Return an optional extent flags field from our lookup functions and wire up
callers to treat unwritten regions as holes for the purpose of returning
zeros to the user.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Due to the size of our group bitmaps, we'll never have a leaf node extent
record with more than 16 bits worth of clusters. Split e_clusters up so that
leaf nodes can get a flags field where we can mark unwritten extents.
Interior nodes whose length references all the child nodes beneath it can't
split their e_clusters field, so we use a union to preserve sizing there.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

We need to fill holes during a splice write. Provide our own splice write
actor which can call ocfs2_file_buffered_write() with a splice-specific
callback.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Do this instead of filemap_fdatawrite() - this way we sync only the
range between i_size and the cluster boundary.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Since we don't zero on extend anymore, truncate needs to be fixed up to zero
the part of a file between i_size and and end of it's cluster. Otherwise a
subsequent extend could expose bad data.

This introduced a new helper, which can be used in ocfs2_write().
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

ocfs2_get_block() didn't understand sparse files, fix that. Also remove some
code that isn't really useful anymore. We can fix up
ocfs2_direct_IO_get_blocks() at the same time.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

These are no longer used, and can't handle file systems with sparse file
allocation.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Unfortunately, ocfs2 can no longer make use of generic_file_aio_write_nlock()
because allocating writes will require zeroing of pages adjacent to the I/O
for cluster sizes greater than page size.

Implement a custom file write here, which can order page locks for zeroing.
This also has the advantage that cluster locks can easily be ordered outside
of the page locks.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

This will be turned back on once we can do allocation in ->page_mkwrite().
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Right now, file allocation for ocfs2 is done within ocfs2_extend_file(),
which is either called from ->setattr() (for an i_size change), or at the
top of ocfs2_file_aio_write().

Inodes on file systems with sparse file support will want to do their
allocation during the actual write call.

In either case the cluster locking decisions are the same. We abstract out
that code into a new function, ocfs2_lock_allocators() which will be used by
a later patch to enable writing to sparse files.

This also provides a nice cleanup of ocfs2_extend_allocation().
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

For ocfs2_truncate_file(), we eliminate the "simple" truncate case which no
longer exists since i_size is not tied to i_clusters. In
ocfs2_extend_file(), we skip the allocation / page zeroing code for file
systems which understand sparse files.

The core truncate code is changed to do a bottom up tree traversal. This
gets abstracted out into it's own function. To make things more readable,
most of the special case handling for in-inode extents from
ocfs2_do_truncate() is also removed.

Though write support for sparse files comes in a later patch, we at least
update ocfs2_prepare_inode_for_write() to skip allocation for sparse files.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

The code in extent_map.c is not prepared to deal with a subtree being
rotated between lookups. This can happen when filling holes in sparse files.
Instead of a lengthy patch to update the code (which would likely lose the
benefit of caching subtree roots), we remove most of the algorithms and
implement a simple path based lookup. A less ambitious extent caching scheme
will be added in a later patch.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Introduce tree rotations into the b-tree code. This will allow ocfs2 to
support sparse files. Much of the added code is designed to be generic (in
the ocfs2 sense) so that it can later be re-used to implement large
extended attributes.

This patch only adds the rotation code and does minimal updates to callers
of the extent api.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

There are two checks in there (one for inode newness, one for other mounted
nodes) which are unnecessary, so remove them. The DLM will allow the trylock
in either case without any messaging overhead.

Removing these makes ocfs2_request_delete() a one liner function, so just
move the trylock out one level into ocfs2_query_inode_wipe().
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Remove node messaging code that becomes unused with the delete inode vote
removal.

[Removed even more cruft which I spotted during review --Mark]
Signed-off-by: NTiger Yang <tiger.yang@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Ocfs2 currently does cluster-wide node messaging to check the open state of
an inode during delete. This patch removes that mechanism in favor of an
inode cluster lock which is taken at shared read when an inode is first read
and dropped in clear_inode(). This allows a deleting node to test the
liveness of an inode by attempting to take an exclusive lock.
Signed-off-by: NTiger Yang <tiger.yang@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

We don't want to print anything at all in ocfs2_lookup() when getting an
error from ocfs2_iget() - it could be something as innocuous as a signal
being detected in the dlm.

ocfs2_permission() should filter on -ENOENT which ocfs2_meta_lock() can
return if the inode was deleted on another node.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

We have noticed panic() hanging leading us to a situation in which
the node, while otherwise dead, is still disk heartbeating. This
leads to a hung cluster as the other nodes are waiting for this
node to stop disk heartbeating. This situation is only resolved
by power resetting the box.
Signed-off-by: NSunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Signed-off-by: NSunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

We don't want the extent map and uptodate cache destruction in
ocfs2_meta_lock_update() on a local mount, so skip that.

This fixes several bugs with uptodate being cleared on buffers and extent
maps being corrupted.
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

In dlm_migrate_all_locks(), we currently call cond_resched_lock() after
processing each lockres in a hash bucket. Move it outside the loop so as to
call it only after the entire hash bucket has been processed.
Signed-off-by: NSunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

There is a possibility that dlm_remaster_locks could overwride node->state
with DLM_RECO_NODE_DATA_REQUESTED after dlm_reco_data_done_handler sets the
node->state to DLM_RECO_NODE_DATA_DONE. This could lead to recovery getting
stuck and requires a cluster reboot. Synchronize with dlm_reco_state_lock
spinlock.
Signed-off-by: NSrinivas Eeda <srinivas.eeda@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

In dlm_migrate_lockres(), we check upfront whether the lockres is a
candidate for migration. This patch encapsulates that code in a separate
function so that dlm_empty_lockres() can also use it during umount. This
patch addresses the umount process spinning problem.
Signed-off-by: NSunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

During umount, the umount thread migrates the lockres' and the dlm_thread
frees the empty lockres'. Due to a race, the reference counting on the
lockres goes awry leading to extra puts.
Signed-off-by: NSunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

__dlm_lockres_unused() expects the caller to take the lockres spinlock.
Signed-off-by: NSunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

In some circumstances, this was causing us to reference freed memory.
Signed-off-by: NSunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Under load, OCFS2 would crash in invalidate_inode_pages2_range() because
invalidate_complete_page2() was unable to invalidate a page.  It would
appear that JBD is holding on to the page.  ext3 has a specific
->releasepage() handler to cover this case.

Steal ext3's ->releasepage(), ->invalidatepage(), and ->migratepage(), as
they appear completely appropriate for OCFS2.
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

This means that a build-up and a teardown could race which would result in a
double-kthread_stop().

Protect the setting and clearing of hr_task with o2hb_live_lock, as it's not
a common thing and not performance critical.
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

If ocfs2_register_hb_callbacks() succeeds on its first callback but fails
its second, it doesn't release the first on the way out. Fix that.

While we're at it, o2hb_unregister_callback() never returns anything but
0, so let's make it void.
Signed-off-by: NJoel Becker <joel.becker@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

        heirarchical, hierachical -> hierarchical
        heirarchy, hierachy -> hierarchy
Signed-off-by: NUwe Kleine-König <zeisberg@informatik.uni-freiburg.de>
Signed-off-by: NAdrian Bunk <bunk@stusta.de>

The semantic effect of insert_at_head is that it would allow new registered
sysctl entries to override existing sysctl entries of the same name.  Which is
pain for caching and the proc interface never implemented.

I have done an audit and discovered that none of the current users of
register_sysctl care as (excpet for directories) they do not register
duplicate sysctl entries.

So this patch simply removes the support for overriding existing entries in
the sys_sysctl interface since no one uses it or cares and it makes future
enhancments harder.
Signed-off-by: NEric W. Biederman <ebiederm@xmission.com>
Acked-by: NRalf Baechle <ralf@linux-mips.org>
Acked-by: NMartin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: David Howells <dhowells@redhat.com>
Cc: "Luck, Tony" <tony.luck@intel.com>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Andi Kleen <ak@muc.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Corey Minyard <minyard@acm.org>
Cc: Neil Brown <neilb@suse.de>
Cc: "John W. Linville" <linville@tuxdriver.com>
Cc: James Bottomley <James.Bottomley@steeleye.com>
Cc: Jan Kara <jack@ucw.cz>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Cc: Mark Fasheh <mark.fasheh@oracle.com>
Cc: David Chinner <dgc@sgi.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Patrick McHardy <kaber@trash.net>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

ocfs2 was did not have the binary number it uses under CTL_FS registered in
sysctl.h.  Register it to avoid future conflicts, and change the name of the
definition to be in line with the rest of the sysctl numbers.
Signed-off-by: NEric W. Biederman <ebiederm@xmission.com>
Acked-by: NMark Fasheh <mark.fasheh@oracle.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This patch is inspired by Arjan's "Patch series to mark struct
file_operations and struct inode_operations const".

Compile tested with gcc & sparse.
Signed-off-by: NJosef 'Jeff' Sipek <jsipek@cs.sunysb.edu>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Many struct inode_operations in the kernel can be "const".  Marking them const
moves these to the .rodata section, which avoids false sharing with potential
dirty data.  In addition it'll catch accidental writes at compile time to
these shared resources.
Signed-off-by: NArjan van de Ven <arjan@linux.intel.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

Many struct file_operations in the kernel can be "const".  Marking them const
moves these to the .rodata section, which avoids false sharing with potential
dirty data.  In addition it'll catch accidental writes at compile time to
these shared resources.
Signed-off-by: NArjan van de Ven <arjan@linux.intel.com>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

As was already pointed out Mathieu Avila on Thu, 07 Sep 2006 03:15:25 -0700
that OCFS2 is expecting bio_add_page() to add pages to BIOs in an easily
predictable manner.

That is not true, especially for devices with own merge_bvec_fn().

Therefore OCFS2's heartbeat code is very likely to fail on such devices.

Move the bio_put() call into the bio's bi_end_io() function. This makes the
whole idea of trying to predict the behaviour of bio_add_page() unnecessary.
Removed compute_max_sectors() and o2hb_compute_request_limits().
Signed-off-by: NPhilipp Reisner <philipp.reisner@linbit.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

When there is a lot of multithreaded I/O usage, two threads can collide
while sending out a message to the other nodes. This is due to the lack of
locking between threads while sending out the messages.

When a connected TCP send(), sendto(), or sendmsg() arrives in the Linux
kernel, it eventually comes through tcp_sendmsg(). tcp_sendmsg() protects
itself by acquiring a lock at invocation by calling lock_sock().
tcp_sendmsg() then loops over the buffers in the iovec, allocating
associated sk_buff's and cache pages for use in the actual send. As it does
so, it pushes the data out to tcp for actual transmission. However, if one
of those allocation fails (because a large number of large sends is being
processed, for example), it must wait for memory to become available. It
does so by jumping to wait_for_sndbuf or wait_for_memory, both of which
eventually cause a call to sk_stream_wait_memory(). sk_stream_wait_memory()
contains a code path that calls sk_wait_event(). Finally, sk_wait_event()
contains the call to release_sock().

The following patch adds a lock to the socket container in order to
properly serialize outbound requests.

From: Zhen Wei <zwei@novell.com>
Acked-by: NJeff Mahoney <jeffm@suse.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

Currently the ocfs2 dlm has no timeout during dlm join domain. While this is
not a problem in normal operation, this does become an issue if, say, the
other node is refusing to let the node join the domain because of a stuck
recovery. This patch adds a 90 sec timeout.
Signed-off-by: NSunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

These messages can easily be activated using the mlog infrastructure
and don't need to be enabled by default.
Signed-off-by: NSunil Mushran <sunil.mushran@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>

There is a small window where a joining node may not see the node(s) that
just died but are still part of the domain. To fix this, we must disallow
join requests if the joining node has a different node map.

A new field node_map is added to dlm_query_join_request to send the current
nodes nodemap along with join request. On the receiving end the nodes that
are part of the cluster verifies if this new node sees all the nodes that
are still part of the cluster. They disallow the join if the maps mismatch.
Signed-off-by: NSrinivas Eeda <srinivas.eeda@oracle.com>
Signed-off-by: NMark Fasheh <mark.fasheh@oracle.com>