recovery.c add a brelse to deal with gfs2_replay_read_block being called
twice on the same block.

add a dput to drop the ref count on the root inode.
This was causing lingering glocks and thus causing
a mount failure to hang.

Fix a endian conversion macro that was was swizzling
16bits when it should have been swizzling 32.
Signed-off-by: NRussell Cattelan <cattelan@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

When recoveries are aborted by other recoveries we can get replies to
status or names requests that we've given up on.  This can cause problems
if we're making another request and receive an old reply.  Add a sequence
number to status/names requests and reject replies that don't match.  A
field already exists for the seq number that's used in other message
types.
Signed-off-by: NDavid Teigland <teigland@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

To aid debugging, it's useful to be able to see what nodeid the dlm is
waiting on for a message reply.
Signed-off-by: NDavid Teigland <teigland@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

When the debug buffer has filled up, break from the loop and return the
correct number of bytes that have been written.
Signed-off-by: NDavid Teigland <teigland@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

When we abort one recovery to do another, break out of the ping_members()
routine more quickly, and wake up the dlm_recoverd thread more quickly
instead of waiting for it to time out.
Signed-off-by: NDavid Teigland <teigland@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

Print the violating name length in the assertion.
Signed-off-by: NDavid Teigland <teigland@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

In some cases we can enter write page without there being buffers
attached to the page. In this case the function to add gfs2_bufdata
to the buffers fails sliently causing further failures down the
stack.

This fix ensures that we always add buffers in writepage if they
didn't already exist (mmap is one way to trigger this).
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

This patch fixes the userland DLM unlock code so that it correctly returns the
address of the userland lock status block in its completion AST.

It fixes bug #201348

Patrick
Signed-Off-By: NPatrick Caulfield <pcaulfie@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

The previous patch didn't change all the gfs_ to gfs2_ so
this is the remainder.
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

When the result of a posix lock request is read it needs to be matched up
with the correct waiting request.  The owner field needs to be used in the
comparison since more than one process may be waiting for locks on the
same file.
Signed-off-by: NDavid Teigland <teigland@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

Use the gfs2_ prefix on the register/unregister functions for the lock
modules.  The gfs_ prefix was left from an old idea on how to share these
with gfs1.
Signed-off-by: NDavid Teigland <teigland@redhat.com>
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

I saw an oops down this path when trying to create a new file on a UDF
filesystem which was internally marked as readonly, but mounted rw:

udf_create
        udf_new_inode
                new_inode
                        alloc_inode
                        	udf_alloc_inode
                udf_new_block
                        returns EIO due to readonlyness
                iput (on error)
                        udf_put_inode
                                udf_discard_prealloc
                                        udf_next_aext
                                                udf_current_aext
                                                        udf_get_fileshortad
                                                                OOPS

the udf_discard_prealloc() path was examining uninitialized fields of the
udf inode.

udf_discard_prealloc() already has this code to short-circuit the discard
path if no extents are preallocated:

        if (UDF_I_ALLOCTYPE(inode) == ICBTAG_FLAG_AD_IN_ICB ||
                inode->i_size == UDF_I_LENEXTENTS(inode))
        {
                return;
        }

so if we initialize UDF_I_LENEXTENTS(inode) = 0 earlier in udf_new_inode,
we won't try to free the (not) preallocated blocks, since this will match
the i_size = 0 set when the inode was initialized.
Signed-off-by: NEric Sandeen <sandeen@sandeen.net>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

reiserfs_write_full_page does zero bytes in the file past eof, but it may
call get_block on those buffers as well.  On machines where the page size
is larger than the blocksize, this can result in mmaped files incorrectly
growing up to a block boundary during writepage.

The fix is to avoid calling get_block for any blocks that are entirely past
eof
Signed-off-by: NChris Mason <mason@suse.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

The correct lock ordering is inode lock -> BKL
Signed-off-by: NChris Mason <mason@suse.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

In bugzilla #6941, Jens Kilian reported:

"The function befs_utf2nls (in fs/befs/linuxvfs.c) writes a 0 byte past the
end of a block of memory allocated via kmalloc(), leading to memory
corruption.  This happens only for filenames which are pure ASCII and a
multiple of 4 bytes in length.  [...]

Without DEBUG_SLAB, this leads to further corruption and hard lockups; I
believe this is the bug which has made kernels later than 2.6.8 unusable
for me.  (This must be due to changes in memory management, the bug has
been in the BeFS driver since the time it was introduced (AFAICT).)

Steps to reproduce:
Create a directory (in BeOS, naturally :-) with files named, e.g.,
"1", "22", "333", "4444", ...  Mount it in Linux and do an "ls" or "find""

This patch implements the suggested fix. Credits to Jens Kilian for
debugging the problem and finding the right fix.
Signed-off-by: NDiego Calleja <diegocg@gmail.com>
Cc: Jens Kilian <jjk@acm.org>
Cc: <stable@kernel.org>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Fixes an i_mutex-inside-i_mutex lockdep nasty.
Signed-off-by: NAlexander Zarochentsev <zam@namesys.com>
Cc: <reiserfs-dev@namesys.com>
Cc: Hans Reiser <reiser@namesys.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

ufs_get_locked_page is called twice in ufs code, one time in ufs_truncate
path(we allocated last block), and another time when fragments are
reallocated.  In ideal world in the second case on allocation/free block
layer we should not know that things like `truncate' exists, but now with
such crutch like ufs_get_locked_page we can (or should?) skip truncated
pages.
Signed-off-by: NEvgeniy Dushistov <dushistov@mail.ru>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

As discussed earlier:
http://lkml.org/lkml/2006/6/28/136
this patch fixes such issue:

`ufs_get_locked_page' takes page from cache
after that `vmtruncate' takes page and deletes it from cache
`ufs_get_locked_page' locks page, and reports about EIO error.

Also because of find_lock_page always return valid page or NULL, we have no
need to check it if page not NULL.
Signed-off-by: NEvgeniy Dushistov <dushistov@mail.ru>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Mmapped files were able to trigger a lock ordering bug. Private
maps do not need to take the glock so early on. Shared maps do
unfortunately, however we can get around that by adding a flag
into the flags for the struct gfs2_file. This only works because
we are taking an exclusive lock at this point, so we know that
nobody else can be racing with us.

Fixes Red Hat bugzilla: #201196
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

We never actually set the b_done field any more; it's always zero.
Signed-off-by: NJ. Bruce Fields <bfields@citi.umich.edu>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>
(cherry picked from af8412d4283ef91356e65e0ed9b025b376aebded commit)

nfs_writedata_free() and nfs_readdata_free() can now become static.
Signed-off-by: NAdrian Bunk <bunk@stusta.de>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>
(cherry picked from 5e1ce40f0c3c8f67591aff17756930d7a18ceb1a commit)

In one of the error paths of nfs_path, it may return with dcache_lock still
held; fix this by adding and using a new error path Elong_unlock which unlocks
dcache_lock.
Signed-off-by: NJosh Triplett <josh@freedesktop.org>
Signed-off-by: NTrond Myklebust <Trond.Myklebust@netapp.com>
(cherry picked from f4b90b43677fb23297c56802c3056fc304f988d9 commit)

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

When an object is created via a symlink into an audited directory, audit misses
the event due to not having collected the inode data for the directory.  Modify
__audit_inode_child() to copy the parent inode data if a parent wasn't found in
audit_names[].
Signed-off-by: NAmy Griffis <amy.griffis@hp.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

When the specified path is an existing file or when it is a symlink, audit
collects the wrong inode number, which causes it to miss the open() event.
Adding a second hook to the open() path fixes this.

Also add audit_copy_inode() to consolidate some code.
Signed-off-by: NAmy Griffis <amy.griffis@hp.com>
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

This was a nasty bug which resulted in corruption of hash tables
in the directory code with larger directories. We forgot to
increment a pointer in the read/write routines internal to the
directory code.
Signed-off-by: NSteven Whitehouse <swhiteho@redhat.com>

Based on a bug report from Russ Ross <russruss@gmail.com>

According to the spec:

"The remove request asks the file server both to remove the file
 represented by fid and to clunk the fid, even if the remove fails."

but the Linux client seems to expect the fid to be valid after a failed
remove attempt.  Specifically, I'm getting this behavior when attempting to
remove a non-empty directory.
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Signed-off-by: NRuss Ross <russross@gmail.com>
Signed-off-by: NEric Van Hensbergen <ericvh@gmail.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

For files other than IFREG, nobh option doesn't make sense.  Modifications
to them are journalled and needs buffer heads to do that.  Without this
patch, we get kernel oops in page_buffers().
Signed-off-by: NBadari Pulavarty <pbadari@us.ibm.com>
Cc: <stable@kernel.org>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Commit 7b2fd697427e73c81d5fa659efd91bd07d303b0e in the historical GIT tree
stopped calling the readdir member of a file_operations struct with the big
kernel lock held, and fixed up all the readdir functions to do their own
locking.  However, that change added calls to unlock_kernel() in
vxfs_readdir, but no call to lock_kernel().  Fix this by adding a call to
lock_kernel().
Signed-off-by: NJosh Triplett <josh@freedesktop.org>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Signed-off-by: NMiklos Szeredi <miklos@szeredi.hu>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

It is entirely possible (though rare) that jiffies half-wraps around, while a
dentry/inode remains in the cache.  This could mean that the dentry/inode is
not invalidated for another half wraparound-time.

To get around this problem, use 64-bit jiffies.  The only problem with this is
that dentry->d_time is 32 bits on 32-bit archs.  So use d_fsdata as the high
32 bits.  This is an ugly hack, but far simpler, than having to allocate
private data just for this purpose.

Since 64-bit jiffies can be assumed never to wrap around, simple comparison
can be used, and a zero time value can represent "invalid".
Signed-off-by: NMiklos Szeredi <miklos@szeredi.hu>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

An attribute and entry timeout of zero should mean, that the entity is
invalidated immediately after the operation.  Previously invalidation only
happened at the next clock tick.

Reported and tested by Craig Davies.
Signed-off-by: NMiklos Szeredi <miklos@szeredi.hu>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

ufs_symlink, in one of its error paths, calls unlock_kernel without ever
having called lock_kernel(); fix this by creating and jumping to a new
label out_notlocked rather than the out label used after calling
lock_kernel().
Signed-off-by: NJosh Triplett <josh@freedesktop.org>
Cc: Evgeniy Dushistov <dushistov@mail.ru>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

If efs_symlink_readpage hits the -ENAMETOOLONG error path, it will call
unlock_kernel without ever having called lock_kernel(); fix this by
creating and jumping to a new label fail_notlocked rather than the fail
label used after calling lock_kernel().
Signed-off-by: NJosh Triplett <josh@freedesktop.org>
Cc: Marcelo Tosatti <marcelo.tosatti@cyclades.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Commit 398c53a757702e1e3a7a2c24860c7ad26acb53ed (in the historical GIT
tree) moved the lock_kernel() in coda_open after the allocation of a
coda_file_info struct, but left an unlock_kernel() in the allocation
failure error path; remove it.
Signed-off-by: NJosh Triplett <josh@freedesktop.org>
Acked-by: NJan Harkes <jaharkes@cs.cmu.edu>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

This is a real deadlock, a nice complex one:
(warning: long explanation follows so that Andrew can have a complete
patch description)

it's an ABCDA deadlock:

A iprune_mutex
B inode->inotify_mutex
C ih->mutex
D dev->ev_mutex

The AB relationship comes straight from invalidate_inodes()

int invalidate_inodes(struct super_block * sb)
{
        int busy;
        LIST_HEAD(throw_away);

        mutex_lock(&iprune_mutex);
        spin_lock(&inode_lock);
        inotify_unmount_inodes(&sb->s_inodes);

where inotify_umount_inodes() takes the
                mutex_lock(&inode->inotify_mutex);

The BC relationship comes directly from inotify_find_update_watch():
s32 inotify_find_update_watch(struct inotify_handle *ih, struct inode *inode,
                              u32 mask)
{
   ...
        mutex_lock(&inode->inotify_mutex);
        mutex_lock(&ih->mutex);

The CD relationship comes from inotify_rm_wd:
inotify_rm_wd does
        mutex_lock(&inode->inotify_mutex);
        mutex_lock(&ih->mutex)
and then calls inotify_remove_watch_locked() which calls
notify_dev_queue_event() which does
	        mutex_lock(&dev->ev_mutex);

(this strictly is a BCD relationship)

The DA relationship comes from the most interesting part:

  [<ffffffff8022d9f2>] shrink_icache_memory+0x42/0x270
  [<ffffffff80240dc4>] shrink_slab+0x11d/0x1c9
  [<ffffffff802b5104>] try_to_free_pages+0x187/0x244
  [<ffffffff8020efed>] __alloc_pages+0x1cd/0x2e0
  [<ffffffff8025e1f8>] cache_alloc_refill+0x3f8/0x821
  [<ffffffff8020a5e5>] kmem_cache_alloc+0x85/0xcb
  [<ffffffff802db027>] kernel_event+0x2e/0x122
  [<ffffffff8021d61c>] inotify_dev_queue_event+0xcc/0x140

inotify_dev_queue_event schedules a kernel_event which does a
kmem_cache_alloc( , GFP_KERNEL) which may try to shrink slabs, including
the inode cache .. which then takes iprune_mutex.

And voila, there is an AB, a BC, a CD relationship (even a direct BCD),
and also now a DA relationship -> a circular type AB-BA deadlock but
involving 4 locks.

The solution is simple: kernel_event() is NOT allowed to use GFP_KERNEL,
but must use GFP_NOFS to not cause recursion into the VFS.
Signed-off-by: NArjan van de Ven <arjan@linux.intel.com>
Acked-by: NIngo Molnar <mingo@elte.hu>
Acked-by: NRobert Love <rml@novell.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Enable mac partition table support per default also for a powermac config.
Signed-off-by: NOlaf Hering <olh@suse.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

We can immediately bail from invalidate_bdev() if the blockdev has no
pagecache.

This solves the huge IPI storms which hald is causing on the big ia64
machines when it polls CDROM drives.
Acked-by: NJes Sorensen <jes@sgi.com>
Cc: <stable@kernel.org>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

A recent commit (7fc90ec9) moved the
call to nfsd_setuser out of the 'find a dentry for a filehandle' branch
of fh_verify so that it would always be called.

This had the unfortunately side-effect of moving *after* the call to
decode_fh, so the prober fsuid was not set when nfsd_acceptable was called,
the 'permission' check did the wrong thing.

This patch moves the nfsd_setuser call back where it was, and add as call
in the other branch of the if.

Cc: "J. Bruce Fields" <bfields@fieldses.org>
Signed-off-by: NNeil Brown <neilb@suse.de>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>