openwrt/staging/blogic.git
18 years ago[GFS2] Fix glock ordering on inode creation
Steven Whitehouse [Wed, 22 Nov 2006 16:13:21 +0000 (11:13 -0500)]
[GFS2] Fix glock ordering on inode creation

The lock order here should be parent -> child rather than
numeric order.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Simplify glops functions
Steven Whitehouse [Mon, 20 Nov 2006 15:37:45 +0000 (10:37 -0500)]
[GFS2] Simplify glops functions

The go_sync callback took two flags, but one of them was set on every
call, so this patch removes once of the flags and makes the previously
conditional operations (on this flag), unconditional.

The go_inval callback took three flags, each of which was set on every
call to it. This patch removes the flags and makes the operations
unconditional, which makes the logic rather more obvious.

Two now unused flags are also removed from incore.h.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Fix Kconfig wrt CRC32
Steven Whitehouse [Mon, 20 Nov 2006 15:04:49 +0000 (10:04 -0500)]
[GFS2] Fix Kconfig wrt CRC32

GFS2 requires the CRC32 library function. This was reported by
Toralf Förster.

Cc: Toralf Förster <toralf.foerster@gmx.de>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Make sentinel dirents compatible with gfs1
Steven Whitehouse [Fri, 17 Nov 2006 17:27:44 +0000 (12:27 -0500)]
[GFS2] Make sentinel dirents compatible with gfs1

When deleting directory entries, we set the inum.no_addr to zero
in a dirent when its the first dirent in a block and thus cannot
be merged into the previous dirent as is the usual case. In gfs1,
inum.no_formal_ino was used instead.

This patch changes gfs2 to set both inum.no_addr and inum.no_formal_ino
to zero. It also changes the test from just looking at inum.no_addr to
look at both inum.no_addr and inum.no_formal_ino and a sentinel is
now considered to be a dirent in which _either_ (or both) of them
is set to zero.

This resolves Red Hat bugzillas: #215809, #211465

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Remove unused function from inode.c
Steven Whitehouse [Thu, 16 Nov 2006 16:08:16 +0000 (11:08 -0500)]
[GFS2] Remove unused function from inode.c

The gfs2_glock_nq_m_atime function is unused in so far as its only
ever called with num_gh = 1, and this falls through to the
gfs2_glock_nq_atime function, so we might as well call that directly.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Remove unused sysfs files
Steven Whitehouse [Thu, 16 Nov 2006 15:58:55 +0000 (10:58 -0500)]
[GFS2] Remove unused sysfs files

Four of the sysfs files are unused and can therefore be removed.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Tidy up bmap & fix boundary bug
Steven Whitehouse [Wed, 15 Nov 2006 20:21:06 +0000 (15:21 -0500)]
[GFS2] Tidy up bmap & fix boundary bug

This moves the locking for bmap into the bmap function itself
rather than using a wrapper function. It also fixes a bug where
the boundary flag was set on the wrong bh. Also the flags on
the mapped bh are reset earlier in the function to ensure that
they are 100% correct on the error path.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Fix memory allocation in glock.c
Steven Whitehouse [Wed, 15 Nov 2006 20:17:03 +0000 (15:17 -0500)]
[GFS2] Fix memory allocation in glock.c

Change from GFP_KERNEL to GFP_NOFS as this was causing a
slow down when trying to push inodes from cache.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[DLM] Fix DLM config
Patrick Caulfield [Wed, 15 Nov 2006 17:29:24 +0000 (12:29 -0500)]
[DLM] Fix DLM config

The attached patch fixes the DLM config so that it selects the chosen network
transport. It should fix the bug where DLM can be left selected when NET gets
unselected. This incorporates all the comments received about this patch.

Cc: Adrian Bunk <bunk@stusta.de>
Cc: Andrew Morton <akpm@osdl.org>
Signed-Off-By: Patrick Caulfield <pcaulfie@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[DLM] clear sbflags on lock master
David Teigland [Fri, 10 Nov 2006 20:16:27 +0000 (14:16 -0600)]
[DLM] clear sbflags on lock master

RH BZ 211622

The ALTMODE flag can be set in the lock master's copy of the lock but
never cleared, so ALTMODE will also be returned in a subsequent conversion
of the lock when it shouldn't be.  This results in lock_dlm incorrectly
switching to the alternate lock mode when returning the result to gfs
which then asserts when it sees the wrong lock state.  The fix is to
propagate the cleared sbflags value to the master node when the lock is
requested.  QA's d_rwrandirectlarge test triggers this bug very quickly.

Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[DLM] do full recover_locks barrier
David Teigland [Wed, 1 Nov 2006 15:31:48 +0000 (09:31 -0600)]
[DLM] do full recover_locks barrier

Red Hat BZ 211914

The previous patch "[DLM] fix aborted recovery during
node removal" was incomplete as discovered with further testing.  It set
the bit for the RS_LOCKS barrier but did not then wait for the barrier.
This is often ok, but sometimes it will cause yet another recovery hang.
If it's a new node that also has the lowest nodeid that skips the barrier
wait, then it misses the important step of collecting and reporting the
barrier status from the other nodes (which is the job of the low nodeid in
the barrier wait routine).

Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[DLM] fix stopping unstarted recovery
David Teigland [Tue, 31 Oct 2006 17:56:08 +0000 (11:56 -0600)]
[DLM] fix stopping unstarted recovery

Red Hat BZ 211914

When many nodes are joining a lockspace simultaneously, the dlm gets a
quick sequence of stop/start events, a pair for adding each node.
dlm_controld in user space sends dlm_recoverd in the kernel each stop and
start event.  dlm_controld will sometimes send the stop before
dlm_recoverd has had a chance to take up the previously queued start.  The
stop aborts the processing of the previous start by setting the
RECOVERY_STOP flag.  dlm_recoverd is erroneously clearing this flag and
ignoring the stop/abort if it happens to take up the start after the stop
meant to abort it.  The fix is to check the sequence number that's
incremented for each stop/start before clearing the flag.

Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[DLM] fix aborted recovery during node removal
David Teigland [Tue, 31 Oct 2006 17:56:01 +0000 (11:56 -0600)]
[DLM] fix aborted recovery during node removal

Red Hat BZ 211914

With the new cluster infrastructure, dlm recovery for a node removal can
be aborted and restarted for a node addition.  When this happens, the
restarted recovery isn't aware that it's doing recovery for the earlier
removal as well as the addition.  So, it then skips the recovery steps
only required when nodes are removed.  This can result in locks not being
purged for failed/removed nodes.  The fix is to check for removed nodes
for which recovery has not been completed at the start of a new recovery
sequence.

Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[DLM] fix requestqueue race
David Teigland [Tue, 31 Oct 2006 17:55:56 +0000 (11:55 -0600)]
[DLM] fix requestqueue race

Red Hat BZ 211914

There's a race between dlm_recoverd (1) enabling locking and (2) clearing
out the requestqueue, and dlm_recvd (1) checking if locking is enabled and
(2) adding a message to the requestqueue.  An order of recoverd(1),
recvd(1), recvd(2), recoverd(2) will result in a message being left on the
requestqueue.  The fix is to have dlm_recvd check if dlm_recoverd has
enabled locking after taking the mutex for the requestqueue and if it has
processing the message instead of queueing it.

Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[DLM] status messages ping-pong between unmounted nodes
David Teigland [Thu, 2 Nov 2006 15:45:56 +0000 (09:45 -0600)]
[DLM] status messages ping-pong between unmounted nodes

Red Hat BZ 213682

If two nodes leave the lockspace (while unmounting the fs in the case of
gfs) after one has sent a STATUS message to the other, STATUS/STATUS_REPLY
messages will then ping-pong between the nodes when neither of them can
find the lockspace in question any longer.  We kill this by not sending
another STATUS message when we get a STATUS_REPLY for an unknown
lockspace.

Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[DLM] res_recover_locks_count not reset when recover_locks is aborted
David Teigland [Thu, 2 Nov 2006 15:49:02 +0000 (09:49 -0600)]
[DLM] res_recover_locks_count not reset when recover_locks is aborted

Red Hat BZ 213684

If a node sends an lkb to the new master (RCOM_LOCK message) during
recovery and recovery is then aborted on both nodes before it gets a
reply, the res_recover_locks_count needs to be reset to 0 so that when the
subsequent recovery comes along and sends the lkb to the new master again
the assertion doesn't trigger that checks that counter is zero.

Signed-off-by: David Teigland <teigland@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[DLM] Add support for tcp communications
Patrick Caulfield [Thu, 2 Nov 2006 16:19:21 +0000 (11:19 -0500)]
[DLM] Add support for tcp communications

The following patch adds a TCP based communications layer
to the DLM which is compile time selectable. The existing SCTP
layer gives the advantage of allowing multihoming, whereas
the TCP layer has been heavily tested in previous versions of
the DLM and is known to be robust and therefore can be used as
a baseline for performance testing.

Signed-off-by: Patrick Caulfield <pcaulfie@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Remove unused zero_readpage from stuffed_readpage
Russell Cattelan [Thu, 9 Nov 2006 16:42:33 +0000 (11:42 -0500)]
[GFS2] Remove unused zero_readpage from stuffed_readpage

Stuffed files only consist of a maximum of
(gfs2 block size - sizeof(struct gfs2_dinode)) bytes. Since the
gfs2 block size is always less than page size, we will never see
a call to stuffed_readpage for anything other than the first page
in the file.

Signed-off-by: Russell Cattelan <cattelan@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Fix race in logging code
Russell Cattelan [Thu, 9 Nov 2006 16:28:08 +0000 (11:28 -0500)]
[GFS2] Fix race in logging code

The log lock is dropped prior to io submittion, but
this exposes a hole in which the log data structures
may be going away due to a truncate.
Store the buffer head in a local pointer prior to
dropping the lock and relay on the buffer_head lock
for consitency on the buffer head.

Signed-Off-By: Russell Cattelan <cattelan@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Remove gfs2_inode_attr_in
Steven Whitehouse [Wed, 8 Nov 2006 20:45:46 +0000 (15:45 -0500)]
[GFS2] Remove gfs2_inode_attr_in

This function wasn't really doing the right thing. There was no need
to update the inode size at this point and the updating of the
i_blocks field has now been moved to the places where di_blocks is
updated. A result of this patch and some those preceeding it is that
unlocking a glock is now a much more efficient process, since there
is no longer any requirement to copy data from the gfs2 inode into
the vfs inode at this point.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Inode number is constant
Steven Whitehouse [Wed, 8 Nov 2006 18:52:05 +0000 (13:52 -0500)]
[GFS2] Inode number is constant

Since the inode number is constant, we don't need to keep updating
it everytime we refresh the other inode fields.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Only set inode flags when required
Steven Whitehouse [Wed, 8 Nov 2006 17:51:06 +0000 (12:51 -0500)]
[GFS2] Only set inode flags when required

We were setting the inode flags from GFS2's flags far too often, even when they
couldn't possibly have changed. This patch reduces the amount of flag
setting going on so that we do it only when the inode is read in or
when the flags have changed. The create case is covered by the "when
the inode is read in" case.

This also fixes a bug where we didn't set S_SYNC correctly.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Fix page lock/glock deadlock
Steven Whitehouse [Wed, 8 Nov 2006 15:26:54 +0000 (10:26 -0500)]
[GFS2] Fix page lock/glock deadlock

This fixes a race between the glock and the page lock encountered
during truncate in gfs2_readpage and gfs2_prepare_write. The gfs2_readpages
function doesn't need the same fix since it only uses a try lock anyway, so
it will fail back to gfs2_readpage in the case of a potential deadlock.

This bug was spotted by Russell Cattelan.

Cc: Russell Cattelan <cattelan@redhat.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Remove unused GL_DUMP flag
Steven Whitehouse [Wed, 8 Nov 2006 14:01:13 +0000 (09:01 -0500)]
[GFS2] Remove unused GL_DUMP flag

There is no way to set the GL_DUMP flag, and in any case the
same thing can be done with systemtap if required for debugging,
so this removes it.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Don't copy meta_header for rgrp in and out
Steven Whitehouse [Tue, 7 Nov 2006 20:14:58 +0000 (15:14 -0500)]
[GFS2] Don't copy meta_header for rgrp in and out

The meta_header for an ondisk rgrp never changes, so there is no point
copying it in and back out to disk. Also there is no reason to keep
a copy for each rgrp in memory.

The code already checks to ensure that the header is correct before
it calls the routine to copy the data in, so that we don't even need
to check whether its correct on disk in the functions in ondisk.c

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Tidy up 0 initialisations in inode.c
Steven Whitehouse [Thu, 2 Nov 2006 16:59:28 +0000 (11:59 -0500)]
[GFS2] Tidy up 0 initialisations in inode.c

We don't need to use endian conversions for 0 initialisations
when creating a new on-disk inode.

Cc: Christoph Hellwig <hch@infradead.org>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Shrink gfs2_inode (8) - i_vn
Steven Whitehouse [Wed, 1 Nov 2006 21:05:38 +0000 (16:05 -0500)]
[GFS2] Shrink gfs2_inode (8) - i_vn

This shrinks the size of the gfs2_inode by 8 bytes by
replacing the version counter with a one bit valid/invalid
flag.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Shrink gfs2_inode (7) - di_payload_format
Steven Whitehouse [Thu, 2 Nov 2006 01:09:14 +0000 (20:09 -0500)]
[GFS2] Shrink gfs2_inode (7) - di_payload_format

This is almost never used. Its there for backward
compatibility with GFS1. It doesn't need its own
field since it can always be calculated from the
inode mode & flags. This saves a bit more space
in the gfs2_inode.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Shrink gfs2_inode (6) - di_atime/di_mtime/di_ctime
Steven Whitehouse [Wed, 1 Nov 2006 19:35:17 +0000 (14:35 -0500)]
[GFS2] Shrink gfs2_inode (6) - di_atime/di_mtime/di_ctime

Remove the di_[amc]time fields and use inode->i_[amc]time
fields instead. This saves 24 bytes from the gfs2_inode.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Shrink gfs2_inode (5) - di_nlink
Steven Whitehouse [Wed, 1 Nov 2006 19:04:17 +0000 (14:04 -0500)]
[GFS2] Shrink gfs2_inode (5) - di_nlink

Remove the di_nlink field in favour of inode->i_nlink and
update the nlink handling to use the proper macros. This
saves 4 bytes.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Shrink gfs2_inode (4) - di_uid/di_gid
Steven Whitehouse [Wed, 1 Nov 2006 18:23:29 +0000 (13:23 -0500)]
[GFS2] Shrink gfs2_inode (4) - di_uid/di_gid

Remove duplicate di_uid/di_gid fields in favour of using
inode->i_uid/inode->i_gid instead. This saves 8 bytes.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Shrink gfs2_inode (3) - di_mode
Steven Whitehouse [Wed, 1 Nov 2006 17:22:46 +0000 (12:22 -0500)]
[GFS2] Shrink gfs2_inode (3) - di_mode

This removes the duplicate di_mode field in favour of using the
inode->i_mode field. This saves 4 bytes.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Shrink gfs2_inode (2) - di_major/di_minor
Steven Whitehouse [Wed, 1 Nov 2006 02:45:08 +0000 (21:45 -0500)]
[GFS2] Shrink gfs2_inode (2) - di_major/di_minor

This removes the device numbers from this structure by using
inode->i_rdev instead. It also cleans up the code in gfs2_mknod.
It results in shrinking the gfs2_inode by 8 bytes.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Shrink gfs2_inode (1) - di_header/di_num
Steven Whitehouse [Wed, 1 Nov 2006 15:34:15 +0000 (10:34 -0500)]
[GFS2] Shrink gfs2_inode (1) - di_header/di_num

The metadata header doesn't need to be stored in the incore
struct gfs2_inode since its constant, and this patch removes it.
Also, there is already a field for the inode's number in the
struct gfs2_inode, so we don't need one in struct gfs2_dinode_host
as well.

This saves 28 bytes of space in the struct gfs2_inode.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Change argument to gfs2_dinode_print
Steven Whitehouse [Wed, 1 Nov 2006 00:00:24 +0000 (19:00 -0500)]
[GFS2] Change argument to gfs2_dinode_print

Change argument for gfs2_dinode_print in order to prepare
for removal of duplicate fields between struct inode and
struct gfs2_dinode_host.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Move gfs2_dinode_in to inode.c
Steven Whitehouse [Tue, 31 Oct 2006 20:28:00 +0000 (15:28 -0500)]
[GFS2] Move gfs2_dinode_in to inode.c

gfs2_dinode_in() is only ever called from one place, so move it
to that place (in inode.c) and make it static.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Change argument to gfs2_dinode_in
Steven Whitehouse [Tue, 31 Oct 2006 20:22:10 +0000 (15:22 -0500)]
[GFS2] Change argument to gfs2_dinode_in

This is a preliminary patch to enable the removal of fields
in gfs2_dinode_host which are duplicated in struct inode.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Change argument of gfs2_dinode_out
Steven Whitehouse [Tue, 31 Oct 2006 20:07:05 +0000 (15:07 -0500)]
[GFS2] Change argument of gfs2_dinode_out

Everywhere this was called, a struct gfs2_inode was available,
but despite that, it was always called with a struct gfs2_dinode
as an argument. By making this change it paves the way to start
eliminating fields duplicated between the kernel's struct inode
and the struct gfs2_dinode.

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] gfs2 __user misannotation fix
Al Viro [Sat, 14 Oct 2006 03:49:23 +0000 (23:49 -0400)]
[GFS2] gfs2 __user misannotation fix

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] gfs2 misc endianness annotations
Al Viro [Sat, 14 Oct 2006 14:46:30 +0000 (10:46 -0400)]
[GFS2] gfs2 misc endianness annotations

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] split and annotate gfs2_quota_change
Al Viro [Sat, 14 Oct 2006 03:46:46 +0000 (23:46 -0400)]
[GFS2] split and annotate gfs2_quota_change

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] split and annotate gfs2_statfs_change
Al Viro [Sat, 14 Oct 2006 03:43:19 +0000 (23:43 -0400)]
[GFS2] split and annotate gfs2_statfs_change

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] split and annotate gfs2_quota
Al Viro [Sat, 14 Oct 2006 03:31:55 +0000 (23:31 -0400)]
[GFS2] split and annotate gfs2_quota

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] split and annotate gfs2_inum
Al Viro [Sat, 14 Oct 2006 02:51:24 +0000 (22:51 -0400)]
[GFS2] split and annotate gfs2_inum

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] split and annotate gfs_rindex
Al Viro [Sat, 14 Oct 2006 02:51:24 +0000 (22:51 -0400)]
[GFS2] split and annotate gfs_rindex

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] split and annotate gfs2_meta_header
Al Viro [Sat, 14 Oct 2006 01:57:23 +0000 (21:57 -0400)]
[GFS2] split and annotate gfs2_meta_header

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] Fix crc32 calculation in recovery.c
Steven Whitehouse [Tue, 31 Oct 2006 19:44:50 +0000 (14:44 -0500)]
[GFS2] Fix crc32 calculation in recovery.c

Commit "[GFS2] split and annotate gfs2_log_head" resulted in an incorrect
checksum calculation for log headers. This patch corrects the
problem without resorting to copying the whole log header as
the previous code used to.

Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] split and annotate gfs2_log_head
Al Viro [Sat, 14 Oct 2006 01:47:13 +0000 (21:47 -0400)]
[GFS2] split and annotate gfs2_log_head

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] split and annotate gfs2_inum_range
Al Viro [Sat, 14 Oct 2006 01:29:46 +0000 (21:29 -0400)]
[GFS2] split and annotate gfs2_inum_range

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] split and annotate gfs2_rgrp
Al Viro [Sat, 14 Oct 2006 01:07:22 +0000 (21:07 -0400)]
[GFS2] split and annotate gfs2_rgrp

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] fields of gfs2_sb_host are host-endian
Al Viro [Sat, 14 Oct 2006 01:02:41 +0000 (21:02 -0400)]
[GFS2] fields of gfs2_sb_host are host-endian

... and several could be killed, but that's another story.
Annotate scalar ones, kill completely unused.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] split gfs2_sb
Al Viro [Sat, 14 Oct 2006 00:45:02 +0000 (20:45 -0400)]
[GFS2] split gfs2_sb

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] gfs2_dinode_host fields are host-endian
Al Viro [Sat, 14 Oct 2006 00:33:01 +0000 (20:33 -0400)]
[GFS2] gfs2_dinode_host fields are host-endian

Annotated scalar fields, dropped unused ones.  Note that
it's not at all obvious that we want to convert all of them
to host-endian...

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years ago[GFS2] split gfs2_dinode into on-disk and host variants
Al Viro [Sat, 14 Oct 2006 00:11:25 +0000 (20:11 -0400)]
[GFS2] split gfs2_dinode into on-disk and host variants

The latter is used as part of gfs2-private part of struct inode.
It actually stores a lot of fields differently; for now the
declaration is just cloned, inode field is swtiched and changes
propagated.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
18 years agoLinux 2.6.19
Linus Torvalds [Wed, 29 Nov 2006 21:57:37 +0000 (13:57 -0800)]
Linux 2.6.19

It's all good.

18 years ago[PATCH] r8169: Fix iteration variable sign
Francois Romieu [Wed, 29 Nov 2006 21:21:33 +0000 (22:21 +0100)]
[PATCH] r8169: Fix iteration variable sign

This changes the type of variable "i" in rtl8169_init_one()
from "unsigned int" to "int". "i" is checked for < 0 later,
which can never happen for "unsigned". This results in broken
error handling.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: Francois Romieu <romieu@fr.zoreil.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years agoMerge branch 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm
Linus Torvalds [Wed, 29 Nov 2006 21:44:00 +0000 (13:44 -0800)]
Merge branch 'for-linus' of /home/rmk/linux-2.6-arm

* 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm:
  [ARM] 3943/1: share declaration of struct pxa2xx_udc_mach_info between multiple platforms
  [ARM] Export smp_call_function()
  [ARM] Add PM_LEGACY defaults

18 years ago[ARM] 3943/1: share declaration of struct pxa2xx_udc_mach_info between multiple platforms
Milan Svoboda [Wed, 29 Nov 2006 11:09:52 +0000 (12:09 +0100)]
[ARM] 3943/1: share declaration of struct pxa2xx_udc_mach_info between multiple platforms

Move declaration of struct pxa2xx_udc_mach_info from
include/asm-arm/arch-pxa/udc.h to new file
include/asm-arm/mach/udc_pxa2xx.h.

This allow us to use this structure with
multiple platforms - pxa and ixp4xx. USB
device controller used in pxa25x is the same
as controller used in ixp4xx.

Signed-off-by: Milan Svoboda <msvoboda@ra.rockwell.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[MIPS] Do topology_init even on uniprocessor kernels.
Ralf Baechle [Wed, 29 Nov 2006 15:04:08 +0000 (15:04 +0000)]
[MIPS] Do topology_init even on uniprocessor kernels.

Otherwise CPU 0 doesn't show up in sysfs which breaks some software.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
Linus Torvalds [Wed, 29 Nov 2006 05:28:14 +0000 (21:28 -0800)]
Merge /pub/scm/linux/kernel/git/davem/net-2.6

* master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6:
  [NET]: Fix MAX_HEADER setting.
  [NETFILTER]: ipt_REJECT: fix memory corruption
  [NETFILTER]: conntrack: fix refcount leak when finding expectation
  [NETFILTER]: ctnetlink: fix reference count leak
  [NETFILTER]: nf_conntrack: fix the race on assign helper to new conntrack
  [NETFILTER]: nfctnetlink: assign helper to newly created conntrack

18 years ago[NET]: Fix MAX_HEADER setting.
David S. Miller [Wed, 29 Nov 2006 04:53:39 +0000 (20:53 -0800)]
[NET]: Fix MAX_HEADER setting.

MAX_HEADER is either set to LL_MAX_HEADER or LL_MAX_HEADER + 48, and
this is controlled by a set of CONFIG_* ifdef tests.

It is trying to use LL_MAX_HEADER + 48 when any of the tunnels are
enabled which set hard_header_len like this:

dev->hard_header_len = LL_MAX_HEADER + sizeof(struct xxx);

The correct set of tunnel drivers which do this are:

ipip
ip_gre
ip6_tunnel
sit

so make the ifdef test match.

Noticed by Patrick McHardy and with help from Herbert Xu.

Signed-off-by: David S. Miller <davem@davemloft.net>
18 years ago[NETFILTER]: ipt_REJECT: fix memory corruption
Patrick McHardy [Wed, 29 Nov 2006 04:10:21 +0000 (20:10 -0800)]
[NETFILTER]: ipt_REJECT: fix memory corruption

On devices with hard_header_len > LL_MAX_HEADER ip_route_me_harder()
reallocates the skb, leading to memory corruption when using the stale
tcph pointer to update the checksum.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years ago[NETFILTER]: conntrack: fix refcount leak when finding expectation
Yasuyuki Kozakai [Mon, 27 Nov 2006 18:26:46 +0000 (10:26 -0800)]
[NETFILTER]: conntrack: fix refcount leak when finding expectation

All users of __{ip,nf}_conntrack_expect_find() don't expect that
it increments the reference count of expectation.

Signed-off-by: Yasuyuki Kozakai <yasuyuki.kozakai@toshiba.co.jp>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years ago[NETFILTER]: ctnetlink: fix reference count leak
Patrick McHardy [Mon, 27 Nov 2006 18:26:25 +0000 (10:26 -0800)]
[NETFILTER]: ctnetlink: fix reference count leak

When NFA_NEST exceeds the skb size the protocol reference is leaked.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years ago[NETFILTER]: nf_conntrack: fix the race on assign helper to new conntrack
Yasuyuki Kozakai [Mon, 27 Nov 2006 18:25:59 +0000 (10:25 -0800)]
[NETFILTER]: nf_conntrack: fix the race on assign helper to new conntrack

The found helper cannot be assigned to conntrack after unlocking
nf_conntrack_lock. This tries to find helper to assign again.

Signed-off-by: Yasuyuki Kozakai <yasuyuki.kozakai@toshiba.co.jp>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years ago[NETFILTER]: nfctnetlink: assign helper to newly created conntrack
Yasuyuki Kozakai [Mon, 27 Nov 2006 18:25:32 +0000 (10:25 -0800)]
[NETFILTER]: nfctnetlink: assign helper to newly created conntrack

This fixes the bug which doesn't assign helper to newly created
conntrack via nf_conntrack_netlink.

Signed-off-by: Yasuyuki Kozakai <yasuyuki.kozakai@toshiba.co.jp>
Signed-off-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
18 years agoMerge branch 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik...
Linus Torvalds [Wed, 29 Nov 2006 04:54:40 +0000 (20:54 -0800)]
Merge branch 'upstream-linus' of /linux/kernel/git/jgarzik/libata-dev

* 'upstream-linus' of master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  [PATCH] libata: Fixup ata_sas_queuecmd to handle __ata_scsi_queuecmd failure
  [PATCH] ahci: AHCI mode SATA patch for Intel ICH9
  [PATCH] libata: don't schedule EH on wcache on/off if old EH

18 years agoMerge branch 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/selinux-2.6
Linus Torvalds [Wed, 29 Nov 2006 01:32:35 +0000 (17:32 -0800)]
Merge branch 'fixes' of git://git./linux/kernel/git/jmorris/selinux-2.6

* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/selinux-2.6:
  selinux: fix dentry_open() error check

18 years agoMerge branch 'for-linus' of git://one.firstfloor.org/home/andi/git/linux-2.6
Linus Torvalds [Wed, 29 Nov 2006 01:28:41 +0000 (17:28 -0800)]
Merge branch 'for-linus' of git://one.firstfloor.org/home/andi/git/linux-2.6

* 'for-linus' of git://one.firstfloor.org/home/andi/git/linux-2.6:
  [PATCH] x86-64: Use stricter in process stack check for unwinder
  [PATCH] i386: Fix compilation with UP genericarch
  [PATCH] x86-64: Fix warning in io_apic.c
  [PATCH] x86-64: work around gcc4 issue with -Os in Dwarf2 stack unwind
  [PATCH] x86_64: Align data segment to PAGE_SIZE boundary

18 years agoMerge branch 'linus' of master.kernel.org:/pub/scm/linux/kernel/git/perex/alsa
Linus Torvalds [Wed, 29 Nov 2006 01:27:52 +0000 (17:27 -0800)]
Merge branch 'linus' of /linux/kernel/git/perex/alsa

* 'linus' of master.kernel.org:/pub/scm/linux/kernel/git/perex/alsa:
  [ALSA] version 1.0.13
  [ALSA] snd-emu10k1: Fix capture for one variant.
  [ALSA] Fix hang-up at disconnection of usb-audio
  [ALSA] hda: fix typo for xw4400 PCI sub-ID
  [ALSA] hda: fix sigmatel dell system detection
  [ALSA] Enable stereo line input for TAS codec
  [ALSA] rtctimer: handle RTC interrupts with a tasklet

18 years agoMerge branch 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus
Linus Torvalds [Wed, 29 Nov 2006 01:27:11 +0000 (17:27 -0800)]
Merge branch 'upstream' of git://ftp.linux-mips.org/upstream-linus

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
  [MIPS] Fix Bonito bootup message.

18 years ago[PATCH] add missing libsas include to fix s390 compilation.
Dave Jones [Tue, 28 Nov 2006 04:58:11 +0000 (23:58 -0500)]
[PATCH] add missing libsas include to fix s390 compilation.

include/scsi/libsas.h:479: error: field 'smp_req' has incomplete type
include/scsi/libsas.h:480: error: field 'smp_resp' has incomplete type

Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] ecryptfs: fix crypto_alloc_blkcipher() error check
Akinobu Mita [Tue, 28 Nov 2006 20:29:49 +0000 (12:29 -0800)]
[PATCH] ecryptfs: fix crypto_alloc_blkcipher() error check

The return value of crypto_alloc_blkcipher() should be checked by IS_ERR().

Cc: Mike Halcrow <mhalcrow@us.ibm.com>
Cc: Phillip Hellewell <phillip@hellewell.homeip.net>
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] fix create_write_pipe() error check
Akinobu Mita [Tue, 28 Nov 2006 20:29:43 +0000 (12:29 -0800)]
[PATCH] fix create_write_pipe() error check

The return value of create_write_pipe()/create_read_pipe() should be
checked by IS_ERR().

Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] Fix Intel/Sharp command set erase suspend bug
Joakim Tjernlund [Tue, 28 Nov 2006 23:11:52 +0000 (23:11 +0000)]
[PATCH] Fix Intel/Sharp command set erase suspend bug

When we sleep and wait for a suspended operation to be resumed, go
back and check until it's ready -- don't just continue after the first
time we're woken. This can cause file system corruption.

Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
Signed-off-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] bridge: fix possible overflow in get_fdb_entries
Chris Wright [Mon, 20 Nov 2006 23:02:49 +0000 (15:02 -0800)]
[PATCH] bridge: fix possible overflow in get_fdb_entries

Make sure to properly clamp maxnum to avoid overflow

Signed-off-by: Chris Wright <chrisw@sous-sol.org>
Acked-by: Eugene Teo <eteo@redhat.com>
Acked-by: Marcel Holtmann <marcel@holtmann.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[MIPS] Fix Bonito bootup message.
Ralf Baechle [Fri, 24 Nov 2006 12:17:51 +0000 (12:17 +0000)]
[MIPS] Fix Bonito bootup message.

Even when enabling Bonito IOBC coherence the kernel would actually claim
it was disabling it.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
18 years ago[PATCH] x86-64: Use stricter in process stack check for unwinder
Andi Kleen [Tue, 28 Nov 2006 19:12:59 +0000 (20:12 +0100)]
[PATCH] x86-64: Use stricter in process stack check for unwinder

Previously it would check for alignment only, which could break
if the stack pointer was unaligned. Now explicitely check if the
stack pointer is in the stack page of the current process.

Ported from i386.

Signed-off-by: Andi Kleen <ak@suse.de>
18 years ago[PATCH] i386: Fix compilation with UP genericarch
Andi Kleen [Tue, 28 Nov 2006 19:12:59 +0000 (20:12 +0100)]
[PATCH] i386: Fix compilation with UP genericarch

Fix

arch/i386/mach-generic/built-in.o: In function `apicid_to_node':
summit.c:(.text+0x2f): undefined reference to `apicid_2_node'

with CONFIG_GENERICH_ARCH and !CONFIG_SMP
Signed-off-by: Andi Kleen <ak@suse.de>
18 years ago[PATCH] x86-64: Fix warning in io_apic.c
Andi Kleen [Tue, 28 Nov 2006 19:12:59 +0000 (20:12 +0100)]
[PATCH] x86-64: Fix warning in io_apic.c

18 years ago[PATCH] x86-64: work around gcc4 issue with -Os in Dwarf2 stack unwind
Jan Beulich [Tue, 28 Nov 2006 19:12:59 +0000 (20:12 +0100)]
[PATCH] x86-64: work around gcc4 issue with -Os in Dwarf2 stack unwind

This fixes a problem with gcc4 mis-compiling the stack unwind code under
-Os, which resulted in 'stuck' messages whenever an assembly routine was
encountered.

(The second hunk is trivial cleanup.)

Signed-off-by: Jan Beulich <jbeulich@novell.com>
18 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6
Andi Kleen [Tue, 28 Nov 2006 19:12:29 +0000 (20:12 +0100)]
Merge branch 'master' of git://git./linux/kernel/git/torvalds/linux-2.6

18 years ago[PATCH] x86_64: fix 'earlyprintk=...,keep' regression
Ingo Molnar [Tue, 28 Nov 2006 08:14:05 +0000 (09:14 +0100)]
[PATCH] x86_64: fix 'earlyprintk=...,keep' regression

Commit 2c8c0e6b8d7700a990da8d24eff767f9ca223b96 ("[PATCH] Convert x86-64
to early param") broke the earlyprintk=...,keep feature.

This restores that functionality.  Tested on x86_64.  Must-have for
v2.6.19, no risk.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[ALSA] version 1.0.13
Jaroslav Kysela [Tue, 28 Nov 2006 14:07:33 +0000 (15:07 +0100)]
[ALSA] version 1.0.13

Signed-off-by: Jaroslav Kysela <perex@suse.cz>
18 years ago[ALSA] snd-emu10k1: Fix capture for one variant.
James Courtier-Dutton [Sat, 25 Nov 2006 22:02:47 +0000 (22:02 +0000)]
[ALSA] snd-emu10k1: Fix capture for one variant.

Fixes ALSA bug#324

Signed-off-by: James Courtier-Dutton <James@superbug.co.uk>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
18 years ago[ALSA] Fix hang-up at disconnection of usb-audio
Takashi Iwai [Wed, 8 Nov 2006 14:41:29 +0000 (15:41 +0100)]
[ALSA] Fix hang-up at disconnection of usb-audio

Fix hang-up at disconnection of usb-audio devices while accessing PCM.
Don't handle PCM operations any more after shutdown flag is set.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
18 years ago[ALSA] hda: fix typo for xw4400 PCI sub-ID
John W. Linville [Mon, 6 Nov 2006 11:01:53 +0000 (12:01 +0100)]
[ALSA] hda: fix typo for xw4400 PCI sub-ID

The PCI sub-device ID for the HP xw4400 is actually 0x280c.

Signed-off-by: John W. Linville <linville@tuxdriver.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
18 years ago[ALSA] hda: fix sigmatel dell system detection
Matt Porter [Mon, 6 Nov 2006 10:20:38 +0000 (11:20 +0100)]
[ALSA] hda: fix sigmatel dell system detection

Fixes Dell system detection on 9200 codecs. The support
to detect certain Dell machines was merged in the
9205 table where it will be unused on the various Dell
9200-based codec systems. This moves the subsystem IDs
to the correct 9200 table.

Signed-off-by: Matt Porter <mporter@embeddedalley.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
18 years ago[ALSA] Enable stereo line input for TAS codec
Paul Mackerras [Tue, 31 Oct 2006 14:24:45 +0000 (15:24 +0100)]
[ALSA] Enable stereo line input for TAS codec

Despite what the data sheet says in one place, to get stereo input
from input A (line in), we have to clear the 'input B monaural' bit in
the ACR.

Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
18 years ago[ALSA] rtctimer: handle RTC interrupts with a tasklet
Clemens Ladisch [Fri, 27 Oct 2006 08:45:00 +0000 (10:45 +0200)]
[ALSA] rtctimer: handle RTC interrupts with a tasklet

The calls to rtc_control() from inside the interrupt handler can upset
the RTC code, so move our interrupt handling code to a tasklet.

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Jaroslav Kysela <perex@suse.cz>
18 years ago[PATCH] libata: Fixup ata_sas_queuecmd to handle __ata_scsi_queuecmd failure
Brian King [Mon, 20 Nov 2006 19:51:56 +0000 (13:51 -0600)]
[PATCH] libata: Fixup ata_sas_queuecmd to handle __ata_scsi_queuecmd failure

Fixes ata_sas_queuecmd to properly handle a failure from
__ata_scsi_queuecmd.

Signed-off-by: Brian King <brking@us.ibm.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] ahci: AHCI mode SATA patch for Intel ICH9
Jason Gaston [Wed, 22 Nov 2006 00:55:58 +0000 (16:55 -0800)]
[PATCH] ahci: AHCI mode SATA patch for Intel ICH9

This patch adds the Intel ICH9 AHCI controller DID's for SATA support.

Signed-off-by: Jason Gaston <jason.d.gaston@intel.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years ago[PATCH] libata: don't schedule EH on wcache on/off if old EH
Tejun Heo [Wed, 22 Nov 2006 03:39:43 +0000 (12:39 +0900)]
[PATCH] libata: don't schedule EH on wcache on/off if old EH

Do not schedule EH for revalidation on wcache on/off if old EH.  Old
EH cannot handle it and will result in WARN_ON()'s and oops.

This closes bug #7412.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Signed-off-by: Jeff Garzik <jeff@garzik.org>
18 years agoselinux: fix dentry_open() error check
Akinobu Mita [Mon, 27 Nov 2006 06:16:48 +0000 (15:16 +0900)]
selinux: fix dentry_open() error check

The return value of dentry_open() shoud be checked by IS_ERR().

Cc: Stephen Smalley <sds@tycho.nsa.gov>
Cc: James Morris <jmorris@namei.org>
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Signed-off-by: James Morris <jmorris@namei.org>
18 years agoFix 'ALIGN()' macro, take 2
Linus Torvalds [Mon, 27 Nov 2006 03:05:22 +0000 (19:05 -0800)]
Fix 'ALIGN()' macro, take 2

You wouldn't think that doing an ALIGN() macro that aligns something up
to a power-of-two boundary would be likely to have bugs, would you?

But hey, in the wonderful world of mixing integer types, you have to be
careful.  This just makes sure that the alignment is interpreted in the
same type as the thing to be aligned.

Thanks to Roland Dreier, who noticed that the amso1100 driver got broken
by the previous fix (that just extended the mask to "unsigned long", but
was still broken in "unsigned long long" - it just happened to be the
same on 64-bit architectures).

See commit 4c8bd7eeee4c8f157fb61fb64b57500990b42e0e for the history of
bugs here...

Acked-by: Roland Dreier <rdreier@cisco.com>
Cc: Andrew Morton <akpm@osdl.org>
Cc: David Miller <davem@davemloft.net>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years ago[PATCH] Fix incorrent type of flags in <asm/semaphore.h>
Kyle McMartin [Sun, 26 Nov 2006 23:56:56 +0000 (18:56 -0500)]
[PATCH] Fix incorrent type of flags in <asm/semaphore.h>

I still think using BUILD_BUG_ON() is unacceptable, especially given how
vague the error message was.

Signed-off-by: Kyle McMartin <kyle@parisc-linux.org>
[ And I already removed gthe BUILD_BUG_ON() in the previous commit ]
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years agoRevert "[PATCH] Enforce "unsigned long flags;" when spinlocking"
Linus Torvalds [Mon, 27 Nov 2006 00:27:17 +0000 (16:27 -0800)]
Revert "[PATCH] Enforce "unsigned long flags;" when spinlocking"

This reverts commit ee3ce191e8eaa4cc15c51a28b34143b36404c4f5, since it
broke on at least ARM, MIPS and PA-RISC due to complicated header file
dependencies.

Conflicts in include/linux/spinlock.h (due to the "nested" variety
fixes) fixed up by hand.

Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Kyle McMartin <kyle@parisc-linux.org>
Cc: Russell King <rmk+lkml@arm.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
18 years agoMerge master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
Linus Torvalds [Sun, 26 Nov 2006 17:42:40 +0000 (09:42 -0800)]
Merge /pub/scm/linux/kernel/git/davem/net-2.6

* master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6:
  [XFRM] STATE: Fix to respond error to get operation if no matching entry exists.
  [NET]: Re-fix of doc-comment in sock.h
  [6PACK]: Masking bug in 6pack driver.
  [NET]: Fix kfifo_alloc() error check.
  [UDP]: Make udp_encap_rcv use pskb_may_pull
  [NETFILTER]: H.323 conntrack: fix crash with CONFIG_IP_NF_CT_ACCT

18 years ago[ARM] Export smp_call_function()
Russell King [Sat, 25 Nov 2006 20:15:12 +0000 (20:15 +0000)]
[ARM] Export smp_call_function()

smp_call_function() will be used with the MP/core oprofile support
patch.  Export it as _GPL.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
18 years ago[ARM] Add PM_LEGACY defaults
Russell King [Sat, 25 Nov 2006 16:19:23 +0000 (16:19 +0000)]
[ARM] Add PM_LEGACY defaults

Eliminate two warnings:

kernel/power/pm.c:205: warning: 'pm_register' is deprecated (declared at kernel/power/pm.c:64)
kernel/power/pm.c:206: warning: 'pm_send_all' is deprecated (declared at kernel/power/pm.c:180)

by updating defconfig files to contain a sensible PM_LEGACY default.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>