pandora-kernel.git
13 years ago[ARM] OMAP3 clock: remove unnecessary dpll_data dereferences
Paul Walmsley [Wed, 28 Jan 2009 19:08:38 +0000 (12:08 -0700)]
[ARM] OMAP3 clock: remove unnecessary dpll_data dereferences

Remove some clutter from omap2_dpll_round_rate().

linux-omap source commit is 4625dceb8583c02a6d67ededc9f6a8347b6b8cb7.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
13 years agoinput: vsense: add middle click
Urja Rannikko [Wed, 30 Jun 2010 04:18:44 +0000 (04:18 +0000)]
input: vsense: add middle click

Addition to mouse buttons mode; Down is middle click. Reorganized the
the logic a bit.

Signed-off-by: Urja Rannikko <urjaman@gmail.com>
13 years agoASoC: fix CLKX polarity
Grazvydas Ignotas [Wed, 16 Jun 2010 00:49:41 +0000 (03:49 +0300)]
ASoC: fix CLKX polarity

After mass production started it was found that several boards exhibit
noise problems during sound playback. After some investigation it was
determined that CLKX polarity is set incorrectly, and even if most boards
can tolerate this setting, there are some that don't.

Fix polarity setup in the board file. As the clock settings for input and
output now match, merge in and out functions and structures to simplify
code.

13 years agoASoC: add start treshold to prevent underflows in some cases
Grazvydas Ignotas [Thu, 10 Jun 2010 20:34:51 +0000 (23:34 +0300)]
ASoC: add start treshold to prevent underflows in some cases

This is useful for apps that don't rely on audio blocking and
feed fixed amount of samples over time period (like emulators).

13 years agoomap: DMA: Init CDAC to zero
manjugk manjugk [Thu, 4 Mar 2010 07:11:56 +0000 (07:11 +0000)]
omap: DMA: Init CDAC to zero

The register DMA4_CDAC needs to be initialized to zero
before starting DMA transfer.

Cc: Santosh Shilimkar <santosh.shilimkar@ti.com>
Cc: Govindraj R <govindraj.raja@ti.com>
Cc: Kevin Hilman <khilman@deeprootsystems.com>
Reported-by:S, Venkatraman <svenkatr@ti.com>
Signed-off-by: Manjunatha GK <manjugk@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
13 years agoOMAP: dma_chan[lch_head].flag & OMAP_DMA_ACTIVE tested twice in omap_dma_unlink_lch()
Roel Kluin [Thu, 14 Jan 2010 02:10:29 +0000 (18:10 -0800)]
OMAP: dma_chan[lch_head].flag & OMAP_DMA_ACTIVE tested twice in omap_dma_unlink_lch()

The same flag and bits were tested twice.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
13 years agoomap: Fix race condition in omap dma driver
Tao Hu [Wed, 11 Nov 2009 02:55:17 +0000 (18:55 -0800)]
omap: Fix race condition in omap dma driver

The bug could cause irq enable bit of one DMA channel is
cleared/set unexpectedly when 2 (or more) drivers are calling
omap_request_dma()/omap_free_dma() simultaneously

Signed-off-by: Fei Yang <AFY095@motorola.com>
Signed-off-by: Tao Hu <taohu@motorola.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
13 years agoomap: SDMA: Fix omap_stop_dma() API for channel linking
Santosh Shilimkar [Thu, 22 Oct 2009 21:46:31 +0000 (14:46 -0700)]
omap: SDMA: Fix omap_stop_dma() API for channel linking

OMAP sDMA driver API omap_stop_dma() doesn't really stop the dma when used
in linking scenario.

The DMA channel needs to be disabled before resetting the chain.
Also fix clearing of the OMAP_DMA_ACTIVE status in the linked case.

Cc: Hari n <hari.zoom@gmail.com>
Cc: Jarkko Nikula <jhnikula@gmail.com>
Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Signed-off-by: Venkatraman S <svenkatr@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
13 years agoARM: OMAP: Fix race in OMAP2/3 DMA IRQ handling
Juha Yrjola [Thu, 29 Jan 2009 16:57:12 +0000 (08:57 -0800)]
ARM: OMAP: Fix race in OMAP2/3 DMA IRQ handling

CSR must be cleared before invoking the callback.

If the callback function starts a new, fast DMA transfer on the same
channel, the completion status might lost if CSR is cleared after
the callback invocation.

Signed-off-by: Juha Yrjola <juha.yrjola@solidboot.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
13 years agoDSS2: make pan, set var not to wait for GO bits
Grazvydas Ignotas [Sun, 6 Jun 2010 21:51:19 +0000 (00:51 +0300)]
DSS2: make pan, set var not to wait for GO bits

In case we want to run faster than LCD refresh rate, we'll
eventually hit already set GO bit, and waiting for it to
clear will not allow to go faster then refresh, which is not
always desirable. Disable this behavior, anyone wanting to
sync can always use FBIO_WAITFORVSYNC.

14 years agoARM: OMAP: gptimer min_delta_ns corrected Release-2010-05/1
Aaro Koskinen [Thu, 29 Jan 2009 16:57:17 +0000 (08:57 -0800)]
ARM: OMAP: gptimer min_delta_ns corrected

When 32 kHz timer is used the min_delta_ns should be initialized so
that it reflects the timer programming cost. A write to the timer
device will be usually posted, but it takes roughly 3 cycles before
it is effective. If the timer is reprogrammed before that, the CPU
will stall until the previous write completes. This was pointed out by
Richard Woodruff.

Since the lower bound for min_delta_ns is 1000, the change is visible
only with tick rates less than 3 MHz.

Also note that the old value is incorrect for 32 kHz also due to
a rounding error, and it can cause the timer queue to hang (due to
clockevent code trying to program the timer with zero ticks).

Signed-off-by: Aaro Koskinen <Aaro.Koskinen@nokia.com>
Reviewed-by: Richard Woodruff <r-woodruff2@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
14 years agoARM: OMAP2: drop redundant pending write check for gptimer
Paul Walmsley [Thu, 11 Dec 2008 01:36:33 +0000 (17:36 -0800)]
ARM: OMAP2: drop redundant pending write check for gptimer

omap_dm_timer_write_reg() already waits for pending writes to complete,
so the extra wait in omap_dm_timer_set_load() is superfluous.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Richard Woodruff <r-woodruff2@ti.com>
Acked-by: Kevin Hilman <khilman@deeprootsystems.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
14 years agoMerge branch 'pandora-27-omap1' into pandora-27-omap1_rework
Grazvydas Ignotas [Fri, 7 May 2010 14:03:11 +0000 (17:03 +0300)]
Merge branch 'pandora-27-omap1' into pandora-27-omap1_rework

14 years agoMerge branch 'master' of git://git.infradead.org/users/dedekind/ubifs-v2.6.27 into...
Grazvydas Ignotas [Fri, 7 May 2010 13:58:24 +0000 (16:58 +0300)]
Merge branch 'master' of git://git.infradead.org/users/dedekind/ubifs-v2.6.27 into pandora-27-omap1_rework

14 years agoMerge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux...
Grazvydas Ignotas [Fri, 7 May 2010 13:52:18 +0000 (16:52 +0300)]
Merge branch 'master' of git://git./linux/kernel/git/stable/linux-2.6.27.y into pandora-27-omap1_rework

Conflicts:
drivers/usb/musb/davinci.c
security/Kconfig

14 years agogpio-keys: enable autorepeat
Grazvydas Ignotas [Wed, 5 May 2010 23:15:34 +0000 (02:15 +0300)]
gpio-keys: enable autorepeat

do this to be consistent with keypad and to make framebuffer console
more useful.

14 years agoInput: gpio-keys - add flag to allow auto repeat
Dominic Curran [Tue, 28 Oct 2008 02:30:53 +0000 (22:30 -0400)]
Input: gpio-keys - add flag to allow auto repeat

This patch adds a flag to gpio-key driver to turn on the input subsystems
auto repeat feature if needed.

Signed-off-by: Dominic Curran <dcurran@ti.com>
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
14 years agodon't touch i2c until it's ready
Grazvydas Ignotas [Tue, 4 May 2010 14:55:03 +0000 (17:55 +0300)]
don't touch i2c until it's ready

14 years agonew NAND layout and enable UBI in defconfig
Grazvydas Ignotas [Sat, 1 May 2010 21:27:28 +0000 (00:27 +0300)]
new NAND layout and enable UBI in defconfig

14 years agomtd: omap2 fix prefetch mode read issue
Vimal Singh [Thu, 7 Jan 2010 06:46:26 +0000 (12:16 +0530)]
mtd: omap2 fix prefetch mode read issue

There is a bug in nand prefetch read routine, which comes into effect
only if nand device is a 16-bit device (as we have in zoom boards).
This bug is effective only with below combination of conditions:
1. nand deivce, in use, is a 16 bit device
2. nand driver supports 'subpage' read
3. SW ECC is in use

This was not seen old  kernel (ex: .23), because when, in early days,
we tested this (nand prefetch read in LDP boards) there was no
'subpage read' support.
Later when we had subpage read in (.27) kernel, we had hw ecc enabled
always in our internal tree. So, we missed this bug.

This patch fixes the issue.

Signed-off-by: Vimal Singh <vimalsingh@ti.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
14 years agomtd: omap: adding DMA mode support in nand prefetch/post-write
vimal singh [Mon, 13 Jul 2009 10:59:16 +0000 (16:29 +0530)]
mtd: omap: adding DMA mode support in nand prefetch/post-write

This patch adds DMA mode support for nand prefetch/post-write engine.

Signed-off-by: Vimal Singh <vimalsingh@ti.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
14 years agomtd: omap: add support for nand prefetch-read and post-write
vimal singh [Mon, 13 Jul 2009 10:56:24 +0000 (16:26 +0530)]
mtd: omap: add support for nand prefetch-read and post-write

This patch adds prefetch support to access nand flash in mpu mode.
This patch also adds 8-bit nand support (omap_read/write_buf8).
Prefetch can be used for both 8- and 16-bit devices.

Signed-off-by: Vimal Singh <vimalsingh@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
14 years agomtd: nand: fix build failure and incorrect return from omap_wait()
vimal singh [Sat, 27 Jun 2009 05:37:06 +0000 (11:07 +0530)]
mtd: nand: fix build failure and incorrect return from omap_wait()

We need to include jiffies.h manually in some cases, and the status
returned from omap_wait() was broken in two separate ways.

Also add cond_resched() to the loop.

Signed-off-by: Vimal Singh <vimalsingh@ti.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
14 years agomtd: nand: add OMAP2/OMAP3 NAND driver
Vimal Singh [Tue, 12 May 2009 20:47:03 +0000 (13:47 -0700)]
mtd: nand: add OMAP2/OMAP3 NAND driver

This driver is present in the OMAP tree, now pushing it to MTD.

Original author(s):
       Jian Zhang <jzhang@ti.com>

Signed-off-by: Vimal Singh <vimalsingh@ti.com>
Cc: Jian Zhang <jzhang@ti.com>
Cc: Artem Bityutskiy <dedekind@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
14 years agoremove NAND driver to apply mainline version
Grazvydas Ignotas [Sat, 1 May 2010 20:01:48 +0000 (23:01 +0300)]
remove NAND driver to apply mainline version

14 years agoMerge branch 'master' of git://git.infradead.org/users/dedekind/ubifs-v2.6.27 into...
Grazvydas Ignotas [Sat, 1 May 2010 16:00:58 +0000 (19:00 +0300)]
Merge branch 'master' of git://git.infradead.org/users/dedekind/ubifs-v2.6.27 into pandora-27-omap1

14 years agoinput: rename pwrbutton and touchscreen
Grazvydas Ignotas [Mon, 26 Apr 2010 20:46:51 +0000 (23:46 +0300)]
input: rename pwrbutton and touchscreen

14 years agopandora: enable supplies on boot
Grazvydas Ignotas [Mon, 26 Apr 2010 12:29:34 +0000 (15:29 +0300)]
pandora: enable supplies on boot

We don't want to rely on bootloader for this, this would be unneeded
dependency.

14 years agoOMAP3 clock: lessen amount of noisy messages
Artem Bityutskiy [Tue, 12 May 2009 23:34:40 +0000 (17:34 -0600)]
OMAP3 clock: lessen amount of noisy messages

On our system we see the following messages:

Disabling unused clock "gpt2_ick"
Disabling unused clock "gpt3_ick"
Disabling unused clock "gpt4_ick"
Disabling unused clock "gpt5_ick"
...

The messages have KERN_INFO level and if you have serial
console, they normally go there. I do not think it is good
idea to print that much stuff there. Moreover, messages
are not properly prefixed and for mortals it is not
immeadietly clear where they come from.

Let's give them debugging level instead.

Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>
[paul@pwsan.com: trimmed debugging output in patch description]

14 years agoinput: twl4030keypad: use generic name
Grazvydas Ignotas [Mon, 26 Apr 2010 11:53:53 +0000 (14:53 +0300)]
input: twl4030keypad: use generic name

14 years agoinput: vsense: mbuttons functionality, more /proc control, refactoring
Grazvydas Ignotas [Mon, 26 Apr 2010 11:24:21 +0000 (14:24 +0300)]
input: vsense: mbuttons functionality, more /proc control, refactoring

14 years agoupdate defconfig
Grazvydas Ignotas [Tue, 20 Apr 2010 15:16:56 +0000 (18:16 +0300)]
update defconfig

this defconfig is used for test kernels

14 years agoDSS2: update gamma for TD043MTEA1 panel
Grazvydas Ignotas [Tue, 20 Apr 2010 15:14:01 +0000 (18:14 +0300)]
DSS2: update gamma for TD043MTEA1 panel

14 years agopandora: platform data for wl1251
Grazvydas Ignotas [Tue, 20 Apr 2010 13:25:51 +0000 (16:25 +0300)]
pandora: platform data for wl1251

14 years agoAdd a very basic platform driver module to bring up the SDIO WiFi so that the platfor...
David-John Willis [Sun, 30 Aug 2009 18:31:31 +0000 (19:31 +0100)]
Add a very basic platform driver module to bring up the SDIO WiFi so that the platform stuff is no longer hacked into the WL1251 driver (TODO: Clean up and move somewhere more mainline friendly).

14 years agoSDIO patches to put some card into into platform devices (hack) to match WL1251 driver.
David-John Willis [Mon, 22 Jun 2009 19:49:31 +0000 (20:49 +0100)]
SDIO patches to put some card into into platform devices (hack) to match WL1251 driver.

14 years agousb otg: remove PM dependency
Grazvydas Ignotas [Thu, 8 Apr 2010 21:58:25 +0000 (00:58 +0300)]
usb otg: remove PM dependency

It can function fine without it.

14 years agoLinux 2.6.27.46 v2.6.27.46
Greg Kroah-Hartman [Thu, 1 Apr 2010 22:53:33 +0000 (15:53 -0700)]
Linux 2.6.27.46

14 years agohwmon: (coretemp) Add missing newline to dev_warn() message
Dean Nelson [Mon, 29 Mar 2010 20:03:00 +0000 (22:03 +0200)]
hwmon: (coretemp) Add missing newline to dev_warn() message

commit 4d7a5644e4adfafe76c2bd8ee168e3f3b5dae3a8 upstream.

Add missing newline to dev_warn() message string. This is more of an issue
with older kernels that don't automatically add a newline if it was missing
from the end of the previous line.

Signed-off-by: Dean Nelson <dnelson@redhat.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: fix usbfs regression
Alan Stern [Sat, 6 Mar 2010 20:04:03 +0000 (15:04 -0500)]
USB: fix usbfs regression

commit 7152b592593b9d48b33f8997b1dfd6df9143f7ec upstream.

This patch (as1352) fixes a bug in the way isochronous input data is
returned to userspace for usbfs transfers.  The entire buffer must be
copied, not just the first actual_length bytes, because the individual
packets will be discontiguous if any of them are short.

Reported-by: Markus Rechberger <mrechberger@gmail.com>
Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agotmpfs: cleanup mpol_parse_str()
KOSAKI Motohiro [Tue, 23 Mar 2010 20:35:32 +0000 (13:35 -0700)]
tmpfs: cleanup mpol_parse_str()

commit 926f2ae04f183098cf9a30521776fb2759c8afeb upstream.

mpol_parse_str() made lots 'err' variable related bug.  Because it is ugly
and reviewing unfriendly.

This patch simplifies it.

Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Ravikiran Thirumalai <kiran@scalex86.org>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Acked-by: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodoc: add the documentation for mpol=local
KOSAKI Motohiro [Tue, 23 Mar 2010 20:35:33 +0000 (13:35 -0700)]
doc: add the documentation for mpol=local

commit 5574169613b40b85d6f4c67208fa4846b897a0a1 upstream.

commit 3f226aa1c (mempolicy: support mpol=local tmpfs mount option) added
new mpol=local mount option.  but it didn't add a documentation.

This patch does it.

Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Ravikiran Thirumalai <kiran@scalex86.org>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Acked-by: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agotmpfs: handle MPOL_LOCAL mount option properly
KOSAKI Motohiro [Tue, 23 Mar 2010 20:35:31 +0000 (13:35 -0700)]
tmpfs: handle MPOL_LOCAL mount option properly

commit 12821f5fb942e795f8009ece14bde868893bd811 upstream.

commit 71fe804b6d5 (mempolicy: use struct mempolicy pointer in
shmem_sb_info) added mpol=local mount option.  but its feature is broken
since it was born.  because such code always return 1 (i.e.  mount
failure).

This patch fixes it.

Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Ravikiran Thirumalai <kiran@scalex86.org>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Acked-by: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agotmpfs: mpol=bind:0 don't cause mount error.
KOSAKI Motohiro [Tue, 23 Mar 2010 20:35:30 +0000 (13:35 -0700)]
tmpfs: mpol=bind:0 don't cause mount error.

commit d69b2e63e9172afb4d07c305601b79a55509ac4c upstream.

Currently, following mount operation cause mount error.

% mount -t tmpfs -ompol=bind:0 none /tmp

Because commit 71fe804b6d5 (mempolicy: use struct mempolicy pointer in
shmem_sb_info) corrupted MPOL_BIND parse code.

This patch restore the needed one.

Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Ravikiran Thirumalai <kiran@scalex86.org>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Acked-by: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agotmpfs: fix oops on mounts with mpol=default
Ravikiran G Thirumalai [Tue, 23 Mar 2010 20:35:28 +0000 (13:35 -0700)]
tmpfs: fix oops on mounts with mpol=default

commit 413b43deab8377819aba1dbad2abf0c15d59b491 upstream.

Fix an 'oops' when a tmpfs mount point is mounted with the mpol=default
mempolicy.

Upon remounting a tmpfs mount point with 'mpol=default' option, the mount
code crashed with a null pointer dereference.  The initial problem report
was on 2.6.27, but the problem exists in mainline 2.6.34-rc as well.  On
examining the code, we see that mpol_new returns NULL if default mempolicy
was requested.  This 'NULL' mempolicy is accessed to store the node mask
resulting in oops.

The following patch fixes it.

Signed-off-by: Ravikiran Thirumalai <kiran@scalex86.org>
Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Christoph Lameter <cl@linux-foundation.org>
Cc: Mel Gorman <mel@csn.ul.ie>
Acked-by: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agob44 WOL setup: one-bit-off stack corruption kernel panic fix
Stanislav Brabec [Wed, 9 Dec 2009 05:00:22 +0000 (21:00 -0800)]
b44 WOL setup: one-bit-off stack corruption kernel panic fix

commit e0188829cb724e7d12a2d4e343b368ff1d6e1471 upstream.

About 50% of shutdowns of b44 Ethernet adapter ends by kernel panic
with kernels compiled with stack-protector.

Checking b44_magic_pattern() return values, one call of
b44_magic_pattern() returns 127. It means, that set_bit(128, pmask)
was called on line 1509. It means that bit 0 of 17th byte of pmask was
overwritten. But pmask has only 16 bytes. Stack corruption happens.

It seems that set_bit() on line 1509 always writes one bit off.

The fix does not only solve the stack corruption, but also makes Wake
On LAN working on my onboard B44 on Asus A7V-333X mainboard.

It seems that this problem affects all kernel versions since commit
725ad800 ([PATCH] b44: add wol for old nic) on 2006-06-20.

Signed-off-by: Stanislav Brabec <sbrabec@suse.cz>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoV4L/DVB (13961): em28xx-dvb: fix memleak in dvb_fini()
Francesco Lavra [Thu, 31 Dec 2009 11:47:11 +0000 (08:47 -0300)]
V4L/DVB (13961): em28xx-dvb: fix memleak in dvb_fini()

commit 19f48cb105b7fa18d0dcab435919a3a29b7a7c4c upstream.

this patch fixes a memory leak which occurs when an em28xx card with DVB
extension is unplugged or its DVB extension driver is unloaded. In
dvb_fini(), dev->dvb must be freed before being set to NULL, as is done
in dvb_init() in case of error.
Note that this bug is also present in the latest stable kernel release.

Signed-off-by: Francesco Lavra <francescolavra@interfree.it>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agobonding: ignore updelay param when there is no active slave
Jiri Pirko [Fri, 24 Apr 2009 03:57:29 +0000 (03:57 +0000)]
bonding: ignore updelay param when there is no active slave

commit 41f8910040639eb106b1a5b5301aab79ecde4940 upstream.

Pointed out by Sean E. Millichamp.

Quote from Documentation/networking/bonding.txt:
"Note that when a bonding interface has no active links, the
driver will immediately reuse the first link that goes up, even if the
updelay parameter has been specified (the updelay is ignored in this
case).  If there are slave interfaces waiting for the updelay timeout
to expire, the interface that first went into that state will be
immediately reused.  This reduces down time of the network if the
value of updelay has been overestimated, and since this occurs only in
cases with no connectivity, there is no additional penalty for
ignoring the updelay."

This patch actually changes the behaviour in this way.

Signed-off-by: Jiri Pirko <jpirko@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Cc: Jean Delvare <jdelvare@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agocoredump: suppress uid comparison test if core output files are pipes
Neil Horman [Fri, 5 Mar 2010 21:44:16 +0000 (13:44 -0800)]
coredump: suppress uid comparison test if core output files are pipes

commit 76595f79d76fbe6267a51b3a866a028d150f06d4 upstream.

Modify uid check in do_coredump so as to not apply it in the case of
pipes.

This just got noticed in testing.  The end of do_coredump validates the
uid of the inode for the created file against the uid of the crashing
process to ensure that no one can pre-create a core file with different
ownership and grab the information contained in the core when they
shouldn' tbe able to.  This causes failures when using pipes for a core
dumps if the crashing process is not root, which is the uid of the pipe
when it is created.

The fix is simple.  Since the check for matching uid's isn't relevant for
pipes (a process can't create a pipe that the uermodehelper code will open
anyway), we can just just skip it in the event ispipe is non-zero

Reverts a pipe-affecting change which was accidentally made in

: commit c46f739dd39db3b07ab5deb4e3ec81e1c04a91af
: Author:     Ingo Molnar <mingo@elte.hu>
: AuthorDate: Wed Nov 28 13:59:18 2007 +0100
: Commit:     Linus Torvalds <torvalds@woody.linux-foundation.org>
: CommitDate: Wed Nov 28 10:58:01 2007 -0800
:
:     vfs: coredumping fix

Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: maximilian attems <max@stro.at>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agox86, ia32_aout: do not kill argument mapping
Jiri Slaby [Wed, 10 Feb 2010 19:55:16 +0000 (20:55 +0100)]
x86, ia32_aout: do not kill argument mapping

commit 318f6b228ba88a394ef560efc1bfe028ad5ae6b6 upstream.

Do not set current->mm->mmap to NULL in 32-bit emulation on 64-bit
load_aout_binary after flush_old_exec as it would destroy already
set brpm mapping with arguments.

Introduced by b6a2fea39318e43fee84fa7b0b90d68bed92d2ba
mm: variable length argument support
where the argument mapping in bprm was added.

[ hpa: this is a regression from 2.6.22... time to kill a.out? ]

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
LKML-Reference: <1265831716-7668-1-git-send-email-jslaby@suse.cz>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ollie Wild <aaw@google.com>
Cc: x86@kernel.org
Signed-off-by: H. Peter Anvin <hpa@zytor.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agofix LOOKUP_FOLLOW on automount "symlinks"
Al Viro [Tue, 16 Feb 2010 18:09:36 +0000 (18:09 +0000)]
fix LOOKUP_FOLLOW on automount "symlinks"

commit ac278a9c505092dd82077a2446af8f9fc0d9c095 upstream.

Make sure that automount "symlinks" are followed regardless of LOOKUP_FOLLOW;
it should have no effect on them.

Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoKVM: x86: check for cr3 validity in ioctl_set_sregs
Marcelo Tosatti [Thu, 16 Apr 2009 11:30:44 +0000 (08:30 -0300)]
KVM: x86: check for cr3 validity in ioctl_set_sregs

commit 59839dfff5eabca01cc4e20b45797a60a80af8cb upstream.

Matt T. Yourst notes that kvm_arch_vcpu_ioctl_set_sregs lacks validity
checking for the new cr3 value:

"Userspace callers of KVM_SET_SREGS can pass a bogus value of cr3 to
the kernel. This will trigger a NULL pointer access in gfn_to_rmap()
when userspace next tries to call KVM_RUN on the affected VCPU and kvm
attempts to activate the new non-existent page table root.

This happens since kvm only validates that cr3 points to a valid guest
physical memory page when code *inside* the guest sets cr3. However, kvm
currently trusts the userspace caller (e.g. QEMU) on the host machine to
always supply a valid page table root, rather than properly validating
it along with the rest of the reloaded guest state."

http://sourceforge.net/tracker/?func=detail&atid=893831&aid=2687641&group_id=180599

Check for a valid cr3 address in kvm_arch_vcpu_ioctl_set_sregs, triple
fault in case of failure.

Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoparisc: isa-eeprom - Fix loff_t usage
Michael Buesch [Mon, 20 Jul 2009 22:58:44 +0000 (22:58 +0000)]
parisc: isa-eeprom - Fix loff_t usage

commit 6b4dbcd86a9d464057fcc7abe4d0574093071fcc upstream.

loff_t is a signed type. If userspace passes a negative ppos, the "count"
range check is weakened. "count"s bigger than HPEE_MAX_LENGTH will pass the check.
Also, if ppos is negative, the readb(eisa_eeprom_addr + *ppos) will poke in random
memory.

Signed-off-by: Michael Buesch <mb@bu3sch.de>
Signed-off-by: Helge Deller <deller@gmx.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agotc: Fix unitialized kernel memory leak
Eric Dumazet [Wed, 2 Sep 2009 02:40:09 +0000 (02:40 +0000)]
tc: Fix unitialized kernel memory leak

commit 16ebb5e0b36ceadc8186f71d68b0c4fa4b6e781b upstream.

Three bytes of uninitialized kernel memory are currently leaked to user

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Reviewed-by: Jiri Pirko <jpirko@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodrm/r128: Add test for initialisation to all ioctls that require it
Ben Hutchings [Sun, 23 Aug 2009 15:59:04 +0000 (16:59 +0100)]
drm/r128: Add test for initialisation to all ioctls that require it

commit 7dc482dfeeeefcfd000d4271c4626937406756d7 upstream.

Almost all r128's private ioctls require that the CCE state has
already been initialised.  However, most do not test that this has
been done, and will proceed to dereference a null pointer.  This may
result in a security vulnerability, since some ioctls are
unprivileged.

This adds a macro for the common initialisation test and changes all
ioctl implementations that require prior initialisation to use that
macro.

Also, r128_do_init_cce() does not test that the CCE state has not
been initialised already.  Repeated initialisation may lead to a crash
or resource leak.  This adds that test.

Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoKVM: VMX: Check cpl before emulating debug register access
Avi Kivity [Tue, 1 Sep 2009 09:03:25 +0000 (12:03 +0300)]
KVM: VMX: Check cpl before emulating debug register access

commit 0a79b009525b160081d75cef5dbf45817956acf2 upstream.

Debug registers may only be accessed from cpl 0.  Unfortunately, vmx will
code to emulate the instruction even though it was issued from guest
userspace, possibly leading to an unexpected trap later.

Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoext4: Avoid null pointer dereference when decoding EROFS w/o a journal
Theodore Ts'o [Tue, 28 Jul 2009 03:09:47 +0000 (23:09 -0400)]
ext4: Avoid null pointer dereference when decoding EROFS w/o a journal

commit 78f1ddbb498283c2445c11b0dfa666424c301803 upstream.

We need to check to make sure a journal is present before checking the
journal flags in ext4_decode_error().

Signed-off-by: Eric Sesterhenn <eric.sesterhenn@lsexperts.de>
Signed-off-by: "Theodore Ts'o" <tytso@mit.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoKVM: x86 emulator: limit instructions to 15 bytes
Avi Kivity [Tue, 24 Nov 2009 11:20:15 +0000 (13:20 +0200)]
KVM: x86 emulator: limit instructions to 15 bytes

commit eb3c79e64a70fb8f7473e30fa07e89c1ecc2c9bb upstream

[ <cebbert@redhat.com>: backport to 2.6.27 ]

While we are never normally passed an instruction that exceeds 15 bytes,
smp games can cause us to attempt to interpret one, which will cause
large latencies in non-preempt hosts.

Signed-off-by: Avi Kivity <avi@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: EHCI: fix counting of transaction error retries
Alan Stern [Fri, 31 Jul 2009 14:41:40 +0000 (10:41 -0400)]
USB: EHCI: fix counting of transaction error retries

commit ef4638f955f2c4a667c8af20769d03f5ed3781ca upstream.

This patch (as1274) simplifies the counting of transaction-error
retries.  Now we will count up from 0 to QH_XACTERR_MAX instead of
down from QH_XACTERR_MAX to 0.

The patch also fixes a small bug: qh->xacterr was not getting
initialized for interrupt endpoints.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Tested-by: Matthijs Kooijman <matthijs@stdin.nl>
Cc: Reinoud Koornstra <koornstra@hp.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: usbfs: properly clean up the as structure on error paths
Linus Torvalds [Tue, 16 Feb 2010 20:35:07 +0000 (12:35 -0800)]
USB: usbfs: properly clean up the as structure on error paths

commit ddeee0b2eec2a51b0712b04de4b39e7bec892a53 upstream.

I notice that the processcompl_compat() function seems to be leaking the
'struct async *as' in the error paths.

I think that the calling convention is fundamentally buggered. The
caller is the one that did the "reap_as()" to get the as thing, the
caller should be the one to free it too.

Freeing it in the caller also means that it very clearly always gets
freed, and avoids the need for any "free in the error case too".

From: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Marcus Meissner <meissner@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUSB: usbfs: only copy the actual data received
Greg KH [Mon, 15 Feb 2010 17:37:46 +0000 (09:37 -0800)]
USB: usbfs: only copy the actual data received

commit d4a4683ca054ed9917dfc9e3ff0f7ecf74ad90d6 upstream.

We need to only copy the data received by the device to userspace, not
the whole kernel buffer, which can contain "stale" data.

Thanks to Marcus Meissner for pointing this out and testing the fix.

Reported-by: Marcus Meissner <meissner@suse.de>
Tested-by: Marcus Meissner <meissner@suse.de>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoserial: 8250: add serial transmitter fully empty test
Dick Hollenbeck [Wed, 9 Dec 2009 20:31:34 +0000 (12:31 -0800)]
serial: 8250: add serial transmitter fully empty test

commit bca476139d2ded86be146dae09b06e22548b67f3 upstream.

When controlling an industrial radio modem it can be necessary to
manipulate the handshake lines in order to control the radio modem's
transmitter, from userspace.

The transmitter should not be turned off before all characters have been
transmitted.  serial8250_tx_empty() was reporting that all characters were
transmitted before they actually were.

===

Discovered in parallel with more testing and analysis by Kees Schoenmakers
as follows:

I ran into an NetMos 9835 serial pci board which behaves a little
different than the standard.  This type of expansion board is very common.

"Standard" 8250 compatible devices clear the 'UART_LST_TEMT" bit together
with the "UART_LSR_THRE" bit when writing data to the device.

The NetMos device does it slightly different

I believe that the TEMT bit is coupled to the shift register.  The problem
is that after writing data to the device and very quickly after that one
does call serial8250_tx_empty, it returns the wrong information.

My patch makes the test more robust (and solves the problem) and it does
not affect the already correct devices.

Alan:

  We may yet need to quirk this but now we know which chips we have a
  way to do that should we find this breaks some other 8250 clone with
  dodgy THRE.

Signed-off-by: Dick Hollenbeck <dick@softplc.com>
Signed-off-by: Alan Cox <alan@linux.intel.com>
Cc: Kees Schoenmakers <k.schoenmakers@sigmae.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoi2c: Do not use device name after device_unregister
Thadeu Lima de Souza Cascardo [Sun, 17 Jan 2010 18:05:58 +0000 (19:05 +0100)]
i2c: Do not use device name after device_unregister

In Linus' tree:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=c556752109794a5ff199b80a1673336b4df8433a

dev_dbg outputs dev_name, which is released with device_unregister. This bug
resulted in output like this:

i2c Xy2�0: adapter [SMBus I801 adapter at 1880] unregistered

The right output would be:
i2c i2c-0: adapter [SMBus I801 adapter at 1880] unregistered

Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@holoscopio.com>
Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agodrivers/char/mem.c: avoid OOM lockup during large reads from /dev/zero
Salman Qazi [Thu, 4 Jun 2009 22:20:39 +0000 (15:20 -0700)]
drivers/char/mem.c: avoid OOM lockup during large reads from /dev/zero

commit 730c586ad5228c339949b2eb4e72b80ae167abc4 upstream.

While running 20 parallel instances of dd as follows:

  #!/bin/bash
  for i in `seq 1 20`; do
           dd if=/dev/zero of=/export/hda3/dd_$i bs=1073741824 count=1 &
  done
  wait

on a 16G machine, we noticed that rather than just killing the processes,
the entire kernel went down.  Stracing dd reveals that it first does an
mmap2, which makes 1GB worth of zero page mappings.  Then it performs a
read on those pages from /dev/zero, and finally it performs a write.

The machine died during the reads.  Looking at the code, it was noticed
that /dev/zero's read operation had been changed by
557ed1fa2620dc119adb86b34c614e152a629a80 ("remove ZERO_PAGE") from giving
zero page mappings to actually zeroing the page.

The zeroing of the pages causes physical pages to be allocated to the
process.  But, when the process exhausts all the memory that it can, the
kernel cannot kill it, as it is still in the kernel mode allocating more
memory.  Consequently, the kernel eventually crashes.

To fix this, I propose that when a fatal signal is pending during
/dev/zero read operation, we simply return and let the user process die.

Signed-off-by: Salman Qazi <sqazi@google.com>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
[ Modified error return and comment trivially.  - Linus]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agosched: wakeup preempt when small overlap
Peter Zijlstra [Sat, 20 Sep 2008 21:38:02 +0000 (23:38 +0200)]
sched: wakeup preempt when small overlap

commit 15afe09bf496ae10c989e1a375a6b5da7bd3e16e upstream.

Lin Ming reported a 10% OLTP regression against 2.6.27-rc4.

The difference seems to come from different preemption agressiveness,
which affects the cache footprint of the workload and its effective
cache trashing.

Aggresively preempt a task if its avg overlap is very small, this should
avoid the task going to sleep and find it still running when we schedule
back to it - saving a wakeup.

Reported-by: Lin Ming <ming.m.lin@intel.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agosched: fine-tune SD_SIBLING_INIT
Ingo Molnar [Fri, 7 Nov 2008 15:09:23 +0000 (16:09 +0100)]
sched: fine-tune SD_SIBLING_INIT

commit 52c642f33b14bfa1b00ef2b68296effb34a573f3 upstream.

fine-tune the HT sched-domains parameters as well.

On a HT capable box, this increases lat_ctx performance from 23.87
usecs to 1.49 usecs:

 # before

 $ ./lat_ctx -s 0 2

   "size=0k ovr=1.89
    2 23.87

 # after

 $ ./lat_ctx -s 0 2

   "size=0k ovr=1.84
     2 1.49

Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agosched: fine-tune SD_MC_INIT
Mike Galbraith [Fri, 7 Nov 2008 14:26:50 +0000 (15:26 +0100)]
sched: fine-tune SD_MC_INIT

commit 14800984706bf6936bbec5187f736e928be5c218 upstream.

Tune SD_MC_INIT the same way as SD_CPU_INIT:
unset SD_BALANCE_NEWIDLE, and set SD_WAKE_BALANCE.

This improves vmark by 5%:

vmark         132102 125968 125497 messages/sec    avg 127855.66    .984
vmark         139404 131719 131272 messages/sec    avg 134131.66   1.033

Signed-off-by: Mike Galbraith <efault@gmx.de>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agosky2: Set SKY2_HW_RAM_BUFFER in sky2_init
Mike McCormack [Mon, 21 Sep 2009 04:08:52 +0000 (04:08 +0000)]
sky2: Set SKY2_HW_RAM_BUFFER in sky2_init

commit 74a61ebf653c6abe459f228eb40e9f24f7ef1fb7 upstream.

The SKY2_HW_RAM_BUFFER bit in hw->flags was checked in sky2_mac_init(),
 before being set later in sky2_up().

Setting SKY2_HW_RAM_BUFFER in sky2_init() where other hw->flags are set
 should avoid this problem recurring.

Signed-off-by: Mike McCormack <mikem@ring3k.org>
Acked-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agox86: fix csum_ipv6_magic asm memory clobber
Samuel Thibault [Thu, 1 Oct 2009 22:44:02 +0000 (15:44 -0700)]
x86: fix csum_ipv6_magic asm memory clobber

commit 392d814daf460a9564d29b2cebc51e1ea34e0504 upstream.

Just like ip_fast_csum, the assembly snippet in csum_ipv6_magic needs a
memory clobber, as it is only passed the address of the buffer, not a
memory reference to the buffer itself.

This caused failures in Hurd's pfinetv4 when we tried to compile it with
gcc-4.3 (bogus checksums).

Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Acked-by: "David S. Miller" <davem@davemloft.net>
Cc: Andi Kleen <andi@firstfloor.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agotcp: fix CONFIG_TCP_MD5SIG + CONFIG_PREEMPT timer BUG()
Robert Varga [Wed, 16 Sep 2009 06:49:21 +0000 (23:49 -0700)]
tcp: fix CONFIG_TCP_MD5SIG + CONFIG_PREEMPT timer BUG()

commit 657e9649e745b06675aa5063c84430986cdc3afa upstream.

I have recently came across a preemption imbalance detected by:

<4>huh, entered ffffffff80644630 with preempt_count 00000102, exited with 00000101?
<0>------------[ cut here ]------------
<2>kernel BUG at /usr/src/linux/kernel/timer.c:664!
<0>invalid opcode: 0000 [1] PREEMPT SMP

with ffffffff80644630 being inet_twdr_hangman().

This appeared after I enabled CONFIG_TCP_MD5SIG and played with it a
bit, so I looked at what might have caused it.

One thing that struck me as strange is tcp_twsk_destructor(), as it
calls tcp_put_md5sig_pool() -- which entails a put_cpu(), causing the
detected imbalance. Found on 2.6.23.9, but 2.6.31 is affected as well,
as far as I can tell.

Signed-off-by: Robert Varga <nite@hq.alert.sk>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agosit: fix off-by-one in ipip6_tunnel_get_prl
Sascha Hlusiak [Tue, 29 Sep 2009 11:27:05 +0000 (11:27 +0000)]
sit: fix off-by-one in ipip6_tunnel_get_prl

commit 298bf12ddb25841804f26234a43b89da1b1c0e21 upstream.

When requesting all prl entries (kprl.addr == INADDR_ANY) and there are
more prl entries than there is space passed from userspace, the existing
code would always copy cmax+1 entries, which is more than can be handled.

This patch makes the kernel copy only exactly cmax entries.

Signed-off-by: Sascha Hlusiak <contact@saschahlusiak.de>
Acked-By: Fred L. Templin <Fred.L.Templin@boeing.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agonet: unix: fix sending fds in multiple buffers
Miklos Szeredi [Fri, 11 Sep 2009 18:31:45 +0000 (11:31 -0700)]
net: unix: fix sending fds in multiple buffers

commit 8ba69ba6a324b13e1190fc31e41954d190fd4f1d upstream.

Kalle Olavi Niemitalo reported that:

  "..., when one process calls sendmsg once to send 43804 bytes of
  data and one file descriptor, and another process then calls recvmsg
  three times to receive the 16032+16032+11740 bytes, each of those
  recvmsg calls returns the file descriptor in the ancillary data.  I
  confirmed this with strace.  The behaviour differs from Linux
  2.6.26, where reportedly only one of those recvmsg calls (I think
  the first one) returned the file descriptor."

This bug was introduced by a patch from me titled "net: unix: fix inflight
counting bug in garbage collector", commit 6209344f5.

And the reason is, quoting Kalle:

  "Before your patch, unix_attach_fds() would set scm->fp = NULL, so
  that if the loop in unix_stream_sendmsg() ran multiple iterations,
  it could not call unix_attach_fds() again.  But now,
  unix_attach_fds() leaves scm->fp unchanged, and I think this causes
  it to be called multiple times and duplicate the same file
  descriptors to each struct sk_buff."

Fix this by introducing a flag that is cleared at the start and set
when the fds attached to the first buffer.  The resulting code should
work equivalently to the one on 2.6.26.

Reported-by: Kalle Olavi Niemitalo <kon@iki.fi>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoax25: Fix possible oops in ax25_make_new
Jarek Poplawski [Sun, 27 Sep 2009 10:57:02 +0000 (10:57 +0000)]
ax25: Fix possible oops in ax25_make_new

commit 8c185ab6185bf5e67766edb000ce428269364c86 upstream.

In ax25_make_new, if kmemdup of digipeat returns an error, there would
be an oops in sk_free while calling sk_destruct, because sk_protinfo
is NULL at the moment; move sk->sk_destruct initialization after this.

BTW of reported-by: Bernard Pidoux F6BVP <f6bvp@free.fr>

Signed-off-by: Jarek Poplawski <jarkao2@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoEHCI: fix bug in keeping track of resuming ports
Alan Stern [Mon, 8 Feb 2010 14:43:22 +0000 (09:43 -0500)]
EHCI: fix bug in keeping track of resuming ports

This patch fixes a bug caused by backporting commit
cec3a53c7fe794237b582e8e77fc0e48465e65ee (USB: EHCI & UHCI: fix race
between root-hub suspend and port resume) to 2.6.27.stable without
also backporting commit eafe5b99f2135488b21cf17a262c54997c44f784 (USB:
EHCI: fix remote-wakeup support for ARC/TDI core).  This extracts the
necessary changes from the earlier patch and backports them.

The symptom of the bug is that the system will fail to suspend more
than once.  The problem is caused by setting ehci->reset_done[i] but
never clearing it.  When ehci_bus_suspend() sees a nonzero value
there, it assumes this means the port is in the middle of resuming so
it aborts the bus suspend.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Cc: Corey Wright <undefined@pobox.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agor8169: Fix receive buffer length when MTU is between 1515 and 1536
Raimonds Cicans [Fri, 13 Nov 2009 10:52:19 +0000 (10:52 +0000)]
r8169: Fix receive buffer length when MTU is between 1515 and 1536

commit 8812304cf1110ae16b0778680f6022216cf4716a upstream.

In r8169 driver MTU is used to calculate receive buffer size.
Receive buffer size is used to configure hardware incoming packet filter.

For jumbo frames:
Receive buffer size = Max frame size = MTU + 14 (ethernet header) + 4
(vlan header) + 4 (ethernet checksum) = MTU + 22

Bug:
driver for all MTU up to 1536 use receive buffer size 1536

As you can see from formula, this mean all IP packets > 1536 - 22
(for vlan tagged, 1536 - 18 for not tagged) are dropped by hardware
filter.

Example:

host_good>  ifconfig eth0 mtu 1536
host_r8169> ifconfig eth0 mtu 1536
host_good>  ping host_r8169
Ok
host_good>  ping -s 1500 host_r8169
Fail
host_good>  ifconfig eth0 mtu 7000
host_r8169> ifconfig eth0 mtu 7000
host_good>  ping -s 1500 host_r8169
Ok

Bonus: got rid of magic number 8

Signed-off-by: Raimonds Cicans <ray@apollo.lv>
Signed-off-by: David S. Miller <davem@davemloft.net>
Cc: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agohwmon: (lm78) Fix I/O resource conflict with PNP
Jean Delvare [Tue, 9 Feb 2010 17:33:29 +0000 (18:33 +0100)]
hwmon: (lm78) Fix I/O resource conflict with PNP

This fix is the combination of the following two upstream patches:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=197027e6ef830d60e10f76efc8d12bf3b6c35db5
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=47c15532ddcd6818f51cb15f914d63864b3ee9ab

Only request I/O ports 0x295-0x296 instead of the full I/O address
range. This solves a conflict with PNP resources on a few motherboards.

Also request the I/O ports individually during device detection,
otherwise the PNP resource may cause the request (and thus the
detection) fail.

Signed-off-by: Jean Delvare <khali@linux-fr.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoprintk: robustify printk, fix #2
Peter Zijlstra [Wed, 20 Aug 2008 07:31:26 +0000 (09:31 +0200)]
printk: robustify printk, fix #2

commit fa33507a22623b3bd543b15a21c362cf364b6cff upstream.

Dmitry Adamushko reported:

> [*] btw., with DEBUG being enabled, pr_debug() generates [1] when
> debug_smp_processor_id() is used (CONFIG_DEBUG_PREEMPT).
>
> the problem seems to be caused by the following commit:
> commit b845b517b5e3706a3729f6ea83b88ab85f0725b0
> Author: Peter Zijlstra <a.p.zijlstra@chello.nl>
> Date:   Fri Aug 8 21:47:09 2008 +0200
>
>     printk: robustify printk
>
>
> wake_up_klogd() -> __get_cpu_var() -> smp_processor_id()
>
> and that's being called from release_console_sem() which is, in turn,
> said to be "may be called from any context" [2]
>
> and in this case, it seems to be called from some non-preemptible
> context (although, it can't be printk()...
> although, I haven't looked carefully yet).
>
> Provided [2], __get_cpu_var() is perhaps not the right solution there.
>
>
> [1]
>
> [ 7697.942005] BUG: using smp_processor_id() in preemptible [00000000] code: syslogd/3542
> [ 7697.942005] caller is wake_up_klogd+0x1b/0x50
> [ 7697.942005] Pid: 3542, comm: syslogd Not tainted 2.6.27-rc3-tip-git #2
> [ 7697.942005] Call Trace:
> [ 7697.942005]  [<ffffffff8036b398>] debug_smp_processor_id+0xe8/0xf0
> [ 7697.942005]  [<ffffffff80239d3b>] wake_up_klogd+0x1b/0x50
> [ 7697.942005]  [<ffffffff8023a047>] release_console_sem+0x1e7/0x200
> [ 7697.942005]  [<ffffffff803c0f17>] do_con_write+0xb7/0x1f30
> [ 7697.942005]  [<ffffffff8020d920>] ? show_trace+0x10/0x20
> [ 7697.942005]  [<ffffffff8020dc42>] ? dump_stack+0x72/0x80
> [ 7697.942005]  [<ffffffff8036392d>] ? __ratelimit+0xbd/0xe0
> [ 7697.942005]  [<ffffffff8036b398>] ? debug_smp_processor_id+0xe8/0xf0
> [ 7697.942005]  [<ffffffff80239d3b>] ? wake_up_klogd+0x1b/0x50
> [ 7697.942005]  [<ffffffff8023a047>] ? release_console_sem+0x1e7/0x200
> [ 7697.942005]  [<ffffffff803c2de9>] con_write+0x19/0x30
> [ 7697.942005]  [<ffffffff803b37b6>] write_chan+0x276/0x3c0
> [ 7697.942005]  [<ffffffff80232b20>] ? default_wake_function+0x0/0x10
> [ 7697.942005]  [<ffffffff804cb872>] ? _spin_lock_irqsave+0x22/0x50
> [ 7697.942005]  [<ffffffff803b1334>] tty_write+0x194/0x260
> [ 7697.942005]  [<ffffffff803b3540>] ? write_chan+0x0/0x3c0
> [ 7697.942005]  [<ffffffff803b14a4>] redirected_tty_write+0xa4/0xb0
> [ 7697.942005]  [<ffffffff803b1400>] ? redirected_tty_write+0x0/0xb0
> [ 7697.942005]  [<ffffffff802a88c2>] do_loop_readv_writev+0x52/0x80
> [ 7697.942005]  [<ffffffff802a939d>] do_readv_writev+0x1bd/0x1d0
> [ 7697.942005]  [<ffffffff802a93e9>] vfs_writev+0x39/0x60
> [ 7697.942005]  [<ffffffff802a9870>] sys_writev+0x50/0x90
> [ 7697.942005]  [<ffffffff8020bb3b>] system_call_fastpath+0x16/0x1b

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Reported-by: Dmitry Adamushko <dmitry.adamushko@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Cc: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoresource: add helpers for fetching rlimits
Jiri Slaby [Thu, 19 Nov 2009 16:16:37 +0000 (17:16 +0100)]
resource: add helpers for fetching rlimits

commit 3e10e716abf3c71bdb5d86b8f507f9e72236c9cd upstream.

We want to be sure that compiler fetches the limit variable only
once, so add helpers for fetching current and maximal resource
limits which do that.

Add them to sched.h (instead of resource.h) due to circular dependency
 sched.h->resource.h->task_struct
Alternative would be to create a separate res_access.h or similar.

Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Cc: James Morris <jmorris@namei.org>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agofutex: Handle user space corruption gracefully
Thomas Gleixner [Tue, 2 Feb 2010 10:40:27 +0000 (11:40 +0100)]
futex: Handle user space corruption gracefully

commit 51246bfd189064079c54421507236fd2723b18f3 upstream.

If the owner of a PI futex dies we fix up the pi_state and set
pi_state->owner to NULL. When a malicious or just sloppy programmed
user space application sets the futex value to 0 e.g. by calling
pthread_mutex_init(), then the futex can be acquired again. A new
waiter manages to enqueue itself on the pi_state w/o damage, but on
unlock the kernel dereferences pi_state->owner and oopses.

Prevent this by checking pi_state->owner in the unlock path. If
pi_state->owner is not current we know that user space manipulated the
futex value. Ignore the mess and return -EINVAL.

This catches the above case and also the case where a task hijacks the
futex by setting the tid value and then tries to unlock it.

Reported-by: Jermome Marchand <jmarchan@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Darren Hart <dvhltc@us.ibm.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agofutex: Handle futex value corruption gracefully
Thomas Gleixner [Wed, 3 Feb 2010 08:33:05 +0000 (09:33 +0100)]
futex: Handle futex value corruption gracefully

commit 59647b6ac3050dd964bc556fe6ef22f4db5b935c upstream.

The WARN_ON in lookup_pi_state which complains about a mismatch
between pi_state->owner->pid and the pid which we retrieved from the
user space futex is completely bogus.

The code just emits the warning and then continues despite the fact
that it detected an inconsistent state of the futex. A conveniant way
for user space to spam the syslog.

Replace the WARN_ON by a consistency check. If the values do not match
return -EINVAL and let user space deal with the mess it created.

This also fixes the missing task_pid_vnr() when we compare the
pi_state->owner pid with the futex value.

Reported-by: Jermome Marchand <jmarchan@redhat.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Darren Hart <dvhltc@us.ibm.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoFix race in tty_fasync() properly
Linus Torvalds [Sun, 7 Feb 2010 18:11:23 +0000 (10:11 -0800)]
Fix race in tty_fasync() properly

commit 80e1e823989ec44d8e35bdfddadbddcffec90424 upstream.

This reverts commit 703625118069 ("tty: fix race in tty_fasync") and
commit b04da8bfdfbb ("fnctl: f_modown should call write_lock_irqsave/
restore") that tried to fix up some of the fallout but was incomplete.

It turns out that we really cannot hold 'tty->ctrl_lock' over calling
__f_setown, because not only did that cause problems with interrupt
disables (which the second commit fixed), it also causes a potential
ABBA deadlock due to lock ordering.

Thanks to Tetsuo Handa for following up on the issue, and running
lockdep to show the problem.  It goes roughly like this:

 - f_getown gets filp->f_owner.lock for reading without interrupts
   disabled, so an interrupt that happens while that lock is held can
   cause a lockdep chain from f_owner.lock -> sighand->siglock.

 - at the same time, the tty->ctrl_lock -> f_owner.lock chain that
   commit 703625118069 introduced, together with the pre-existing
   sighand->siglock -> tty->ctrl_lock chain means that we have a lock
   dependency the other way too.

So instead of extending tty->ctrl_lock over the whole __f_setown() call,
we now just take a reference to the 'pid' structure while holding the
lock, and then release it after having done the __f_setown.  That still
guarantees that 'struct pid' won't go away from under us, which is all
we really ever needed.

Reported-and-tested-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Acked-by: Greg Kroah-Hartman <gregkh@suse.de>
Acked-by: Américo Wang <xiyou.wangcong@gmail.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoFix potential crash with sys_move_pages
Linus Torvalds [Sat, 6 Feb 2010 00:16:50 +0000 (16:16 -0800)]
Fix potential crash with sys_move_pages

commit 6f5a55f1a6c5abee15a0e878e5c74d9f1569b8b0 upstream.

We incorrectly depended on the 'node_state/node_isset()' functions
testing the node range, rather than checking it explicitly.  That's not
reliable, even if it might often happen to work.  So do the proper
explicit test.

Reported-by: Marcus Meissner <meissner@suse.de>
Acked-and-tested-by: Brice Goglin <Brice.Goglin@inria.fr>
Acked-by: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoUBI: fix volume creation input checking
Mika Westerberg [Tue, 26 Jan 2010 15:47:05 +0000 (17:47 +0200)]
UBI: fix volume creation input checking

commit c5ce5b46af76f52dea21f467397d24c4ae6cb3ff upstream.

Do not use an unchecked variable UBI_IOCMKVOL ioctl.

Signed-off-by: Mika Westerberg <ext-mika.1.westerberg@nokia.com>
Signed-off-by: Artem Bityutskiy <Artem.Bityutskiy@nokia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
14 years agoASoC: don't allow refcount to go below 0
Grazvydas Ignotas [Mon, 15 Mar 2010 19:28:25 +0000 (21:28 +0200)]
ASoC: don't allow refcount to go below 0

On certain underrun contitions we get exess stop triggers, which
were messing up state and causing new start triggers to be ignored.

14 years agoASoC: OMAP: Set minimum buffer size constraint for McBSP2 in OMAP3
Jarkko Nikula [Fri, 27 Mar 2009 13:32:01 +0000 (15:32 +0200)]
ASoC: OMAP: Set minimum buffer size constraint for McBSP2 in OMAP3

McBSP2 in OMAP3 has 1 ksample (1k x 32 bit) internal FIFO. During
initial playback startup, this FIFO is keeping the DMA request active
until the FIFO is full.

So now if ALSA buffer size is smaller, DMA is looping around it while
filling up the HW FIFO, generating burst of interrupts as well and SW
doesn't have any change to fill enough data.

Signed-off-by: Jarkko Nikula <jarkko.nikula@nokia.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
14 years agoOMAP: Make sgx_fck programmable with clk_set_rate()
Grazvydas Ignotas [Sun, 28 Feb 2010 23:32:52 +0000 (01:32 +0200)]
OMAP: Make sgx_fck programmable with clk_set_rate()

The 3.01.00.02 SGX/PVR driver wants it to be configurable,
or else just fails.

14 years agoOMAP3 clock: use pr_debug() rather than pr_info() in some clock change code
Paul Walmsley [Tue, 12 May 2009 23:27:10 +0000 (17:27 -0600)]
OMAP3 clock: use pr_debug() rather than pr_info() in some clock change code

The CORE DPLL M2 frequency change code should use pr_debug(), not
pr_info(), for its debug messages.  Same with
omap2_clksel_round_rate_div().  While here, convert a few printk(KERN_ERR ..
into pr_err().

Signed-off-by: Paul Walmsley <paul@pwsan.com>
14 years agopandora input: change key mappings again, on ED's request
Grazvydas Ignotas [Sat, 27 Feb 2010 20:39:53 +0000 (22:39 +0200)]
pandora input: change key mappings again, on ED's request

When will this end? :(

14 years agopower_supply_sysfs: Handle -ENODATA in a special way
Anton Vorontsov [Wed, 2 Dec 2009 21:24:51 +0000 (00:24 +0300)]
power_supply_sysfs: Handle -ENODATA in a special way

There are cases when some device can not report any meaningful value,
e.g. TWL4030 charger can report voltage only when charging is
active.

In these cases drivers will return -ENODATA, and we shouldn't flood
kernel log with error messages.

Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
14 years agopower_supply: bq27x00: fix voltage and current units
Grazvydas Ignotas [Sat, 27 Feb 2010 14:15:40 +0000 (16:15 +0200)]
power_supply: bq27x00: fix voltage and current units

The chip returns voltage and current in mV and mA, but
power supply class uses uV and uA, so add missing conversion.

Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>
14 years agopower_supply: bq27x00: add status and time properties
Grazvydas Ignotas [Sat, 27 Feb 2010 13:54:49 +0000 (15:54 +0200)]
power_supply: bq27x00: add status and time properties

The BQ27x00 series of chips can report time-to-empty and
time-to-full, so let's add corresponding properties.
Also report charge status based on status flag register.

Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>
14 years agopandora input: allow changing game button keycodes
Grazvydas Ignotas [Thu, 25 Feb 2010 20:52:24 +0000 (22:52 +0200)]
pandora input: allow changing game button keycodes

Do this to allow toggling between keyboard-like and gamepad
keycodes as doing both simultaneouly confuses some programs.

14 years agogpio_keys: add ability to reserve keys
Grazvydas Ignotas [Thu, 25 Feb 2010 20:48:27 +0000 (22:48 +0200)]
gpio_keys: add ability to reserve keys

Allow reserving additional keys in case we want to change
some keycodes during runtime.

14 years agoomap_hsmmc: enable verbose errors
Grazvydas Ignotas [Wed, 17 Feb 2010 16:08:44 +0000 (18:08 +0200)]
omap_hsmmc: enable verbose errors

14 years agoomap_hsmmc: backport error handling fixes from mainline
Grazvydas Ignotas [Wed, 17 Feb 2010 15:59:01 +0000 (17:59 +0200)]
omap_hsmmc: backport error handling fixes from mainline

14 years agoASoC: Fix handling of DAPM suspend work
Mark Brown [Tue, 14 Oct 2008 12:58:36 +0000 (13:58 +0100)]
ASoC: Fix handling of DAPM suspend work

Since we can query the playback stream power state directly we do not
need to infer if it is powered up from the timer being scheduled.  Doing
this avoids problems that previously existed with streams being
incorrectly determined to be powered up caused when the timer is
scheduled when streams are closed after being partially set up.

Reported-by: Nobin Mathew <nobin.mathew@gmail.com>
Reported-by: Jukka Hynninen <ext-jukka.hynninen@vaisala.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
14 years agohook the bq27500 fuel gauge
Grazvydas Ignotas [Sat, 13 Feb 2010 16:36:49 +0000 (18:36 +0200)]
hook the bq27500 fuel gauge

14 years agopower_supply: bq27x00: add BQ27500 support
Grazvydas Ignotas [Fri, 12 Feb 2010 20:47:18 +0000 (22:47 +0200)]
power_supply: bq27x00: add BQ27500 support

BQ27500 is newer fuel gauge chip from TI with slightly changed
register layout and some different register semantics. Add new i2c
ID for it and handle differences where needed. Also rename bq27200
functions to bq27x00, as they are no longer bq27200 specific.

Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>
14 years agopower_supply: bq27x00: fix temperature conversion
Grazvydas Ignotas [Fri, 12 Feb 2010 20:00:20 +0000 (22:00 +0200)]
power_supply: bq27x00: fix temperature conversion

The power supply class requires tenths of degree Celsius.

Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>
14 years agopower_supply: bq27x00: remove unused struct fields
Grazvydas Ignotas [Fri, 12 Feb 2010 21:31:41 +0000 (23:31 +0200)]
power_supply: bq27x00: remove unused struct fields

These are probably leftover from old TI code.

Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>