Now that the ktrace_enter() code is using atomics, the non-power-of-2
buffer sizes - which require modulus operations to get the index - are
showing up as using substantial CPU in the profiles.
Force the buffer sizes to be rounded up to the nearest power of two and
use masking rather than modulus operations to convert the index counter to
the buffer index. This reduces ktrace_enter overhead to 8% of a CPU time,
and again almost halves the trace intensive test runtime.
SGI-PV: 977546
SGI-Modid: xfs-linux-melb:xfs-kern:30538a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
ktrace_enter() is consuming vast amounts of CPU time due to the use of a
single global lock for protecting buffer index increments. Change it to
use per-buffer atomic counters - this reduces ktrace_enter() overhead
during a trace intensive test on a 4p machine from 58% of all CPU time to
12% and halves test runtime.
SGI-PV: 977546
SGI-Modid: xfs-linux-melb:xfs-kern:30537a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
XFS changes the c/mtime of an inode when truncating it to the same size.
The c/mtime is only supposed to change if the size is changed. Not to be
confused with ftruncate, where the c/mtime is supposed to be changed even
if the size is not changed.
The Linux VFS encodes this semantic difference in the flags it sends down
to ->setattr, which XFS currently ignores. We need to make XFS pay
attention to the VFS flags and hence Do The Right Thing.
SGI-PV: 977547
SGI-Modid: xfs-linux-melb:xfs-kern:30536a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
As Dave pointed out after the export ops changes we now always encode the
parent into the filehandle for regular files, but it's not actually needed
when the filesystem is export with no_subtree_check. This one-liner fixes
xfs_fs_encode_fh to skip encoding the parent unless nessecary.
SGI-PV: 976035
SGI-Modid: xfs-linux-melb:xfs-kern:30535a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
We can just use xfs_ilock/xfs_iunlock instead and get rid of the ugly
bhv_vrwlock_t.
SGI-PV: 976035
SGI-Modid: xfs-linux-melb:xfs-kern:30533a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Instead of of xfs_get_dir_entry use a macro to get the xfs_inode from the
dentry in the callers and grab the reference manually.
Only grab the reference once as it's fine to keep it over the dmapi calls.
(And even that reference is actually superflous in Linux but I'll leave
that for another patch)
SGI-PV: 976035
SGI-Modid: xfs-linux-melb:xfs-kern:30531a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Cleanup the unneeded intermediate vnode step in the flushing helpers and
go directly from the xfs_inode to the struct address_space.
SGI-PV: 976035
SGI-Modid: xfs-linux-melb:xfs-kern:30530a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
- use proper goto based unwinding instead of the current mess of
multiple conditionals
- rename ip to inode because that's the normal convention for Linux
inodes while ip is the convention for xfs_inodes
- remove unlikely checks for the default_acl - branches marked unlikely
might lead to extreme branch bredictor slowdons if taken and for some
workloads a default acl is quite common
- properly indent the switch statements
- remove xfs_has_fs_struct as nfsd has a fs_struct in any semi-recent
kernel
SGI-PV: 976035
SGI-Modid: xfs-linux-melb:xfs-kern:30529a
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Now that we update the log tail LSN less frequently on transaction
completion, we pass the contention straight to the global log state lock
(l_iclog_lock) during transaction completion.
We currently have to take this lock to decrement the iclog reference
count. there is a reference count on each iclog, so we need to take þhe
global lock for all refcount changes.
When large numbers of processes are all doing small trnasctions, the iclog
reference counts will be quite high, and the state change that absolutely
requires the l_iclog_lock is the except rather than the norm.
Change the reference counting on the iclogs to use atomic_inc/dec so that
we can use atomic_dec_and_lock during transaction completion and avoid the
need for grabbing the l_iclog_lock for every reference count decrement
except the one that matters - the last.
SGI-PV: 975671
SGI-Modid: xfs-linux-melb:xfs-kern:30505a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
When hundreds of processors attempt to commit transactions at the same
time, they can contend on the AIL lock when updating the tail LSN held in
the in-core log structure.
At the moment, the tail LSN is only needed when actually writing out an
iclog, so it really does not need to be updated on every single
transaction completion - only those that result in switching iclogs and
flushing them to disk.
The result is that we reduce the number of times we need to grab the AIL
lock and the log grant lock by up to two orders of magnitude on large
processor count machines. The problem has previously been hidden by AIL
lock contention walking the AIL list which was recently solved and
uncovered this issue.
SGI-PV: 975671
SGI-Modid: xfs-linux-melb:xfs-kern:30504a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Remove open coded checks for the whether the inode is clean and replace
them with an inlined function.
SGI-PV: 977461
SGI-Modid: xfs-linux-melb:xfs-kern:30503a
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_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>
When pdflush is writing back inodes, it can get stuck on inode cluster
buffers that are currently under I/O. This occurs when we write data to
multiple inodes in the same inode cluster at the same time.
Effectively, delayed allocation marks the inode dirty during the data
writeback. Hence if the inode cluster was flushed during the writeback of
the first inode, the writeback of the second inode will block waiting for
the inode cluster write to complete before writing it again for the newly
dirtied inode.
Basically, we want to avoid this from happening so we don't block pdflush
and slow down all of writeback. Hence we introduce a non-blocking async
inode flush flag that pdflush uses. If this flag is set, we use
non-blocking operations (e.g. try locks) whereever we can to avoid
blocking or extra I/O being issued.
SGI-PV: 970925
SGI-Modid: xfs-linux-melb:xfs-kern:30501a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
The only difference between the functions is one passes an inode for the
lookup, the other passes an inode number. However, they don't do the same
validity checking or set all the same state on the buffer that is returned
yet they should.
Factor the functions into a common implementation.
SGI-PV: 970925
SGI-Modid: xfs-linux-melb:xfs-kern:30500a
Signed-off-by: David Chinner <dgc@sgi.com>
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>
On a forced shutdown, xfs_finish_reclaim() will skip flushing the inode.
If the inode flush lock is not already held and there is an outstanding
xfs_iflush_done() then we might free the inode prematurely. By acquiring
and releasing the flush lock we will synchronise with xfs_iflush_done().
SGI-PV: 909874
SGI-Modid: xfs-linux-melb:xfs-kern:30468a
Signed-off-by: Lachlan McIlroy <lachlan@sgi.com>
Signed-off-by: David Chinner <dgc@sgi.com>
As pointed out by Sergey Vlasov, UDF implements its own version of
the CRC ITU-T V.41. Convert it to use the one in the library.
Signed-off-by: Bob Copeland <me@bobcopeland.com>
Cc: Sergey Vlasov <vsu@altlinux.ru>
Signed-off-by: Jan Kara <jack@suse.cz>
Fix two compilation warnings (and actual bugs in message formatting)
when UDF debugging is turned on.
Signed-off-by: Sebastian Manciulea <manciuleas@yahoo.com>
Signed-off-by: Jan Kara <jack@suse.cz>
Fix mapping of blocks using VAT when it is stored in an inode.
UDF_I(inode)->i_data already points to the beginning of VAT header so there's
no need to add udf_ext0_offset(inode).
Signed-off-by: Sebastian Manciulea <manciuleas@yahoo.com>
Signed-off-by: Jan Kara <jack@suse.cz>
This patch implements parsing of metadata partitions and reading of Metadata
File thus allowing to read UDF 2.50 media. Error resilience is implemented
through accessing the Metadata Mirror File in case the data the Metadata File
cannot be read. The patch is based on the original patch by Sebastian Manciulea
<manciuleas@yahoo.com> and Mircea Fedoreanu <mirceaf_spl@yahoo.com>.
Signed-off-by: Sebastian Manciulea <manciuleas@yahoo.com>
Signed-off-by: Mircea Fedoreanu <mirceaf_spl@yahoo.com>
Signed-off-by: Jan Kara <jack@suse.cz>
According to OSTA UDF specification, only anchor blocks and primary volume
descriptors are placed on media relative to the last session. All other block
numbers are absolute (in the partition or the whole media). This seems to be
confirmed by multisession media created by other systems.
Signed-off-by: Sebastian Manciulea <manciuleas@yahoo.com>
Signed-off-by: Jan Kara <jack@suse.cz>
As we don't properly support writing to pseudooverwrite partition (we should
add entries to VAT and relocate blocks instead of just writing them), mount
filesystems with such partition as read-only.
Signed-off-by: Jan Kara <jack@suse.cz>
We didn't handle VAT packed inside the inode - we tried to call udf_block_map()
on such file which lead to strange results at best. Add proper handling of
packed VAT as we do it with other packed files.
Signed-off-by: Jan Kara <jack@suse.cz>
UDF media with VAT could have never worked because udf_fill_inode() didn't
know about case FILE_TYPE_VAT20. Fix this.
Signed-off-by: Jan Kara <jack@suse.cz>
We incorrectly (way to strictly) checked version of VAT on loading and thus
refuse to mount correct media. There are just two format versions - below 2.0
and above 2.0 and we understand both. So update the version check accordingly.
Signed-off-by: Jan Kara <jack@suse.cz>
Add <last block>+1 and <last block>-1 to a list of blocks which can be the
real last recorded block on a UDF media. Sebastian Manciulea
<manciuleas@yahoo.com> claims this helps some drive + media combinations
he is able to test.
Signed-off-by: Jan Kara <jack@suse.cz>
UDF anchor block detection is complicated by several things - there are several
places where the anchor point can be, some of them relative to the last
recorded block which some devices report wrongly. Moreover some devices on some
media seem to have 7 spare blocks sectors for every 32 blocks (at least as far
as I understand the old code) so we have to count also with that possibility.
This patch splits anchor block detection into several functions so that it is
clearer what we actually try to do. We fix several bugs of the type "for such
and such media, we fail to check block blah" as a result of the cleanup.
Signed-off-by: Jan Kara <jack@suse.cz>
This patch move processing of UDF virtual partitions close to the place
where other partition types are processed. As a result we now also
properly fill in partition access type.
Signed-off-by: Jan Kara <jack@suse.cz>
Report error when we fail to allocate memory for a bitmap and properly
release allocated memory and inodes for all the partitions in case of
mount failure and umount.
Signed-off-by: Jan Kara <jack@suse.cz>
Cleanup processing of volume descriptor sequence so that it is more readable,
make code handle errors (e.g. media problems) better.
Signed-off-by: Jan Kara <jack@suse.cz>
According to ECMA 167 rev. 3 (see 3/8.4.2.1), Anchor Volume Descriptor
Pointer should be recorded at two or more anchor points located at sectors
256, N, N - 256, where N - is a largest logical sector number at volume
space.
So we should always try to detect N on UDF volume before trying to find
Anchor Volume Descriptor (i.e. calling to udf_find_anchor()).
That said, all this patch does is updates the s_last_block even if the
udf_vrs() returns positive value.
Originally written and tested by Yuri Per, ported on latest mainline by me.
Signed-off-by: Yuri Per <Yuri.Per@acronis.com>
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Cc: Max Lyadvinsky <Max.Lyadvinsky@acronis.com>
Cc: Vladimir Simonov <Vladimir.Simonov@acronis.com>
Cc: Andrew Neporada <Andrew.Neporada@acronis.com>
Cc: Kirill Korotaev <dev@openvz.org>
Signed-off-by: Jan Kara <jack@suse.cz>
There are several places in UDF where we declared temporary arrays of
UDF_NAME_LEN bytes on stack. This is not nice to stack usage so this patch
changes those places to use kmalloc() instead. Also clean up bail-out paths
in those functions when we are changing them.
Signed-off-by: Jan Kara <jack@suse.cz>
We don't have to check whether a directory entry already exists in a directory
when creating a new one since we've already checked that earlier by lookup and
we are holding directory i_mutex all the time.
Signed-off-by: Jan Kara <jack@suse.cz>
reorganize few code blocks in super.c which
were needlessly indented (and hard to read):
so change from:
rettype fun()
{
init;
if (sth) {
long block of code;
}
}
to:
rettype fun()
{
init;
if (!sth)
return;
long block of code;
}
or
from:
rettype fun2()
{
init;
while (sth) {
init2();
if (sth2) {
long block of code;
}
}
}
to:
rettype fun2()
{
init;
while (sth) {
init2();
if (!sth2)
continue;
long block of code;
}
}
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Jan Kara <jack@suse.cz>
remove now unneeded kernel_timestamp type with conversion functions
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Jan Kara <jack@suse.cz>
* kernel_timestamp type was almost unused - only callers of udf_stamp_to_time
and udf_time_to_stamp used it, so let these functions handle endianness
internally and don't clutter code with conversions
* rename udf_stamp_to_time to udf_disk_stamp_to_time
and udf_time_to_stamp to udf_time_to_disk_stamp
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Jan Kara <jack@suse.cz>
block cannot be less than 0, because it's sector_t,
so remove unneeded checks
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Jan Kara <jack@suse.cz>
- translate udf_file_entry_alloc_offset macro into function
- translate udf_ext0_offset macro into function
- add comment about crypticly named fields in struct udf_inode_info
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Jan Kara <jack@suse.cz>
- move all brelse(ibh) after main if, because it's called
on every path except one where ibh is null
- move variables to the most inner blocks
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Jan Kara <jack@suse.cz>
replace all:
little_endian_variable = cpu_to_leX(leX_to_cpu(little_endian_variable) +
expression_in_cpu_byteorder);
with:
leX_add_cpu(&little_endian_variable, expression_in_cpu_byteorder);
sparse didn't generate any new warning with this patch
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com
Signed-off-by: Jan Kara <jack@suse.cz>
- remove one indentation level by little code reorganization
- convert "if (smth) BUG();" to "BUG_ON(smth);"
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Jan Kara <jack@suse.cz>