7 years agoARM: tegra: cardhu: open drain pin need not to set tristate
Laxman Dewangan [Tue, 24 Apr 2012 11:23:38 +0000]
ARM: tegra: cardhu: open drain pin need not to set tristate

When open drian pin is set as gpio-input, the pin is set as
tristate and hence need not to set this again tristate from
pinmux controller.
Setting the pin in normal in pinmux controller and then
- setting HIGH by gpio-input and pull-up so that pin is
  tristated through gpio controller.
- Setting LOW by gpio-output and drive to LOW. As pin is in
  normal state in the pinmux, the output will be set to LOW.

bug 973591
bug 969182

Change-Id: Ia9518f79987c9562bb57f95a468bdc5b5e143b87
Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Reviewed-on: http://git-master/r/98434
Reviewed-by: Rakesh Bodla <rbodla@nvidia.com>
Tested-by: Rakesh Bodla <rbodla@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Venkat Moganty <vmoganty@nvidia.com>

7 years agoARM: tegra: cardhu: pm299: off core-power in LP0
Laxman Dewangan [Tue, 24 Apr 2012 10:09:58 +0000]
ARM: tegra: cardhu: pm299: off core-power in LP0

Making core power to be off in LP0.

bug 973979

Change-Id: I0c8f44fce2204395530c63fd0056f98a3c4ef115
Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Reviewed-on: http://git-master/r/98413
Reviewed-by: Automatic_Commit_Validation_User

7 years agopower: max17048: configure hibernate mode
Syed Rafiuddin [Fri, 20 Apr 2012 11:20:10 +0000]
power: max17048: configure hibernate mode

configure fuel-gauge to enter and exit hibernate mode

Change-Id: Ifaa471a4b796fc1aa2b30f109091227eb19cf6ae
Signed-off-by: Syed Rafiuddin <srafiuddin@nvidia.com>
Reviewed-on: http://git-master/r/97900
Reviewed-by: Simone Willett <swillett@nvidia.com>
Tested-by: Simone Willett <swillett@nvidia.com>

7 years agoarm: tegra: pm: use writel instead pmc_32kwritel
Shridhar Rasal [Wed, 18 Apr 2012 04:49:57 +0000]
arm: tegra: pm: use writel instead pmc_32kwritel

For update of PMC_CTRL register delay is not required.
Replacing pmc_32kwritel by writel to improve cluster switch time.

bug 954247

Change-Id: Ic39c6fafd606321d549cf26e4cfe662f462b9bdc
Signed-off-by: Shridhar Rasal <srasal@nvidia.com>
Reviewed-on: http://git-master/r/97229
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Bharat Nihalani <bnihalani@nvidia.com>

7 years agokernel: nvhost: add OOM checks
Anton Kondratenko [Tue, 17 Apr 2012 20:52:42 +0000]
kernel: nvhost: add OOM checks

This change adds check for out of memory conditions
after memory allocations

Bug 967504

Change-Id: Icafc16528880ea376dd69a023570b85c25e3d057
Signed-off-by: Anton Kondratenko <akondratenko@nvidia.com>
Reviewed-on: http://git-master/r/97113
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>

7 years agoARM: tegra: dc: support YUV422 format
Jon Mayo [Fri, 13 May 2011 18:55:16 +0000]
ARM: tegra: dc: support YUV422 format

Change-Id: I8aa25b03fe6801882b65209cb1a6e125ef27ac2c
Signed-off-by: Michael I. Gold <gold@nvidia.com>
Reviewed-on: http://git-master/r/98319
Reviewed-by: Rohan Somvanshi <rsomvanshi@nvidia.com>
Tested-by: Rohan Somvanshi <rsomvanshi@nvidia.com>

7 years agoARM: tegra: provide fixed mapping for PCIe host
Peter De Schrijver [Wed, 25 Apr 2012 12:43:57 +0000]
ARM: tegra: provide fixed mapping for PCIe host

Provide a fixed mapping for the PCIe host registers. This reduces the pressure
on the VMALLOC area significantly.

bug 969392

Change-Id: I80ea0dd5e81a005f86a26eb47aea00d78e9e0ad2
Signed-off-by: Peter De Schrijver <pdeschrijver@nvidia.com>
Signed-off-by: Manoj Chourasia <mchourasia@nvidia.com>
Reviewed-on: http://git-master/r/96748
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Juha Tukkinen <jtukkinen@nvidia.com>
Reviewed-by: Bharat Nihalani <bnihalani@nvidia.com>



Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>

7 years agoARM: tegra: pcie: fix pcie resume issue
Shridhar Rasal [Wed, 11 Apr 2012 10:54:47 +0000]
ARM: tegra: pcie: fix pcie resume issue

Its observed that PCIE all clocks enabled on resume.
Follow up resume and suspend only if any port added

bug 943712

Change-Id: I0644aad8a4994726451cda094f2607eb8398aadf
Signed-off-by: Shridhar Rasal <srasal@nvidia.com>
Reviewed-on: http://git-master/r/95836
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Bharat Nihalani <bnihalani@nvidia.com>

7 years agoARM: tegra2: dvfs: Update DVFS rails statistic for Tegra2
Joshua Cha [Mon, 23 Apr 2012 01:13:47 +0000]
ARM: tegra2: dvfs: Update DVFS rails statistic for Tegra2

LP2 state of Tegra2 is considered into rail statistic.

Change-Id: Iab2e0fe25ecb8feca1f4aa1040ce5020e6dcf584
Signed-off-by: Joshua Cha <joshuac@nvidia.com>
Reviewed-on: http://git-master/r/98118
Reviewed-by: Rohan Somvanshi <rsomvanshi@nvidia.com>
Tested-by: Rohan Somvanshi <rsomvanshi@nvidia.com>

7 years agovideo: tegra: dc: add tracing information
Jon Mayo [Tue, 14 Feb 2012 21:12:04 +0000]
video: tegra: dc: add tracing information

add trace_printk to log useful debug information.

Bug 870685

Change-Id: I29c0b1600f234ebb06d19c8b6c713b16f6e7643c
Signed-off-by: Jon Mayo <jmayo@nvidia.com>
Reviewed-on: http://git-master/r/89204
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Prashant Malani <pmalani@nvidia.com>

7 years agomedia: video: tegra: tegra_camera: re-arch power and clock
Jihoon Bang [Mon, 26 Mar 2012 21:24:22 +0000]
media: video: tegra: tegra_camera: re-arch power and clock

There are a couple of issues found in tegra_camera.

1. clock enable/disable is controlled by user space.
   -> If client process crashes, there is no way to disable clock.
2. power enable/disable is associated with clock enable.
   -> There is no reason to relate power with clock.
   -> There is only one regulator for this driver.
   -> As same as #1, it may leave power up when client process crashes.
3. driver allows multiple clients to access.
   -> This is not the case for this driver.

This changes addresses the problems described above.

Bug 948780

Change-Id: Ie534771327175f56cf0e138f1c07096ddba470a8
Signed-off-by: Jihoon Bang <jbang@nvidia.com>
Reviewed-on: http://git-master/r/92386
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: David Schalig <dschalig@nvidia.com>
Reviewed-by: Dan Willemsen <dwillemsen@nvidia.com>

7 years agoARM: tegra: dvfs: update DVFS table for Kai
Ankit Pashiney [Tue, 10 Apr 2012 00:24:06 +0000]
ARM: tegra: dvfs: update DVFS table for Kai

update DVFS table for kai.
bug 945613

Signed-off-by: Ankit Pashiney <apashiney@nvidia.com>

Change-Id: If4d2e814bc01df1d14ae05b8f5c557d387ae19e4
Reviewed-on: http://git-master/r/95463
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Ankit Pashiney <apashiney@nvidia.com>
Tested-by: Ankit Pashiney <apashiney@nvidia.com>
Reviewed-by: Yu-Huan Hsu <yhsu@nvidia.com>

7 years agomedia: common camera header
Erik Lilliebjerg [Tue, 3 Apr 2012 05:45:49 +0000]
media: common camera header

- Add valid flag to GPIO runtime structure so that scratch
  flag can be used for other purposes.

Bug 852480

Change-Id: Ie5e36b733fe67b504e83efcf606be264725c987c
Signed-off-by: Erik Lilliebjerg <elilliebjerg@nvidia.com>
Reviewed-on: http://git-master/r/94156
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Bharat Nihalani <bnihalani@nvidia.com>

7 years agoarm: tegra: xmm: check return value of autopm_get
Shawn Joo [Mon, 12 Mar 2012 22:47:36 +0000]
arm: tegra: xmm: check return value of autopm_get

check return value of usb_autopm_get_interface().
if return value is not success, usb_autopm_put_interface() should
not be called to make up the reference count.

Bug 936094

Signed-off-by: Shawn Joo <sjoo@nvidia.com>
Reviewed-on: http://git-master/r/85974
(cherry picked from commit aad2bf5c8fef639465c4bb895b73a23c3f0c0403)

Change-Id: Ifd1deb1e0953eae1fd8d41f48989b650d6951fed
Reviewed-on: http://git-master/r/97647
Reviewed-by: Shawn Joo <sjoo@nvidia.com>
Tested-by: Shawn Joo <sjoo@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Steve Lin <stlin@nvidia.com>

7 years agovideo: tegra: nvavp: Add force clock stay on API
Hyung Taek Ryoo [Mon, 16 Apr 2012 18:33:49 +0000]
video: tegra: nvavp: Add force clock stay on API

Add nvavp_force_clock_stay_on ioctl which provides way for user-mode driver
to stay on AVP clock state. This change is to fix LP0 resume fail during
Widevine playback. Since VDE/BSEV clocks are used by OTF driver in secure
world during closing sesssion, the change makes VDE/BSEV clocks running
while entering LP0.

Bug 960130
Bug 961015

Change-Id: I7eaaa9a33537a72b6ae0a016372bc513fef532e2
Reviewed-on: http://git-master/r/96302
Reviewed-by: Hyung Taek Ryoo <hryoo@nvidia.com>
Tested-by: Hyung Taek Ryoo <hryoo@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Karan Jhavar <kjhavar@nvidia.com>
Reviewed-by: Dan Willemsen <dwillemsen@nvidia.com>

7 years agoASoC: Tegra: update codec name in dai link
Ravindra Lokhande [Wed, 25 Apr 2012 08:29:44 +0000]
ASoC: Tegra: update codec name in dai link

wm8753 codec name is changed from wm8753-codec to wm8753.
Updated machine driver to use correct name.

bug 971663

Change-Id: I087cd64583bc0fcde822937511ef278bbd4c4f15
Signed-off-by: Ravindra Lokhande <rlokhande@nvidia.com>
Reviewed-on: http://git-master/r/98677
Reviewed-by: Bharat Nihalani <bnihalani@nvidia.com>
Tested-by: Bharat Nihalani <bnihalani@nvidia.com>

7 years agonet: wireless: bcmdhd: sanity check input param in dhd_wlfc_init()
Varun Wadekar [Tue, 24 Apr 2012 11:13:35 +0000]
net: wireless: bcmdhd: sanity check input param in dhd_wlfc_init()

Bug 974673

Change-Id: I085603243c936a8235667ce5162989143429e1af
Signed-off-by: Varun Wadekar <vwadekar@nvidia.com>
Reviewed-on: http://git-master/r/98433
Reviewed-by: Rakesh Kumar <krakesh@nvidia.com>

7 years agoLinux 3.4-rc4
Linus Torvalds [Sat, 21 Apr 2012 21:47:52 +0000]
Linux 3.4-rc4

7 years agosparc32,leon: add notify_cpu_starting()
Yong Zhang [Thu, 19 Apr 2012 20:28:32 +0000]
sparc32,leon: add notify_cpu_starting()

Otherwise cpu_active_mask will not set, which lead to other issue.

Signed-off-by: Yong Zhang <yong.zhang0@gmail.com>
Signed-off-by: Konrad Eisele <konrad@gaisler.com>
Reviewed-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

7 years agodmaengine: Kconfig: fix Atmel at_hdmac entry
Nicolas Ferre [Thu, 15 Mar 2012 10:31:58 +0000]
dmaengine: Kconfig: fix Atmel at_hdmac entry

Remove SoC dependency and make it generic for every Atmel ARM AT91. That will
allow to select this driver for newer chips. Keep dependency on AT91 because of
the use of an header file located in include/mach directory.
Modify the comment to reflect this.

Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Acked-by: Vinod Koul <vinod.koul@linux.intel.com>

7 years agoUSB: gadget/at91_udc: add gpio_to_irq() function to vbus interrupt
Nicolas Ferre [Mon, 16 Apr 2012 08:58:12 +0000]
USB: gadget/at91_udc: add gpio_to_irq() function to vbus interrupt

Now that we are using irqdomains, we need to convert GPIO pins to Linux
IRQ numbers using the gpio_to_irq() function.
This call is added to request/free_irq calls.

Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Acked-by: Felipe Balbi <balbi@ti.com>
Cc: linux-usb@vger.kernel.org

7 years agoUSB: ohci-at91: change annotations for probe/remove functions
Nicolas Ferre [Fri, 6 Apr 2012 12:35:38 +0000]
USB: ohci-at91: change annotations for probe/remove functions

Add __devinit and __devexit on *_probe() and *_remove() functions
with proper modification of struct platform_driver.

Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Cc: linux-usb@vger.kernel.org

7 years agoleds-atmel-pwm.c: Make pwmled_probe() __devinit
Nicolas Ferre [Fri, 6 Apr 2012 10:52:53 +0000]
leds-atmel-pwm.c: Make pwmled_probe() __devinit

Commit 892a884 (leds: convert led platform drivers to module_platform_driver)
is omitting the section mismatch error: so change annotation of the probe
function to __devinit instead of __init.

Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Cc: Richard Purdie <richard.purdie@linuxfoundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>

7 years agoARM: at91: fix at91sam9261ek Ethernet dm9000 irq
Jean-Christophe PLAGNIOL-VILLARD [Tue, 10 Apr 2012 15:32:44 +0000]
ARM: at91: fix at91sam9261ek Ethernet dm9000 irq

You need to setup the dm9000 irq via gpio_to_irq() since
d0fbda9add (ARM: at91/gpio: drop PIN_BASE).

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Cc: stable <stable@vger.kernel.org> [3.2+]

7 years agoARM: at91: fix rm9200ek flash size
Jean-Christophe PLAGNIOL-VILLARD [Tue, 10 Apr 2012 14:02:31 +0000]
ARM: at91: fix rm9200ek flash size

The flash size is 8MiB.

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>

7 years agoARM: at91: remove empty at91_init_serial function
Joachim Eastwood [Fri, 6 Apr 2012 12:36:19 +0000]
ARM: at91: remove empty at91_init_serial function

The real function is long gone and the empty one will generate warnings
when configured without Atmel serial:
  arch/arm/mach-at91/at91rm9200_devices.c:1176: warning: 'struct at91_uart_config' declared inside parameter list
  arch/arm/mach-at91/at91rm9200_devices.c:1176: warning: its scope is only this definition or declaration, which is probably not what you want

Signed-off-by: Joachim Eastwood <joachim.eastwood@jotron.com>
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>

7 years agoARM: at91: fix typo in at91_pmc_base assembly declaration
Nicolas Ferre [Tue, 10 Apr 2012 12:30:24 +0000]
ARM: at91: fix typo in at91_pmc_base assembly declaration

Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>

7 years agoARM: at91: Export at91_matrix_base
Joachim Eastwood [Sat, 7 Apr 2012 17:30:23 +0000]
ARM: at91: Export at91_matrix_base

at91_matrix_* macro's are used by at91_udc usb gadget driver,
which can be built as module, therefore we need to export the
variable containing matrix base address.

Signed-off-by: Joachim Eastwood <joachim.eastwood@jotron.com>
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>

7 years agoARM: at91: Export at91_pmc_base
Joachim Eastwood [Sat, 7 Apr 2012 17:30:22 +0000]
ARM: at91: Export at91_pmc_base

After commit b55149529d26 (ARM: at91/PMC: make register base soc independent)
building atmel_usba_udc as a module fails with following message
  ERROR: "at91_pmc_base" [drivers/usb/gadget/atmel_usba_udc.ko] undefined!
  make[1]: *** [__modpost] Error 1
  make: *** [modules] Error 2

Export symbol to allow driver to be built as a module again.

Signed-off-by: Joachim Eastwood <joachim.eastwood@jotron.com>
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>

7 years agoARM: at91: Export at91_ramc_base
Joachim Eastwood [Fri, 30 Mar 2012 21:03:50 +0000]
ARM: at91: Export at91_ramc_base

After commit f363c40 (ARM: at91: make sdram/ddr register base soc independent)
building at91_cf as a module fails with:
  ERROR: "at91_ramc_base" [drivers/pcmcia/at91_cf.ko] undefined!
  make[1]: *** [__modpost] Error 1
  make: *** [modules] Error 2

Export at91_ramc_base symbol to allow drivers using at91_ramc_*
functions to be built as modules again.

Signed-off-by: Joachim Eastwood <joachim.eastwood@jotron.com>
[nicolas.ferre@atmel.com: modify slightly commit message]
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>

7 years agoARM: at91: Export at91_st_base
Joachim Eastwood [Wed, 4 Apr 2012 17:15:15 +0000]
ARM: at91: Export at91_st_base

After commit 5e9cf5e (ARM: at91: make ST (System Timer) soc independent)
building at91rm9200_wdt as a module fails with following message
  ERROR: "at91_st_base" [drivers/watchdog/at91rm9200_wdt.ko] undefined!
  make[1]: *** [__modpost] Error 1
  make: *** [modules] Error 2

Export symbol to allow wdt driver to be built as a module again.

Signed-off-by: Joachim Eastwood <joachim.eastwood@jotron.com>
[nicolas.ferre@atmel.com: use EXPORT_SYMBOL_GPL()]
Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com>

7 years agoARM: ux500: update defconfig
Linus Walleij [Fri, 13 Apr 2012 08:18:36 +0000]
ARM: ux500: update defconfig

Some new drivers and changed Kconfig dependencies for the v3.4
kernel affecting the U8500 defconfig:

- The SOC config options are now brought in by default
- No need to explicitly select misc devices anymore
- I2C is selected by default
- We now have support for charging from the AB8500 so compile
  in this
- The regulator framework needs to be explictly selected

Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

7 years agoARM: ux500: Fix unmet direct dependency
Lee Jones [Fri, 13 Apr 2012 13:08:01 +0000]
ARM: ux500: Fix unmet direct dependency

A recent change to a Kconfig configuration saw MACH_U8500
remove TPS6105X selection. In doing so Kconfig stopped
selecting REGULATORS, which is still required. This patch
sees UX500_SOC_DB8500 explicitly select it instead.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

7 years agoARM: ux500: wake secondary cpu via resched
Jonas Aaberg [Thu, 22 Dec 2011 08:22:56 +0000]
ARM: ux500: wake secondary cpu via resched

Wake secondary cpu via resched instead of
"Unknown IPI message 0x1"

Signed-off-by: Jonas Aaberg <jonas.aberg@stericsson.com>
Reviewed-by: Rickard Andersson <rickard.andersson@stericsson.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

7 years agoARM: imx: Fix imx5 idle logic bug
Robert Lee [Mon, 16 Apr 2012 23:37:48 +0000]
ARM: imx: Fix imx5 idle logic bug

The imx5_idle() check of the tzic_eanble_wake() return value uses
incorrect (inverted) logic causing all attempt to idle to fail.

Signed-off-by: Robert Lee <rob.lee@linaro.org>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

7 years agoARM: imx27-dt: Fix build due to removal of irq_domain_add_simple()
Fabio Estevam [Thu, 12 Apr 2012 01:12:09 +0000]
ARM: imx27-dt: Fix build due to removal of irq_domain_add_simple()

commit 6b783f7c (irq_domain: Remove irq_domain_add_simple()
replaced irq_domain_add_simple with irq_domain_add_legacy()

Implement this conversion so that imx27-dt can be built again.

Reported-by: Chris Ball <cjb@laptop.org>
Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

7 years agoARM: imx_v4_v5_defconfig: Add support for CONFIG_REGULATOR_FIXED_VOLTAGE
Fabio Estevam [Wed, 11 Apr 2012 23:24:39 +0000]
ARM: imx_v4_v5_defconfig: Add support for CONFIG_REGULATOR_FIXED_VOLTAGE


Without this option the mx27_3ds cannot have the external Ethernet functional
due to the need of smsc regulators.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>

7 years agoARM: OMAP1: DMTIMER: fix broken timer clock source selection
Paul Walmsley [Wed, 11 Apr 2012 00:36:02 +0000]
ARM: OMAP1: DMTIMER: fix broken timer clock source selection

DMTIMER source selection on OMAP1 is broken.  omap1_dm_timer_set_src()
tries to use __raw_{read,write}l() to read from and write to physical
addresses, but those functions take virtual addresses.

sparse caught this:

arch/arm/mach-omap1/timer.c:50:13: warning: incorrect type in argument 1 (different base types)
arch/arm/mach-omap1/timer.c:50:13:    expected void const volatile [noderef] <asn:2>*<noident>
arch/arm/mach-omap1/timer.c:50:13:    got unsigned int
arch/arm/mach-omap1/timer.c:52:9: warning: incorrect type in argument 1 (different base types)
arch/arm/mach-omap1/timer.c:52:9:    expected void const volatile [noderef] <asn:2>*<noident>
arch/arm/mach-omap1/timer.c:52:9:    got unsigned int

Fix by using omap_{read,writel}(), just like the other users of the
MOD_CONF_CTRL_1 register in the OMAP1 codebase.  Of course, in the long term,
removing omap_{read,write}l() is the appropriate thing to do; but
this will take some work to do this cleanly.

Looks like this was caused by 97933d6 (ARM: OMAP1: dmtimer: conversion
to platform devices) that dangerously moved code and changed it in
the same patch.

Signed-off-by: Paul Walmsley <paul@pwsan.com>
Cc: Tarun Kanti DebBarma <tarun.kanti@ti.com>
Cc: stable@vger.kernel.org
[tony@atomide.com: updated comments to include the breaking commit]
Signed-off-by: Tony Lindgren <tony@atomide.com>

7 years agoARM: OMAP: serial: Fix the ocp smart idlemode handling bug
Santosh Shilimkar [Fri, 13 Apr 2012 17:55:04 +0000]
ARM: OMAP: serial: Fix the ocp smart idlemode handling bug

The current serial UART code, while fidling with ocp idlemode bits,
forget about the smart idle wakeup bit even if it is supported by
UART IP block. This will lead to missing the module wakeup on OMAP's
where the smart idle wakeup is supported.

This was the root cause of the console sluggishness issue, I have been
observing on OMAP4 devices and also can be potential reason for some
other UART wakeup issues.

Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Acked-by: Kevin Hilman <khilman@ti.com>
Acked-by: Govindraj.R <govindraj.raja@ti.com>
Reviewed-by: Paul Walmsley <paul@pwsan.com>
Cc: stable@vger.kernel.org
Signed-off-by: Tony Lindgren <tony@atomide.com>

7 years agoARM: OMAP2+: UART: Fix incorrect population of default uart pads
Govindraj.R [Tue, 17 Apr 2012 17:35:47 +0000]
ARM: OMAP2+: UART: Fix incorrect population of default uart pads

Commit (7496ba3  ARM: OMAP2+: UART: Add default mux for all uarts)
wrongly added muxing of default pads for all uarts. This causes
breakage on multiple boards using uart pins for alternate functions.

For example, on zoom3 random oopses can be seen with nfsroot as
the smsc911x ethernet FIFO timings on GPMC bus are controlled
by gpmc_wait2 and gpmc_wait3 pins. This means we can't mux these
pads to uart4 functionality as commit 7496ba3 was doing.

Not all boards tend to use all uarts and most of unused uart pins
are muxed for other purpose. This commit breaks the modules which
where trying to use unused uart pins on their boards.

So remove the default pad muxing. Note that this is not a complete
fix, as we now rely on bootloader set muxing for the uart wake-up
events. Further patching is needed to enable wake-up events for
uarts that are already muxed to uart mode.

Cc: Felipe Balbi <balbi@ti.com>
Cc: Kevin Hilman <khilman@ti.com>
Acked-by: Russ Dill <russ.dill@gmail.com>
Reported-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Govindraj.R <govindraj.raja@ti.com>
[tony@atomide.com: updated comments to describe oops on zoom3]
Signed-off-by: Tony Lindgren <tony@atomide.com>

7 years agoARM: OMAP: sram: fix BUG in dpll code for !PM case
Grazvydas Ignotas [Fri, 6 Apr 2012 21:53:21 +0000]
ARM: OMAP: sram: fix BUG in dpll code for !PM case

_omap3_sram_configure_core_dpll is called when SDRC is reprogrammed,
which is done regardless of CONFIG_PM setting, so we always need it's
setup code too. Without this, we hit a BUG() on OMAP3 when kernel is
built without CONFIG_PM:

Reprogramming SDRC clock to 332000000 Hz
------------[ cut here ]------------
kernel BUG at arch/arm/plat-omap/sram.c:342!
Internal error: Oops - BUG: 0 [#1] ARM
[<c001c694>] (omap3_configure_core_dpll+0x68/0x6c) from [<c001b2dc>] (omap3_core_dpll_m2_set_rate+0x1)
[<c001b2dc>] (omap3_core_dpll_m2_set_rate+0x138/0x1b0) from [<c001a478>] (omap2_clk_set_rate+0x14/0x2)
[<c001a478>] (omap2_clk_set_rate+0x14/0x20) from [<c001c9dc>] (clk_set_rate+0x54/0x74)
[<c001c9dc>] (clk_set_rate+0x54/0x74) from [<c022b9c8>] (omap_sdrc_init+0x70/0x90)
[<c022b9c8>] (omap_sdrc_init+0x70/0x90) from [<c022f178>] (omap3pandora_init+0x11c/0x164)
[<c022f178>] (omap3pandora_init+0x11c/0x164) from [<c022849c>] (customize_machine+0x20/0x28)
[<c022849c>] (customize_machine+0x20/0x28) from [<c0225810>] (do_one_initcall+0xa0/0x16c)
[<c0225810>] (do_one_initcall+0xa0/0x16c) from [<c02259e0>] (kernel_init+0x104/0x1ac)
[<c02259e0>] (kernel_init+0x104/0x1ac) from [<c0009cec>] (kernel_thread_exit+0x0/0x8)

Signed-off-by: Grazvydas Ignotas <notasas@gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

7 years agoARM: OMAP2/3: VENC hwmods: Remove OCPIF_SWSUP_IDLE flag from VENC slave interface
Archit Taneja [Fri, 13 Apr 2012 08:01:55 +0000]
ARM: OMAP2/3: VENC hwmods: Remove OCPIF_SWSUP_IDLE flag from VENC slave interface

The clocks for all DSS slave interfaces were recently changed to "dss_ick" on
OMAP2 and OMAP3, this clock can be autoidled by PRCM. The VENC interface
previously had "dss_54m_fck" as it's clock which couldn't be autoidled, and
hence the OCPIF_SWSUP_IDLE flag was needed.

Remove the OCPIF_SWSUP_IDLE flag from VENC interfaces as it's clock is
now "dss_ick".  This allows the PRCM hardware to autoidle the VENC
interface clocks when they are not active, rather than relying on the
software to do it, which can keep the interface clocks active

Signed-off-by: Archit Taneja <archit@ti.com>
[paul@pwsan.com: add a short description of the fix to the commit log]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

7 years agoARM: OMAP2+: hwmod: Revert "ARM: OMAP2+: hwmod: Make omap_hwmod_softreset wait for...
Paul Walmsley [Fri, 13 Apr 2012 11:08:43 +0000]
ARM: OMAP2+: hwmod: Revert "ARM: OMAP2+: hwmod: Make omap_hwmod_softreset wait for reset status"

This reverts commit f9a2f9c3fa76eec55928e8e06f3094c8f01df7cb.  This
commit caused a regression in the I2C hwmod reset on OMAP2/3/4,
logging messages similar to these during boot:

[    0.200378] omap_hwmod: i2c1: softreset failed (waited 10000 usec)
[    0.222076] omap_hwmod: i2c2: softreset failed (waited 10000 usec)

While the original patch was intended to fix some reset-related timing
issues, it's believed that these problems were actually fixed by
commit 2800852a079504f35f88e44faf5c9c96318c0cca ("ARM: OMAP2+: hwmod:
Restore sysc after a reset"):


Cc: Rajendra Nayak <rnayak@ti.com>
Signed-off-by: Paul Walmsley <paul@pwsan.com>

7 years agoARM: OMAP2+: hwmod: add softreset delay field and OMAP4 data
Fernando Guzman Lugo [Fri, 13 Apr 2012 11:08:03 +0000]
ARM: OMAP2+: hwmod: add softreset delay field and OMAP4 data

Due to HW limitation, some IPs should not be accessed just after a
softreset. Since the current hwmod sequence is accessing the sysconfig
register just after the reset, it might lead to OCP bus error in
that case.

Add a new field in the sysconfig structure to specify a delay in usecs
needed after doing a softreset.

In the case of the ISS and FDIF modules, the L3 OCP port will be
disconnected upon a SW reset. That issue was confirmed with HW simulation
and an errata should be available soon. The HW recommendation to avoid
that is to wait for 100 OCP clk cycles, before accessing the IP.

Considering the worse case (OPP50), the L3 bus will run at 100 MHz,
so a 1 usec delay is needed. Add an x2 margin to be safe.

Acked-by: Benoit Cousson <b-cousson@ti.com>
Signed-off-by: Fernando Guzman Lugo <fernando.lugo@ti.com>
[paul@pwsan.com: dropped FDIF change for now since the hwmod data is not
 yet upstream; the FDIF change will need to be added later once the FDIF
 data is merged]
Signed-off-by: Paul Walmsley <paul@pwsan.com>

7 years agoARM: OMAP1: mux: add missing include
Aaro Koskinen [Sun, 8 Apr 2012 20:23:30 +0000]
ARM: OMAP1: mux: add missing include

Fix the following build breakage in v3.4-rc2 that happens

arch/arm/mach-omap1/mux.c:89:1: error: 'FUNC_MUX_CTRL_9' undeclared here (not in a function)
arch/arm/mach-omap1/mux.c:89:1: error: 'PULL_DWN_CTRL_2' undeclared here (not in a function)
arch/arm/mach-omap1/mux.c:93:1: error: 'FUNC_MUX_CTRL_C' undeclared here (not in a function)

Signed-off-by: Aaro Koskinen <aaro.koskinen@iki.fi>
Acked-by: Jarkko Nikula <jarkko.nikula@bitmer.com>
[tony@atomide.com: updated comments]
Signed-off-by: Tony Lindgren <tony@atomide.com>

7 years agoARM: bcmring: fix UART declarations
Paul Gortmaker [Mon, 16 Apr 2012 19:38:28 +0000]
ARM: bcmring: fix UART declarations

This error appeared in the bcmring_defconfig build:

CC arch/arm/mach-bcmring/core.o
arch/arm/mach-bcmring/core.c:55: error: macro "AMBA_APB_DEVICE" requires 6 arguments, but only 5 given
arch/arm/mach-bcmring/core.c:55: warning: type defaults to 'int' in declaration of 'AMBA_APB_DEVICE'
arch/arm/mach-bcmring/core.c:56: error: macro "AMBA_APB_DEVICE" requires 6 arguments, but only 5 given
arch/arm/mach-bcmring/core.c:56: warning: type defaults to 'int' in declaration of 'AMBA_APB_DEVICE'
arch/arm/mach-bcmring/core.c:134: error: 'uartA_device' undeclared here (not in a function)
arch/arm/mach-bcmring/core.c:135: error: 'uartB_device' undeclared here (not in a function)
make[2]: *** [arch/arm/mach-bcmring/core.o] Error 1

It appeared as of commit 8ede1ae65e61282cddba39bde4142be3885a6f5a

    "ARM: amba: bcmring: use common amba device initializers"

Note that in include/linux/amba/bus.h we have:
   #define AMBA_APB_DEVICE(name, busid, id, base, irqs, data) ...

There is an a --> A case error in the busid and a missing zero
placeholder for the id field.

Cc: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
[olof: reworded patch subject]
Signed-off-by: Olof Johansson <olof@lixom.net>

7 years agommc: core: Do not pre-claim host in suspend
Ulf Hansson [Thu, 19 Apr 2012 09:55:25 +0000]
mmc: core: Do not pre-claim host in suspend

Since SDIO drivers may want to do some SDIO operations in their suspend
callback functions, we must not keep the host claimed when calling them.

Daniel Drake reported that libertas_sdio encountered a deadlock in its
suspend function.

Signed-off-by: Ulf Hansson <ulf.hansson@stericsson.com>
Tested-by: Daniel Drake <dsd@laptop.org>
[stable@: please apply to 3.2-stable and 3.3-stable]
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Chris Ball <cjb@laptop.org>

7 years agommc: dw_mmc: prevent NULL dereference for dma_ops
Jaehoon Chung [Wed, 18 Apr 2012 06:42:31 +0000]
mmc: dw_mmc: prevent NULL dereference for dma_ops

Now, dma_ops is assumed that use the IDMAC.  But if dma_ops is assigned
the pdata->dma_ops, we didn't ensure that callback function is defined.

If the callback isn't defined, then we should run in PIO mode.

Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Acked-by: Will Newton <will.newton@imgtec.com>
Signed-off-by: Chris Ball <cjb@laptop.org>

7 years agommc: unbreak sdhci-esdhc-imx on i.MX25
Eric Bénard [Wed, 18 Apr 2012 00:30:20 +0000]
mmc: unbreak sdhci-esdhc-imx on i.MX25

This was broken by me in 37865fe91582582a6f6c00652f6a2b1ff71f8a78
("mmc: sdhci-esdhc-imx: fix timeout on i.MX's sdhci") where more
extensive tests would have shown that read or write of data to the
card were failing (even if the partition table was correctly read).

Signed-off-by: Eric Bénard <eric@eukrea.com>
Acked-by: Wolfram Sang <w.sang@pengutronix.de>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Chris Ball <cjb@laptop.org>

7 years agommc: cd-gpio: Include header to pickup exported symbol prototypes
H Hartley Sweeten [Tue, 17 Apr 2012 20:03:38 +0000]
mmc: cd-gpio: Include header to pickup exported symbol prototypes

Include the linux/mmc/cd-gpio.h header to pickup the prototypes
for the two exported symbols.

This quiets the sparse warnings:

warning: symbol 'mmc_cd_gpio_request' was not declared. Should it be static?
warning: symbol 'mmc_cd_gpio_free' was not declared. Should it be static?

Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Acked-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Acked-by: Venkatraman S <svenkatr@ti.com>
Signed-off-by: Chris Ball <cjb@laptop.org>

7 years agommc: sdhci: refine non-removable card checking for card detection
Daniel Drake [Mon, 9 Apr 2012 23:14:20 +0000]
mmc: sdhci: refine non-removable card checking for card detection

Commit c79396c191bc19 ("mmc: sdhci: prevent card detection activity
for non-removable cards") disables card detection where the cards
are marked as non-removable.

This makes sense, but the implementation detail of calling
mmc_card_is_removable() causes some problems, because
mmc_card_is_removable() is overloaded with CONFIG_MMC_UNSAFE_RESUME

In the OLPC XO case, we need CONFIG_MMC_UNSAFE_RESUME because our root
filesystem is stored on SD, but we also have external SD card slots
where we want automatic card detection.

Refine the check to only apply to hosts marked as MMC_CAP_NONREMOVABLE,
which is defined to mean that the card is *really* nonremovable. This
could be revisited in future if we find a way to improve

Signed-off-by: Daniel Drake <dsd@laptop.org>
Acked-by: Chuanxiao Dong <chuanxiao.dong@intel.com>
[stable@: please apply to 3.3-stable]
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Chris Ball <cjb@laptop.org>

7 years agommc: dw_mmc: Fix switch from DMA to PIO
Seungwon Jeon [Tue, 10 Apr 2012 00:53:32 +0000]
mmc: dw_mmc: Fix switch from DMA to PIO

When dw_mci_pre_dma_transfer returns failure in some reasons,
dw_mci_submit_data will prepare to switch the PIO mode from DMA.
After switching to PIO mode, DMA(IDMAC in particular) is still
enabled. This makes the corruption in handling interrupt and
the driver lock-up.

Signed-off-by: Seungwon Jeon <tgih.jun@samsung.com>
Acked-by: Will Newton <will.newton@imgtec.com>
Signed-off-by: Chris Ball <cjb@laptop.org>

7 years agommc: remove MMC bus legacy suspend/resume method
Chuanxiao Dong [Wed, 25 Apr 2012 06:29:21 +0000]
mmc: remove MMC bus legacy suspend/resume method

MMC bus is using legacy suspend/resume method, which is not compatible if
runtime pm callbacks are used. In this scenario, MMC bus suspend/resume
callbacks cannot be called when system entering S3. So change to use the
new defined dev_pm_ops for system sleeping mode.

Tested on AM335x Platform. Solves major issue/crash reported at

Signed-off-by: Chuanxiao Dong <chuanxiao.dong@intel.com>
Tested-by: Hebbar, Gururaja <gururaja.hebbar@ti.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Acked-by: Ulf Hansson <ulf.hansson@stericsson.com>
Signed-off-by: Chris Ball <cjb@laptop.org>



Change-Id: I0cdd31a9d15f10a6b6df08664969bf4df6afd96a

7 years agommc: omap_hsmmc: Get rid of of_have_populated_dt() usage
Rajendra Nayak [Wed, 11 Apr 2012 10:03:13 +0000]
mmc: omap_hsmmc: Get rid of of_have_populated_dt() usage

of_have_populated_dt() is not expected to be used in drivers but
instead only in early platform init code.
Drivers on the other hand should rely on dev->of_node or of_match_device().
Besides usage of of_have_populated_dt() also throws up build error as below
which was reported by Balaji TK, when omap_hsmmc is built as a module.

ERROR: "allnodes" [drivers/mmc/host/omap_hsmmc.ko] undefined!
make[1]: *** [__modpost] Error 1
make: *** [modules] Error 2

So get rid of all of_have_populated_dt() usage in omap_hsmmc driver and
instead use dev->of_node to make the same dicisions as earlier.

Signed-off-by: Rajendra Nayak <rnayak@ti.com>
Reported-by: Benoit Cousson <b-cousson@ti.com>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Acked-by: Rob Herring <rob.herring@calxeda.com>
Reviewed-by: Balaji T K <balajitk@ti.com>
Signed-off-by: Chris Ball <cjb@laptop.org>

7 years agommc: omap_hsmmc: build fix for CONFIG_OF=y and CONFIG_MMC_OMAP_HS=m
Chris Ball [Tue, 10 Apr 2012 13:57:36 +0000]
mmc: omap_hsmmc: build fix for CONFIG_OF=y and CONFIG_MMC_OMAP_HS=m

Commit 46856a68dc ("mmc: omap_hsmmc: Convert hsmmc driver to use device tree")
introduced in 3.4-rc1 has a missing semi-colon, causing:

drivers/mmc/host/omap_hsmmc.c:1745: error: expected ',' or ';' before 'extern'

Reported-by: Russell King <linux@arm.linux.org.uk>
Signed-off-by: Chris Ball <cjb@laptop.org>

7 years agommc: fixes for eMMC v4.5 sanitize operation
Adrian Hunter [Thu, 5 Apr 2012 11:45:48 +0000]
mmc: fixes for eMMC v4.5 sanitize operation

eMMC v4.5 sanitize operation erases all copies of unmapped
data.  However trim or erase operations must be used first
to unmap the required sectors.  That was not being done.

Fixes apply to linux 3.2 on.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: <stable@vger.kernel.org>
Acked-by: Jaehoon Chung <jh80.chung@samsung.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>

7 years agommc: fixes for eMMC v4.5 discard operation
Adrian Hunter [Thu, 5 Apr 2012 11:45:47 +0000]
mmc: fixes for eMMC v4.5 discard operation

eMMC v4.5 discard operation is significantly different from the
existing trim operation because it is not guaranteed to work with
the new sanitize operation.  Consequently mmc_can_trim() is
separated from mmc_can_discard().

Also the new discard operation does not result in the sectors being
set to all-zeros, so discard_zeroes_data must not be set.

In addition, the new discard has the same timeout as trim, but from
v4.5 trim is defined to use the hc timeout.  The timeout calculation
is adjusted accordingly.

Fixes apply to linux 3.2 on.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
Cc: <stable@vger.kernel.org>
Acked-by: Jaehoon Chung <jh80.chung@samsung.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Chris Ball <cjb@laptop.org>

7 years ago[media] xc5000: support 32MHz & 31.875MHz xtal using the 41.024.5 firmware
Michael Krufky [Mon, 16 Apr 2012 17:59:32 +0000]
[media] xc5000: support 32MHz & 31.875MHz xtal using the 41.024.5 firmware

Rather than loading firmware specific for the xtal frequency, just use
the standard firmware and set the xtal frequency after firmware upload.

The modified firmware will never be released, so we're better off
merging this now rather than waiting for v3.5.

Signed-off-by: Michael Krufky <mkrufky@kernellabs.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

7 years ago[media] V4L: mt9m032: fix compilation breakage
Guennadi Liakhovetski [Wed, 18 Apr 2012 06:59:01 +0000]
[media] V4L: mt9m032: fix compilation breakage

Fix the following compilation failure:

linux-2.6/drivers/media/video/mt9m032.c: In function '__mt9m032_get_pad_crop':
linux-2.6/drivers/media/video/mt9m032.c:337: error: implicit declaration of function 'v4l2_subdev_get_try_crop'
linux-2.6/drivers/media/video/mt9m032.c:337: warning: return makes pointer from integer without a cast
linux-2.6/drivers/media/video/mt9m032.c: In function '__mt9m032_get_pad_format':
linux-2.6/drivers/media/video/mt9m032.c:359: error: implicit declaration of function 'v4l2_subdev_get_try_format'
linux-2.6/drivers/media/video/mt9m032.c:359: warning: return makes pointer from integer without a cast
linux-2.6/drivers/media/video/mt9m032.c: In function 'mt9m032_probe':
linux-2.6/drivers/media/video/mt9m032.c:767: error: 'struct v4l2_subdev' has no member named 'entity'
linux-2.6/drivers/media/video/mt9m032.c:826: error: 'struct v4l2_subdev' has no member named 'entity'
linux-2.6/drivers/media/video/mt9m032.c: In function 'mt9m032_remove':
linux-2.6/drivers/media/video/mt9m032.c:842: error: 'struct v4l2_subdev' has no member named 'entity'
make[4]: *** [drivers/media/video/mt9m032.o] Error 1

by adding a dependency on VIDEO_V4L2_SUBDEV_API.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

7 years ago[media] V4L: DocBook: Fix typos in the multi-plane formats description
Sylwester Nawrocki [Mon, 9 Apr 2012 12:31:56 +0000]
[media] V4L: DocBook: Fix typos in the multi-plane formats description

Signed-off-by: Sylwester Nawrocki <sylvester.nawrocki@gmail.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

7 years ago[media] V4L: mt9m032: fix two dead-locks
Guennadi Liakhovetski [Sun, 8 Apr 2012 19:31:24 +0000]
[media] V4L: mt9m032: fix two dead-locks

Fix a copy-paste typo and a nested locking function call in mt9m032.

Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

7 years ago[media] rc-core: set mode for winbond-cir
David Härdeman [Sun, 8 Apr 2012 09:13:04 +0000]
[media] rc-core: set mode for winbond-cir

Setting the correct mode is required by rc-core or scancodes won't be
generated (which isn't very user-friendly).

This one-line fix should be suitable for 3.4-rc2.

Signed-off-by: David Härdeman <david@hardeman.nu>
Cc: stable@kernel.org
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

7 years ago[media] drxk: Does not unlock mutex if sanity check failed in scu_command()
Alexey Khoroshilov [Thu, 5 Apr 2012 21:53:20 +0000]
[media] drxk: Does not unlock mutex if sanity check failed in scu_command()

If sanity check fails in scu_command(), goto error leads to unlock of
an unheld mutex. The check should not fail in reality, but it nevertheless
worth fixing.

Found by Linux Driver Verification project (linuxtesting.org).

Signed-off-by: Alexey Khoroshilov <khoroshilov@ispras.ru>
Cc: stable@kernel.org
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

7 years ago[media] dvb_frontend: Fix a regression when switching back to DVB-S
Mauro Carvalho Chehab [Tue, 17 Apr 2012 21:32:19 +0000]
[media] dvb_frontend: Fix a regression when switching back to DVB-S

There are some softwares (Kaffeine and likely xine) that uses a
DVBv5 call to switch to DVB-S2, but expects that a DVBv3 call to
switch back to DVB-S. Well, this is not right, as a DVBv3 call
doesn't know anything about delivery systems.

However, as, by accident, this used to work, we need to restore its
behavior, in order to avoid regressions with those softwares.

Reported on this Fedora 16 bugzilla:

Reported-by: Dieter Roever <Dieter.Roever@gmx.de>
Cc: stable@kernel.org # for version 3.3
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

7 years agomfd: Fix modular builds of rc5t583 regulator support
Paul Gortmaker [Mon, 16 Apr 2012 19:24:32 +0000]
mfd: Fix modular builds of rc5t583 regulator support

The combination of commit 1b1247dd75aa5cf5fae54a3bec7280046e9c7957

    "mfd: Add support for RICOH PMIC RC5T583"

and commit 6ffc3270210efa2bea526953a142ffc908f5bd86

    "regulator: Add support for RICOH PMIC RC5T583 regulator"

are causing the i386 allmodconfig builds to fail with this:

  ERROR: "rc5t583_update" [drivers/regulator/rc5t583-regulator.ko] undefined!
  ERROR: "rc5t583_set_bits" [drivers/regulator/rc5t583-regulator.ko] undefined!
  ERROR: "rc5t583_clear_bits" [drivers/regulator/rc5t583-regulator.ko] undefined!
  ERROR: "rc5t583_read" [drivers/regulator/rc5t583-regulator.ko] undefined!

and this:

  ERROR: "rc5t583_ext_power_req_config" [drivers/regulator/rc5t583-regulator.ko] undefined!

For the 1st four, make the simple ops static inline, instead of
polluting the namespace with trivial exports.  For the last one,

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>

7 years agomfd: Fix asic3_gpio_to_irq
Dmitry Artamonow [Thu, 12 Apr 2012 11:33:34 +0000]
mfd: Fix asic3_gpio_to_irq

Assumption that irq numbers of asic3 gpios start at
IRQ_BOARD_START is certainly wrong - driver may as well
use any other base for its irqs (consider for example
the imaginary case of two ASIC3 chips onboard)

Furthermore, some platforms even don't have IRQ_BOARD_START
defined, so driver will fail to build on them:
drivers/mfd/asic3.c: In function 'asic3_gpio_to_irq':
drivers/mfd/asic3.c:530: error: 'IRQ_BOARD_START' undeclared (first use in this function)
drivers/mfd/asic3.c:530: error: (Each undeclared identifier is reported only once
drivers/mfd/asic3.c:530: error: for each function it appears in.)

Fix it by using irq_base from driver data.

Signed-off-by: Dmitry Artamonow <mad_soft@inbox.ru>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>

7 years agoARM: OMAP3: USB: Fix the EHCI ULPI PHY reset issue
Keshava Munegowda [Mon, 19 Mar 2012 06:42:47 +0000]
ARM: OMAP3: USB: Fix the EHCI ULPI PHY reset issue

It is observed that the echi ports of 3430 sdp board
are not working due to the random timing of programming
the associated GPIOs of the ULPI PHYs of the EHCI for reset.
If the PHYs are reset at during usbhs core driver, host ports will
not work because EHCI driver is loaded after the resetting PHYs.
The PHYs should be in reset state while initializing the EHCI
The code which does the GPIO pins associated with the PHYs
are programmed to reset is moved from the USB host core driver
to EHCI driver.

Signed-off-by: Keshava Munegowda <keshava_mgowda@ti.com>
Reviewed-by: Partha Basak <parthab@india.ti.com>
Acked-by: Felipe Balbi <balbi@ti.com>
Tested-by: Igor Grinberg <grinberg@compulab.co.il>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>

7 years agomfd: Convert twl6040 to i2c driver, and separate it from twl core
Peter Ujfalusi [Tue, 3 Apr 2012 08:56:51 +0000]
mfd: Convert twl6040 to i2c driver, and separate it from twl core

Complete the separation of the twl6040 from the twl core since
it is a separate chip, not part of the twl6030 PMIC.

Make the needed Kconfig changes for the depending drivers at the
same time to avoid breaking the kernel build (vibra, ASoC components).

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Reviewed-by: Mark Brown <broonie@opensource.wolfsonicro.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Acked-by: Dmitry Torokhov <dtor@mail.ru>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>

7 years agomfd : Fix dbx500 compilation error
Daniel Lezcano [Mon, 2 Apr 2012 15:55:48 +0000]
mfd : Fix dbx500 compilation error

The ux500 default config enables the db5500 and the db8500.
The incoming cpuidle driver uses the 'prcmu_enable_wakeups'
and the 'prcmu_set_power_state' functions but these ones
are defined but not implemented for the db5500, leading to
an unresolved symbol error at link time. In order to compile,
we have to disable the db5500 support which is not acceptable
for the default config.

I noticed there are also some other functions which are
defined but not implemented.

This patch fix this by removing the functions definitions
and move out of the config section the empty functions which
are normally used when the DB550 config is disabled.
Only the functions which are not implemented are concerned
by this modification.

Signed-off-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>

7 years agokill mm argument of vm_munmap()
Al Viro [Sat, 21 Apr 2012 01:57:04 +0000]
kill mm argument of vm_munmap()

it's always current->mm

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

7 years agoperfmon: kill some helpers and arguments
Al Viro [Sat, 21 Apr 2012 01:53:35 +0000]
perfmon: kill some helpers and arguments

pfm_vm_munmap() is simply vm_munmap() and pfm_remove_smpl_mapping()
always get current as the first argument.

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

7 years agoaio: don't bother with unmapping when aio_free_ring() is coming from exit_aio()
Al Viro [Sat, 21 Apr 2012 01:49:41 +0000]
aio: don't bother with unmapping when aio_free_ring() is coming from exit_aio()

... since exit_mmap() is coming and it will munmap() everything anyway.
In all other cases aio_free_ring() has ctx->mm == current->mm; moreover,
all other callers of vm_munmap() have mm == current->mm, so this will
allow us to get rid of mm argument of vm_munmap().

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

7 years agoVM: add "vm_mmap()" helper function
Linus Torvalds [Sat, 21 Apr 2012 00:13:58 +0000]
VM: add "vm_mmap()" helper function

This continues the theme started with vm_brk() and vm_munmap():
vm_mmap() does the same thing as do_mmap(), but additionally does the
required VM locking.

This uninlines (and rewrites it to be clearer) do_mmap(), which sadly
duplicates it in mm/mmap.c and mm/nommu.c.  But that way we don't have
to export our internal do_mmap_pgoff() function.

Some day we hopefully don't have to export do_mmap() either, if all
modular users can become the simpler vm_mmap() instead.  We're actually
very close to that already, with the notable exception of the (broken)
use in i810, and a couple of stragglers in binfmt_elf.

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

7 years agoVM: add "vm_munmap()" helper function
Linus Torvalds [Fri, 20 Apr 2012 23:20:01 +0000]
VM: add "vm_munmap()" helper function

Like the vm_brk() function, this is the same as "do_munmap()", except it
does the VM locking for the caller.

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

7 years agoVM: add "vm_brk()" helper function
Linus Torvalds [Fri, 20 Apr 2012 22:35:40 +0000]
VM: add "vm_brk()" helper function

It does the same thing as "do_brk()", except it handles the VM locking

It turns out that all external callers want that anyway, so we can make
do_brk() static to just mm/mmap.c while at it.

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

7 years agopinctrl: implement pinctrl_check_ops
Stephen Warren [Fri, 16 Mar 2012 20:54:25 +0000]
pinctrl: implement pinctrl_check_ops

Most code assumes that the pinctrl ops are present. Validate this when
registering a pinctrl driver. Remove the one place in the code that
was checking whether one of these non-optional ops was present.

Signed-off-by: Stephen Warren <swarren@wwwdotorg.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

7 years agopinctrl: include <linux/bug.h> to prevent compile errors
Stephen Warren [Fri, 16 Mar 2012 20:54:23 +0000]
pinctrl: include <linux/bug.h> to prevent compile errors

Macros in <linux/pinctrl/machine.h> call ARRAY_SIZE(), the definition of
which eventually calls BUILD_BUG_ON_ZERO(), which is defined in
<linux/bug.h>. Include that so that every .c file using the pinctrl macros
doesn't have to do that itself.

Signed-off-by: Stephen Warren <swarren@wwwdotorg.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

7 years agopinctrl: fix compile error if not select PINMUX support
Dong Aisheng [Thu, 5 Apr 2012 09:07:23 +0000]
pinctrl: fix compile error if not select PINMUX support

The pinctrl_register_mappings is defined in core.c, so change the dependent

The compile error message is:
drivers/pinctrl/core.c:886: error: redefinition of 'pinctrl_register_mappings'
include/linux/pinctrl/machine.h:160: note: previous definition of 'pinctrl_register_mappings' was here
make[2]: *** [drivers/pinctrl/core.o] Error 1
make[1]: *** [drivers/pinctrl] Error 2
make: *** [drivers] Error 2

Acked-by: Stephen Warren <swarren@wwwdotorg.org>
Signed-off-by: Dong Aisheng <dong.aisheng@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

7 years agodrivers/tty/amiserial.c: add missing tty_unlock
Julia Lawall [Thu, 19 Apr 2012 16:12:40 +0000]
drivers/tty/amiserial.c: add missing tty_unlock

tty_unlock is used on all other exits from the function.

Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Acked-by: Jiri Slaby <jslaby@suse.cz>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agopch_uart: Fix dma channel unallocated issue
Tomoya MORINAGA [Thu, 12 Apr 2012 01:47:50 +0000]
pch_uart: Fix dma channel unallocated issue

This driver anticipates pch_uart_verify_port() is not called
during installation.
However, actually pch_uart_verify_port() is called during
As a result, memory access violation occurs like below.

0. initial value: use_dma=0
1. starup()
    - dma channel is not allocated because use_dma=0
2. pch_uart_verify_port()
    - Set use_dma=1
3. UART processing acts DMA mode because use_dma=1
     - memory access violation occurs!

This patch fixes the issue.

Whenever pch_uart_verify_port() is called and then
dma channel is not allocated, the channel should be allocated.

Signed-off-by: Tomoya MORINAGA <tomoya.rohm@gmail.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agoARM: clps711x: serial driver hungs are a result of call disable_irq within ISR
Alexander Shiyan [Tue, 27 Mar 2012 08:22:49 +0000]
ARM: clps711x: serial driver hungs are a result of call disable_irq within ISR

Since 2.6.30-rc1 clps711x serial driver hungs system. This is a result
of call disable_irq from ISR. synchronize_irq waits for end of interrupt
and goes to infinite loop. This patch fix this problem.

Signed-off-by: Alexander Shiyan <shc_work@mail.ru>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agoUSB: serial: cp210x: Fixed usb_control_msg timeout values
Yuri Matylitski [Fri, 20 Apr 2012 09:38:32 +0000]
USB: serial: cp210x: Fixed usb_control_msg timeout values

Fixed too small hardcoded timeout values for usb_control_msg
in driver for SiliconLabs cp210x-based usb-to-serial adapters.

Signed-off-by: Yuri Matylitski <ym@tekinsoft.com>
Acked-by: Kirill A. Shutemov <kirill@shutemov.name>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agoUSB: ehci-tegra: don't call set_irq_flags(IRQF_VALID)
Stephen Warren [Wed, 25 Apr 2012 06:26:06 +0000]
USB: ehci-tegra: don't call set_irq_flags(IRQF_VALID)

This call is not needed; the IRQ controller should (and does) set up
interrupts correctly. set_irq_flags() isn't exported to modules, to
this also fixes compilation of ehci-tegra.c as a module.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>



Change-Id: I841bdceb6505f9b7089b0446d910da12fdbb27bb

7 years agoUSB: yurex: Fix missing URB_NO_TRANSFER_DMA_MAP flag in urb
Tomoki Sekiyama [Thu, 29 Mar 2012 23:51:36 +0000]
USB: yurex: Fix missing URB_NO_TRANSFER_DMA_MAP flag in urb

Current probing code is setting URB_NO_TRANSFER_DMA_MAP flag into a wrong urb
structure, and this causes BUG_ON with some USB host implementations.
This patch fixes the issue.

Signed-off-by: Tomoki Sekiyama <tomoki.sekiyama@gmail.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agoUSB: yurex: Remove allocation of coherent buffer for setup-packet buffer
Tomoki Sekiyama [Thu, 29 Mar 2012 23:51:28 +0000]
USB: yurex: Remove allocation of coherent buffer for setup-packet buffer

Removes allocation of coherent buffer for the control-request setup-packet
buffer from the yurex driver. Using coherent buffers for setup-packet is
obsolete and does not work with some USB host implementations.

Signed-off-by: Tomoki Sekiyama <tomoki.sekiyama@gmail.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agodrivers/usb/misc/usbtest.c: add kfrees
Julia Lawall [Sun, 25 Mar 2012 19:08:32 +0000]
drivers/usb/misc/usbtest.c: add kfrees

Free the two previously allocated buffers before exiting the function in an
error case.

Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agoUSB: ehci-fsl: Fix kernel crash on mpc5121e
Anatolij Gustschin [Tue, 10 Apr 2012 08:48:11 +0000]
USB: ehci-fsl: Fix kernel crash on mpc5121e

Since commit 28c56ea1431421dec51b7b229369e991481453df
(powerpc/usb: fix bug of kernel hang when initializing usb)
the kernel crashes on mpc5121e. mpc5121e doesn't have system interface
registers, accessing this register address space cause the machine check
exception and a kernel crash:
[    1.294596] ehci_hcd: USB 2.0 'Enhanced' Host Controller (EHCI) Driver
[    1.316491] fsl-ehci fsl-ehci.0: Freescale On-Chip EHCI Host Controller
[    1.337334] fsl-ehci fsl-ehci.0: new USB bus registered, assigned bus number 1
[    1.358548] Machine check in kernel mode.
[    1.375917] Caused by (from SRR1=49030): Transfer error ack signal
[    1.395505] Oops: Machine check, sig: 7 [#1]
[    1.413113] MPC5121 ADS
[    1.428718] Modules linked in:
[    1.444841] NIP: c026efc4 LR: c0278b50 CTR: 00000000
[    1.463342] REGS: df837ba0 TRAP: 0200   Not tainted  (3.3.0-08839-gb5174fa)
[    1.484083] MSR: 00049030 <EE,ME,IR,DR>  CR: 42042022  XER: 20000000
[    1.504099] TASK = df834000[1] 'swapper' THREAD: df836000
[    1.509667] GPR00: 1c000000 df837c50 df834000 df9d74e0 00000003 00000010 00000000 00000000
[    1.531650] GPR08: 00000020 00000000 c037cdd8 e1088000 22042028 1001a69c 00000000 00000000
[    1.553762] GPR16: 1ffbce70 00000000 1fef5b28 1fef3e08 00000000 00000000 1ffcbc7c c045b264
[    1.575824] GPR24: 0000008b 00000002 c04a7dd0 e1088000 df33c960 df9d74e0 00000000 df9d7400
[    1.612295] NIP [c026efc4] ehci_fsl_setup_phy+0x110/0x124
[    1.632454] LR [c0278b50] ehci_fsl_setup+0x29c/0x304
[    1.652065] Call Trace:
[    1.668923] [df837c50] [c0278a40] ehci_fsl_setup+0x18c/0x304 (unreliable)
[    1.690332] [df837c70] [c025cba4] usb_add_hcd+0x1f0/0x66c
[    1.710377] [df837cb0] [c0277ab8] ehci_fsl_drv_probe+0x180/0x308
[    1.731322] [df837ce0] [c01fc7a8] platform_drv_probe+0x20/0x30
[    1.752202] [df837cf0] [c01fb0ac] driver_probe_device+0x8c/0x214
[    1.773491] [df837d10] [c01f956c] bus_for_each_drv+0x6c/0xa8
[    1.794279] [df837d40] [c01fafdc] device_attach+0xb4/0xd8
[    1.814574] [df837d60] [c01fa44c] bus_probe_device+0xa4/0xb4
[    1.835343] [df837d80] [c01f87a8] device_add+0x52c/0x5dc
[    1.855462] [df837dd0] [c01fcd58] platform_device_add+0x124/0x1d0
[    1.876558] [df837df0] [c036dcec] fsl_usb2_device_register+0xa0/0xd4
[    1.897512] [df837e10] [c036df28] fsl_usb2_mph_dr_of_probe+0x208/0x264
[    1.918253] [df837e90] [c01fc7a8] platform_drv_probe+0x20/0x30
[    1.938300] [df837ea0] [c01fb0ac] driver_probe_device+0x8c/0x214
[    1.958511] [df837ec0] [c01fb2f0] __driver_attach+0xbc/0xc0
[    1.978088] [df837ee0] [c01f9608] bus_for_each_dev+0x60/0x9c
[    1.997589] [df837f10] [c01fab88] driver_attach+0x24/0x34
[    2.016757] [df837f20] [c01fa744] bus_add_driver+0x1ac/0x274
[    2.036339] [df837f50] [c01fb898] driver_register+0x88/0x150
[    2.056052] [df837f70] [c01fcabc] platform_driver_register+0x68/0x78
[    2.076650] [df837f80] [c0446500] fsl_usb2_mph_dr_driver_init+0x18/0x28
[    2.097734] [df837f90] [c0003988] do_one_initcall+0x148/0x1b0
[    2.117934] [df837fc0] [c042d89c] kernel_init+0xfc/0x190
[    2.137667] [df837ff0] [c000d2c4] kernel_thread+0x4c/0x68
[    2.157240] Instruction dump:
[    2.174119] 90050004 4e800020 2f840003 419e0014 2f840004 409eff64 6400c000 4bffff5c
[    2.196000] 64001000 7c0004ac 812b0500 0c090000 <4c00012c> 61290200 7c0004ac 912b0500
[    2.218100] ---[ end trace 21659aedb84ad816 ]---
[    2.237089]
[    3.232940] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000007
[    3.232954]
[    3.271575] Rebooting in 1 seconds..

Check pdata->have_sysif_regs flag before accessing system interface

Signed-off-by: Anatolij Gustschin <agust@denx.de>
Cc: Shengzhou Liu <Shengzhou.Liu@freescale.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agouwb: fix error handling
Oliver Neukum [Wed, 18 Apr 2012 08:05:55 +0000]
uwb: fix error handling

Fatal errors such as a device disconnect must not trigger
error handling. The error returns must be checked.

Signed-off-by: Oliver Neukum <oneukum@suse.de>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agouwb: fix use of del_timer_sync() in interrupt
Oliver Neukum [Mon, 16 Apr 2012 13:28:28 +0000]
uwb: fix use of del_timer_sync() in interrupt

del_timer_sync() cannot be used in interrupt.
Replace it with del_timer() and a flag

Signed-off-by: Oliver Neukum <oneukum@suse.de>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agoEHCI: always clear the STS_FLR status bit
Alan Stern [Wed, 18 Apr 2012 15:33:00 +0000]
EHCI: always clear the STS_FLR status bit

This patch (as1544) fixes a problem affecting some EHCI controllers.
They can generate interrupts whenever the STS_FLR status bit is turned
on, even though that bit is masked out in the Interrupt Enable

Since the driver doesn't use STS_FLR anyway, the patch changes the
interrupt routine to clear that bit whenever it is set, rather than
leaving it alone.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Reported-and-tested-by: Tomoya MORINAGA <tomoya.rohm@gmail.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agoEHCI: fix criterion for resuming the root hub
Alan Stern [Tue, 17 Apr 2012 19:24:15 +0000]
EHCI: fix criterion for resuming the root hub

This patch (as1542) changes the criterion ehci-hcd uses to tell when
it needs to resume the controller's root hub.  A resume is needed when
a port status change is detected, obviously, but only if the root hub
is currently suspended.

Right now the driver tests whether the root hub is running, and that
is not the correct test.  In particular, if the controller has died
then the root hub should not be restarted.  In addition, some buggy
hardware occasionally requires the root hub to be running and
sending out SOF packets even while it is nominally supposed to be

In the end, the test needs to be changed.  Rather than checking whether
the root hub is currently running, the driver will now check whether
the root hub is currently suspended.  This will yield the correct
behavior in all cases.

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
CC: Peter Chen <B29397@freescale.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agoUSB: sierra: avoid QMI/wwan interface on MC77xx
Bjørn Mork [Tue, 17 Apr 2012 19:37:29 +0000]
USB: sierra: avoid QMI/wwan interface on MC77xx

These devices have a number of non serial interfaces as well.  Use
the existing "Direct IP" blacklist to prevent binding to interfaces
which are handled by other drivers.

We also extend the "Direct IP" blacklist with with interfaces only
seen in "QMI" mode, assuming that these devices use the same
interface numbers for serial interfaces both in "Direct IP" and in
"QMI" mode.

Signed-off-by: Bjørn Mork <bjorn@mork.no>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agousb: usbtest: avoid integer overflow in alloc_sglist()
Xi Wang [Mon, 9 Apr 2012 19:48:45 +0000]
usb: usbtest: avoid integer overflow in alloc_sglist()

A large `nents' from userspace could overflow the allocation size,
leading to memory corruption.

| alloc_sglist()
| usbtest_ioctl()

Use kmalloc_array() to avoid the overflow.

Signed-off-by: Xi Wang <xi.wang@gmail.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agousb: usbtest: avoid integer overflow in test_ctrl_queue()
Xi Wang [Mon, 9 Apr 2012 19:48:55 +0000]
usb: usbtest: avoid integer overflow in test_ctrl_queue()

Avoid overflowing context.count = param->sglen * param->iterations,
where both `sglen' and `iterations' are from userspace.

| test_ctrl_queue()
| usbtest_ioctl()

Keep -EOPNOTSUPP for error code.

Signed-off-by: Xi Wang <xi.wang@gmail.com>
Acked-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agoUSB: fix deadlock in bConfigurationValue attribute method
Alan Stern [Tue, 17 Apr 2012 19:22:39 +0000]
USB: fix deadlock in bConfigurationValue attribute method

This patch (as154) fixes a self-deadlock that occurs when userspace
writes to the bConfigurationValue sysfs attribute for a hub with
children.  The task tries to lock the bandwidth_mutex at a time when
it already owns the lock:

The attribute's method calls usb_set_configuration(),
which calls usb_disable_device() with the bandwidth_mutex

usb_disable_device() unregisters the existing interfaces,
which causes the hub driver to be unbound.

The hub_disconnect() routine calls hub_quiesce(), which
calls usb_disconnect() for each of the hub's children.

usb_disconnect() attempts to acquire the bandwidth_mutex
around a call to usb_disable_device().

The solution is to make usb_disable_device() acquire the mutex for
itself instead of requiring the caller to hold it.  Then the mutex can
cover only the bandwidth deallocation operation and not the region
where the interfaces are unregistered.

This has the potential to change system behavior slightly when a
config change races with another config or altsetting change.  Some of
the bandwidth released from the old config might get claimed by the
other config or altsetting, make it impossible to restore the old
config in case of a failure.  But since we don't try to recover from
config-change failures anyway, this doesn't matter.

[This should be marked for stable kernels that contain the commit
fccf4e86200b8f5edd9a65da26f150e32ba79808 "USB: Free bandwidth when
usb_disable_device is called."
That commit was marked for stable kernels as old as 2.6.32.]

Signed-off-by: Alan Stern <stern@rowland.harvard.edu>
Signed-off-by: Sarah Sharp <sarah.a.sharp@linux.intel.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

7 years agousb: gadget: eliminate NULL pointer dereference (bugfix)
Andrzej Pietrasiewicz [Wed, 28 Mar 2012 07:30:50 +0000]
usb: gadget: eliminate NULL pointer dereference (bugfix)

usb: gadget: eliminate NULL pointer dereference (bugfix)

This patch fixes a bug which causes NULL pointer dereference in
ffs_ep0_ioctl. The bug happens when the FunctionFS is not bound (either
has not been bound yet or has been bound and then unbound) and can be
reproduced with running the following commands:

$ insmod g_ffs.ko
$ mount -t functionfs func /dev/usbgadget
$ ./null

where null.c is:

#include <fcntl.h>
#include <linux/usb/functionfs.h>

int main(void)
int fd = open("/dev/usbgadget/ep0", O_RDWR);

return 0;

Signed-off-by: Andrzej Pietrasiewicz <andrzej.p@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Cc: stable@vger.kernel.org
Signed-off-by: Felipe Balbi <balbi@ti.com>

7 years agousb: gadget: uvc: Remove non-required locking from 'uvc_queue_next_buffer' routine
Bhupesh Sharma [Fri, 23 Mar 2012 16:53:13 +0000]
usb: gadget: uvc: Remove non-required locking from 'uvc_queue_next_buffer' routine

This patch removes the non-required spinlock acquire/release calls on
'queue->irqlock' from 'uvc_queue_next_buffer' routine.

This routine is called from 'video->encode' function (which translates to
either 'uvc_video_encode_bulk' or 'uvc_video_encode_isoc') in 'uvc_video.c'.
As, the 'video->encode' routines are called with 'queue->irqlock' already held,
so acquiring a 'queue->irqlock' again in 'uvc_queue_next_buffer' routine causes
a spin lock recursion.

Signed-off-by: Bhupesh Sharma <bhupesh.sharma@st.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>

7 years agousb: gadget: rndis: fix Missing req->context assignment
Lukasz Majewski [Fri, 10 Feb 2012 08:54:51 +0000]
usb: gadget: rndis: fix Missing req->context assignment

It is crucial to assign each req->context value to struct rndis.

The problem happens for multi function gadget (g_multi) when multiple
functions are calling common usb_composite_dev control request.

It might happen that *_setup method from one usb function will
alter some fields of this common request issued by other USB

Signed-off-by: Lukasz Majewski <l.majewski@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>

7 years agousb: musb: omap: fix the error check for pm_runtime_get_sync
Shubhrajyoti D [Thu, 22 Mar 2012 07:18:06 +0000]
usb: musb: omap: fix the error check for pm_runtime_get_sync

pm_runtime_get_sync returns a signed integer. In case of errors
it returns a negative value. This patch fixes the error check
by making it signed instead of unsigned thus preventing register
access if get_sync_fails. Also passes the error cause to the
debug message.

Cc: stable@vger.kernel.org
Cc: Kishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: Shubhrajyoti D <shubhrajyoti@ti.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>

7 years agousb: gadget: udc-core: fix asymmetric calls in remove_driver
Kishon Vijay Abraham I [Wed, 21 Mar 2012 16:04:30 +0000]
usb: gadget: udc-core: fix asymmetric calls in remove_driver

During modprobe of gadget driver, pullup is called after
udc_start. In order to make the exit path symmetric when
removing a gadget driver, call pullup before ->udc_stop.

This is needed to avoid issues with PM where udc_stop
disables the module completely (put IP in reset state,
cut functional and interface clocks, and so on), which
prevents us from accessing the IP's address space,
thus creating the possibility of an abort exception
when we try to access IP's address space after clocks
are off.

Cc: stable@vger.kernel.org
Signed-off-by: Partha Basak <p-basak2@ti.com>
Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: Felipe Balbi <balbi@ti.com>