Merge branch 'upstream-fixes'

Merge branch 'master'

Merge branch 'release' of git://git./linux/kernel/git/aegl/linux-2.6

[IPV6]: Address autoconfiguration does not work after device down/up cycle

If you set network interface down and up again, the IPv6 address
autoconfiguration does not work. 'ip addr' shows that the link-local
address is in tentative state. We don't even react to periodical router
advertisements.

During NETDEV_DOWN we clear IF_READY, and we don't set it back in
NETDEV_UP. While starting to perform DAD on the link-local address, we
notice that the device is not in IF_READY, and we abort autoconfiguration
process (which would eventually send router solicitations).

Acked-by: Juha-Matti Tapio <jmtapio@verkkotelakka.net>
Acked-by: YOSHIFUJI Hideaki <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

[PATCH] x86-64: Add sys_unshare

Add unshare syscall for x86-64

ppoll/pselect are not ready yet, but add reservations.

Signed-off-by: Andi Kleen <ak@suse.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Merge master.kernel.org:/home/rmk/linux-2.6-mmc

Merge master.kernel.org:/home/rmk/linux-2.6-serial

Merge master.kernel.org:/home/rmk/linux-2.6-arm

Manual conflict merge of arch/arm/Kconfig

[MMC] Remove extra character in AU1XXX MMC Kconfig entry

An obvious vi fat finger on my part.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[ARM] 3278/1: OMAP: 3/3 Fix low-level io init for omap2 boards

Patch from Tony Lindgren

This patch fixes the low-level IO init for omap2 boards.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[ARM] 3280/1: OMAP: 2/3 Fix low-level io init for omap1 boards

Patch from Tony Lindgren

This patch fixes the low-level IO init for omap1 boards.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[ARM] 3279/1: OMAP: 1/3 Fix low-level io init

Patch from Tony Lindgren

This patch adds the missing cache flushes to common low-level
init that are needed to access the IO region. These flushes
are normally done at the end of devicemaps_init(), but we
need to detect the OMAP core type early.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[ARM] 3299/1: S3C24XX - fix irq range on adc device

Patch from Ben Dooks

Change the IRQ resource range for the ADC device
to be two distinct IRQs

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

[SERIAL] 8250_pci: add new PCI serial card support

This patch adds new PCI serial card support.

Signed-off-by: Yoichi Yuasa <yoichi_yuasa@tripeaks.co.jp>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[SERIAL] ip22zilog: Whitespace cleanup.

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

[SERIAL] PPC32 CPM_UART: update to utilize the new TTY flip API

This replaces old direct usage of tty->flip stuff with relative flip API
calls.

Signed-off-by: Vitaly Bordug <vbordug@ru.mvista.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[SERIAL] Fix compile error in 8250_au1x00.c

The DB1550 actually doesn't have a UART2.  Remove it from the list.

Signed-off-by: Jordan Crouse <jordan.crouse@amd.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[SERIAL] 8250 serial console update uart_8250_port ier

On some embedded PowerPC (MPC834x) systems an extra byte would some
times be required to flush data out of the fifo.
serial8250_console_write() was updating the IER in hardware without
also updating the copy in uart_8250_port. This causes issues functions
like serial8250_start_tx() and __stop_tx() to misbehave.

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

[ARM] 3283/1: S3C2400 - defines the number of serial ports

Patch from Lucas Correia Villa Real

This patch defines the number of serial ports on the S3C2400.

Signed-off-by: Lucas Correia Villa Real <lucasvr@gobolinux.org>
Signed-off-by: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[ARM] 3313/1: Use OSC4 instead of OSC1 for CLCD

Patch from Catalin Marinas

Because of a type, OSC1 was used for setting the display clock instead of
OSC4. This patch fixes it.

Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[ARM] 3311/1: clean up include/asm-arm/mutex.h

Patch from Nicolas Pitre

Since:

if (unlikely(__res || __ex_flag))

produces worse code on ARM than:

if (unlikely(__res | __ex_flag))

I therefore made it more explicit:

__res |= __ex_flag;
if (unlikely(__res != 0))

so it is not seen as a typo again.

Also made everything static inline rather than macros for better readability
(both produce the same code after all).

And finally added missing \t from multi-line assembly code.

Signed-off-by: Nicolas Pitre <nico@cam.org>
Acked-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[ARM] 3310/1: add a comment about the possible __kuser_cmpxchg transient false
 negative

Patch from Nicolas Pitre

The pre ARMv5 implementation can be aborted if an exception occurs in
the middle of it.  Because of that, the ARMv6 implementation doesn't
re-attempt the operation on a failed strex either.  Let's make this
transient nature of such a false positive more explicit in the
definition.

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

[ARM] 3309/1: disable the pre-ARMv5 NPTL kernel helper in the non MMU case

Patch from Nicolas Pitre

The cmpxchg emulation on pre-ARMv5 relies on user code executed from a
kernel address.  If the operation cannot complete atomically, it is
aborted from the usr_entry macro by clearing the Z flag.  This clearing
of the Z flag is done whenever the user pc is above TASK_SIZE.

However this "pc >= TASK_SIZE" test cannot work in the non MMU case.
Worse: the current code will corrupt the Z flag on every entry to the
kernel.

Let's disable it in the non MMU case for now.  Using NPTL on non MMU
targets needs to be worked out anyway.

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

[ARM] 3308/1: old ABI compat: struct sockaddr_un

Patch from Nicolas Pitre

struct sockaddr_un loses its padding with EABI.  Since the size of the
structure is used as a validation test in unix_mkname(), we need to
change the length argument to 110 whenever it is 112.

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

[ARM] Experimental config options should have (EXPERIMENTAL)

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

[ARM] 3307/1: old ABI compat: mark it experimental

Patch from Nicolas Pitre

Although OABI_COMPAT works fine in most cases, it is still experimental
and could be for ever since it is nearly impossible to handle
everything, e.g. ioctls.

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

[ARM] 3306/1: S3C24XX - update defconfig

Patch from Ben Dooks

Bring s3c2410 defconfig up to date

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

[ARM] 3305/1: Minor typographical and spelling fixes in Konfig

Patch from Martin Michlmayr

Minor typographical and spelling fixes in Konfig

Signed-off-by: Martin Michlmayr <tbm@cyrius.com>
---

 Kconfig |    8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[ARM] 3304/1: Add help descriptions to ARCH config items that don't have one

Patch from Martin Michlmayr

Add help descriptions to ARCH config items that don't have one.

Signed-off-by: Martin Michlmayr <tbm@cyrius.com>
---

 Kconfig               |   32 ++++++++++++++++++++++++++++++--
 mach-clps711x/Kconfig |    2 ++
 2 files changed, 32 insertions(+), 2 deletions(-)
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[ARM] 3303/1: S3C24XX - add clock enable usage counting

Patch from Ben Dooks

Move to using an enable count for the shared clocks
and protect the clock system using a mutex instead
of just disabling IRQs during the clock update.

Since there is little more code in the path for
non-shared clocks, the enable and disable calls
use the same code for each.

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

[ARM] 3302/1: make pci=firmware the default for ixp2000

Patch from Lennert Buytenhek

Most ixp2000 boards don't actually work if pci=firmware isn't used, so
the defconfig isn't really the right place to specify this.  Instead of
specifying it in the defconfigs, make the relevant board code take care
of setting pci=firmware.

Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Deepak Saxena <dsaxena@plexity.net>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[ARM] 3301/1: remove unnecessary clock default from ixdp2801 defconfig

Patch from Lennert Buytenhek

The ixdp2x01_clock is already 50MHz by default, so no need to
override it with 50MHz in the ixdp2801 defconfig as is done now,
which is confusing as well.

Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Deepak Saxena <dsaxena@plexity.net>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[ARM] 3300/1: make ixdp2x01 co-exist with other ixp2000 machine types

Patch from Lennert Buytenhek

The ixdp2x01 pci init call doesn't check whether it's really running
on an ixdp2x01, making it impossible to compile one kernel that works
on both the ixdp2x01 and another ixp2000 board.

Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Deepak Saxena <dsaxena@plexity.net>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

[IA64] prevent sn2 specific code to be run in generic kernels

Prevent SN2 specific code to be executed on non SN2 platforms when
running a generic kernel.

Signed-off-by: Jes Sorensen <jes@sgi.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>

Merge branch 'upstream' of git://ftp.linux-mips.org/upstream-linus

[MIPS] Add 'const' to readb and friends

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---

[MIPS] Sparse: Add __user tags to syscall.c

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---

[MIPS] Wire up new syscalls.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---

[MIPS] Make do_signal return void.

It's return value is ignored everywhere.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---

[MIPS] Add support for TIF_RESTORE_SIGMASK.

Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
---

Merge branch 'for-linus2' of git://git./linux/kernel/git/viro/bird

Merge branch 'for-linus' of git://brick.kernel.dk/data/git/linux-2.6-block

[PATCH] alpha: set cpu_possible_map much earlier

All the percpu data structure walkers want cpu_possible_map to be
initialized early, but alpha instead populated "hwrpb_cpu_present_mask"
early in setup_smp(), and then initialized cpu_possible_map only much
later.

Thanks go to Heiko Carstens and Dipankar Sarma for noticing.

This fixes it and we can get rid of hwrpb_cpu_present_mask entirely.

Signed-off-by: Linus Torvalds <torvalds@osdl.org>

[PATCH] SLOB=y && SMP=y fix

fix CONFIG_SLOB=y (when CONFIG_SMP=y): get rid of the 'align' parameter
from its __alloc_percpu() implementation. Boot-tested on x86.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

[PATCH] block: implement elv_insert and use it (fix ordcolor flipping bug)

q->ordcolor must only be flipped on initial queueing of a hardbarrier
request.

Constructing ordered sequence and requeueing used to pass through
__elv_add_request() which flips q->ordcolor when it sees a barrier
request.

This patch separates out elv_insert() from __elv_add_request() and uses
elv_insert() when constructing ordered sequence and requeueing.
elv_insert() inserts the given request at the specified position and
does nothing else.

Signed-off-by: Tejun Heo <htejun@gmail.com>
Acked-by: Jens Axboe <axboe@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

ppc: fix up trivial Kconfig config selection

Quoth BenH:
  "Ok, looks like I forgot to update the Kconfig for the new i2c driver,
   it should select I2C_POWERMAC instead.  Do you want a new patch or
   can you just fix it there ?"

Signed-off-by: Linus Torvalds <torvalds@osdl.org>

[PATCH] blk: Fix SG_IO ioctl failure retry looping

When issuing an SG_IO ioctl through sd that resulted in an unrecoverable
error, a nearly infinite retry loop was discovered. This is due to the
fact that the block layer SG_IO code is not setting up rq->retries. This
patch also fixes up the sg_scsi_ioctl path.

Signed-off-by: Brian King <brking@us.ibm.com>
Signed-off-by: Jens Axboe <axboe@suse.de>

[PATCH] sh: lvalues abuse in arch/sh/boards/renesas/rts7751r2d/io.c

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

[PATCH] m32r_sio iomem annotations

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

[PATCH] mark HISAX_AMD7930 as broken

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

[PATCH] type-safe min() in prism54

we do min() on u8 and small integer constant; cast the latter to u8.

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

[PATCH] powerpc: Thermal control for dual core G5s

This patch adds a windfarm module, windfarm_pm112, for the dual core G5s
(both 2 and 4 core models), keeping the machine from getting into
vacuum-cleaner mode ;) For proper credits, the patch was initially
written by Paul Mackerras, and slightly reworked by me to add overtemp
handling among others. The patch also removes the sysfs attributes from
windfarm_pm81 and windfarm_pm91 and instead adds code to the windfarm
core to automagically expose attributes for sensor & controls.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

[PATCH] s390 __get_user() bogus warnings removal

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

[PATCH] synclink_gt is PCI-only

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

[PATCH] __user annotations in powerpc thread_info

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

[PATCH] fix iomem annotations in dart_iommu

it's int __iomem *, not int * __iomem...

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

[PATCH] s390 misc __user annotations

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

[PATCH] compat_ioctl __user annotations

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

[PATCH] sn3 iomem annotations and fixes

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

[PATCH] powerpc signal __user annotations

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

[PATCH] drivers/media/video __user annotations and fixes

* compat_alloc_user_space() returns __user pointer
* copying between two userland areas is copy_in_user(), not copy_from_user()
* dereferencing userland pointers is bad
* so's get_user() from local variables

... plus usual __user annotations

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

[PATCH] powermac pci iomem annotations

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

[PATCH] s390x compat __user annotations

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

[PATCH] mips: namespace pollution - mem_... -> __mem_... in io.h

A pile of internal functions use only inside mips io.h has names starting
with mem_... and clashing with names in drivers; renamed to __mem_....

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

[PATCH] arm: fix dependencies for MTD_XIP

MTD_XIP depends on having working asm/mtd-xip.h; it's not just per-architecture
(arm-only, as current Kconfig would have it), but actually per-subarch as
well.  Introduced a new symbol (ARCH_MTD_XIP) set by arch Kconfig; MTD_XIP
depends on it.

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

Merge /pub/scm/linux/kernel/git/davem/sparc-2.6

[PATCH] powerpc: Fix sound driver use of i2c

The PowerMac sound drivers used to rely on a "bug" of the i2c-keywest
driver that implemented I2C_SMBUS_BLOCK_DATA incorrectly, that is it did
what I2C_SMBUS_I2C_BLOCK_DATA should have done. The new i2c-powermac
driver that replaces keywest has this bug fixed, thus the sound drivers
must be fixed too.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Merge git://git./linux/kernel/git/paulus/powerpc-merge

[SPARC64]: Update defconfig.

Do not enable CONFIG_LOCALVERSION_AUTO by default.
When doing kernel development it just leaves a ton
of crap around.

Signed-off-by: David S. Miller <davem@davemloft.net>

[SPARC]: Wire up sys_unshare().

Also, the Solaris syscall table is sized differrently,
and does not go beyond entry 255, so trim off the excess
entries.

Signed-off-by: David S. Miller <davem@davemloft.net>

[PATCH] umount_tree() decrements mount count on wrong dentry

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

[PATCH] nfsroot port= parameter fix [backport of 2.4 fix]

Direct backport of 2.4 fix that didn't get propagated to 2.6; original
comment follows:
<quote>
   When I specify the NFS port for nfsroot (e.g.,
   nfsroot=<dir>,port=2049), the
   kernel uses the wrong port. In my case it tries to use 264 (0x108)
   instead
   of 2049 (0x801).

   This patch adds the missing htons().

   Eric
</quote>

Patch got applied in 2.4.21-pre6.  Author: Eric Lammerts (<eric@lammerts.org>,
AFAICS).

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

[PATCH] arch/x86_64/pci/mmconfig.c NULL noise removal

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

[PATCH] net/ipv6/mcast.c NULL noise removal

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

[PATCH] __user annotations of video_spu_palette

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

[PATCH] amd64 time.c __iomem annotations

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

[PATCH] bogus extern in low_i2c.c

extern in function definition is an odd thing..

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

[PATCH] eeh_driver NULL noise removal

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

[PATCH] sg gfp_t annotations

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

[PATCH] scsi_transport_iscsi gfp_t annotations

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

[PATCH] cmm NULL noise removal, __user annotations

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

[PATCH] drivers/edac/i82875p_edac.c __user annotations

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

[PATCH] fix __user annotations in drivers/base/memory.c

sysfs store doesn't deal with userland pointers

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

[PATCH] drivers/char/watchdog/sbc_epx_c3.c __user annotations

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

[PATCH] dvb NULL noise removal

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

[PATCH] kernel/sys.c NULL noise removal

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

[PATCH] timer.c NULL noise removal

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

[PATCH] ipv4 NULL noise removal

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

[PATCH] fix __user annotations in fs/select.c

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

[PATCH] missing include in ser_a2232

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

[PATCH] fallout from ptrace consolidation patch: cris/arch-v10

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

[PATCH] drivers/scsi/mac53c94.c __iomem annotations

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

[PATCH] ppc: last_task_.... is defined only on non-SMP

... so it should be exported only on non-SMP.

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

[PATCH] restore power-off on sparc32

Damn you, Eric

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

[PATCH] fix breakage in ocp.c

it's ocp_device_...., not ocp_driver_....

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

[PATCH] missing includes in drivers/net/mv643xx_eth.c

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

[PATCH] drive_info removal outside of arch/i386

drive_info is used only by hd.c and that happens under #ifdef __i386__.

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

[PATCH] bogus asm/delay.h includes

asm/delay.h is non-portable; linux/delay.h should be used in generic code.

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

[PATCH] remove bogus asm/bug.h includes.

A bunch of asm/bug.h includes are both not needed (since it will get
pulled anyway) and bogus (since they are done too early).  Removed.

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

Merge /pub/scm/linux/kernel/git/gregkh/driver-2.6