Patch from Dmitry Mishin <dim@openvz.org>:

Replace add_timer() by mod_timer() in dst_run_gc
in order to avoid BUG message.

       CPU1                            CPU2
dst_run_gc()  entered           dst_run_gc() entered
spin_lock(&dst_lock)                   .....
del_timer(&dst_gc_timer)         fail to get lock
       ....                         mod_timer() <--- puts 
                                                 timer back
                                                 to the list
add_timer(&dst_gc_timer) <--- BUG because timer is in list already.

Found during OpenVZ internal testing.

At first we thought that it is OpenVZ specific as we
added dst_run_gc(0) call in dst_dev_event(),
but as Alexey pointed to me it is possible to trigger
this condition in mainstream kernel.

F.e. timer has fired on CPU2, but the handler was preeempted
by an irq before dst_lock is tried.
Meanwhile, someone on CPU1 adds an entry to gc list and
starts the timer.
If CPU2 was preempted long enough, this timer can expire
simultaneously with resuming timer handler on CPU1, arriving
exactly to the situation described.
Signed-off-by: NDmitry Mishin <dim@openvz.org>
Signed-off-by: NKirill Korotaev <dev@openvz.org>
Signed-off-by: NAlexey Kuznetsov <kuznet@ms2.inr.ac.ru>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Need to check some more cases in IPX receive.  If the skb is purely
fragments, the IPX header needs to be extracted. The function
pskb_may_pull() may in theory invalidate all the pointers in the skb,
so references to ipx header must be refreshed.
Signed-off-by: NStephen Hemminger <shemminger@osdl.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The ->set_mac_address handlers expect a pointer to a
sockaddr which contains the MAC address, whereas
IFLA_ADDRESS provides just the MAC address itself.

So whip up a sockaddr to wrap around the netlink
attribute for the ->set_mac_address call.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix a subtle race condition between tg3_start_xmit() and tg3_tx()
discovered by Herbert Xu <herbert@gondor.apana.org.au>:

CPU0					CPU1
tg3_start_xmit()
	if (tx_ring_full) {
		tx_lock
					tg3_tx()
						if (!netif_queue_stopped)
		netif_stop_queue()
		if (!tx_ring_full)
						update_tx_ring 
			netif_wake_queue()
		tx_unlock
	}

Even though tx_ring is updated before the if statement in tg3_tx() in
program order, it can be re-ordered by the CPU as shown above.  This
scenario can cause the tx queue to be stopped forever if tg3_tx() has
just freed up the entire tx_ring.  The possibility of this happening
should be very rare though.

The following changes are made:

1. Add memory barrier to fix the above race condition.

2. Eliminate the private tx_lock altogether and rely solely on
netif_tx_lock.  This eliminates one spinlock in tg3_start_xmit()
when the ring is full.

3. Because of 2, use netif_tx_lock in tg3_tx() before calling
netif_wake_queue().

4. Change TX_BUFFS_AVAIL to an inline function with a memory barrier.
Herbert and David suggested using the memory barrier instead of
volatile.

5. Check for the full wake queue condition before getting
netif_tx_lock in tg3_tx().  This reduces the number of unnecessary
spinlocks when the tx ring is full in a steady-state condition.

6. Update version to 3.65.
Signed-off-by: NMichael Chan <mchan@broadcom.com>
Acked-by: NHerbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Do not count retransmitted segments.
Signed-off-by: NWei Yongjun <yjwei@nanjing-fnst.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Mirror the bug fix from fill_packet_ipv4()
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NChen-Li Tien <cltien@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

From: Kirill Korotaev <dev@sw.ru>

During OpenVZ stress testing we found that UDP traffic with random src
can generate too much excessive rt hash growing leading finally to OOM
and kernel panics.

It was found that for 4GB i686 system (having 1048576 total pages and
  225280 normal zone pages) kernel allocates the following route hash:
syslog: IP route cache hash table entries: 262144 (order: 8, 1048576
bytes) => ip_rt_max_size = 4194304 entries, i.e.  max rt size is
4194304 * 256b = 1Gb of RAM > normal_zone

Attached the patch which removes HASH_HIGHMEM flag from
alloc_large_system_hash() call.
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch will linearize and check there is enough data.
It handles the pprop case as well as avoiding a whole audit of
the routing code.
Signed-off-by: NStephen Hemminger <shemminger@osdl.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

All caller of netdev_alloc_skb need to assign skb->dev shortly
afterwards.  Move it into common code.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

* 'for-linus' of master.kernel.org:/pub/scm/linux/kernel/git/dtor/input: (24 commits)
  Input: ati_remote - use msec instead of jiffies
  Input: ati_remote - add missing input_sync()
  Input: ati_remote - relax permissions sysfs module parameters
  Input: ati_remote - make filter time a module parameter
  Input: atkbd - restore repeat rate when resuming
  Input: trackpoint - activate protocol when resuming
  Input: logips2pp - fix button mapping for MX300
  Input: keyboard - change to use kzalloc
  Input: serio/gameport - check whether driver core calls succeeded
  Input: spaceball - make 4000FLX Lefty work
  Input: keyboard - simplify emulate_raw() implementation
  Input: keyboard - remove static variable and clean up initialization
  Input: hiddev - use standard list implementation
  Input: add missing handler->start() call
  Input: HID - fix potential out-of-bound array access
  Input: fix list iteration in input_release_device()
  Input: iforce - add Trust Force Feedback Race Master support
  Input: iforce - check array bounds before accessing elements
  Input: libps2 - warn instead of oopsing when passed bad arguments
  Input: fm801-gp - fix use after free
  ...

* master.kernel.org:/pub/scm/linux/kernel/git/perex/alsa:
  [ALSA] Don't reject O_RDWR at opening PCM OSS with read/write-only device
  [ALSA] snd-emu10k1: Implement support for Audigy 2 ZS [SB0353]
  [ALSA] add MAINTAINERS entry for snd-aoa
  [ALSA] aoa: platform function gpio: ignore errors from functions that don't exist
  [ALSA] make snd-powermac load even when it can't bind the device
  [ALSA] aoa: fix toonie codec
  [ALSA] aoa: feature gpio layer: fix IRQ access
  [ALSA] Conversions from kmalloc+memset to k(z|c)alloc
  [ALSA] snd-emu10k1: Fixes ALSA bug#2190

* master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6:
  [LAPB]: Fix windowsize check
  [TCP]: Fixes IW > 2 cases when TCP is application limited
  [PKT_SCHED] RED: Fix overflow in calculation of queue average
  [LLX]: SOCK_DGRAM interface fixes
  [PKT_SCHED]: Return ENOENT if qdisc module is unavailable
  [BRIDGE]: netlink status fix

While busy-waiting for completion, check the hardware after scheduling;
don't schedule and then immediately check the _timeout_.  If the yield()
took a long time (as it does on my OLPC prototype board when it's busy),
we'd report a timeout even though the hardware was now ready.

This fixes it, and also switches the yield() for a cond_resched() because
we don't actually want to be _that_ nice about it.  I see nice
tightly-packed SMBus transactions now, rather than waiting for milliseconds
between successive phases.

Actually, we shouldn't be busy-waiting here at all.  We should be using
interrupts.  That's an exercise for another day though.
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>
Cc: Christer Weinigel <wingel@nano-system.com>
Cc: <Jordan.Crouse@amd.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

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>

The recent fixups in futex.c need to be applied to futex_compat.c too.  Fixes
a hang reported by Olaf.
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Cc: Olaf Hering <olh@suse.de>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

A patch in -mm kernel correct the parsing of "address resources" of pnpacpi.
Before we assumed it was memory only, but it could be also IO.

But this change show an hidden bug : some resources could be producer type
that are not handled by pnp layer.  So we should ignore the producer
resources.

This patch fixes bug 6292 (http://bugzilla.kernel.org/show_bug.cgi?id=6292).
Some devices like PNP0A03 have 0xd00-0xffff and 0x0-0xcf7 as IO producer
resources.

Before correcting "address resources" parsing, it was seen as memory and was
harmless, because nobody tried to reserve this memory range as it should be
IO.

With the correction it become IO resources, and make failed all others device
that want to register IO in this range and use pnp layer (like a ISA sound
card).

The solution is to ignore producer resources
Signed-off-by: NMatthieu CASTET <castet.matthieu@free.fr>
Signed-off-by: NUwe Bugla <uwe.bugla@gmx.de>
Cc: Bjorn Helgaas <bjorn.helgaas@hp.com>
Cc: Adam Belay <ambx1@neo.rr.com>
Cc: "Brown, Len" <len.brown@intel.com>
Acked-by: NShaohua Li <shaohua.li@intel.com>
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>

This patch is for collision check enhancement for memory hot add.

It's better to do resouce collision check before doing memory hot add,
which will touch memory management structures.

And add_section() should check section exists or not before calling
sparse_add_one_section(). (sparse_add_one_section() will do another
check anyway. but checking in memory_hotplug.c will be easy to understand.)
Signed-off-by: NKAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: keith mannthey <kmannth@us.ibm.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

both of acpi_memory_enable_device() and acpi_memory_add_device() may evaluate
_CRS method.

We should avoid evaluate device's resource twice if we could get it
successfully in past.
Signed-off-by: NKAMEZWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Keith Mannthey <kmannth@gmail.com>
Cc: Yasunori Goto <y-goto@jp.fujitsu.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

add_memory() does all necessary check to avoid collision.  then, acpi layer
doesn't have to check region by itself.

(*) pfn_valid() just returns page struct is valid or not. It returns 0
    if a section has been already added even is ioresource is not added.
    ioresource collision check in mm/memory_hotplug.c can do more precise
    collistion check.
    added enabled bit check just for sanity check..
Signed-off-by: NKAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Keith Mannthey <kmannth@gmail.com>
Cc: Yasunori Goto <y-goto@jp.fujitsu.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: "Brown, Len" <len.brown@intel.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

find_next_system_ram() is used to find available memory resource at onlining
newly added memory.  This patch fixes following problem.

find_next_system_ram() cannot catch this case.

Resource:      (start)-------------(end)
Section :                (start)-------------(end)
Signed-off-by: NKAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Keith Mannthey <kmannth@gmail.com>
Cc: Yasunori Goto <y-goto@jp.fujitsu.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

find_next_system_ram() returns valid memory range which meets requested area,
only used by memory-hot-add.

This function always rewrite requested resource even if returned area is not
fully fit in requested one.  And sometimes the returnd resource is larger than
requested area.  This annoyes the caller.  This patch changes the returned
value to fit in requested area.
Signed-off-by: NKAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Keith Mannthey <kmannth@gmail.com>
Cc: Yasunori Goto <y-goto@jp.fujitsu.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

ioresouce handling code in memory hotplug allows not-aligned memory hot add.
But when memmap and other memory structures are initialized, parameters should
be aligned.  (if not aligned, initialization of mem_map will do wrong, it
assumes parameters are aligned.) This patch fix it.

And this patch allows ioresource collision check to handle -EEXIST.
Signed-off-by: NKAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Keith Mannthey <kmannth@gmail.com>
Cc: Yasunori Goto <y-goto@jp.fujitsu.com>
Cc: Dave Hansen <haveblue@us.ibm.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>

At least Maxtor OneTouch III require a "start stop unit" command after auto
spin-down before the next access can proceed.  This patch activates the
responsible code in scsi_mod for all Maxtor SBP-2 disks.
https://bugzilla.novell.com/show_bug.cgi?id=183011

Maybe that should be done for all SBP-2 disks, but better be cautious.
Signed-off-by: NStefan Richter <stefanr@s5r6.in-berlin.de>
Cc: Jody McIntyre <scjody@modernduck.com>
Cc: Ben Collins <bcollins@ubuntu.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

While helping someone to submit a patch to the stable branch, I noticed
that the stable branch is not listed in the MAINTAINERS file.  This was
after I went there to look for the email addresses for the stable branch
list (stable@kernel.org).

This patch adds the stable branch to the maintainers file so that people
can find where to send patches when they have a fix for the stable team.
Signed-off-by: NSteven Rostedt <rostedt@goodmis.org>
Signed-off-by: NChris Wright <chrisw@sous-sol.org>
Cc: Greg KH <greg@kroah.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Clean up proc file removal in sq module for superh arch.  currently on a
failed module load or on module unload a proc file is left registered which
can cause a random memory execution or oopses if read after unload.  This
patch cleans up that deregistration.
Signed-off-by: NNeil Horman <nhorman@tuxdriver.com>
Acked-by: NPaul Mundt <lethal@linux-sh.org>
Cc: Kazumoto Kojima <kkojima@rr.iij4u.or.jp>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

* MODE_MASK is unused in eicon driver.
* Conflicts with a ptrace stuff on arm.

drivers/isdn/hardware/eicon/divasync.h:259:1: warning: "MODE_MASK" redefined
include2/asm/ptrace.h:48:1: warning: this is the location of the previous definition
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Acked-by: NKarsten Keil <kkeil@suse.de>
Acked-by: NArmin Schindler <armin@melware.de>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

A set of tty line discipline cleanup patches were introduced before the
dawn of time, in kernel version 2.4.21.  This patch performs that cleanup
for the hvsi driver.

The hvsi driver is used only on IBM pSeries PowerPC boxes.  The driver was
originally written by Hollis Blanchard, who has delegated maintainership to
me.  So this my first and maybe only patch in this official new role,
because this driver is otherwise bug-free :-)

Alan: "Actually its also a bug fix, tty->ldisc should be locked by refcounting
and the helpers do this for you."
Signed-off-by: NLinas Vepstas <linas@austin.ibm.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Acked-by: NAlan Cox <alan@redhat.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Under certain rare circumstances, it appears that there can be be a
NULL-pointer deref when a user fiddles with terminal emeulation programs while
outpu is being sent to the console.  This patch checks for and avoids a
NULL-pointer deref.
Signed-off-by: NHollis Blanchard <hollisbl@austin.ibm.com>
Signed-off-by: NLinas Vepstas <linas@austin.ibm.com>
Cc: Paul Fulghum <paulkf@microgate.com>
Acked-by: NAlan Cox <alan@lxorguk.ukuu.org.uk>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Cc: Andi Kleen <ak@muc.de>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Signed-off-by: NMaxime Bizon <mbizon@freebox.fr>
Acked-by: NSimon Horman <horms@verge.net.au>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

If we don't find the item we are lookng for, we allocate a new one, and
then grab the lock again and search to see if it has been added while we
did the alloc.  If it had been added we need to 'cache_put' the newly
created item that we are never going to use.  But as it hasn't been
initialised properly, putting it can cause an oops.

So move the ->init call earlier to that it will always be fully initilised
if we have to put it.

Thanks to Philipp Matthias Hahn <pmhahn@svs.Informatik.Uni-Oldenburg.de>
for reporting the problem.
Signed-off-by: NNeil Brown <neilb@suse.de>
Cc: <stable@kernel.org>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

Removes many, many "declared inside parameter list" warnings on parisc.
Signed-off-by: NAlexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>

The POSIX_FADV_NOREUSE hint means "the application will use this range of the
file a single time".  It seems to be intended that the implementation will use
this hint to perform drop-behind of that part of the file when the application
gets around to reading or writing it.

However for reasons which aren't obvious (or sane?) I mapped
POSIX_FADV_NOREUSE onto POSIX_FADV_WILLNEED.  ie: it does readahead.

That's daft.  So for now, make POSIX_FADV_NOREUSE a no-op.

This is a non-back-compatible change.  If someone was using POSIX_FADV_NOREUSE
to perform readahead, they lose.  The likelihood is low.

If/when we later implement POSIX_FADV_NOREUSE things will get interesting - to
do it fully we'll need to maintain file offset/length ranges and peform all
sorts of complex tricks, and managing the lifetime of those ranges' data
structures will be interesting..

A sensible implementation would probably ignore the file range and would
simply mark the entire file as needing some form of drop-behind treatment.

Cc: Michael Kerrisk <mtk-manpages@gmx.net>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NLinus Torvalds <torvalds@osdl.org>