entire stream to another AG. Other inodes in the same
stream are moved to the new AG on their next allocation (i.e. lazy
update).
Stream objects are kept in a cache and hold a reference on the inode.
Hence the inode cannot be reclaimed while there is an outstanding stream
reference. This means that on unlink we need to remove the stream
association and we also need to flush all the associations on certain
events that want to reclaim all unreferenced inodes (e.g. filesystem
freeze).
SGI-PV: 964469
SGI-Modid: xfs-linux-melb:xfs-kern:29096a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Barry Naujok <bnaujok@sgi.com>
Signed-off-by: Donald Douwsma <donaldd@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Signed-off-by: Vlad Apostolov <vapo@sgi.com>
The remount readonly path can fail to writeback properly because we still
have active transactions after calling xfs_quiesce_fs(). Further
investigation shows that this path is broken in the same ways that the xfs
freeze path was broken so fix it the same way.
SGI-PV: 964464
SGI-Modid: xfs-linux-melb:xfs-kern:28869a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Tim Shimmin <tes@sgi.com>
When we have a couple of hundred transactions on the fly at once, they all
typically modify the on disk superblock in some way.
create/unclink/mkdir/rmdir modify inode counts, allocation/freeing modify
free block counts.
When these counts are modified in a transaction, they must eventually lock
the superblock buffer and apply the mods. The buffer then remains locked
until the transaction is committed into the incore log buffer. The result
of this is that with enough transactions on the fly the incore superblock
buffer becomes a bottleneck.
The result of contention on the incore superblock buffer is that
transaction rates fall - the more pressure that is put on the superblock
buffer, the slower things go.
The key to removing the contention is to not require the superblock fields
in question to be locked. We do that by not marking the superblock dirty
in the transaction. IOWs, we modify the incore superblock but do not
modify the cached superblock buffer. In short, we do not log superblock
modifications to critical fields in the superblock on every transaction.
In fact we only do it just before we write the superblock to disk every
sync period or just before unmount.
This creates an interesting problem - if we don't log or write out the
fields in every transaction, then how do the values get recovered after a
crash? the answer is simple - we keep enough duplicate, logged information
in other structures that we can reconstruct the correct count after log
recovery has been performed.
It is the AGF and AGI structures that contain the duplicate information;
after recovery, we walk every AGI and AGF and sum their individual
counters to get the correct value, and we do a transaction into the log to
correct them. An optimisation of this is that if we have a clean unmount
record, we know the value in the superblock is correct, so we can avoid
the summation walk under normal conditions and so mount/recovery times do
not change under normal operation.
One wrinkle that was discovered during development was that the blocks
used in the freespace btrees are never accounted for in the AGF counters.
This was once a valid optimisation to make; when the filesystem is full,
the free space btrees are empty and consume no space. Hence when it
matters, the "accounting" is correct. But that means the when we do the
AGF summations, we would not have a correct count and xfs_check would
complain. Hence a new counter was added to track the number of blocks used
by the free space btrees. This is an *on-disk format change*.
As a result of this, lazy superblock counters are a mkfs option and at the
moment on linux there is no way to convert an old filesystem. This is
possible - xfs_db can be used to twiddle the right bits and then
xfs_repair will do the format conversion for you. Similarly, you can
convert backwards as well. At some point we'll add functionality to
xfs_admin to do the bit twiddling easily....
SGI-PV: 964999
SGI-Modid: xfs-linux-melb:xfs-kern:28652a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Christoph Hellwig