btrfs_next_leaf() can return -errno, and we should propagate
it to userspace.
This also simplifies how we walk the btree path.
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
btrfs_next_leaf() can return -errno, and we should propagate
it to userspace.
This also simplifies how we walk the btree path.
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
The extent_io code can take cached pointers into the extent state trees,
and these can make lookups much faster in common operations. The
caching only happens when specific bits are set that prevent merging
and splitting of the extent state.
A help function was added to uncache the state, and it was testing
the same set of conditionals. This can leak in very strange corner
cases where the lock bit goes away unexpectedly.
The uncaching should be unconditional. Once we have a ref on the
extent we should always give it up.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The inode was used in an earlier version of the code, but it isn't
used anymore.
Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
In fs/nfs/super.c::nfs_fs_mount() we test for a NULL 'data':
...
if (data == NULL || mntfh == NULL)
goto out_free_fh;
...
and then further down in the function we test 'data' again:
...
nfs_fscache_get_super_cookie(
s, data ? data->fscache_uniq : NULL, NULL);
...
this second check is just dead code since there is no way 'data' could
possibly be NULL here.
We also rely on a non-NULL 'data' in more than one location between these
two tests, further proving the point that the second test is bogus.
This patch removes the dead code.
Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* git://git.kernel.org/pub/scm/linux/kernel/git/sfrench/cifs-2.6:
cifs: don't allow mmap'ed pages to be dirtied while under writeback (try #3)
[CIFS] Warn on requesting default security (ntlm) on mount
[CIFS] cifs: clarify the meaning of tcpStatus == CifsGood
cifs: wrap received signature check in srv_mutex
cifs: clean up various nits in unicode routines (try #2)
cifs: clean up length checks in check2ndT2
cifs: set ra_pages in backing_dev_info
cifs: fix broken BCC check in is_valid_oplock_break
cifs: always do is_path_accessible check in cifs_mount
various endian fixes to cifs
Elminate sparse __CHECK_ENDIAN__ warnings on port conversion
Max share size is too small
Allow user names longer than 32 bytes
cifs: replace /proc/fs/cifs/Experimental with a module parm
cifs: check for private_data before trying to put it
nfs_scan_commit() is called with the inode->i_lock held, but it then
calls __mark_inode_dirty() while still holding the lock. This causes
a deadlock.
Push the inode->i_lock into nfs_scan_commit() so it can protect only
the parts of the code it needs to and can be dropped before the call
to __mark_inode_dirty() to avoid the deadlock.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Tested-by: Will Simoneau <simoneau@ele.uri.edu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This reverts commit 93f1c20bc8.
It turns out that libmount misparses it because it adds a '-' character
in the uuid string, which libmount then incorrectly confuses with the
separator string (" - ") at the end of all the optional arguments.
Upstream libmount (in the util-linux tree) has been fixed, but until
that fix actually percolates up to users, we'd better not expose this
change in the kernel.
Let's revisit this later (possibly by exposing the UUID without any '-'
characters in it, avoiding the user-space bug).
Reported-by: Dave Jones <davej@redhat.com>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Karel Zak <kzak@redhat.com>
Cc: Ram Pai <linuxram@us.ibm.com>
Cc: Miklos Szeredi <mszeredi@suse.cz>
Cc: Eric Sandeen <sandeen@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This is more or less the same patch as before, but with some merge
conflicts fixed up.
If a process has a dirty page mapped into its page tables, then it has
the ability to change it while the client is trying to write the data
out to the server. If that happens after the signature has been
calculated then that signature will then be wrong, and the server will
likely reset the TCP connection.
This patch adds a page_mkwrite handler for CIFS that simply takes the
page lock. Because the page lock is held over the life of writepage and
writepages, this prevents the page from becoming writeable until
the write call has completed.
With this, we can also remove the "sign_zero_copy" module option and
always inline the pages when writing.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
Warn once if default security (ntlm) requested. We will
update the default to the stronger security mechanism
(ntlmv2) in 2.6.41. Kerberos is also stronger than
ntlm, but more servers support ntlmv2 and ntlmv2
does not require an upcall, so ntlmv2 is a better
default.
Reviewed-by: Jeff Layton <jlayton@redhat.com>
CC: Suresh Jayaraman <sjayaraman@suse.de>
Reviewed-by: Shirish Pargaonkar <shirishp@us.ibm.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
When the TCP_Server_Info is first allocated and connected, tcpStatus ==
CifsGood means that the NEGOTIATE_PROTOCOL request has completed and the
socket is ready for other calls. cifs_reconnect however sets tcpStatus
to CifsGood as soon as the socket is reconnected and the optional
RFC1001 session setup is done. We have no clear way to tell the
difference between these two states, and we need to know this in order
to know whether we can send an echo or not.
Resolve this by adding a new statusEnum value -- CifsNeedNegotiate. When
the socket has been connected but has not yet had a NEGOTIATE_PROTOCOL
request done, set it to this value. Once the NEGOTIATE is done,
cifs_negotiate_protocol will set tcpStatus to CifsGood.
This also fixes and cleans the logic in cifs_reconnect and
cifs_reconnect_tcon. The old code checked for specific states when what
it really wants to know is whether the state has actually changed from
CifsNeedReconnect.
Reported-and-Tested-by: JG <jg@cms.ac>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
While testing my patchset to fix asynchronous writes, I hit a bunch
of signature problems when testing with signing on. The problem seems
to be that signature checks on receive can be running at the same
time as a process that is sending, or even that multiple receives can
be checking signatures at the same time, clobbering the same data
structures.
While we're at it, clean up the comments over cifs_calculate_signature
and add a note that the srv_mutex should be held when calling this
function.
This patch seems to fix the problems for me, but I'm not clear on
whether it's the best approach. If it is, then this should probably
go to stable too.
Cc: stable@kernel.org
Cc: Shirish Pargaonkar <shirishpargaonkar@gmail.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
Minor revision to the original patch. Don't abuse the __le16 variable
on the stack by casting it to wchar_t and handing it off to char2uni.
Declare an actual wchar_t on the stack instead. This fixes a valid
sparse warning.
Fix the spelling of UNI_ASTERISK. Eliminate the unneeded len_remaining
variable in cifsConvertToUCS.
Also, as David Howells points out. We were better off making
cifsConvertToUCS *not* use put_unaligned_le16 since it means that we
can't optimize the mapped characters at compile time. Switch them
instead to use cpu_to_le16, and simply use put_unaligned to set them
in the string.
Reported-and-acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
Thus spake David Howells:
The code that follows this:
remaining = total_data_size - data_in_this_rsp;
if (remaining == 0)
return 0;
else if (remaining < 0) {
generates better code if you drop the 'remaining' variable and compare
the values directly.
Clean it up per his recommendation...
Reported-and-acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
Commit 522440ed made cifs set backing_dev_info on the mapping attached
to new inodes. This change caused a fairly significant read performance
regression, as cifs started doing page-sized reads exclusively.
By virtue of the fact that they're allocated as part of cifs_sb_info by
kzalloc, the ra_pages on cifs BDIs get set to 0, which prevents any
readahead. This forces the normal read codepaths to use readpage instead
of readpages causing a four-fold increase in the number of read calls
with the default rsize.
Fix it by setting ra_pages in the BDI to the same value as that in the
default_backing_dev_info.
Fixes https://bugzilla.kernel.org/show_bug.cgi?id=31662
Cc: stable@kernel.org
Reported-and-Tested-by: Till <till2.schaefer@uni-dortmund.de>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
The BCC is still __le16 at this point, and in any case we need to
use the get_bcc_le macro to make sure we don't hit alignment
problems.
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
Currently, we skip doing the is_path_accessible check in cifs_mount if
there is no prefixpath. I have a report of at least one server however
that allows a TREE_CONNECT to a share that has a DFS referral at its
root. The reporter in this case was using a UNC that had no prefixpath,
so the is_path_accessible check was not triggered and the box later hit
a BUG() because we were chasing a DFS referral on the root dentry for
the mount.
This patch fixes this by removing the check for a zero-length
prefixpath. That should make the is_path_accessible check be done in
this situation and should allow the client to chase the DFS referral at
mount time instead.
Cc: stable@kernel.org
Reported-and-Tested-by: Yogesh Sharma <ysharma@cymer.com>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
make modules C=2 M=fs/cifs CF=-D__CHECK_ENDIAN__
Found for example:
CHECK fs/cifs/cifssmb.c
fs/cifs/cifssmb.c:728:22: warning: incorrect type in assignment (different base types)
fs/cifs/cifssmb.c:728:22: expected unsigned short [unsigned] [usertype] Tid
fs/cifs/cifssmb.c:728:22: got restricted __le16 [usertype] <noident>
fs/cifs/cifssmb.c:1883:45: warning: incorrect type in assignment (different base types)
fs/cifs/cifssmb.c:1883:45: expected long long [signed] [usertype] fl_start
fs/cifs/cifssmb.c:1883:45: got restricted __le64 [usertype] start
fs/cifs/cifssmb.c:1884:54: warning: restricted __le64 degrades to integer
fs/cifs/cifssmb.c:1885:58: warning: restricted __le64 degrades to integer
fs/cifs/cifssmb.c:1886:43: warning: incorrect type in assignment (different base types)
fs/cifs/cifssmb.c:1886:43: expected unsigned int [unsigned] fl_pid
fs/cifs/cifssmb.c:1886:43: got restricted __le32 [usertype] pid
In checking new smb2 code for missing endian conversions, I noticed
some endian errors had crept in over the last few releases into the
cifs code (symlink, ntlmssp, posix lock, and also a less problematic warning
in fscache). A followon patch will address a few smb2 endian
problems.
Reviewed-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
Ports are __be16 not unsigned short int
Eliminates the remaining fixable endian warnings:
~/cifs-2.6$ make modules C=1 M=fs/cifs CF=-D__CHECK_ENDIAN__
CHECK fs/cifs/connect.c
fs/cifs/connect.c:2408:23: warning: incorrect type in assignment (different base types)
fs/cifs/connect.c:2408:23: expected unsigned short *sport
fs/cifs/connect.c:2408:23: got restricted __be16 *<noident>
fs/cifs/connect.c:2410:23: warning: incorrect type in assignment (different base types)
fs/cifs/connect.c:2410:23: expected unsigned short *sport
fs/cifs/connect.c:2410:23: got restricted __be16 *<noident>
fs/cifs/connect.c:2416:24: warning: incorrect type in assignment (different base types)
fs/cifs/connect.c:2416:24: expected unsigned short [unsigned] [short] <noident>
fs/cifs/connect.c:2416:24: got restricted __be16 [usertype] <noident>
fs/cifs/connect.c:2423:24: warning: incorrect type in assignment (different base types)
fs/cifs/connect.c:2423:24: expected unsigned short [unsigned] [short] <noident>
fs/cifs/connect.c:2423:24: got restricted __be16 [usertype] <noident>
fs/cifs/connect.c:2326:23: warning: incorrect type in assignment (different base types)
fs/cifs/connect.c:2326:23: expected unsigned short [unsigned] sport
fs/cifs/connect.c:2326:23: got restricted __be16 [usertype] sin6_port
fs/cifs/connect.c:2330:23: warning: incorrect type in assignment (different base types)
fs/cifs/connect.c:2330:23: expected unsigned short [unsigned] sport
fs/cifs/connect.c:2330:23: got restricted __be16 [usertype] sin_port
fs/cifs/connect.c:2394:22: warning: restricted __be16 degrades to integer
Signed-off-by: Steve French <sfrench@us.ibm.com>
In several places the sequence (set_extent_uptodate, unlock_extent) is used.
This leads to a duplicate lookup of the extent state. This patch lets
set_extent_uptodate return a cached extent_state which can be passed to
unlock_extent_cached.
The occurences of the above sequences are updated to use the cache. Only
end_bio_extent_readpage is updated that it first gets a cached state to
pass it to the readpage_end_io_hook as the prototype requested and is later
on being used for set/unlock.
Signed-off-by: Arne Jansen <sensille@gmx.net>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
I've been working on making our O_DIRECT latency not suck and I noticed we were
taking the trans_mutex in btrfs_end_transaction. So to do this we convert
num_writers and use_count to atomic_t's and just decrement them in
btrfs_end_transaction. Instead of deleting the transaction from the trans list
in put_transaction we do that in btrfs_commit_transaction() since that's the
only time it actually needs to be removed from the list. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
We artificially limited the user name to 32 bytes, but modern servers handle
larger. Set the maximum length to a reasonable 256, and make the user name
string dynamically allocated rather than a fixed size in session structure.
Also clean up old checkpatch warning.
Signed-off-by: Steve French <sfrench@us.ibm.com>
This flag currently only affects whether we allow "zero-copy" writes
with signing enabled. Typically we map pages in the pagecache directly
into the write request. If signing is enabled however and the contents
of the page change after the signature is calculated but before the
write is sent then the signature will be wrong. Servers typically
respond to this by closing down the socket.
Still, this can provide a performance benefit so the "Experimental" flag
was overloaded to allow this. That's really not a good place for this
option however since it's not clear what that flag does.
Move that flag instead to a new module parameter that better describes
its purpose. That's also better since it can be set at module insertion
time by configuring modprobe.d.
Reviewed-by: Suresh Jayaraman <sjayaraman@suse.de>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
cifs_close doesn't check that the filp->private_data is non-NULL before
trying to put it. That can cause an oops in certain error conditions
that can occur on open or lookup before the private_data is set.
Reported-by: Ben Greear <greearb@candelatech.com>
CC: Stable <stable@kernel.org>
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: Steve French <sfrench@us.ibm.com>
We create two subvolumes (meego_root and meego_home) in
btrfs root directory. And set meego_root as default mount
subvolume. After we remount btrfs, meego_root is mounted
to top directory by default. Then when we try to mount
meego_home (subvol=meego_home) to a subdirectory, it failed.
The problem is when default mount subvolume is set to
meego_root, we search meego_home in meego_root but can not find
it. So the solution is to add a new mount option (subvolrootid)
to specify subvol id of root and search subvol name in it. For
our case, now we can use "-o subvolrootid=0,subvol=meego_home)
to mount meego_home.
Detail information can be found in meego bugzilla:
https://bugs.meego.com/show_bug.cgi?id=15055
Signed-off-by: Zhong, Xin <xin.zhong@intel.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Apparently it is ok to submit a read to an IDE device with the same target page
for different offsets. This is what Windows does under qemu. The problem is
under DIO we expect them to be different buffers for checksumming reasons, and
so this sort of thing will result in checksum errors, when in reality the file
is fine. So when reading, check to make sure that all iov bases are different,
and if they aren't fall back to buffered mode, since that will work out right.
Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This patch fixes memory leaks in btrfs_new_inode().
Signed-off-by: Yoshinori Sano <yoshinori.sano@gmail.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
* 'for-linus' of git://oss.sgi.com/xfs/xfs:
xfs: use proper interfaces for on-stack plugging
xfs: fix xfs_debug warnings
xfs: fix variable set but not used warnings
xfs: convert log tail checking to a warning
xfs: catch bad block numbers freeing extents.
xfs: push the AIL from memory reclaim and periodic sync
xfs: clean up code layout in xfs_trans_ail.c
xfs: convert the xfsaild threads to a workqueue
xfs: introduce background inode reclaim work
xfs: convert ENOSPC inode flushing to use new syncd workqueue
xfs: introduce a xfssyncd workqueue
xfs: fix extent format buffer allocation size
xfs: fix unreferenced var error in xfs_buf.c
Also, applied patch from Tony Luck that fixes ia64:
xfs_destroy_workqueues() should not be tagged with__exit
in the branch before merging.
ia64 throws away .exit sections for the built-in CONFIG case, so routines
that are used in other circumstances should not be tagged as __exit.
Signed-off-by: Tony Luck <tony.luck@intel.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:
ext4: fix data corruption regression by reverting commit 6de9843dab
ext4: Allow indirect-block file to grow the file size to max file size
ext4: allow an active handle to be started when freezing
ext4: sync the directory inode in ext4_sync_parent()
ext4: init timer earlier to avoid a kernel panic in __save_error_info
jbd2: fix potential memory leak on transaction commit
ext4: fix a double free in ext4_register_li_request
ext4: fix credits computing for indirect mapped files
ext4: remove unnecessary [cm]time update of quota file
jbd2: move bdget out of critical section
Revert commit 6de9843dab, since it
caused a data corruption regression with BitTorrent downloads. Thanks
to Damien for discovering and bisecting to find the problem commit.
https://bugzilla.kernel.org/show_bug.cgi?id=32972
Reported-by: Damien Grassart <damien@grassart.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
We can create 4402345721856 byte file with indirect block mapping.
However, if we grow an indirect-block file to the size with ftruncate(),
we can see an ext4 warning. The following patch fixes this problem.
How to reproduce:
# dd if=/dev/zero of=/mnt/mp1/hoge bs=1 count=0 seek=4402345721856
0+0 records in
0+0 records out
0 bytes (0 B) copied, 0.000221428 s, 0.0 kB/s
# tail -n 1 /var/log/messages
Nov 25 15:10:27 test kernel: EXT4-fs warning (device sda8): ext4_block_to_path:345: block 1074791436 > max in inode 12
Signed-off-by: Kazuya Mio <k-mio@sx.jp.nec.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
ext4_journal_start_sb() should not prevent an active handle from being
started due to s_frozen. Otherwise, deadlock is easy to happen, below
is a situation.
================================================
freeze | truncate
================================================
| ext4_ext_truncate()
freeze_super() | starts a handle
sets s_frozen |
| ext4_ext_truncate()
| holds i_data_sem
ext4_freeze() |
waits for updates |
| ext4_free_blocks()
| calls dquot_free_block()
|
| dquot_free_blocks()
| calls ext4_dirty_inode()
|
| ext4_dirty_inode()
| trys to start an active
| handle
|
| block due to s_frozen
================================================
Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Reported-by: Amir Goldstein <amir73il@users.sf.net>
Reviewed-by: Jan Kara <jack@suse.cz>
Reviewed-by: Andreas Dilger <adilger@dilger.ca>
ext4 has taken the stance that, in the absence of a journal,
when an fsync/fdatasync of an inode is done, the parent
directory should be sync'ed if this inode entry is new.
ext4_sync_parent(), which implements this, does indeed sync
the dirent pages for parent directories, but it does not
sync the directory *inode*. This patch fixes this.
Also now return error status from ext4_sync_parent().
I tested this using a power fail test, which panics a
machine running a file server getting requests from a
client. Without this patch, on about every other test run,
the server is missing many, many files that had been synced.
With this patch, on > 6 runs, I see zero files being lost.
Google-Bug-Id: 4179519
Signed-off-by: Curt Wohlgemuth <curtw@google.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
* 'bugfixes' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6:
NFS: Change initial mount authflavor only when server returns NFS4ERR_WRONGSEC
NFS: Fix a signed vs. unsigned secinfo bug
Revert "net/sunrpc: Use static const char arrays"
Apparently it is ok to submit a read to an IDE device with the same target page
for different offsets. This is what Windows does under qemu. The problem is
under DIO we expect them to be different buffers for checksumming reasons, and
so this sort of thing will result in checksum errors, when in reality the file
is fine. So when reading, check to make sure that all iov bases are different,
and if they aren't fall back to buffered mode, since that will work out right.
Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
In btrfs_get_block_direct we call btrfs_get_extent to lookup the extent for the
range that we are looking for. If we don't find an extent, btrfs_get_extent
will insert a extent_map for that area and mark it as a hole. So it does the
job of allocating a new extent map and inserting it into the io tree. But if
we're creating a new extent we free it up and redo all of that work. So instead
pass the em to btrfs_new_extent_direct(), and if it will work just allocate the
disk space and set it up properly and bypass the freeing/allocating of a new
extent map and the expensive operation of inserting the thing into the io_tree.
Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
When looking at our DIO performance Chris said that for small IO's doing the
async submit stuff tends to be more overhead than it's worth. With this on top
of my other fixes I get about a 17-20% speedup doing a sequential dd with 4k
IO's. Basically if we don't have to split the bio for the map length it's small
enough to be directly submitted, otherwise go back to the async submit. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
We have been unconditionally allocating a new bio and re-adding all pages from
our original bio to the new bio. This is needed if our original bio is larger
than our stripe size, but if it is smaller than the stripe size then there is no
need to do this. So check the map length and if we are under that then go ahead
and submit the original bio. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
In the DIO code we often don't update the i_disk_size because the i_size isn't
updated until after the DIO is completed, so basically we are allocating a path,
doing a search, and updating the inode item for no reason since nothing changed.
btrfs_ordered_update_i_size will return 1 if it didn't update i_disk_size, so
only run btrfs_update_inode if btrfs_ordered_update_i_size returns 0. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Instead of calling kmap_atomic for every thing we set in the inode item, map the
entire inode item at the start and unmap it at the end. This makes a sequential
dd of 400mb O_DIRECT something like 1% faster. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
I saw a lockup where we kept getting into this start transaction->commit
transaction loop because of enospce. The fact is if we fail to make our
reservation, we've tried _everything_ several times, so we only need to try and
commit the transaction once, and if that doesn't work then we really are out of
space and need to just exit. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Currently we don't handle running out of space in the cache, so to fix this we
keep track of how far in the cache we are. Then we only dirty the pages if we
successfully modify all of them, otherwise if we have an error or run out of
space we can just drop them and not worry about the vm writing them out.
Thanks,
Tested-by Johannes Hirte <johannes.hirte@fem.tu-ilmenau.de>
Signed-off-by: Josef Bacik <josef@redhat.com>
* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs-2.6:
quota: Don't write quota info in dquot_commit()
ext3: Fix writepage credits computation for ordered mode
Add proper blk_start_plug/blk_finish_plug pairs for the two places where
we issue buffer I/O, and remove the blk_flush_plug in xfs_buf_lock and
xfs_buf_iowait, given that context switches already flush the per-process
plugging lists.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>
For a CONFIG_XFS_DEBUG=n build gcc complains about statements with no
effect in xfs_debug:
fs/xfs/quota/xfs_qm_syscalls.c: In function 'xfs_qm_scall_trunc_qfiles':
fs/xfs/quota/xfs_qm_syscalls.c:291:3: warning: statement with no effect
The reason for that is that the various new xfs message functions have a
return value which is never used, and in case of the non-debug build
xfs_debug the macro evaluates to a plain 0 which produces the above
warnings. This can be fixed by turning xfs_debug into an inline function
instead of a macro, but in addition to that I've also changed all the
message helpers to return void as we never use their return values.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Alex Elder <aelder@sgi.com>
GCC 4.6 now warnings about variables set but not used. Fix the trivially
fixable warnings of this sort.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Alex Elder <aelder@sgi.com>
On the Power platform, the log tail debug checks fire excessively
causing the system to panic early in testing. The debug checks are
known to be racy, though on x86_64 there is no evidence that they
trigger at all.
We want to keep the checks active on debug systems to alert us to
problems with log space accounting, but we need to reduce the impact
of a racy check on testing on the Power platform.
As a result, convert the ASSERT conditions to warnings, and
allow them to fire only once per filesystem mount. This will prevent
false positives from interfering with testing, whilst still
providing us with the indication that they may be a problem with log
space accounting should that occur.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
A fuzzed filesystem crashed a kernel when freeing an extent with a
block number beyond the end of the filesystem. Convert all the debug
asserts in xfs_free_extent() to active checks so that we catch bad
extents and return that the filesytsem is corrupted rather than
crashing.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
When we are short on memory, we want to expedite the cleaning of
dirty objects. Hence when we run short on memory, we need to kick
the AIL flushing into action to clean as many dirty objects as
quickly as possible. To implement this, sample the lsn of the log
item at the head of the AIL and use that as the push target for the
AIL flush.
Further, we keep items in the AIL that are dirty that are not
tracked any other way, so we can get objects sitting in the AIL that
don't get written back until the AIL is pushed. Hence to get the
filesystem to the idle state, we might need to push the AIL to flush
out any remaining dirty objects sitting in the AIL. This requires
the same push mechanism as the reclaim push.
This patch also renames xfs_trans_ail_tail() to xfs_ail_min_lsn() to
match the new xfs_ail_max_lsn() function introduced in this patch.
Similarly for xfs_trans_ail_push -> xfs_ail_push.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Alex Elder <aelder@sgi.com>
This patch rearranges the location of functions in xfs_trans_ail.c
to remove the need for forward declarations of those functions in
preparation for adding new functions without the need for forward
declarations.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Alex Elder <aelder@sgi.com>
Similar to the xfssyncd, the per-filesystem xfsaild threads can be
converted to a global workqueue and run periodically by delayed
works. This makes sense for the AIL pushing because it uses
variable timeouts depending on the work that needs to be done.
By removing the xfsaild, we simplify the AIL pushing code and
remove the need to spread the code to implement the threading
and pushing across multiple files.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
Background inode reclaim needs to run more frequently that the XFS
syncd work is run as 30s is too long between optimal reclaim runs.
Add a new periodic work item to the xfs syncd workqueue to run a
fast, non-blocking inode reclaim scan.
Background inode reclaim is kicked by the act of marking inodes for
reclaim. When an AG is first marked as having reclaimable inodes,
the background reclaim work is kicked. It will continue to run
periodically untill it detects that there are no more reclaimable
inodes. It will be kicked again when the first inode is queued for
reclaim.
To ensure shrinker based inode reclaim throttles to the inode
cleaning and reclaim rate but still reclaim inodes efficiently, make it kick the
background inode reclaim so that when we are low on memory we are
trying to reclaim inodes as efficiently as possible. This kick shoul
d not be necessary, but it will protect against failures to kick the
background reclaim when inodes are first dirtied.
To provide the rate throttling, make the shrinker pass do
synchronous inode reclaim so that it blocks on inodes under IO. This
means that the shrinker will reclaim inodes rather than just
skipping over them, but it does not adversely affect the rate of
reclaim because most dirty inodes are already under IO due to the
background reclaim work the shrinker kicked.
These two modifications solve one of the two OOM killer invocations
Chris Mason reported recently when running a stress testing script.
The particular workload trigger for the OOM killer invocation is
where there are more threads than CPUs all unlinking files in an
extremely memory constrained environment. Unlike other solutions,
this one does not have a performance impact on performance when
memory is not constrained or the number of concurrent threads
operating is <= to the number of CPUs.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
On of the problems with the current inode flush at ENOSPC is that we
queue a flush per ENOSPC event, regardless of how many are already
queued. Thi can result in hundreds of queued flushes, most of
which simply burn CPU scanned and do no real work. This simply slows
down allocation at ENOSPC.
We really only need one active flush at a time, and we can easily
implement that via the new xfs_syncd_wq. All we need to do is queue
a flush if one is not already active, then block waiting for the
currently active flush to complete. The result is that we only ever
have a single ENOSPC inode flush active at a time and this greatly
reduces the overhead of ENOSPC processing.
On my 2p test machine, this results in tests exercising ENOSPC
conditions running significantly faster - 042 halves execution time,
083 drops from 60s to 5s, etc - while not introducing test
regressions.
This allows us to remove the old xfssyncd threads and infrastructure
as they are no longer used.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
All of the work xfssyncd does is background functionality. There is
no need for a thread per filesystem to do this work - it can al be
managed by a global workqueue now they manage concurrency
effectively.
Introduce a new gglobal xfssyncd workqueue, and convert the periodic
work to use this new functionality. To do this, use a delayed work
construct to schedule the next running of the periodic sync work
for the filesystem. When the sync work is complete, queue a new
delayed work for the next running of the sync work.
For laptop mode, we wait on completion for the sync works, so ensure
that the sync work queuing interface can flush and wait for work to
complete to enable the work queue infrastructure to replace the
current sequence number and wakeup that is used.
Because the sync work does non-trivial amounts of work, mark the
new work queue as CPU intensive.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
When formatting an inode item, we have to allocate a separate buffer
to hold extents when there are delayed allocation extents on the
inode and it is in extent format. The allocation size is derived
from the in-core data fork representation, which accounts for
delayed allocation extents, while the on-disk representation does
not contain any delalloc extents.
As a result of this mismatch, the allocated buffer can be far larger
than needed to hold the real extent list which, due to the fact the
inode is in extent format, is limited to the size of the literal
area of the inode. However, we can have thousands of delalloc
extents, resulting in an allocation size orders of magnitude larger
than is needed to hold all the real extents.
Fix this by limiting the size of the buffer being allocated to the
size of the literal area of the inodes in the filesystem (i.e. the
maximum size an inode fork can grow to).
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Alex Elder <aelder@sgi.com>
When attempting an initial mount, we should only attempt other
authflavors if AUTH_UNIX receives a NFS4ERR_WRONGSEC error.
This allows other errors to be passed back to userspace programs.
Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* 'for-linus' of git://git.infradead.org/ubifs-2.6:
UBI: do not select KALLSYMS_ALL
UBI: do not compare array with NULL
UBI: check if we are in RO mode in the erase routine
UBIFS: fix debugging failure in dbg_check_space_info
UBIFS: fix error path in dbg_debugfs_init_fs
UBIFS: unify error path dbg_debugfs_init_fs
UBIFS: do not select KALLSYMS_ALL
UBIFS: fix assertion warnings
UBIFS: fix oops on error path in read_pnode
UBIFS: do not read flash unnecessarily
rpc_authflavor_t is cast from an unsigned int, but the
initial code tried to use it as a signed int. I fix
this by passing an rpc_authflavor_t pointer around, and
returning signed integers from functions.
Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Reported-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
During mount, when we fail to open journal inode or root inode, the
__save_error_info will mod_timer. But actually s_err_report isn't
initialized yet and the kernel oops. The detailed information can
be found https://bugzilla.kernel.org/show_bug.cgi?id=32082.
The best way is to check whether the timer s_err_report is initialized
or not. But it seems that in include/linux/timer.h, we can't find a
good function to check the status of this timer, so this patch just
move the initializtion of s_err_report earlier so that we can avoid
the kernel panic. The corresponding del_timer is also added in the
error path.
Reported-by: Sami Liedes <sliedes@cc.hut.fi>
Signed-off-by: Tao Ma <boyu.mt@taobao.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
There is potential memory leak of journal head in function
jbd2_journal_commit_transaction. The problem is that JBD2 will not
reclaim the journal head of commit record if error occurs or journal
is abotred.
I use the following script to reproduce this issue, on a RHEL6
system. I found it very easy to reproduce with async commit enabled.
mount /dev/sdb /mnt -o journal_checksum,journal_async_commit
touch /mnt/xxx
echo offline > /sys/block/sdb/device/state
sync
umount /mnt
rmmod ext4
rmmod jbd2
Removal of the jbd2 module will make slab complaining that
"cache `jbd2_journal_head': can't free all objects".
Signed-off-by: Zhang Huan <zhhuan@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/axboe/linux-2.6-block:
ide: always ensure that blk_delay_queue() is called if we have pending IO
block: fix request sorting at unplug
dm: improve block integrity support
fs: export empty_aops
ide: ide_requeue_and_plug() reinstate "always plug" behaviour
blk-throttle: don't call xchg on bool
ufs: remove unessecary blk_flush_plug
block: make the flush insertion use the tail of the dispatch list
block: get rid of elv_insert() interface
block: dump request state on seeing a corrupted request completion
On an error path in inotify_init1 a normal user can trigger a double
free of struct user. This is a regression introduced by a2ae4cc9a1
("inotify: stop kernel memory leak on file creation failure").
We fix this by making sure that if a group exists the user reference is
dropped when the group is cleaned up. We should not explictly drop the
reference on error and also drop the reference when the group is cleaned
up.
The new lifetime rules are that an inotify group lives from
inotify_new_group to the last fsnotify_put_group. Since the struct user
and inotify_devs are directly tied to this lifetime they are only
changed/updated in those two locations. We get rid of all special
casing of struct user or user->inotify_devs.
Signed-off-by: Eric Paris <eparis@redhat.com>
Cc: stable@kernel.org (2.6.37 and up)
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
With the ->sync_page() hook gone, we have a few users that
add their own static address_space_operations without any
functions defined.
fs/inode.c already has an empty_aops that it uses for init
purposes. Lets export that and use it in the places where
an otherwise empty aops was defined.
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
We already flush the per-process plugging list when context switching,
so a blk_flush_plug call just before a yield() is not needed.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable:
Btrfs: don't warn in btrfs_add_orphan
Btrfs: fix free space cache when there are pinned extents and clusters V2
Btrfs: Fix uninitialized root flags for subvolumes
btrfs: clear __GFP_FS flag in the space cache inode
Btrfs: fix memory leak in start_transaction()
Btrfs: fix memory leak in btrfs_ioctl_start_sync()
Btrfs: fix subvol_sem leak in btrfs_rename()
Btrfs: Fix oops for defrag with compression turned on
Btrfs: fix /proc/mounts info.
Btrfs: fix compiler warning in file.c
This patch fixes a debugging failure with which looks like this:
UBIFS error (pid 32313): dbg_check_space_info: free space changed from 6019344 to 6022654
The reason for this failure is described in the comment this patch adds
to the code. But in short - 'c->freeable_cnt' may be different before
and after re-mounting, and this is normal. So the debugging code should
make sure that free space calculations do not depend on 'c->freeable_cnt'.
A similar issue has been reported here:
http://lists.infradead.org/pipermail/linux-mtd/2011-April/034647.html
This patch should fix it.
For the -stable guys: this patch is only relevant for kernels 2.6.30
onwards.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Cc: stable@kernel.org [2.6.30+]
The debug interface is substandard and on error returns either
NULL or an error code packed in the pointer. So using "IS_ERR"
for the pointers returned by debugfs function is incorrect.
Instead, we should use IS_ERR_OR_NULL.
This path is an improved vestion of the original patch from
Phil Carmody.
Reported-by: Phil Carmody <ext-phil.2.carmody@nokia.com>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Acked-by: Phil Carmody <ext-phil.2.carmody@nokia.com>
This is just a small clean-up patch which simlifies and unifies the
error path in the dbg_debugfs_init_fs(). We have common error path
for all failure cases in this function except of the very first
case. And this patch makes the first failure case use the same
error path as the other cases by using the 'fname' and 'dent'
variables.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Acked-by: Phil Carmody <ext-phil.2.carmody@nokia.com>
All UBIFS needs is to make sure we stacktraces when UBIFS debugging
is enabled. It is enough to select KALLSYMS for this, KALLSYMS_ALL
is not necessary. Moreover, Randy Dunlap reported that UBIFS causes
the following Kconfig dependency warning:
warning: (UBIFS_FS_DEBUG && LOCKDEP && LATENCYTOP) selects KALLSYMS_ALL
which has unmet direct dependencies (DEBUG_KERNEL && KALLSYMS)
The reason is that KALLSYMS_ALL requires DEBUG_KERNEL and KALLSYMS, so
ideally, to select KALLSYMS_ALL we'd need to select DEBUG_KERNEL and
KALLSYMS first.
This seems to be too much to select. The easiest way to go is to forget
about KALLSYMS_ALL and just select KALLSYMS when UBIFS debugging is
enabled - that should be enough for stackdumps.
Reported-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Acked-by: Randy Dunlap <randy.dunlap@oracle.com>
This patch fixes UBIFS assertion warnings like:
UBIFS assert failed in ubifs_leb_unmap at 135 (pid 29365)
Pid: 29365, comm: integck Tainted: G I 2.6.37-ubi-2.6+ #34
Call Trace:
[<ffffffffa047c663>] ubifs_lpt_init+0x95e/0x9ee [ubifs]
[<ffffffffa04623a7>] ubifs_remount_fs+0x2c7/0x762 [ubifs]
[<ffffffff810f066e>] do_remount_sb+0xb6/0x101
[<ffffffff81106ff4>] ? do_mount+0x191/0x78e
[<ffffffff811070bb>] do_mount+0x258/0x78e
[<ffffffff810da1e8>] ? alloc_pages_current+0xa2/0xc5
[<ffffffff81107674>] sys_mount+0x83/0xbd
[<ffffffff81009a12>] system_call_fastpath+0x16/0x1b
They happen when we re-mount from R/O mode to R/W mode. While
re-mounting, we write to the media, but we still have the c->ro_mount
flag set. The fix is very simple - just clear the flag before
starting re-mounting R/W.
These warnings are caused by the following commit:
2ef13294d2
For -stable guys: this bug was introduced in 2.6.38, this is materieal
for 2.6.38-stable.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Cc: stable@kernel.org [2.6.38]
Thanks to coverity which spotted that UBIFS will oops if 'kmalloc()'
in 'read_pnode()' fails and we dereference a NULL 'pnode' pointer
when we 'goto out'.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Cc: stable@kernel.org
This fix makes the 'dbg_check_old_index()' function return
immediately if debugging is disabled, instead of executing
incorrect 'goto out' which causes UBIFS to:
1. Allocate memory
2. Read the flash
On every commit. OK, we do not commit that often, but it is
still silly to do unneeded I/O anyway.
Credits to coverity for spotting this silly issue.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Cc: stable@kernel.org
When I moved the orphan adding to btrfs_truncate I missed the fact that during
orphan cleanup we just add the orphan items to the orphan list without going
through btrfs_orphan_add, which results in lots of warnings on mount if you have
any orphan items that need to be truncated. Just remove this warning since it's
ok, this will allow all of the normal space accounting take place. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
I noticed a huge problem with the free space cache that was presenting
as an early ENOSPC. Turns out when writing the free space cache out I
forgot to take into account pinned extents and more importantly
clusters. This would result in us leaking free space everytime we
unmounted the filesystem and remounted it.
I fix this by making sure to check and see if the current block group
has a cluster and writing out any entries that are in the cluster to the
cache, as well as writing any pinned extents we currently have to the
cache since those will be available for us to use the next time the fs
mounts.
This patch also adds a check to the end of load_free_space_cache to make
sure we got the right amount of free space cache, and if not make sure
to clear the cache and re-cache the old fashioned way.
Signed-off-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
root_item->flags and root_item->byte_limit are not initialized when
a subvolume is created. This bug is not revealed until we added
readonly snapshot support - now you mount a btrfs filesystem and you
may find the subvolumes in it are readonly.
To work around this problem, we steal a bit from root_item->inode_item->flags,
and use it to indicate if those fields have been properly initialized.
When we read a tree root from disk, we check if the bit is set, and if
not we'll set the flag and initialize the two fields of the root item.
Reported-by: Andreas Philipp <philipp.andreas@gmail.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Tested-by: Andreas Philipp <philipp.andreas@gmail.com>
cc: stable@kernel.org
Signed-off-by: Chris Mason <chris.mason@oracle.com>
the object id of the space cache inode's key is allocated from the relative
root, just like the regular file. So we can't identify space cache inode by
checking the object id of the inode's key, and we have to clear __GFP_FS flag
at the time we look up the space cache inode.
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Free btrfs_trans_handle when join_transaction() fails
in start_transaction()
Signed-off-by: Yoshinori Sano <yoshinori.sano@gmail.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
btrfs_rename() does not release the subvol_sem if the transaction failed to start.
Signed-off-by: Johann Lombardi <johann@whamcloud.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
When we defrag a file, whose size can be fit into an inline extent,
with compression enabled, the compress type is set to be
fs_info->compress_type, which is 0 if the btrfs filesystem is mounted
without compress option. This leads to oops.
Reported-by: Daniel Blueman <daniel.blueman@gmail.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Some mount options are not displayed by /proc/mounts.
This patch displays the option such as compress_type by /proc/mounts.
Ex.
[before]
$ mount | grep sdc2
/dev/sdc2 on /test12 type btrfs (rw,space_cache,compress=lzo)
$ cat /proc/mounts | grep sdc2
/dev/sdc2 /test12 btrfs rw,relatime,compress 0 0
[after]
$ mount | grep sdc2
/dev/sdc2 on /test12 type btrfs (rw,space_cache,compress=lzo)
$ cat /proc/mounts | grep sdc2
/dev/sdc2 /test12 btrfs rw,relatime,compress=lzo,space_cache 0 0
Signed-off-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
While compiling Btrfs, I got following messages:
CC [M] fs/btrfs/file.o
fs/btrfs/file.c: In function '__btrfs_buffered_write':
fs/btrfs/file.c:909: warning: 'ret' may be used uninitialized in this function
CC [M] fs/btrfs/tree-defrag.o
This patch fixes compiler warning.
Signed-off-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
In ext4_register_li_request, we malloc a ext4_li_request and
inserts it into ext4_li_info->li_request_list. In case of any
error later, we free it in the end. But if we have some error
in ext4_run_lazyinit_thread, the whole li_request_list will be
dropped and freed in it. So we will double free this ext4_li_request.
This patch just sets elr to NULL after it is inserted to the list
so that the latter kfree won't double free it.
Signed-off-by: Tao Ma <boyu.mt@taobao.com>
Reviewed-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@kernel.org
When writing a contiguous set of blocks, two indirect blocks could be
needed depending on how the blocks are aligned, so we need to increase
the number of credits needed by one.
[ Also fixed a another bug which could further underestimate the
number of journal credits needed by 1; the code was using integer
division instead of DIV_ROUND_UP() -- tytso]
Signed-off-by: Yongqiang Yang <xiaoqiangnk@gmail.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Cc: stable@kernel.org
It is not necessary to update [cm]time of quota file on each quota
file write and it wastes journal space and IO throughput with inode
writes. So just remove the updating from ext4_quota_write() and only
update times when quotas are being turned off. Userspace cannot get
anything reliable from quota files while they are used by the kernel
anyway.
Signed-off-by: Jan Kara <jack@suse.cz>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
bdget() should not be called when we hold spinlocks since
it might sleep.
Reviewed-by: Jan Kara <jack@suse.cz>
Signed-off-by: Zhu Yanhai <gaoyang.zyh@taobao.com>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
There's no reason to write quota info in dquot_commit(). The writing is a
relict from the old days when we didn't have dquot_acquire() and
dquot_release() and thus dquot_commit() could have created / removed quota
structures from the file. These days dquot_commit() only updates usage counters
/ limits in quota structure and thus there's no need to write quota info.
This also fixes an issue with journaling filesystem which didn't reserve
enough space in the transaction for write of quota info (it could have been
dirty at the time of dquot_commit() because of a race with other operation
changing it).
CC: stable@kernel.org
Reported-and-tested-by: Lukas Czerner <lczerner@redhat.com>
Signed-off-by: Jan Kara <jack@suse.cz>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ryusuke/nilfs2:
nilfs2: fix whitespace coding style issues
nilfs2: fix oops due to a bad aops initialization
nilfs2: fix data loss in mmap page write for hole blocks
* git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client:
libceph: Create a new key type "ceph".
libceph: Get secret from the kernel keys api when mounting with key=NAME.
ceph: Move secret key parsing earlier.
libceph: fix null dereference when unregistering linger requests
ceph: unlock on error in ceph_osdc_start_request()
ceph: fix possible NULL pointer dereference
ceph: flush msgr_wq during mds_client shutdown
From the result of a function test of mmap, mmap write to shared pages
turned out to be broken for hole blocks. It doesn't write out filled
blocks and the data will be lost after umount. This is due to a bug
that the target file is not queued for log writer when filling hole
blocks.
Also, nilfs_page_mkwrite function exits normal code path even after
successfully filled hole blocks due to a change of block_page_mkwrite
function; just after nilfs was merged into the mainline,
block_page_mkwrite() started to return VM_FAULT_LOCKED instead of zero
by the patch "mm: close page_mkwrite races" (commit:
b827e496c8). The current nilfs_page_mkwrite() is not handling
this value properly.
This corrects nilfs_page_mkwrite() and will resolve the data loss
problem in mmap write.
[This should be applied to every kernel since 2.6.30 but a fix is
needed for 2.6.37 and prior kernels]
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Tested-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Cc: stable <stable@kernel.org> [2.6.38]
This makes the base64 logic be contained in mount option parsing,
and prepares us for replacing the homebew key management with the
kernel key retention service.
Signed-off-by: Tommi Virtanen <tommi.virtanen@dreamhost.com>
Signed-off-by: Sage Weil <sage@newdream.net>
Fix the incorrect use of igrab() inside the i_lock in NFS and Ceph‥
If we are already holding the i_lock, we have a reference to the
inode so we can safely use ihold() to gain an extra reference. This
avoids hangs due to lock recursion on the i_lock now that the
inode_lock is gone and igrab() uses the i_lock itself.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: linux-fsdevel@vger.kernel.org
Cc: Ryan Mallon <ryan@bluewatersys.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'for-linus' of git://oss.sgi.com/xfs/xfs:
xfs: stop using the page cache to back the buffer cache
xfs: register the inode cache shrinker before quotachecks
xfs: xfs_trans_read_buf() should return an error on failure
xfs: introduce inode cluster buffer trylocks for xfs_iflush
vmap: flush vmap aliases when mapping fails
xfs: preallocation transactions do not need to be synchronous
Fix up trivial conflicts in fs/xfs/linux-2.6/xfs_buf.c due to plug removal.
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ecryptfs/ecryptfs-2.6:
eCryptfs: write lock requested keys
eCryptfs: move ecryptfs_find_auth_tok_for_sig() call before mutex_lock
eCryptfs: verify authentication tokens before their use
eCryptfs: modified size of keysig in the ecryptfs_key_sig structure
eCryptfs: removed num_global_auth_toks from ecryptfs_mount_crypt_stat
eCryptfs: ecryptfs_keyring_auth_tok_for_sig() bug fix
eCryptfs: Unlock page in write_begin error path
ecryptfs: modify write path to encrypt page in writepage
eCryptfs: Remove ECRYPTFS_NEW_FILE crypt stat flag
eCryptfs: Remove unnecessary grow_file() function
* 'for-linus-unmerged' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable: (45 commits)
Btrfs: fix __btrfs_map_block on 32 bit machines
btrfs: fix possible deadlock by clearing __GFP_FS flag
btrfs: check link counter overflow in link(2)
btrfs: don't mess with i_nlink of unlocked inode in rename()
Btrfs: check return value of btrfs_alloc_path()
Btrfs: fix OOPS of empty filesystem after balance
Btrfs: fix memory leak of empty filesystem after balance
Btrfs: fix return value of setflags ioctl
Btrfs: fix uncheck memory allocations
btrfs: make inode ref log recovery faster
Btrfs: add btrfs_trim_fs() to handle FITRIM
Btrfs: adjust btrfs_discard_extent() return errors and trimmed bytes
Btrfs: make btrfs_map_block() return entire free extent for each device of RAID0/1/10/DUP
Btrfs: make update_reserved_bytes() public
btrfs: return EXDEV when linking from different subvolumes
Btrfs: Per file/directory controls for COW and compression
Btrfs: add datacow flag in inode flag
btrfs: use GFP_NOFS instead of GFP_KERNEL
Btrfs: check return value of read_tree_block()
btrfs: properly access unaligned checksum buffer
...
Fix up trivial conflicts in fs/btrfs/volumes.c due to plug removal in
the block layer.
When a hole spans across page boundaries, the next write forces
a read of the block. This could end up reading existing garbage
data from the disk in ocfs2_map_page_blocks. This leads to
non-zero holes. In order to avoid this, mark the writes as new
when the holes span across page boundaries.
Signed-off-by: Goldwyn Rodrigues <rgoldwyn@suse.de>
Signed-off-by: jlbec <jlbec@evilplan.org>
When CONFIG_DEBUG_FS=y and CONFIG_OCFS2_FS_STATS=n, we get the
following warning:
fs/ocfs2/cluster/tcp.c:213:16: warning: ‘o2net_get_func_run_time’
defined but not used
Since o2net_get_func_run_time is only called from
o2net_update_recv_stats, so move it under CONFIG_OCFS2_FS_STATS.
Signed-off-by: Rakib Mullick <rakib.mullick@gmail.com>
Signed-off-by: jlbec <jlbec@evilplan.org>
* 'bugfixes' of git://git.linux-nfs.org/projects/trondmy/nfs-2.6:
NFS: Ensure that rpc_release_resources_task() can be called twice.
NFS: Don't leak RPC clients in NFSv4 secinfo negotiation
NFS: Fix a hang in the writeback path
Recent changes for discard support didn't compile,
this fixes them not to try and % 64 bit numbers.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Using the GFP_HIGHUSER_MOVABLE flag to allocate the metadata's page may cause
deadlock.
Task1
open()
...
btrfs_search_slot()
...
btrfs_cow_block()
...
alloc_page()
wait for reclaiming
shrink_slab()
...
shrink_icache_memory()
...
btrfs_evict_inode()
...
btrfs_search_slot()
If the path is locked by task1, the deadlock happens.
So the btree's page cache is different with the file's page cache, it can not
allocate pages by GFP_HIGHUSER_MOVABLE flag, we must clear __GFP_FS flag in
GFP_HIGHUSER_MOVABLE flag.
Reported-by: Itaru Kitayama <kitayama@cl.bb4u.ne.jp>
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
old_inode is not locked; it's not safe to play with its link
count. Instead of bumping it and calling btrfs_unlink_inode(),
add a variant of the latter that does not do btrfs_drop_nlink()/
btrfs_update_inode(), call it instead of btrfs_inc_nlink()/
btrfs_unlink_inode() and do btrfs_update_inode() ourselves.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Adding the check on the return value of btrfs_alloc_path() to several places.
And, some of callers are modified by this change.
Signed-off-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
btrfs will remove unused block groups after balance.
When a empty filesystem is balanced, the block group with tag "DATA" may be
dropped, and after umount and mount again, it will not find "DATA" space_info
and lead to OOPS.
So we initial the necessary space_infos(DATA, SYSTEM, METADATA) to avoid OOPS.
Reported-by: Daniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
After Josef's patch(commit 3c14874acc),
btrfs will exclude super bytes when reading block groups(by marking a extent
state UPTODATE). However, these bytes do not get freed while balance remove
unused block groups, and we won't process those removed ones any more, when
we do umount and unload the btrfs module, btrfs hits a memory leak.
This patch add the missing free operation.
Reproduce steps:
$ mkfs.btrfs disk
$ mount disk /mnt/btrfs -o loop
$ btrfs filesystem balance /mnt/btrfs
$ umount /mnt/btrfs
$ rmmod btrfs
Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
setflags ioctl should return error when any checks fail.
Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
Reviewed-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
To make Btrfs code more robust, several return value checks where memory
allocation can fail are introduced. I use BUG_ON where I don't know how
to handle the error properly, which increases the number of using the
notorious BUG_ON, though.
Signed-off-by: Yoshinori Sano <yoshinori.sano@gmail.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
When we recover from crash via write-ahead log tree and process
the inode refs, for each btrfs_inode_ref item, we will
1) check if we already have a perfect match in fs/file tree, if
we have, then we're done.
2) search the corresponding back reference in fs/file tree, and
check all the names in this back reference to see if they are
also in the log to avoid conflict corners.
3) recover the logged inode refs to fs/file tree.
In current btrfs, however,
- for 2)'s check, once is enough, since the checked back reference
will remain unchanged after processing all the inode refs belonged
to the key.
- it has no need to do another 1) between 2) and 3).
I've made a small test to show how it improves,
$dd if=/dev/zero of=foobar bs=4K count=1
$sync
$make 100 hard links continuously, like ln foobar link_i
$fsync foobar
$echo b > /proc/sysrq-trigger
after reboot
$time mount DEV PATH
without patch:
real 0m0.285s
user 0m0.001s
sys 0m0.009s
with patch:
real 0m0.123s
user 0m0.000s
sys 0m0.010s
Changelog v1->v2:
- fix double free - pointed by David Sterba
Changelog v2->v3:
- adjust free order
Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
We take an free extent out from allocator, trim it, then put it back,
but before we trim the block group, we should make sure the block group is
cached, so plus a little change to make cache_block_group() run without a
transaction.
Signed-off-by: Li Dongyang <lidongyang@novell.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Callers of btrfs_discard_extent() should check if we are mounted with -o discard,
as we want to make fitrim to work even the fs is not mounted with -o discard.
Also we should use REQ_DISCARD to map the free extent to get a full mapping,
last we only return errors if
1. the error is not a EOPNOTSUPP
2. no device supports discard
Signed-off-by: Li Dongyang <lidongyang@novell.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
btrfs_map_block() will only return a single stripe length, but we want the
full extent be mapped to each disk when we are trimming the extent,
so we add length to btrfs_bio_stripe and fill it if we are mapping for REQ_DISCARD.
Signed-off-by: Li Dongyang <lidongyang@novell.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Make the function public as we should update the reserved extents calculations
after taking out an extent for trimming.
Signed-off-by: Li Dongyang <lidongyang@novell.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
btrfs_link returns EPERM if a cross-subvolume link is attempted.
However, in this case I believe EXDEV to be the more appropriate value.
>From the link(2) man page:
EXDEV oldpath and newpath are not on the same mounted file system. (Linux
permits a file system to be mounted at multiple points, but link()
does not work across different mount points, even if the same file
system is mounted on both.)
This matters because an application may have different behaviors based on
return codes.
Signed-off-by: Mark Fasheh <mfasheh@suse.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Data compression and data cow are controlled across the entire FS by mount
options right now. ioctls are needed to set this on a per file or per
directory basis. This has been proposed previously, but VFS developers
wanted us to use generic ioctls rather than btrfs-specific ones.
According to Chris's comment, there should be just one true compression
method(probably LZO) stored in the super. However, before this, we would
wait for that one method is stable enough to be adopted into the super.
So I list it as a long term goal, and just store it in ram today.
After applying this patch, we can use the generic "FS_IOC_SETFLAGS" ioctl to
control file and directory's datacow and compression attribute.
NOTE:
- The compression type is selected by such rules:
If we mount btrfs with compress options, ie, zlib/lzo, the type is it.
Otherwise, we'll use the default compress type (zlib today).
v1->v2:
- rebase to the latest btrfs.
v2->v3:
- fix a problem, i.e. when a file is set NOCOW via mount option, then this NOCOW
will be screwed by inheritance from parent directory.
Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
In the filesystem context, we must allocate memory by GFP_NOFS,
or we may start another filesystem operation and make kswap thread hang up.
Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This patch is checking return value of read_tree_block(),
and if it is NULL, error processing.
Signed-off-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
On Fri, Mar 18, 2011 at 11:56:53AM -0400, Chris Mason wrote:
> Thanks for fielding this one. Does put_unaligned_le32 optimize away on
> platforms with efficient access? It would be great if we didn't need
> the #ifdef.
(quicktest: assembly output is same for put_unaligned_le32 and direct
assignment on my x86_64)
I was originally following examples in
Documentation/unaligned-memory-access.txt. From other code it seems to me that
the define CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS is intended for larger
portions of code. Macros/wrappers for {put,get}_unaligned* are chosen via
arch/<arch>/include/asm/unaligned.h accordingly, therefore it's safe to use
put_unaligned_le32 without the ifdef.
dave
Signed-off-by: Chris Mason <chris.mason@oracle.com>
This patch changes some BUG_ON() to the error return.
(but, most callers still use BUG_ON())
Signed-off-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
Tracepoints can provide insight into why btrfs hits bugs and be greatly
helpful for debugging, e.g
dd-7822 [000] 2121.641088: btrfs_inode_request: root = 5(FS_TREE), gen = 4, ino = 256, blocks = 8, disk_i_size = 0, last_trans = 8, logged_trans = 0
dd-7822 [000] 2121.641100: btrfs_inode_new: root = 5(FS_TREE), gen = 8, ino = 257, blocks = 0, disk_i_size = 0, last_trans = 0, logged_trans = 0
btrfs-transacti-7804 [001] 2146.935420: btrfs_cow_block: root = 2(EXTENT_TREE), refs = 2, orig_buf = 29368320 (orig_level = 0), cow_buf = 29388800 (cow_level = 0)
btrfs-transacti-7804 [001] 2146.935473: btrfs_cow_block: root = 1(ROOT_TREE), refs = 2, orig_buf = 29364224 (orig_level = 0), cow_buf = 29392896 (cow_level = 0)
btrfs-transacti-7804 [001] 2146.972221: btrfs_transaction_commit: root = 1(ROOT_TREE), gen = 8
flush-btrfs-2-7821 [001] 2155.824210: btrfs_chunk_alloc: root = 3(CHUNK_TREE), offset = 1103101952, size = 1073741824, num_stripes = 1, sub_stripes = 0, type = DATA
flush-btrfs-2-7821 [001] 2155.824241: btrfs_cow_block: root = 2(EXTENT_TREE), refs = 2, orig_buf = 29388800 (orig_level = 0), cow_buf = 29396992 (cow_level = 0)
flush-btrfs-2-7821 [001] 2155.824255: btrfs_cow_block: root = 4(DEV_TREE), refs = 2, orig_buf = 29372416 (orig_level = 0), cow_buf = 29401088 (cow_level = 0)
flush-btrfs-2-7821 [000] 2155.824329: btrfs_cow_block: root = 3(CHUNK_TREE), refs = 2, orig_buf = 20971520 (orig_level = 0), cow_buf = 20975616 (cow_level = 0)
btrfs-endio-wri-7800 [001] 2155.898019: btrfs_cow_block: root = 5(FS_TREE), refs = 2, orig_buf = 29384704 (orig_level = 0), cow_buf = 29405184 (cow_level = 0)
btrfs-endio-wri-7800 [001] 2155.898043: btrfs_cow_block: root = 7(CSUM_TREE), refs = 2, orig_buf = 29376512 (orig_level = 0), cow_buf = 29409280 (cow_level = 0)
Here is what I have added:
1) ordere_extent:
btrfs_ordered_extent_add
btrfs_ordered_extent_remove
btrfs_ordered_extent_start
btrfs_ordered_extent_put
These provide critical information to understand how ordered_extents are
updated.
2) extent_map:
btrfs_get_extent
extent_map is used in both read and write cases, and it is useful for tracking
how btrfs specific IO is running.
3) writepage:
__extent_writepage
btrfs_writepage_end_io_hook
Pages are cirtical resourses and produce a lot of corner cases during writeback,
so it is valuable to know how page is written to disk.
4) inode:
btrfs_inode_new
btrfs_inode_request
btrfs_inode_evict
These can show where and when a inode is created, when a inode is evicted.
5) sync:
btrfs_sync_file
btrfs_sync_fs
These show sync arguments.
6) transaction:
btrfs_transaction_commit
In transaction based filesystem, it will be useful to know the generation and
who does commit.
7) back reference and cow:
btrfs_delayed_tree_ref
btrfs_delayed_data_ref
btrfs_delayed_ref_head
btrfs_cow_block
Btrfs natively supports back references, these tracepoints are helpful on
understanding btrfs's COW mechanism.
8) chunk:
btrfs_chunk_alloc
btrfs_chunk_free
Chunk is a link between physical offset and logical offset, and stands for space
infomation in btrfs, and these are helpful on tracing space things.
9) reserved_extent:
btrfs_reserved_extent_alloc
btrfs_reserved_extent_free
These can show how btrfs uses its space.
Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>
The pointer to the extent buffer for the root of each tree
is protected by a spinlock so that we can safely read the pointer
and take a reference on the extent buffer.
But now that the extent buffers are freed via RCU, we can safely
use rcu_read_lock instead.
Signed-off-by: Chris Mason <chris.mason@oracle.com>
A requested key is write locked in order to prevent modifications on the
authentication token while it is being used.
Signed-off-by: Roberto Sassu <roberto.sassu@polito.it>
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
The ecryptfs_find_auth_tok_for_sig() call is moved before the
mutex_lock(s->tfm_mutex) instruction in order to avoid possible deadlocks
that may occur by holding the lock on the two semaphores 'key->sem' and
's->tfm_mutex' in reverse order.
Signed-off-by: Roberto Sassu <roberto.sassu@polito.it>
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
Authentication tokens content may change if another requestor calls the
update() method of the corresponding key. The new function
ecryptfs_verify_auth_tok_from_key() retrieves the authentication token from
the provided key and verifies if it is still valid before being used to
encrypt or decrypt an eCryptfs file.
Signed-off-by: Roberto Sassu <roberto.sassu@polito.it>
[tyhicks: Minor formatting changes]
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
The size of the 'keysig' array is incremented of one byte in order to make
room for the NULL character. The 'keysig' variable is used, in the function
ecryptfs_generate_key_packet_set(), to find an authentication token with
the given signature and is printed a debug message if it cannot be
retrieved.
Signed-off-by: Roberto Sassu <roberto.sassu@polito.it>
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
This patch removes the 'num_global_auth_toks' field of the
ecryptfs_mount_crypt_stat structure, used to count the number of items in
the 'global_auth_tok_list' list. This variable is not needed because there
are no checks based upon it.
Signed-off-by: Roberto Sassu <roberto.sassu@polito.it>
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
The pointer '(*auth_tok_key)' is set to NULL in case request_key()
fails, in order to prevent its use by functions calling
ecryptfs_keyring_auth_tok_for_sig().
Signed-off-by: Roberto Sassu <roberto.sassu@polito.it>
Cc: <stable@kernel.org>
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
Unlock the page in error path of ecryptfs_write_begin(). This may
happen, for example, if decryption fails while bring the page
up-to-date.
Cc: <stable@kernel.org>
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
Change the write path to encrypt the data only when the page is written to
disk in ecryptfs_writepage. Previously, ecryptfs encrypts the page in
ecryptfs_write_end which means that if there are multiple write requests to
the same page, ecryptfs ends up re-encrypting that page over and over again.
This patch minimizes the number of encryptions needed.
Signed-off-by: Thieu Le <thieule@chromium.org>
[tyhicks: Changed NULL .drop_inode sop pointer to generic_drop_inode]
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
Now that grow_file() is not called in the ecryptfs_create() path, the
ECRYPTFS_NEW_FILE flag is no longer needed. It helped
ecryptfs_readpage() know not to decrypt zeroes that were read from the
lower file in the grow_file() path.
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
When creating a new eCryptfs file, the crypto metadata is written out
and then the lower file was being "grown" with 4 kB of encrypted zeroes.
I suspect that growing the encrypted file was to prevent an information
leak that the unencrypted file was empty. However, the unencrypted file
size is stored, in plaintext, in the metadata so growing the file is
unnecessary.
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
Fix inode.c kernel-doc fatal error: 2 comment sections have the same name:
Error(fs/inode.c:1171): duplicate section name 'Note'
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When m_start returns an error, the seq_file logic will still call m_stop
with that error entry, so we'd better make sure that we check it before
using it as a vma.
Introduced by commit ec6fd8a435 ("report errors in /proc/*/*map*
sanely"), which replaced NULL with various ERR_PTR() cases.
(On ia64, you happen to get a unaligned fault instead of a page fault,
since the address used is generally some random error code like -EPERM)
Reported-by: Anca Emanuel <anca.emanuel@gmail.com>
Reported-by: Tony Luck <tony.luck@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Américo Wang <xiyou.wangcong@gmail.com>
Cc: Stephen Wilson <wilsons@start.ca>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Now that the inode scalability patches have been merged, it is no longer
safe to call igrab() under the inode->i_lock.
Now that we no longer call nfs_clear_request() until the nfs_page is
being freed, we know that we are always holding a reference to the
nfs_open_context, which again holds a reference to the path, and so
the inode cannot be freed until the last nfs_page has been removed
from the radix tree and freed.
We can therefore skip the igrab()/iput() altogether.
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
I noticed that dio_end_io calls the appropriate endio function with an error,
but the endio functions don't actually do anything with that error, they assume
that if there was an error then the bio will not be uptodate. So if we had
checksum failures we would never pass back EIO. So if there is an error in our
endio functions make sure to clear the uptodate flag on the bio. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
When doing direct writes we store the checksums in the ordered sum stuff in the
ordered extent for writing them when the write completes, so we don't even use
the dip->csums array. So if we're writing, don't bother allocating dip->csums
since we won't use it anyway. Thanks,
Signed-off-by: Josef Bacik <josef@redhat.com>
This patch makes the free space cluster refilling code a little easier to
understand, and fixes some things with the bitmap part of it. Currently we
either want to refill a cluster with
1) All normal extent entries (those without bitmaps)
2) A bitmap entry with enough space
The current code has this ugly jump around logic that will first try and fill up
the cluster with extent entries and then if it can't do that it will try and
find a bitmap to use. So instead split this out into two functions, one that
tries to find only normal entries, and one that tries to find bitmaps.
This also fixes a suboptimal thing we would do with bitmaps. If we used a
bitmap we would just tell the cluster that we were pointing at a bitmap and it
would do the tree search in the block group for that entry every time we tried
to make an allocation. Instead of doing that now we just add it to the clusters
group.
I tested this with my ENOSPC tests and xfstests and it survived.
Signed-off-by: Josef Bacik <josef@redhat.com>
Now that the buffer cache has it's own LRU, we do not need to use
the page cache to provide persistent caching and reclaim
infrastructure. Convert the buffer cache to use alloc_pages()
instead of the page cache. This will remove all the overhead of page
cache management from setup and teardown of the buffers, as well as
needing to mark pages accessed as we find buffers in the buffer
cache.
By avoiding the page cache, we also remove the need to keep state in
the page_private(page) field for persistant storage across buffer
free/buffer rebuild and so all that code can be removed. This also
fixes the long-standing problem of not having enough bits in the
page_private field to track all the state needed for a 512
sector/64k page setup.
It also removes the need for page locking during reads as the pages
are unique to the buffer and nobody else will be attempting to
access them.
Finally, it removes the buftarg address space lock as a point of
global contention on workloads that allocate and free buffers
quickly such as when creating or removing large numbers of inodes in
parallel. This remove the 16TB limit on filesystem size on 32 bit
machines as the page index (32 bit) is no longer used for lookups
of metadata buffers - the buffer cache is now solely indexed by disk
address which is stored in a 64 bit field in the buffer.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Alex Elder <aelder@sgi.com>
During mount, we can do a quotacheck that involves a bulkstat pass
on all inodes. If there are more inodes in the filesystem than can
be held in memory, we require the inode cache shrinker to run to
ensure that we don't run out of memory.
Unfortunately, the inode cache shrinker is not registered until we
get to the end of the superblock setup process, which is after a
quotacheck is run if it is needed. Hence we need to register the
inode cache shrinker earlier in the mount process so that we don't
OOM during mount. This requires that we also initialise the syncd
work before we register the shrinker, so we nee dto juggle that
around as well.
While there, make sure that we have set up the block sizes in the
VFS superblock correctly before the quotacheck is run so that any
inodes that are cached as a result of the quotacheck have their
block size fields set up correctly.
Cc: stable@kernel.org
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Alex Elder <aelder@sgi.com>
When inside a transaction and we fail to read a buffer,
xfs_trans_read_buf returns a null buffer pointer and no error.
xfs_do_da_buf() checks the error return, but not the buffer, and as
a result this read failure condition causes a panic when it attempts
to dereference the non-existant buffer.
Make xfs_trans_read_buf() return the same error for this situation
regardless of whether it is in a transaction or not. This means
every caller does not need to check both the error return and the
buffer before proceeding to use the buffer.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Alex Elder <aelder@sgi.com>
There is an ABBA deadlock between synchronous inode flushing in
xfs_reclaim_inode and xfs_icluster_free. xfs_icluster_free locks the
buffer, then takes inode ilocks, whilst synchronous reclaim takes
the ilock followed by the buffer lock in xfs_iflush().
To avoid this deadlock, separate the inode cluster buffer locking
semantics from the synchronous inode flush semantics, allowing
callers to attempt to lock the buffer but still issue synchronous IO
if it can get the buffer. This requires xfs_iflush() calls that
currently use non-blocking semantics to pass SYNC_TRYLOCK rather
than 0 as the flags parameter.
This allows xfs_reclaim_inode to avoid the deadlock on the buffer
lock and detect the failure so that it can drop the inode ilock and
restart the reclaim attempt on the inode. This allows
xfs_ifree_cluster to obtain the inode lock, mark the inode stale and
release it and hence defuse the deadlock situation. It also has the
pleasant side effect of avoiding IO in xfs_reclaim_inode when it
tries to next reclaim the inode as it is now marked stale.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
On 32 bit systems, vmalloc space is limited and XFS can chew through
it quickly as the vmalloc space is lazily freed. This can result in
failure to map buffers, even when there is apparently large amounts
of vmalloc space available. Hence, if we fail to map a buffer, purge
the aliases that have not yet been freed to hopefuly free up enough
vmalloc space to allow a retry to succeed.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
Preallocation and hole punch transactions are currently synchronous
and this is causing performance problems in some cases. The
transactions don't need to be synchronous as we don't need to
guarantee the preallocation is persistent on disk until a
fdatasync, fsync, sync operation occurs. If the file is opened
O_SYNC or O_DATASYNC, only then should the transaction be issued
synchronously.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Alex Elder <aelder@sgi.com>
The release method for mds connections uses a backpointer to the
mds_client, so we need to flush the workqueue of any pending work (and
ceph_connection references) prior to freeing the mds_client. This fixes
an oops easily triggered under UML by
while true ; do mount ... ; umount ... ; done
Also fix an outdated comment: the flush in ceph_destroy_client only flushes
OSD connections out. This bug is basically an artifact of the ceph ->
ceph+libceph conversion.
Signed-off-by: Sage Weil <sage@newdream.net>
* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4: (43 commits)
ext4: fix a BUG in mb_mark_used during trim.
ext4: unused variables cleanup in fs/ext4/extents.c
ext4: remove redundant set_buffer_mapped() in ext4_da_get_block_prep()
ext4: add more tracepoints and use dev_t in the trace buffer
ext4: don't kfree uninitialized s_group_info members
ext4: add missing space in printk's in __ext4_grp_locked_error()
ext4: add FITRIM to compat_ioctl.
ext4: handle errors in ext4_clear_blocks()
ext4: unify the ext4_handle_release_buffer() api
ext4: handle errors in ext4_rename
jbd2: add COW fields to struct jbd2_journal_handle
jbd2: add the b_cow_tid field to journal_head struct
ext4: Initialize fsync transaction ids in ext4_new_inode()
ext4: Use single thread to perform DIO unwritten convertion
ext4: optimize ext4_bio_write_page() when no extent conversion is needed
ext4: skip orphan cleanup if fs has unknown ROCOMPAT features
ext4: use the nblocks arg to ext4_truncate_restart_trans()
ext4: fix missing iput of root inode for some mount error paths
ext4: make FIEMAP and delayed allocation play well together
ext4: suppress verbose debugging information if malloc-debug is off
...
Fi up conflicts in fs/ext4/super.c due to workqueue changes
* 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6: (9356 commits)
[media] rc: update for bitop name changes
fs: simplify iget & friends
fs: pull inode->i_lock up out of writeback_single_inode
fs: rename inode_lock to inode_hash_lock
fs: move i_wb_list out from under inode_lock
fs: move i_sb_list out from under inode_lock
fs: remove inode_lock from iput_final and prune_icache
fs: Lock the inode LRU list separately
fs: factor inode disposal
fs: protect inode->i_state with inode->i_lock
lib, arch: add filter argument to show_mem and fix private implementations
SLUB: Write to per cpu data when allocating it
slub: Fix debugobjects with lockless fastpath
autofs4: Do not potentially dereference NULL pointer returned by fget() in autofs_dev_ioctl_setpipefd()
autofs4 - remove autofs4_lock
autofs4 - fix d_manage() return on rcu-walk
autofs4 - fix autofs4_expire_indirect() traversal
autofs4 - fix dentry leak in autofs4_expire_direct()
autofs4 - reinstate last used update on access
vfs - check non-mountpoint dentry might block in __follow_mount_rcu()
...
NOTE!
This merge commit was created to fix compilation error. The block
tree was merged upstream and removed the 'elv_queue_empty()'
function which the new 'mtdswap' driver is using. So a simple
merge of the mtd tree with upstream does not compile. And the
mtd tree has already be published, so re-basing it is not an option.
To fix this unfortunate situation, I had to merge upstream into the
mtd-2.6.git tree without committing, put the fixup patch on top of
this, and then commit this. The result is that we do not have commits
which do not compile.
In other words, this merge commit "merges" 3 things: the MTD tree, the
upstream tree, and the fixup patch.
This was noticed by users who performed more than 2^32 lock operations
and hence made this counter overflow (eventually leading to
use-after-free's). Setting rq_client to NULL here means that it won't
later get auth_domain_put() when it should be.
Appears to have been introduced in 2.5.42 by "[PATCH] kNFSd: Move auth
domain lookup into svcauth" which moved most of the rq_client handling
to common svcauth code, but left behind this one line.
Cc: Neil Brown <neilb@suse.de>
Cc: stable@kernel.org
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:
fs: simplify iget & friends
fs: pull inode->i_lock up out of writeback_single_inode
fs: rename inode_lock to inode_hash_lock
fs: move i_wb_list out from under inode_lock
fs: move i_sb_list out from under inode_lock
fs: remove inode_lock from iput_final and prune_icache
fs: Lock the inode LRU list separately
fs: factor inode disposal
fs: protect inode->i_state with inode->i_lock
autofs4: Do not potentially dereference NULL pointer returned by fget() in autofs_dev_ioctl_setpipefd()
autofs4 - remove autofs4_lock
autofs4 - fix d_manage() return on rcu-walk
autofs4 - fix autofs4_expire_indirect() traversal
autofs4 - fix dentry leak in autofs4_expire_direct()
autofs4 - reinstate last used update on access
vfs - check non-mountpoint dentry might block in __follow_mount_rcu()
Merge get_new_inode/get_new_inode_fast into iget5_locked/iget_locked
as those were the only callers. Remove the internal ifind/ifind_fast
helpers - ifind_fast only had a single caller, and ifind had two
callers wanting it to do different things. Also clean up the comments
in this area to focus on information important to a developer trying
to use it, instead of overloading them with implementation details.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
First thing we do in writeback_single_inode() is take the i_lock and
the last thing we do is drop it. A caller already holds the i_lock,
so pull the i_lock out of writeback_single_inode() to reduce the
round trips on this lock during inode writeback.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
All that remains of the inode_lock is protecting the inode hash list
manipulation and traversals. Rename the inode_lock to
inode_hash_lock to reflect it's actual function.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Protect the inode writeback list with a new global lock
inode_wb_list_lock and use it to protect the list manipulations and
traversals. This lock replaces the inode_lock as the inodes on the
list can be validity checked while holding the inode->i_lock and
hence the inode_lock is no longer needed to protect the list.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Protect the per-sb inode list with a new global lock
inode_sb_list_lock and use it to protect the list manipulations and
traversals. This lock replaces the inode_lock as the inodes on the
list can be validity checked while holding the inode->i_lock and
hence the inode_lock is no longer needed to protect the list.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Now that inode state changes are protected by the inode->i_lock and
the inode LRU manipulations by the inode_lru_lock, we can remove the
inode_lock from prune_icache and the initial part of iput_final().
instead of using the inode_lock to protect the inode during
iput_final, use the inode->i_lock instead. This protects the inode
against new references being taken while we change the inode state
to I_FREEING, as well as preventing prune_icache from grabbing the
inode while we are manipulating it. Hence we no longer need the
inode_lock in iput_final prior to setting I_FREEING on the inode.
For prune_icache, we no longer need the inode_lock to protect the
LRU list, and the inodes themselves are protected against freeing
races by the inode->i_lock. Hence we can lift the inode_lock from
prune_icache as well.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Introduce the inode_lru_lock to protect the inode_lru list. This
lock is nested inside the inode->i_lock to allow the inode to be
added to the LRU list in iput_final without needing to deal with
lock inversions. This keeps iput_final() clean and neat.
Further, where marking the inode I_FREEING and removing it from the
LRU, move the LRU list manipulation within the inode->i_lock to keep
the list manipulation consistent with iput_final. This also means
that most of the open coded LRU list removal + unused inode
accounting can now use the inode_lru_list_del() wrappers which
cleans the code up further.
However, this locking change means what the LRU traversal in
prune_icache() inverts this lock ordering and needs to use trylock
semantics on the inode->i_lock to avoid deadlocking. In these cases,
if we fail to lock the inode we move it to the back of the LRU to
prevent spinning on it.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
We have a couple of places that dispose of inodes. factor the
disposal into evict() to isolate this code and make it simpler to
peel away the inode_lock from the code.
While doing this, change the logic flow in iput_final() to separate
the different cases that need to be handled to make the transitions
the inode goes through more obvious.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Protect inode state transitions and validity checks with the
inode->i_lock. This enables us to make inode state transitions
independently of the inode_lock and is the first step to peeling
away the inode_lock from the code.
This requires that __iget() is done atomically with i_state checks
during list traversals so that we don't race with another thread
marking the inode I_FREEING between the state check and grabbing the
reference.
Also remove the unlock_new_inode() memory barrier optimisation
required to avoid taking the inode_lock when clearing I_NEW.
Simplify the code by simply taking the inode->i_lock around the
state change and wakeup. Because the wakeup is no longer tricky,
remove the wake_up_inode() function and open code the wakeup where
necessary.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
changes LAYOUTGET and GETDEVICEINFO XDR parsing to:
- not use vmap, which doesn't work on incoherent archs
- use xdr_stream parsing for all xdr
Signed-off-by: Weston Andros Adamson <dros@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
Test NFS_INO_LAYOUTCOMMIT before kzalloc
Mark inode dirty to retry LAYOUTCOMMIT on kzalloc failure.
Add comments.
Signed-off-by: Andy Adamson <andros@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
In fs/autofs4/dev-ioctl.c::autofs_dev_ioctl_setpipefd() we call fget(),
which may return NULL, but we do not explicitly test for that NULL return
so we may end up dereferencing a NULL pointer - bad.
When I originally submitted this patch I had chosen EBUSY as the return
value to use if this happens. Ian Kent was kind enough to explain why that
would most likely be wrong and why EBADF should most likely be used
instead. This version of the patch uses EBADF.
Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The autofs4_lock introduced by the rcu-walk changes has unnecessarily
broad scope. The locking is better handled by the per-autofs super
block lookup_lock.
Signed-off-by: Ian Kent <raven@themaw.net>
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The daemon never needs to block and, in the rcu-walk case an error
return isn't used, so always return zero.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The vfs-scale changes changed the traversal used in
autofs4_expire_indirect() from a list to a depth first tree traversal
which isn't right.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
There is a missing dput() when returning from autofs4_expire_direct()
when we see that the dentry is already a pending mount.
Signed-off-by: Ian Kent <raven@themaw.net>
Acked-by: David Howells <dhowells@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
When direct (and offset) mounts were introduced the the last used
timeout could no longer be updated in ->d_revalidate(). This is
because covered direct mounts would be followed over without calling
the autofs file system. As a result the definition of the busyness
check for all entries was changed to be "actually busy" being an open
file or working directory within the automount. But now we have a call
back in the follow so the last used update on any access can be
re-instated. This requires DCACHE_MANAGE_TRANSIT to always be set.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
When following a mount in rcu-walk mode we must check if the incoming dentry
is telling us it may need to block, even if it isn't actually a mountpoint.
Signed-off-by: Ian Kent <raven@themaw.net>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
When sec=<something> is not presented as a mount option,
we should attempt to determine what security flavor the
server is using.
Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
A submount may use different security than the parent
mount does. We should figure out what sec flavor the
submount uses at mount time.
Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
A later patch will need to perform a lookup using an
alternate client with a different security flavor.
This patch adds support for doing that on NFS v4.
Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
This patch changes nfs4_call_sync() from a macro into a
static inline function. As a macro, the call_sync()
function will not do any type checking and depends
on the sequence arguments always having the same name.
As a function, we get to have type checking and can
rename the arguments if we so choose.
Signed-off-by: Bryan Schumaker <bjschuma@netapp.com>
Signed-off-by: Trond Myklebust <Trond.Myklebust@netapp.com>
* 'for-2.6.39/core' of git://git.kernel.dk/linux-2.6-block: (65 commits)
Documentation/iostats.txt: bit-size reference etc.
cfq-iosched: removing unnecessary think time checking
cfq-iosched: Don't clear queue stats when preempt.
blk-throttle: Reset group slice when limits are changed
blk-cgroup: Only give unaccounted_time under debug
cfq-iosched: Don't set active queue in preempt
block: fix non-atomic access to genhd inflight structures
block: attempt to merge with existing requests on plug flush
block: NULL dereference on error path in __blkdev_get()
cfq-iosched: Don't update group weights when on service tree
fs: assign sb->s_bdi to default_backing_dev_info if the bdi is going away
block: Require subsystems to explicitly allocate bio_set integrity mempool
jbd2: finish conversion from WRITE_SYNC_PLUG to WRITE_SYNC and explicit plugging
jbd: finish conversion from WRITE_SYNC_PLUG to WRITE_SYNC and explicit plugging
fs: make fsync_buffers_list() plug
mm: make generic_writepages() use plugging
blk-cgroup: Add unaccounted time to timeslice_used.
block: fixup plugging stubs for !CONFIG_BLOCK
block: remove obsolete comments for blkdev_issue_zeroout.
blktrace: Use rq->cmd_flags directly in blk_add_trace_rq.
...
Fix up conflicts in fs/{aio.c,super.c}
* 'for-2.6.39' of git://linux-nfs.org/~bfields/linux:
SUNRPC: Remove resource leak in svc_rdma_send_error()
nfsd: wrong index used in inner loop
nfsd4: fix comment and remove unused nfsd4_file fields
nfs41: make sure nfs server return right ca_maxresponsesize_cached
nfsd: fix compile error
svcrpc: fix bad argument in unix_domain_find
nfsd4: fix struct file leak
nfsd4: minor nfs4state.c reshuffling
svcrpc: fix rare race on unix_domain creation
nfsd41: modify the members value of nfsd4_op_flags
nfsd: add proc file listing kernel's gss_krb5 enctypes
gss:krb5 only include enctype numbers in gm_upcall_enctypes
NFSD, VFS: Remove dead code in nfsd_rename()
nfsd: kill unused macro definition
locks: use assign_type()
* git://git.kernel.org/pub/scm/linux/kernel/git/pkl/squashfs-linus:
Squashfs: Use vmalloc rather than kmalloc for zlib workspace
Squashfs: handle corruption of directory structure
Squashfs: wrap squashfs_mount() definition
Squashfs: xz_wrapper doesn't need to include squashfs_fs_i.h anymore
Squashfs: Update documentation to include compression options
Squashfs: Update Kconfig help text to include xz compression
Squashfs: add compression options support to xz decompressor
Squashfs: extend decompressor framework to handle compression options
* 'for-linus' of git://git.open-osd.org/linux-open-osd:
exofs: deprecate the commands pending counter
exofs: Write sbi->s_nextid as part of the Create command
exofs: Add option to mount by osdname
exofs: Override read-ahead to align on stripe_size
exofs: simple fsync race fix
exofs: Optimize read_4_write
exofs: Trivial: fix some indentation and debug prints
exofs: Remove redundant unlikely()
This patch fixes the following UBIFS assertion warning:
UBIFS assert failed in do_readpage at 115 (pid 199)
[<b00321b8>] (unwind_backtrace+0x0/0xdc) from [<af025118>]
(do_readpage+0x108/0x594 [ubifs])
[<af025118>] (do_readpage+0x108/0x594 [ubifs]) from [<af025764>]
(ubifs_write_end+0x1c0/0x2e8 [ubifs])
[<af025764>] (ubifs_write_end+0x1c0/0x2e8 [ubifs]) from
[<b00a0164>] (generic_file_buffered_write+0x18c/0x270)
[<b00a0164>] (generic_file_buffered_write+0x18c/0x270) from
[<b00a08d4>] (__generic_file_aio_write+0x478/0x4c0)
[<b00a08d4>] (__generic_file_aio_write+0x478/0x4c0) from
[<b00a0984>] (generic_file_aio_write+0x68/0xc8)
[<b00a0984>] (generic_file_aio_write+0x68/0xc8) from
[<af024a78>] (ubifs_aio_write+0x178/0x1d8 [ubifs])
[<af024a78>] (ubifs_aio_write+0x178/0x1d8 [ubifs]) from
[<b00d104c>] (do_sync_write+0xb0/0x100)
[<b00d104c>] (do_sync_write+0xb0/0x100) from [<b00d1abc>]
(vfs_write+0xac/0x154)
[<b00d1abc>] (vfs_write+0xac/0x154) from [<b00d1c10>]
(sys_write+0x3c/0x68)
[<b00d1c10>] (sys_write+0x3c/0x68) from [<b002d9a0>]
(ret_fast_syscall+0x0/0x2c)
The 'PG_checked' flag is used to indicate that the page does not
supposedly exist on the media (e.g., a hole or a page beyond the
inode size), so it requires slightly bigger budget, because we have
to account the indexing size increase. And this flag basically
tells that the budget for this page has to be "new page budget".
The "new page budget" is slightly bigger than the "existing page
budget".
The 'do_readpage()' function has the following assertion which
sometimes is hit: 'ubifs_assert(!PageChecked(page))'. Obviously,
the meaning of this assertion is: "I should not be asked to read
a page which does not exist on the media".
However, in 'ubifs_write_begin()' we have a small "trick". Notice,
that VFS may write pages which were not read yet, so the page data
were not loaded from the media to the page cache yet. If VFS tells
that it is going to change only some part of the page, we obviously
have to load it from the media. However, if VFS tells that it is
going to change whole page, we do not read it from the media for
optimization purposes.
However, since we do not read it, we do not know if it exists on
the media or not (a hole, etc). So we set the 'PG_checked' flag
to this page to force bigger budget, just in case.
So 'ubifs_write_begin()' sets 'PG_checked'. Then we are in
'ubifs_write_end()'. And VFS tells us: "hey, for some reasons I
changed my mind and did not change whole page". Frankly, I do not
know why this happens, but I hit this somehow on an ARM platform.
And this is extremely rare.
So in this case UBIFS does the following:
1. Cancels allocated budget.
2. Loads the page from the media by calling 'do_readpage()'.
3. Asks VFS to repeat the whole write operation from the very
beginning (call '->write_begin() again, etc).
And the assertion warning is hit at the step 2 - remember we have
the 'PG_checked' set for this page, and 'do_readpage()' does not
like this. So this patch fixes the problem by adding step 1.5 and
cleaning the 'PG_checked' before calling 'do_readpage()'.
All in all, this patch does not fix any functionality issue, but it
silences UBIFS false positive warning which may happen in very very
rare cases.
And while on it, this patch also improves a commentary which explains
the reasons of setting the 'PG_checked' flag for the page. The old
commentary was a bit difficult to understand.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Simplify UBIFS configuration menu and kill the option to enable self-check
compile-time. We do not really need this because we can do this run-time
using the module parameters or the corresponding sysfs interfaces. And
there is a value in simplifying the kernel configuration menu which becomes
increasingly large.
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>