Commit Graph
15050 Commits
Author SHA1 Message Date
Magnus Damm 46420e49c9 sh: x3 - add ipi vectors
With the intc dual prio register support in place it is now possible
to add the ipi vectors to x3.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:50 +09:00
Magnus Damm f8b40d8ce8 sh: replace sh specific CONFIG_VOYAGERGX with CONFIG_MFD_SM501
This patch replaces all instances of CONFIG_VOYAGERGX with
CONFIG_MFD_SM501. While at it we make sure the r2d code compiles
both with and without SM501.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:50 +09:00
Magnus Damm 2eeec85638 sh: intc - convert board specific r2d code
This patch converts the board specific interrupt code for r2d to make
use of intc. While at it we improve the Kconfig to avoid confusion.

- Two sets of interrupt tables exist - one for R2D-1 and one for R2D-PLUS.
- R2D-1 and R2D-PLUS use the same irq constants.
- R2D-1 has AX88796 support, R2D-PLUS does not hook up that IRQ.
- R2D-PLUS has KEY support, R2D-1 does not hook up that IRQ.
- The number and order of IRQ values are disconnected from register bits.
- Interrupt sources now start from IRQ 100.
- The machvec demux function converts from irlm IRQ 0-14 to IRQ 100++.

Tested on R2D-1 and R2D-PLUS boards.

Version 2 adds CONFIG_RTS7751R2D_1 and CONFIG_RTS7751R2D_PLUS together
with intc structured as __initdata.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:50 +09:00
Magnus Damm 73505b445d sh: intc - rework core code
This patch reworks the intc core, implementing the following features:

- Support dual priority registers - one set and one clear register
- All 8/16/32 bit register combinations are now supported
- Both single mask and single enable bitmap register are supported
- Add code to set interrupt priority
- Speedup sense and priority configuration code
- Allocate data using bootmem, allows intc data structures to be
  __initdata
- Save memory - allocated memory footprint is smaller than intc
  structures

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:50 +09:00
Magnus Damm 6ef5fb2cfc sh: intc - add a clear register to struct intc_prio_reg
We need a secondary register member in struct intc_prio_reg to support
dual priority registers used by ipi on x3.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:49 +09:00
Magnus Damm d6aee69ca1 sh: x3 - fix setup_bootmem_node() compile error with shx3_defconfig
This makes sure the function prototype for setup_bootmem_node() gets
included. The file setup-shx3.c does not compile otherwise for
CONFIG_NUMA=n.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:49 +09:00
Markus Brunner 3ea6bc3de4 sh: Add SH7720 CPU support.
This adds support for the SH7720 (SH3-DSP) CPU.

Signed-off by: Markus Brunner <super.firetwister@gmail.com>
Signed-off by: Mark Jonas <toertel@gmail.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:49 +09:00
Paul Mundt f7f86345fd sh: Kill off duplicate symbol exports on SMP.
synchronize_irq() was being exported twice, kill off the duplicate
export.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:49 +09:00
Paul Mundt d3428e9101 sh: Wire up CSM node for SH-X3.
Now that NODES_SHIFT is bumped up, we can plug in the CSM block as
a separate node, too.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:49 +09:00
Paul Mundt 9904494d6a sh: Bump up NODES_SHIFT for SH-X3.
This was at 1, which is where most subtypes want it. SH-X3 has
more nodes, so bump it up to 3.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:48 +09:00
Paul Mundt af24fdc168 sh: x3proto: Trivial ILSEL build fix.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:48 +09:00
Paul Mundt e7109a96e6 sh: x3proto: Hook up ethernet and USB platform devices.
Now that we have a way of enabling the IRQs, hook up the platform
devices and wrap in to ILSEL for mapping the vectors.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:48 +09:00
Paul Mundt fef9608627 sh: x3proto: ILSEL IRQ support.
This adds functionality for the on-board ILSEL IRQs that chain
IRL mode events. Many on-board devices (ethernet, usb, etc.) rely
on ILSEL IRQs directly.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:48 +09:00
Paul Mundt 94807a33dc sh: Initial support for SH-X3 prototype board.
This adds initial support for the SH-X3 prototype board.
Only simple logic for the IRQ controller and the heartbeat driver
for now.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:48 +09:00
Paul Mundt 8d797cd191 sh: Plug in SH-X3 SCIF0 for sercon earlyprintk.
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:48 +09:00
Paul Mundt 6d64d4256c sh: intc: Fix sense regs oops for IRL IRQs.
IRL doesn't always define sense registers, so don't bother trying to
iterate through the table. This ended up causing an oops on SH-X3
when using IRL mode.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:48 +09:00
Magnus Damm 96290d808f sh: remove intc2 code
There is no point in keeping around the now unused intc2 code.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:48 +09:00
Magnus Damm 48180cab3a sh: intc - convert voyagergx code
This patch converts the sh-specific voyagergx interrupt code to make use
of intc. A lot of "interesting" old cruft gets replaced with intc tables
and some simple demux code.

- All interrupt sources in the sm501 data sheet are now in the header.
- The number and order of IRQ values are disconnected from register bits.
- Interrupt sources now start from IRQ 200.
- set_irq_chained_handler() is now used to hook up the demux function.

In the future it would probably make sense to move the interrupt demuxer
into into the mfd driver, but this is probably a nice step in the right
direction until that happens.

Tested on a R2D-1 board using the serial port hooked up to the sm501.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:48 +09:00
Magnus Damm 51da64264b sh: intc - add single bitmap register support
This patch adds single bitmap register support to intc. The current
code only handles 16 and 32 bit registers where a set bit means
interrupt enabled, but this is easy to extend in the future.

The INTC_IRQ() macro is also added to provide a way to hook in
interrupt controllers for FPGAs in boards or companion chips.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:47 +09:00
Magnus Damm 2635e8558a sh: intc - remove redundant irq code for shmin
This patch removes redundant interrupt code for the shmin board which
is using a sh770x processor and 4 IRQ lines as individual interrupts
(IRQ-mode).

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:47 +09:00
Magnus Damm d59645d6ba sh: intc - remove redundant irq code for sh03, snapgear and titan
This patch removes redundant board specific interrupt code for boards
using sh775x processors and 4 IRQ lines in "Individual Interrupt Mode"
aka IRLM.

Three boards are affected: sh03, snapgear and titan.

The right way to do this is to use cpu specific code provided by intc.
A nice side effect is that sh03 now compiles, board not BROKEN any more.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:47 +09:00
Paul Mundt ad89f87a84 rtc: rtc-sh: Support 4-digit year on SH7705/SH7710/SH7712.
All SH-4 parts have a 4-digit year, while the SH-3 parts typically
only use a 2-digit one. The SH7705, SH7710, and SH7712 SH-3 parts
however opted to extend it to 4-digit and still look and act like
an SH-3 RTC in all other ways.

This adds a capability flag (RTC_CAP_4_DIGIT_YEAR) that these
corner-case CPU subtypes can set in their platform data and cleans
up some of the ifdef mess in the driver as a result.

Reported-by: Markus Brunner <super.firetwister@gmail.com>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:47 +09:00
Paul Mundt 7da3b8ef66 sh: Initial multiple-node support for SH-X3.
Wire up CPU#0 URAM as node 1 on SH-X3.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:47 +09:00
Magnus Damm 1ee010087e sh: intc - add support for x3
This patch converts the cpu specific interrupt setup code for x3 from
intc2 to intc. New vectors are also added to match the preliminary
information.

Use plat_irq_setup_pins() to select between IRQ and IRL mode for IRQ0-3.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:47 +09:00
Magnus Damm 137b53b71c sh: intc - fix IRQ4 and IRQ5 typo on sh3
The intc tables for sh3 currently contain a typo where the bit
fields in IPRD are mixed up for IRQ4 and IRQ5. This patch makes
sure the correct bit fields are used - all according to the
datasheets.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:47 +09:00
Paul Mundt ac38e4d33d sh: Fix cache disabling build failures on nommu.
The cache disabling stuff screwed up some of the sh4 nommu
builds, fix it up again.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:47 +09:00
Magnus Damm c9d03d9a80 sh: simplify se7780 IRQ code
The new intc code handles IRQ3 and IRQ7 in the cpu specific code
already, so there is no reason to duplicate that here.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:47 +09:00
Magnus Damm e29bfbc443 sh: intc - add support for SH7760
This patch converts the cpu specific interrupt setup code for sh7760
from ipr + intc2 to intc. New vectors are also added to match the
information provided by the datasheet.

Vectors for IRQ4-IRQ7 are enabled by default. Use plat_irq_setup_pins()
if pins IRL0-3 should be used in IRLM mode.

The patch also adds the SIM block to the serial port platform data.
Version two of this patch fixes MMCIF problems reported by Manuel Lauss.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Acked-by: Manuel Lauss <mano@roarinelk.homelinux.net>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:46 +09:00
Magnus Damm a0e23267d4 sh: intc - add support for SH7785
This patch converts the cpu specific interrupt setup code for sh7785
from intc2 to intc. New vectors are also added to match the information
provided by the datasheet.

No IRQ/IRL pin vectors are enabled by default. Use plat_irq_setup_pins()
to select between IRL and IRQ mode.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:46 +09:00
Paul Mundt e7bd34a15b sh: Support explicit L1 cache disabling.
This reworks the cache mode configuration in Kconfig, and allows for
explicit selection of write-back/write-through/off configurations.
All of the cache flushing routines are optimized away for the off
case.

Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:46 +09:00
Magnus Damm ac919986d7 sh: intc - avoid SH7710 specific vector on SH7712
This patch makes sure that the sh7710 specific ipsec vector is missing
if building for a sh7712. All according to the datasheet.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:46 +09:00
Magnus Damm 1301e71562 sh: intc - add missing vectors for SH7707
This patch adds a few missing vectors for sh7707. The only interrupt
controller differences between sh7707 and sh7709 seem to be added
vectors for one LCD controller and two PCMCIA slots.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:46 +09:00
Magnus Damm ec58f1f32d sh: intc - add support for SH7706, SH7707, SH7708, SH7709
This patch unifies the cpu specific interrupt setup code for
sh7706, sh7707, sh7708 and sh7709 and moves the code into a new
file called setup-sh770x.c.  It makes sense to share the setup code
between these processors because most hardware blocks are identical
from a software point of view. With this patch the sh770x processors
now have a complete set of vectors that match with the information
provided by the data sheets. This is a big improvement for sh7708.

Vectors for IRQ4 and IRQ5 are enabled by default. Use
plat_irq_setup_pins() if pins IRQ0-3 should be used in IRQ mode.

This patch also unifies the platform device setup code which means
that the rtc driver now has platform data for all sh770x processors.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:46 +09:00
Magnus Damm 28b146c84e sh: intc - add support for SH7710
This patch converts the cpu specific interrupt setup code
for sh7710 from ipr to intc. While at it new vectors are added
to match the information provided by the datasheet. Version two
simplifies the Kconfig part.

Vectors for IRQ4 and IRQ5 are enabled by default. Use
plat_irq_setup_pins() if pins IRQ0-3 should be used in IRQ mode.

This patch also adds sh7710 specific platform data for the rtc
driver. The base address of SCIF1 is adjusted to match the
datasheet.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:45 +09:00
Magnus Damm 70e8be0a4e sh: intc - add support for SH7705
This patch converts the cpu specific interrupt setup code for
sh7705 from ipr to intc. While at it new vectors are added to
match the information provided by the datasheet.

Vectors for IRQ4 and IRQ5 are enabled by default.

Use plat_irq_setup_pins() if pins IRQ0-3 should be used in IRQ mode.

This patch also adds sh7705 specific platform data for the rtc driver.

Signed-off-by: Magnus Damm <damm@igel.co.jp>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2007-09-21 11:57:45 +09:00
Paul Bolle bbc15f46fe [x86 setup] Fix typo in arch/i386/boot/header.S
There's an obvious typo in arch/i386/boot/header.S (in your
linux-2.6-x86setup.git) that I noticed by just studying the code.

Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
2007-09-20 11:06:59 -07:00
H. Peter Anvin 91c4b8cb5a [acpi] Correct the decoding of video mode numbers in wakeup.S
wakeup.S looks at the video mode number from the setup header and
looks to see if it is a VESA mode.  Unfortunately, the decoding is
done incorrectly and it will attempt to frob the VESA BIOS for any
mode number 0x0200 or larger.  Correct this, and remove a bunch of #if
0'd code.

Massive thanks to Jeff Chua for reporting the bug, and suffering
though a large number of experiments in order to track this problem
down.

Cc: Pavel Machek <pavel@ucw.cz>
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
2007-09-20 11:06:58 -07:00
H. Peter Anvin 3f662b3f6e [x86 setup] Present the canonical video mode number to the kernel
Canonicalize the video mode number as presented to the kernel.  The
video mode number may be user-entered (e.g. ASK_VGA), an alias
(e.g. NORMAL_VGA), or a size specification, and that confuses the
suspend wakeup code.

Signed-off-by: H. Peter Anvin <hpa@zytor.com>
2007-09-20 11:06:58 -07:00
Linus Torvalds a88a8eff1e Merge branch 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus
* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus:
  [MIPS] cpu-bugs64.c: GCC 3.3 constraint workaround
  [MIPS] DEC: Initialise ioasic_ssr_lock
2007-09-19 11:45:32 -07:00
Linus Torvalds f15f41383d Merge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc
* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/paulus/powerpc:
  [POWERPC] Fix timekeeping on PowerPC 601
  [POWERPC] Don't expose clock vDSO functions when CPU has no timebase
  [POWERPC] spusched: Fix null pointer dereference in find_victim
2007-09-19 11:38:25 -07:00
Linus Torvalds dbe3ed1c07 x86-64: page faults from user mode are always user faults
Randy Dunlap noticed an interesting "crashme" behaviour on his dual
Prescott Xeon setup, where he gets page faults with the error code
having a zero "user" bit, but the register state points back to user
mode.

This may be a CPU microcode buglet triggered by some strange instruction
pattern that crashme generates, and loading a microcode update seems to
possibly have fixed it.

Regardless, we really should trust the register state more than the
error code, since it's really the register state that determines whether
we can actually send a signal, or whether we're in kernel mode and need
to oops/kill the process in the case of a page fault.

Cc: Randy Dunlap <rdunlap@xenotime.net>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-09-19 11:37:14 -07:00
Maciej W. Rozycki 09abbcffb3 [MIPS] cpu-bugs64.c: GCC 3.3 constraint workaround
Add a workaround to address warnings generated on the "n" constraint by
GCC 3.3 and below.

Signed-off-by: Maciej W. Rozycki <macro@linux-mips.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-09-19 19:33:14 +01:00
Maciej W. Rozycki 6883599943 [MIPS] DEC: Initialise ioasic_ssr_lock
Fix the definition of the ioasic_ssr_lock spinlock to include a proper 
initialisation.

Signed-off-by: Maciej W. Rozycki <macro@linux-mips.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-09-19 19:33:14 +01:00
Jeff Dike 508a92741a uml: fix irqstack crash
This patch fixes a crash caused by an interrupt coming in when an IRQ stack
is being torn down.  When this happens, handle_signal will loop, setting up
the IRQ stack again because the tearing down had finished, and handling
whatever signals had come in.

However, to_irq_stack returns a mask of pending signals to be handled, plus
bit zero is set if the IRQ stack was already active, and thus shouldn't be
torn down.  This causes a problem because when handle_signal goes around
the loop, sig will be zero, and to_irq_stack will duly set bit zero in the
returned mask, faking handle_signal into believing that it shouldn't tear
down the IRQ stack and return thread_info pointers back to their original
values.

This will eventually cause a crash, as the IRQ stack thread_info will
continue pointing to the original task_struct and an interrupt will look
into it after it has been freed.

The fix is to stop passing a signal number into to_irq_stack.  Rather, the
pending signals mask is initialized beforehand with the bit for sig already
set.  References to sig in to_irq_stack can be replaced with references to
the mask.

[akpm@linux-foundation.org: use UL]
Signed-off-by: Jeff Dike <jdike@linux.intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-09-19 11:24:18 -07:00
Jeremy Fitzhardinge 389a3c0249 xen: don't bother trying to set cr4
Xen ignores all updates to cr4, and some versions will kill the domain if
you try to change its value.  Just ignore all changes.

Signed-off-by: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: Andi Kleen <ak@suse.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-09-19 11:24:18 -07:00
Nicolas George 2c392a4f47 uml: use correct type in BLKGETSIZE ioctl
I found a type mismatch in UML that makes host block devices unusable as ubd
devices on x86_64 and other 64 bits systems (segfault of the mm subsystem):

In block/ioctl.c, the following lines show that the BLKGETSIZE ioctl expects
a pointer to a long:

	case BLKGETSIZE:
		if ((bdev->bd_inode->i_size >> 9) > ~0UL)
			return -EFBIG;
		return put_ulong(arg, bdev->bd_inode->i_size >> 9);

In arch/um/os-Linux/file.c, os_file_size calls it with an int.

The ioctl_list man page should be fixed as well.

Cc: Jeff Dike <jdike@addtoit.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-09-19 11:24:17 -07:00
Benjamin Herrenschmidt c27da33969 [POWERPC] Fix timekeeping on PowerPC 601
Recent changes to the timekeeping code broke support for the PowerPC 601
processor which doesn't have the usual timebase facility but a slightly
different thing called (yuck) the RTC.

This fixes it, boot tested on an old 601 based PowerMac 7200.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
2007-09-19 15:26:34 +10:00
Benjamin Herrenschmidt 7b5acbaac3 [POWERPC] Don't expose clock vDSO functions when CPU has no timebase
We forgot to remove the clock_gettime, clock_getres and get_tbfreq vDSO
calls on CPUs that have no timebase such as 601 or 403 (old CPUs that have
different mechanisms and for which the vDSO code will not work properly).
This fixes it.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
2007-09-19 15:26:34 +10:00
Christoph Hellwig c0e7b4aa1c [POWERPC] spusched: Fix null pointer dereference in find_victim
find_victim can dereference a NULL pointer when iterating over the list
of victim spus because list_mutex only guarantees spu->ct to be stable,
but of course not to be non-NULL.

Also fix find_victim to not call spu_unbind_context without list_mutex
because that violates the above guarantee.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Arnd Bergmann <arnd.bergmann@de.ibm.com>
Signed-off-by: Jeremy Kerr <jk@ozlabs.org>
Signed-off-by: Paul Mackerras <paulus@samba.org>
2007-09-19 15:26:29 +10:00
David S. Miller 8a177c4f17 [SPARC64]: Warn user if cpu is ignored.
When NR_CPUS is smaller than the cpu probed, let the user
know that the cpu won't be used.

Suggested by Al Viro.

Signed-off-by: David S. Miller <davem@davemloft.net>
2007-09-16 14:45:06 -07:00