Here is another version, with the incremental patch rolled up, and
added reclaim context annotation to kswapd, and allocation tracing
to slab allocators (which may only ever reach the page allocator
in rare cases, so it is good to put annotations here too).

Haven't tested this version as such, but it should be getting closer
to merge worthy ;)

--
After noticing some code in mm/filemap.c accidentally perform a __GFP_FS
allocation when it should not have been, I thought it might be a good idea to
try to catch this kind of thing with lockdep.

I coded up a little idea that seems to work. Unfortunately the system has to
actually be in __GFP_FS page reclaim, then take the lock, before it will mark
it. But at least that might still be some orders of magnitude more common
(and more debuggable) than an actual deadlock condition, so we have some
improvement I hope (the concept is no less complete than discovery of a lock's
interrupt contexts).

I guess we could even do the same thing with __GFP_IO (normal reclaim), and
even GFP_NOIO locks too... but filesystems will have the most locks and fiddly
code paths, so let's start there and see how it goes.

It *seems* to work. I did a quick test.

=================================
[ INFO: inconsistent lock state ]
2.6.28-rc6-00007-ged313489-dirty #26
---------------------------------
inconsistent {in-reclaim-W} -> {ov-reclaim-W} usage.
modprobe/8526 [HC0[0]:SC0[0]:HE1:SE1] takes:
 (testlock){--..}, at: [<ffffffffa0020055>] brd_init+0x55/0x216 [brd]
{in-reclaim-W} state was registered at:
  [<ffffffff80267bdb>] __lock_acquire+0x75b/0x1a60
  [<ffffffff80268f71>] lock_acquire+0x91/0xc0
  [<ffffffff8070f0e1>] mutex_lock_nested+0xb1/0x310
  [<ffffffffa002002b>] brd_init+0x2b/0x216 [brd]
  [<ffffffff8020903b>] _stext+0x3b/0x170
  [<ffffffff80272ebf>] sys_init_module+0xaf/0x1e0
  [<ffffffff8020c3fb>] system_call_fastpath+0x16/0x1b
  [<ffffffffffffffff>] 0xffffffffffffffff
irq event stamp: 3929
hardirqs last  enabled at (3929): [<ffffffff8070f2b5>] mutex_lock_nested+0x285/0x310
hardirqs last disabled at (3928): [<ffffffff8070f089>] mutex_lock_nested+0x59/0x310
softirqs last  enabled at (3732): [<ffffffff8061f623>] sk_filter+0x83/0xe0
softirqs last disabled at (3730): [<ffffffff8061f5b6>] sk_filter+0x16/0xe0

other info that might help us debug this:
1 lock held by modprobe/8526:
 #0:  (testlock){--..}, at: [<ffffffffa0020055>] brd_init+0x55/0x216 [brd]

stack backtrace:
Pid: 8526, comm: modprobe Not tainted 2.6.28-rc6-00007-ged313489-dirty #26
Call Trace:
 [<ffffffff80265483>] print_usage_bug+0x193/0x1d0
 [<ffffffff80266530>] mark_lock+0xaf0/0xca0
 [<ffffffff80266735>] mark_held_locks+0x55/0xc0
 [<ffffffffa0020000>] ? brd_init+0x0/0x216 [brd]
 [<ffffffff802667ca>] trace_reclaim_fs+0x2a/0x60
 [<ffffffff80285005>] __alloc_pages_internal+0x475/0x580
 [<ffffffff8070f29e>] ? mutex_lock_nested+0x26e/0x310
 [<ffffffffa0020000>] ? brd_init+0x0/0x216 [brd]
 [<ffffffffa002006a>] brd_init+0x6a/0x216 [brd]
 [<ffffffffa0020000>] ? brd_init+0x0/0x216 [brd]
 [<ffffffff8020903b>] _stext+0x3b/0x170
 [<ffffffff8070f8b9>] ? mutex_unlock+0x9/0x10
 [<ffffffff8070f83d>] ? __mutex_unlock_slowpath+0x10d/0x180
 [<ffffffff802669ec>] ? trace_hardirqs_on_caller+0x12c/0x190
 [<ffffffff80272ebf>] sys_init_module+0xaf/0x1e0
 [<ffffffff8020c3fb>] system_call_fastpath+0x16/0x1b
Signed-off-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

This modifies the timer code in a way to allow lockdep to detect
deadlocks resulting from a lock being taken in the timer function
as well as around the del_timer_sync() call.
Signed-off-by: Johannes Berg <johannes@sipsolutions.net>

Conflicts:
	fs/btrfs/locking.c

* git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable: (37 commits)
  Btrfs: Make sure dir is non-null before doing S_ISGID checks
  Btrfs: Fix memory leak in cache_drop_leaf_ref
  Btrfs: don't return congestion in write_cache_pages as often
  Btrfs: Only prep for btree deletion balances when nodes are mostly empty
  Btrfs: fix btrfs_unlock_up_safe to walk the entire path
  Btrfs: change btrfs_del_leaf to drop locks earlier
  Btrfs: Change btrfs_truncate_inode_items to stop when it hits the inode
  Btrfs: Don't try to compress pages past i_size
  Btrfs: join the transaction in __btrfs_setxattr
  Btrfs: Handle SGID bit when creating inodes
  Btrfs: Make btrfs_drop_snapshot work in larger and more efficient chunks
  Btrfs: Change btree locking to use explicit blocking points
  Btrfs: hash_lock is no longer needed
  Btrfs: disable leak debugging checks in extent_io.c
  Btrfs: sort references by byte number during btrfs_inc_ref
  Btrfs: async threads should try harder to find work
  Btrfs: selinux support
  Btrfs: make btrfs acls selectable
  Btrfs: Catch missed bios in the async bio submission thread
  Btrfs: fix readdir on 32 bit machines
  ...

The addition of filename encryption caused a regression in unencrypted
filename symlink support.  ecryptfs_copy_filename() is used when dealing
with unencrypted filenames and it reported that the new, copied filename
was a character longer than it should have been.

This caused the return value of readlink() to count the NULL byte of the
symlink target.  Most applications don't care about the extra NULL byte,
but a version control system (bzr) helped in discovering the bug.
Signed-off-by: Tyler Hicks <tyhicks@linux.vnet.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

* 'x86/fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/frob/linux-2.6-roland:
  x86-64: fix int $0x80 -ENOSYS return

One of my past fixes to this code introduced a different new bug.
When using 32-bit "int $0x80" entry for a bogus syscall number,
the return value is not correctly set to -ENOSYS.  This only happens
when neither syscall-audit nor syscall tracing is enabled (i.e., never
seen if auditd ever started).  Test program:

	/* gcc -o int80-badsys -m32 -g int80-badsys.c
	   Run on x86-64 kernel.
	   Note to reproduce the bug you need auditd never to have started.  */

	#include <errno.h>
	#include <stdio.h>

	int
	main (void)
	{
	  long res;
	  asm ("int $0x80" : "=a" (res) : "0" (99999));
	  printf ("bad syscall returns %ld\n", res);
	  return res != -ENOSYS;
	}

The fix makes the int $0x80 path match the sysenter and syscall paths.
Reported-by: Dmitry V. Levin <ldv@altlinux.org>
Signed-off-by: Roland McGrath <roland@redhat.com>

* 'to-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/frob/linux-2.6-roland:
  elf core dump: fix get_user use

The elf_core_dump() code does its work with set_fs(KERNEL_DS) in force,
so vma_dump_size() needs to switch back with set_fs(USER_DS) to safely
use get_user() for a normal user-space address.

Checking for VM_READ optimizes out the case where get_user() would fail
anyway.  The vm_file check here was already superfluous given the control
flow earlier in the function, so that is a cleanup/optimization unrelated
to other changes but an obvious and trivial one.
Reported-by: Gerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: Roland McGrath <roland@redhat.com>

This is a modification of a patch by Bill Pemberton <wfp5p@virginia.edu>

nobh_write_end() could call attach_nobh_buffers() with head == NULL.
This would result in a trap when attach_nobh_buffers() attempted to
access bh->b_this_page.

This can be illustrated by running the writev01 testcase from LTP on jfs.

This error was introduced by commit 5b41e74a "vfs: fix data leak in
nobh_write_end()".  That patch did not take into account that if
PageMappedToDisk() is true upon entry to nobh_write_begin(), then no
buffers will be allocated for the page.  In that case, we won't have to
worry about a failed write leaving unitialized data in the page.

Of course, head != NULL implies !page_has_buffers(page), so no need to
test both.
Signed-off-by: Dave Kleikamp <shaggy@linux.vnet.ibm.com>
Cc: Bill Pemberton <wfp5p@virginia.edu>
Cc: Dmitri Monakhov <dmonakhov@openvz.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
  ALSA: hda - Add missing COEF initialization for ALC887
  ALSA: hda - Add missing initialization for ALC272
  sound: usb-audio: handle wMaxPacketSize for FIXED_ENDPOINT devices
  ALSA: hda - Fix misc workqueue issues
  ALSA: hda - Add quirk for FSC Amilo Xi2550

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6:
  ieee1394: dv1394: move deprecation message from module init to file open
  firewire: core: Remove card from list of cards when enable fails

This fixes the shortlog attribution e.g. for 106757b3Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Acked-by: Sascha Hauer <s.hauer@pengutronix.de>
Acked-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

I created commit 7971db5a on a machine
where I forgot to set user.name and user.email before.  The default
values were not optimal.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Acked-by: Wolfram Sang <w.sang@pengutronix.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

I happened to forked lots of processes, and hit NULL pointer dereference.
It is because in copy_process() after checking max_threads, 0 is returned
but not -EAGAIN.

The bug is introduced by "CRED: Detach the credentials from task_struct"
(commit f1752eec).
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: James Morris <jmorris@namei.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

The S_ISGID check in btrfs_new_inode caused an oops during subvol creation
because sometimes the dir is null.
Signed-off-by: Chris Mason <chris.mason@oracle.com>

* 'for-linus' of git://neil.brown.name/md:
  md: Ensure an md array never has too many devices.
  md: Fix a bug in linear.c causing which_dev() to return the wrong device.
  md: Allow read error in a single drive raid1 to be passed up.

On many Linux installations, the dv1394 driver will be auto-loaded
whenever an AV/C device (e.g. camcorder or audio device) is plugged in.
An irritating message would then appear in the kernel log.

Defer this message to until a dv1394 character device file is actually
used by a program.  Also include the program name in the message and
update the message slightly.
Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de>

Signed-off-by: Takashi Iwai <tiwai@suse.de>

ALC272 needs EAPD for speaker outputs as well as other similar ALC
codecs.
Signed-off-by: Takashi Iwai <tiwai@suse.de>

For audio devices that do not have proper audio descriptors (e.g.,
Edirol UA-20), we use hardcoded parameters from our quirks list.
However, we must still read the maximum packet size from the standard
endpoint descriptor; otherwise, we might use packets that are too big
and therefore rejected by the USB core.
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Cc: <stable@kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

Each different metadata format supported by md supports a
different maximum number of devices.
We really should be enforcing this maximum in the kernel, but
we aren't quite doing that properly.

We currently only enforce it at the 'hot_add' point, which is an
older interface which is not used by current userspace.

We need to also enforce it at 'add_new_disk' time for active arrays
and at 'do_md_run' time when starting a new array.

So move the test from 'hot_add' into 'bind_rdev_to_array' which is
called from both 'hot_add' and 'add_new_disk, and add a new
test in 'analyse_sbs' which is called from 'do_md_run'.

This bug (or missing feature) has been around "forever" and so
the patch is suitable for any -stable that is currently maintained.

Cc: stable@kernel.org
Signed-off-by: NeilBrown <neilb@suse.de>

ab5bd5cb introduced the following
bug in linear software raid for large arrays on 32 bit machines:

which_dev() computes the device holding a given sector by shifting
down the sector number to a 32 bit range, dividing by the array
spacing and looking up the resulting index in the hash table of
the array.

Because the computed index might be slightly too small, a loop at
the end of which_dev() increases the index until the given sector
actually falls into the range of the device associated with that index.

The changes of the above mentioned commit caused this loop to check
whether the _index_ rather than the sector number is small enough,
effectively bypassing the loop and thus possibly returning the wrong
device.

As reported by Simon Kirby, this leads to errors such as

	linear_make_request: Sector 2340486136 out of bounds on dev sdi: 156301312 sectors, offset 2109870464

Fix this bug by introducing a local variable for the index so that
the variable containing the passed sector is left unchanged.

Cc: stable@kernel.org
Signed-off-by: Andre Noll <maan@systemlinux.org>
Signed-off-by: NeilBrown <neilb@suse.de>

If a raid1 only has a single working device and gets a read error, 
we choose to simply return that error up to the filesystem (or whatever)
rather than failing the whole array.

However the codes doesn't quite do that.  We attempt a readbalance
which allocates the same drive, so we retry the read - indefinitely. 

Instead:  If read_balance in the error case chooses the same drive that just
failed, treat it as a failure and don't retry.
Signed-off-by: NeilBrown <neilb@suse.de>

Prevent kprobes from catching spurious faults which will cause infinite
recursive page-fault and memory corruption by stack overflow.
Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com>
Cc: <stable@kernel.org>		[2.6.28.x]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

... and yes, gcc is insane enough to eat that without complaint.
We probably want sparse to scream on those...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mfasheh/ocfs2:
  Revert "configfs: Silence lockdep on mkdir(), rmdir() and configfs_depend_item()"

* 'sh/for-2.6.29' of git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6:
  sh: Fix up T-bit error handling in SH-4A mutex fastpath.
  sh: Fix up spurious syscall restarting.
  sh: fcnvds fix with denormalized numbers on SH-4 FPU.
  sh: Only reserve memory under CONFIG_ZERO_PAGE_OFFSET when it != 0.
  sh: Handle calling csum_partial with misaligned data
  sh: ap325rxa: Enable ov772x in defconfig.
  sh: ap325rxa: Add ov772x support.
  sh: ap325rxa: control camera power toggling.
  sh: mach-migor: Enable ov772x and tw9910 in defconfig.

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6:
  Revert "tcp: Always set urgent pointer if it's beyond snd_nxt"
  ipv6: Copy cork options in ip6_append_data
  udp: Fix UDP short packet false positive
  gianfar: Fix potential soft reset race
  gianfar: Fix BD_LENGTH_MASK definition
  cxgb3: Fix lro switch
  iwlwifi: save PCI state before suspend, restore after resume
  iwlwifi: clean key table in iwl_clear_stations_table

This reverts commit 64ff3b93.

Jeff Chua reports that it breaks rlogin for him.
Signed-off-by: David S. Miller <davem@davemloft.net>

As the options passed to ip6_append_data may be ephemeral, we need
to duplicate it for corking.  This patch applies the simplest fix
which is to memdup all the relevant bits.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

The UDP header pointer assignment must happen after calling
pskb_may_pull().  As pskb_may_pull() can potentially alter the SKB
buffer.

This was exposted by running multicast traffic through the NIU driver,
as it won't prepull the protocol headers into the linear area on
receive.
Signed-off-by: Jesper Dangaard Brouer <hawk@comx.dk>
Signed-off-by: David S. Miller <davem@davemloft.net>

lseek() further than length of the file will leave stale ->index
(second-to-last during iteration). Next seq_read() will not notice
that ->f_pos is big enough to return 0, but will print last item
as if ->f_pos is pointing to it.

Introduced in commit cb510b81
aka "seq_file: more atomicity in traverse()".
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

In 2.6.25 some /proc files were converted to use the seq_file
infrastructure.  But seq_files do not correctly support pread(), which
broke some usersapce applications.

To handle pread correctly we can't assume that f_pos is where we left it
in seq_read.  So move traverse() so that we can eventually use it in
seq_read and do thus some day support pread().
Signed-off-by: Eric Biederman <ebiederm@xmission.com>
Cc: Paul Turner <pjt@google.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

A missing type cast results in writing way beyond the end of a kzalloc()'d
memory segment resulting in slab corruption. But it seems like the better
solution is to define ->recv_msg_slots as a 'void *' rather than a
'struct xpc_notify_mq_msg_uv *' and add the type cast.
Signed-off-by: Dean Nelson <dcn@sgi.com>
Cc: <stable@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Do usual do {} while (0) dance, otherwise

fs/gfs2/util.c:99: error: expected expression before 'else'
drivers/scsi/lpfc/lpfc_sli.c:363: error: expected expression before 'else'
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Acked-by: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Richard Henderson <rth@twiddle.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

If we return directly with -EPERM then lock_kernel() is still held.

This was found with a code checker (http://repo.or.cz/w/smatch.git/).

[akpm@linux-foundation.org: fix another such path - missed func_exit()]
Signed-off-by: Dan Carpenter <error27@gmail.com>
Cc: <R.E.Wolff@BitWizard.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>