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

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Transforms old policer to use generic statistic via TCA_STATS2
and maintain backward compatibility via TCA_STATS. Adds a new
API for classifiers to invoke the dumping process.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Dumps the statistic common to all action modules via the newly
introduced TLV type TCA_ACT_STATS in tcf_action_copy_stats but
allows the corresponding module to dump its own statistic by
implementing the get_stats callback.
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Based upon a patch from Herbert Xu:

> Hi Ingo:
> 
> Your recent fix to dev_queue_xmit moved the local_bh_disable to
> include stuff like skb_linearize and skb_checksum_help.  These
> are potentially expensive operations.  Since they don't need to
> run with preempt off, we could simply move the local_bh_enable
> up instead.

No, this breaks the normal return path.

How about this instead?
Signed-off-by: Tommy S. Christensen <tommy.christensen@tpack.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch allows tcpdiag to support ipv6 work as a module when itself
is also a module.

I was planning for a more extensible method but now I think the effort
would be better directed at a new generic socket netlink interface than
the TCP-specific tcpdiag.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

Signed-off-by: Hideaki YOSHIFUJI <yoshfuji@linux-ipv6.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

The new atyfb has the return value for the Power Management routine
wrong. This prevents sleep from working on PowerBooks with mach64
video chips. This fixes it.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

XFree (and X.org) has both always been broken in their conversion of the DPMS
blanking mode when passing to the kernel ioctl FBIOBLANK (in fbdev mode).

This patch makes sure that at least the value passed by userland stays in the
legal range, which has the side effect that an X DPMS POWERDOWN request will
result as an fbdev VESA_POWERDOWN request at the fbdev level now, instead of
an out-of-range value.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Small tweaks to the checksumming routines
Fix a corner case in csum_partial
Committed-by: Randolph Chung <tausq@parisc-linux.org>

Superio cleanup
Committed-by: Grant Grundler <grundler@parisc-linux.org>

The default interruption handler "handle_interruption" does not properly
check to see if the faulting space is the same as the users space.
The problem lies in the fact that if a fault happens on the gateway we
will not deliver signals to the process, the process will not die, and we
may continue handling the same fault in a loop forever.  Any malicious
user code can crash the kernel by jumping into the gateway page at an
inopportune address.

The solution is to check if the user has the right privilege and if the
spaces match, both faulting and currently active. The best action is to
force the process back to it's own space of execution at address zero, and
let it take a SIGSEGV. The process can never recover from this because it
happens immediately after the return from the interrupt via rfi or rfir.
Committed-by: Carlos O'Donell <carlos@parisc-linux.org>

Add new syscalls:
tkill, waitid, fadvise64_64, set_tid_address, tgkill, mbind,
[sg]et_mempolicy, {,l,f}{set,get,list,remove}xattr,
timer_{create,settime,gettime,getoverrun,delete},
clock_{settime,gettime,getres,nanosleep}
Committed-by: Matthew Wilcox <willy@parisc-linux.org>

PTRACE_GETEVENTMSG was not 64-bit clean, fix it
Committed-by: Randolph Chung <tausq@parisc-linux.org>

remove the now unused lcopy_*_user routines
Committed-by: Randolph Chung <tausq@parisc-linux.org>

Stop exporting the old memcpy symbols
Committed-by: Randolph Chung <tausq@parisc-linux.org>

dump the stack on a BUG()
Committed-by: Randolph Chung <tausq@parisc-linux.org>

register cpus with sysfs
Committed-by: Randolph Chung <tausq@parisc-linux.org>

 - new memcpy routine, replaces memcpy, bcopy, copy_{from,to,in}_user
   implementations
 - workaround gcc-3.0 limitations for asm()
 - Pip and Rufus pointed out the flaw in the current misaligned memcopy
   routines.  Randolph now owes lots of Milk bones.
 - updates to work with testsuite
 - sparse annotations
 - Improve copy performance for small and unaligned copies
 - Enable merge optimization for non-PA20 builds too
Committed-by: Randolph Chung <tausq@parisc-linux.org>
Committed-by: James Bottomley <jejb@parisc-linux.org>

fix CONFIG_DISCONTIGMEM support
Committed-by: Randolph Chung <tausq@parisc-linux.org>

 - Fix Astro/Elroy LMMIO space reporting in /proc/iomem.
 - Output in /proc/iomem include PCI bus number and "Extra LMMIO" range.
 - LBA now reports proper addresses routed down from SBA.
 - add mercury_cfg_ops. This fixes the tg3 init delay problem.
   PDC took ~20ms per PCI cfg write. tg3 performs ~1600 writes (~30 seconds
   per NIC) during init. Alternative is the relocate PDC to host ram.
   But that would burn ~4MB of RAM. This is cheaper.
 - add PCI_F_EXTEND to asm/pci.h - replaces private definitions
Committed-by: Grant Grundler <grundler@parisc-linux.org>

Add memory clobber to userspace syscall wrapper provided by kernel,
this matches glibc definition.
Committed-by: Carlos O'Donell <carlos@parisc-linux.org>

export global fixup routines
Committed-by: Randolph Chung <tausq@parisc-linux.org>

Revamp exceptions support to use fixup sections
Committed-by: Randolph Chung <tausq@parisc-linux.org>

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

- linux/8250.h is replaced by linux/serial_8250.h
- we now pass serial port information via a platform device.

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

Patch from Ben Dooks

Fixes mis-labelled timer variables, and incldes the
clock header to allow the code to compile. This
was not caught on previous rounds as all the
s3c2440 machine support is pending the finalisation
of the new s3c2440 uart driver (arriving soon)

Signed-off-by: Ben Dooks 

Patch from Ben Dooks

Select the PCLK based TIMER4 to provide the system timer tick
if there is no machine specific timer setup specified. This
should make it easier to add new machines, as well as cut down
on the code required in time.c when they are added.

Signed-off-by: Ben Dooks 

Patch from George G. Davis

The Lubbock PCMCIA socket driver, drivers/pcmcia/pxa2xx_lubbock.c,
lacks a MODULE_LICENSE declaration. This patch merely adds the missing
MODULE_LICENSE declaration to eliminate tainted kernel warnings when
this driver is built as a module. I have intentionally not used the
Dual MPL/GPL license in this case because 1) current linux-2.6.10-rc1
PXA PCMCIA drivers use only the GPL license, 2) the original source
files on which the PXA PCMCIA drivers are based, i.e. the SA11xx PCMCIA
drivers, state: "If you wish to allow the use of your version of this
file only under the terms of the GPL and not to allow others to use your
version of this file under the MPL, indicate your decision by deleting
the provisions above and replace them with the notice and other
provisions required by the GPL."


Signed-off-by:: George G. Davis 

Patch from Nicolas Pitre

This allows to set the default state of GPIOs before they're enabled 
as output, eliminating spurious level transitions.

Signed-off-by: Nicolas Pitre 

Here's the patch that supercedes the one I asked you to drop.  It changes
the fan speed logic so that the fans will go incrementally fast instead of
using just two steps.  Also, it starts and stops both fans instead of just
one on Powerbooks 15" and 17" (functionality asked by most of the Albooks
users on Debian's ML, and justified by the fact that fan 1 and fan 2 do not
correspond to CPU and GPU, as we previously thought).
Signed-off-by: Colin Leroy <colin@colino.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>