Instead of reading in the resource groups when gfs2 is checking
for free space to allocate from, gfs2 can store the necessary infromation
in the resource group's lvb.  Also, instead of searching for unlinked
inodes in every resource group that's checked for free space, gfs2 can
store the number of unlinked but inodes in the lvb, and only check for
unlinked inodes if it will find some.

The first time a resource group is locked, the lvb must initialized.
Since this involves counting the unlinked inodes in the resource group,
this takes a little extra time.  But after that, if the resource group
is locked with GL_SKIP, the buffer head won't be read in unless it's
actually needed.

Enabling the resource groups lvbs is done via the rgrplvb mount option.  If
this option isn't set, the lvbs will still be set and updated, but they won't
be verfied or used by the filesystem.  To safely turn on this option, all of
the nodes mounting the filesystem must be running code with this patch, and
the filesystem must have been completely unmounted since they were updated.
Signed-off-by: NBenjamin Marzinski <bmarzins@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

For the glocks and glstats seq_files, which are exposed via debugfs
we should cache the most recent hash bucket, along with the offset
into that bucket. This allows us to restart from that point, rather
than having to begin at the beginning each time.

This is an idea from Eric Dumazet, however I've slightly extended it
so that if the position from which we are due to start is at any
point beyond the last cached point, we start from the last cached
point, plus whatever is the appropriate offset. I don't really expect
people to be lseeking around these files, but if they did so with only
positive offsets, then we'd still get some of the benefit of using a
cached offset.

With my simple test of around 200k entries in the file, I'm seeing
an approx 10x speed up.

Cc: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

As per Al Viro's suggestion, this increases the buffer size used
for these two files. This provides a speed up of slightly less than
8x (i.e. proportional to the buffer size) for cases when we have
large numbers of glocks.

Cc: Al Viro <viro@ZenIV.linux.org.uk>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

When we read an invalid block from the journal, we should not call
withdraw, but simply print a message and return an error. It is
up to the caller to then handle that error. In the case of mount
that means a failed mount, rather than a withdraw (requiring a
reboot). In the case of recovering another nodes journal then
we return an error via the uevent.
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

This patch adds support for the "top dir" flag. Currently this is unused
but a subsequent patch is planned which will add support for the
Orlov allocation policy when allocating subdirectories in a parent
with this flag set.

In order to ensure backward compatible behaviour, mkfs.gfs2 does
not currently tag the root directory with this flag, it must always be
set manually.
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

This patch moves the ancillary quota data structures into the
block reservations structure. This saves GFS2 some time and
effort in allocating and deallocating the qadata structure.
Signed-off-by: NBob Peterson <rpeterso@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

This patch lengthens the lifespan of the reservations structure for
inodes. Before, they were allocated and deallocated for every write
operation. With this patch, they are allocated when the first write
occurs, and deallocated when the last process closes the file.
It's more efficient to do it this way because it saves GFS2 a lot of
unnecessary allocates and frees. It also gives us more flexibility
for the future: (1) we can now fold the qadata structure back into
the structure and save those alloc/frees, (2) we can use this for
multi-block reservations.
Signed-off-by: NBob Peterson <rpeterso@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

Cyrill Gorcunov reports that I broke the fdinfo files with commit
30a08bf2 ("proc: move fd symlink i_mode calculations into
tid_fd_revalidate()"), and he's quite right.

The tid_fd_revalidate() function is not just used for the <tid>/fd
symlinks, it's also used for the <tid>/fdinfo/<fd> files, and the
permission model for those are different.

So do the dynamic symlink permission handling just for symlinks, making
the fdinfo files once more appear as the proper regular files they are.

Of course, Al Viro argued (probably correctly) that we shouldn't do the
symlink permission games at all, and make the symlinks always just be
the normal 'lrwxrwxrwx'.  That would have avoided this issue too, but
since somebody noticed that the permissions had changed (which was the
reason for that original commit 30a08bf2 in the first place), people
do apparently use this feature.

[ Basically, you can use the symlink permission data as a cheap "fdinfo"
  replacement, since you see whether the file is open for reading and/or
  writing by just looking at st_mode of the symlink.  So the feature
  does make sense, even if the pain it has caused means we probably
  shouldn't have done it to begin with. ]
Reported-and-tested-by: NCyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

The config options in the Kconfig file (with _CODEPAGE_ in the name)
didn't match the config option name in the Makefile (no _CODEPAGE_).

And both of them were of the hard-to-read MACXYZZY variety, which made
them hard to parse for normal humans: MACROMAN easily reads as "macro
man", not as "Mac Roman".

So rename the options to be consistent, and be NLS_MAC_xyzzy.  Rename
the files to be mac-xyzzy.c too, and drop the "nls" part entirely (it's
already in the directory name).
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

These were debug things which snuck through.
Reported-by: NYinghai Lu <yinghai@kernel.org>
Cc: Vladimir Serbinenko <phcoder@gmail.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Reviewed-by: NJeff Layton <jlayton@redhat.com>
Signed-off-by: NPavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: NSteve French <sfrench@us.ibm.com>

Signed-off-by: NPavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: NSteve French <sfrench@us.ibm.com>

Signed-off-by: NPavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: NSteve French <sfrench@us.ibm.com>

when cifs_reconnect sets maxBuf to 0 and we try to calculate a size
of memory we need to store locks.
Signed-off-by: NPavel Shilovsky <pshilovsky@samba.org>
Signed-off-by: NSteve French <sfrench@us.ibm.com>

Everyone either defines it in arch thread_info.h or has TIF_RESTORE_SIGMASK
and picks default set_restore_sigmask() in linux/thread_info.h.  Kill the
ifdefs, slap #error in linux/thread_info.h to catch breakage when new ones
get merged.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

NFSv4 can't do reliable opens in d_revalidate, since it cannot know whether a
mount needs to be followed or not.  It does check d_mountpoint() on the dentry,
which can result in a weird error if the VFS found that the mount does not in
fact need to be followed, e.g.:

  # mount --bind /mnt/nfs /mnt/nfs-clone
  # echo something > /mnt/nfs/tmp/bar
  # echo x > /tmp/file
  # mount --bind /tmp/file /mnt/nfs-clone/tmp/bar
  # cat  /mnt/nfs/tmp/bar
  cat: /mnt/nfs/tmp/bar: Not a directory

Which should, by any sane filesystem, result in "something" being printed.

So instead do the open in f_op->open() and in the unlikely case that the cached
dentry turned out to be invalid, drop the dentry and return EOPENSTALE to let
the VFS retry.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
CC: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

NFS optimizes away d_revalidates for last component of open.  This means that
open itself can find the dentry stale.

This patch allows the filesystem to return EOPENSTALE and the VFS will retry the
lookup on just the last component if possible.

If the lookup was done using RCU mode, including the last component, then this
is not possible since the parent dentry is lost.  In this case fall back to
non-RCU lookup.  Currently this is not used since NFS will always leave RCU
mode.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

If open fails, don't put the file.  This allows it to be reused if open needs to
be retried.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Copy __dentry_open() into nameidata_to_filp().
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Move put_filp() out to __dentry_open(), the only caller now.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Split __dentry_open() into two functions:

  do_dentry_open() - does most of the actual work, doesn't put file on failure
  open_check_o_direct() - after a successful open, checks direct_IO method

This will allow i_op->atomic_open to do just the file initialization and leave
the direct_IO checking to the VFS.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Now the post lookup code can be shared between O_CREAT and plain opens since
they are essentially the same.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

This allows this code to be shared between O_CREAT and plain opens.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

This allows this code to be shared between O_CREAT and plain opens.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Check for ENOTDIR before finishing open.  This allows this code to be shared
between O_CREAT and plain opens.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

This will allow this code to be used in RCU mode.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

This will allow this code to be used in RCU mode.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Use helper variable instead of path->dentry->d_inode before complete_walk().
This will allow this code to be used in RCU mode.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Copy walk_component() into do_lookup().
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Allow returning from do_last() with LOOKUP_RCU still set on the "out:" and
"exit:" labels.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Split do_lookup() into two functions:

  lookup_fast() - does cached lookup without i_mutex
  lookup_slow() - does lookup with i_mutex

Both follow managed dentries.

The new functions are needed by atomic_open.
Signed-off-by: NMiklos Szeredi <mszeredi@suse.cz>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Btrfs had been doing it's own file_update_time so we could catch ENOSPC
properly, so just update our btrfs_update_time to work with the new stuff and
then we'll be fancy later.  Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>

Btrfs has to make sure we have space to allocate new blocks in order to modify
the inode, so updating time can fail.  We've gotten around this by having our
own file_update_time but this is kind of a pain, and Christoph has indicated he
would like to make xfs do something different with atime updates.  So introduce
->update_time, where we will deal with i_version an a/m/c time updates and
indicate which changes need to be made.  The normal version just does what it
has always done, updates the time and marks the inode dirty, and then
filesystems can choose to do something different.

I've gone through all of the users of file_update_time and made them check for
errors with the exception of the fault code since it's complicated and I wasn't
quite sure what to do there, also Jan is going to be pushing the file time
updates into page_mkwrite for those who have it so that should satisfy btrfs and
make it not a big deal to check the file_update_time() return code in the
generic fault path. Thanks,
Signed-off-by: NJosef Bacik <josef@redhat.com>

This patch stops reiserfs using the VFS 'write_super()' method along with the
s_dirt flag, because they are on their way out.

The whole "superblock write-out" VFS infrastructure is served by the
'sync_supers()' kernel thread, which wakes up every 5 (by default) seconds and
writes out all dirty superblock using the '->write_super()' call-back.  But the
problem with this thread is that it wastes power by waking up the system every
5 seconds, even if there are no diry superblocks, or there are no client
file-systems which would need this (e.g., btrfs does not use
'->write_super()'). So we want to kill it completely and thus, we need to make
file-systems to stop using the '->write_super()' VFS service, and then remove
it together with the kernel thread.
Signed-off-by: NArtem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

The 'journal_mark_dirty()' function currently first marks the superblock as
dirty by setting 's_dirt' to 1, then does various sanity checks and returns,
then actuall does all the magic with the journal.

This is not an ideal order, though. It makes more sense to first do all the
checks, then do all the internal stuff, and at the end notify the VFS that the
superblock is now dirty.

This patch moves the 's_dirt = 1' assignment from the very beginning of this
function to the very end.
Signed-off-by: NArtem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

The 'reiserfs_resize()' function marks the superblock as dirty by assigning 1
to 's_dirt' and then calls 'journal_mark_dirty()' which does the same. Thus,
we can remove the assignment from 'reiserfs_resize()'.
Signed-off-by: NArtem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

Turn 'reiserfs_flush_old_commits()' into a void function because the callers
do not cares about what it returns anyway.

We are going to remove the 'sb->s_dirt' field completely and this patch is a
small step towards this direction.
Signed-off-by: NArtem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

We have the reiserfs superblock pointer in the 'sbi' variable in this
function, no need to use the 'REISERFS_SB(s)' macro which is the same.
This is jut a small clean-up.
Signed-off-by: NArtem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

after all, 0 bytes and 0 pages is the same thing...
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

When truncating a file, we unmap pages from userspace first, as that's
usually more efficient than relying, page by page, on the fallback in
truncate_inode_page() - particularly if the file is mapped many times.

Do the same when punching a hole: 3.4 added truncate_pagecache_range()
to do the unmap and trunc, so use it in ext4_ext_punch_hole(), instead
of calling truncate_inode_pages_range() directly.
Signed-off-by: NHugh Dickins <hughd@google.com>
Signed-off-by: N"Theodore Ts'o" <tytso@mit.edu>