This gets rid of a bug found in some IRQ handling logic, after tripping
a debug assertion.  Basically, a recent patch called the wrong routine to
unlink a QH.  Net result, it wasn't allowing for the case that some other
QH was already being unlinked.  This patch uses the correct routine; the
names are confusingly similar, and the effect is often identical.

The consequence of using the wrong routine was that the driver could
lose one of the pending unlinks (probably wedging some activity) and
treat the other one as completed before it was safe to do so (which
probably wouldn't oops, but could cause other nasty corruption).

From: Brian Murphy <brian@murphy.dk>
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

This patch is clearly needed for us to be in compliance with the USB spec.
It adds the mandated recovery-time delay following a port reset.
Regardless of anything else we do to alter the device initialization
sequence, this is necessary.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

Turns out that a workaround for a different EHCI chip trips up at
least one NForce4 board.  Neither controller can multiply right.
Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>

From: Art Haas

The current GCC cvs code does not like the include/linux/fb.h file:

In file included from drivers/video/aty/atyfb_base.c:63:
include/linux/fb.h:865: error: array type has incomplete element type

This error is due to recent changes in GCC. A thread discussing this
change can be found by following the link below:

http://gcc.gnu.org/ml/gcc/2005-02/msg00053.html

The patch moves the array declaration after the definition of the
fb_modelist structure, and with this small change GCC is happy once
again.
Signed-off-by: Antonino Daplas <adaplas@pol.net>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

It appears that access to the PLL registers of the radeon chip is
unreliable while the card is in "legacy VGA" text mode. I don't have a
good explanation yet, it might be the BIOS mucking around behind my
back. This fixes the lockup by moving the code that enables/disables the
dynamic power management to after the mode is set.

I'm still waiting for a proper explanation from ATI...
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

into ppc970.osdl.org:/home/torvalds/v2.6/linux

CMSG reception needs to occur even if msg_name is not
set.
Signed-off-by: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch adds a net_device argument to ifdown.  After all,
it's a bit silly to notify someone of an ifdown event without
telling them what which device it was for :)
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

We checked the wrong byte, causing the touchpad
to lose sync if an absolute packet is received
after a relative packet with negative Y displacement.
Signed-off-by: Dmitry Torokhov <dtor_core@ameritech.net>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Noted while going through the n_tty code: that buffer check
used to check against the size of the temporary kernel buffer,
but since the tty layer was changed to use kernel buffers
though-out, the kernel buffer array became just a pointer, and
the check was limiting the opost blocksize to the size of a
pointer, which makes no sense.

Just remove it, since now the whole buffer is always in kernel
space.

into ppc970.osdl.org:/home/torvalds/v2.6/linux

The NTTY code can get confused by 4kB chunks, apparently
because n_tty_receive_room() will claim to have more room
than n_tty_receive_buf() can actually accept.

Until somebody figures out what the real n_tty_receive_room()
logic should be, let's just limit it to a safe 2kB.

Thanks go to Andreas Schwab for finding a test-case.

Signed-off-by: Olaf Hering <olh@suse.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

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

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

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

cg3_blank() needs to clear the video enable register bit
to blank the screen, not set it.
Signed-off-by: David S. Miller <davem@davemloft.net>

into nuts.davemloft.net:/disk1/BK/sparc-2.6

into ppc970.osdl.org:/home/torvalds/v2.6/linux

into ppc970.osdl.org:/home/torvalds/v2.6/linux

Add some missing __iomem annotations for Integrator machines.
Signed-off-by: Russell King <rmk@arm.linux.org.uk>

This adds some missing annotations found by making the raw IO
operations check their pointer type.
Signed-off-by: Russell King <rmk@arm.linux.org.uk>

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

Patch from Ben Dooks

Fix the following:
arch/arm/mach-ixp2000/pci.c:148: warning: `cli' is deprecated (declared at include/linux/interrupt.h:65)
arch/arm/mach-ixp2000/pci.c:161: warning: `sti' is deprecated (declared at include/linux/interrupt.h:69)
arch/arm/mach-ixp2000/pci.c:178: warning: `cli' is deprecated (declared at include/linux/interrupt.h:65)
arch/arm/mach-ixp2000/pci.c:191: warning: `sti' is deprecated (declared at include/linux/interrupt.h:69)
By replacing cli() wht local_irq_save() and sti() with local_irq_restore().
This patch has not been tested.

Signed-off-by: Ben Dooks
Signed-off-by: Russell King

Patch from Ben Dooks

Fix the following:
arch/arm/mach-ixp4xx/common.c:305: warning: initialization from incompatible pointer type

Signed-off-by: Ben Dooks
Signed-off-by: Russell King

It is defined as a _IOW() which is erroneous, it should
have been defined as _IORW() but that cannot be changed
now without breaking all existing applications using this
ioctl.
Signed-off-by: David S. Miller <davem@davemloft.net>

this oneliner fixes the situation that I can enter a year to blacklist
ACPI devices even if ACPI is completely disabled.
Signed-off-by: Rolf Eike Beer <eike-kernel@sf-tec.de>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

When the lockfs patches went in an important bit got lost, the call in
generic_file_write to put newly incoming writers to sleep when a filesystem
is frozen.  Nathan added back the call in the now separate XFS write patch,
and the patch for the generic code is below:
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

The size of LAN bridged messages was not being returned properly from the
function that calculated address sizes.
Signed-off-by: Corey Minyard <minyard@acm.org>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

unifdef complains about the space before #ifndef.
Signed-off-by: Olaf Hering <olh@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

isdn_net_get_locked_lp is doing a local_bh_enable with hard IRQs disabled. 
This is not allowed.

The following patch fixes the problem by removing the unnecessary
local_bh_enable while the hard IRQs are disabled.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

into ppc970.osdl.org:/home/torvalds/v2.6/linux

Patch from Thomas Gleixner

Remove NULL initializers of static variables.

Signed-off-by: Thomas Gleixner
Signed-off-by: Russell King

Patch from Thomas Gleixner

Trivial build fix for shannon platform

Signed-off-by: Thomas Gleixner
Signed-off-by: Russell King

Patch from Thomas Gleixner

Trivial build fix for badge4 platform

Signed-off-by: Thomas Gleixner
Signed-off-by: Russell King

into ppc970.osdl.org:/home/torvalds/v2.6/linux

into pobox.com:/garz/repo/net-drivers-2.6

Fixes MDIO delay. Please apply.
Signed-off-by: Ralph Siemsen <ralphs@netwinder.org>
Signed-off-by: Matt Porter <mporter@kernel.crashing.org>
Signed-off-by: Jeff Garzik <jgarzik@pobox.com>