GitHub/mt8127/android_kernel_alcatel_ttab.git
11 years agomfd: lpc_sch: Accomodate partial population of the MFD devices
Darren Hart [Fri, 8 Feb 2013 23:20:36 +0000 (15:20 -0800)]
mfd: lpc_sch: Accomodate partial population of the MFD devices

The current probe aborts if any of the 3 base address registers are
disabled. On a TunnelCreek system I am working on, this resulted in the
SMBIOS and GPIO devices being removed when it couldn't read the base
address for the watchdog timer.

This patch accommodates partial population of the lpc_sch_cells array and
only aborts if all the base address registers are disabled. A max size
array is allocated and the individual device cells are added to it after
their base addresses are successfully determined. This simplifies the
code a bit by removing the need for the separate tunnelcreek cells array
and combining some of the add/remove logic.

Cc: Grant Likely <grant.likely@secretlab.ca>,
Cc: Denis Turischev <denis@compulab.co.il>,
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
Cc: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: da9052-i2c: Staticize da9052_i2c_fix()
Fabio Estevam [Mon, 11 Feb 2013 20:48:01 +0000 (18:48 -0200)]
mfd: da9052-i2c: Staticize da9052_i2c_fix()

da9052_i2c_fix() is only used locally, so let it be static.

Fix the following sparse warning:

drivers/mfd/da9052-i2c.c:63:5: warning: symbol 'da9052_i2c_fix' was not declared. Should it be static?

Cc: Ashish Jangam <ashish.jangam@kpitcummins.com>
Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: syscon: Fix sparse warning
Fabio Estevam [Mon, 11 Feb 2013 20:48:00 +0000 (18:48 -0200)]
mfd: syscon: Fix sparse warning

Fix the following sparse warnings:

drivers/mfd/syscon.c:40:15: warning: symbol 'syscon_node_to_regmap' was not declared. Should it be static?
drivers/mfd/syscon.c:56:15: warning: symbol 'syscon_regmap_lookup_by_compatible' was not declared. Should it be static?
drivers/mfd/syscon.c:72:15: warning: symbol 'syscon_regmap_lookup_by_phandle' was not declared. Should it be static?

Cc: Dong Aisheng <dong.aisheng@linaro.org>
Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: twl-core: Fix kernel panic on boot
Jon Hunter [Mon, 11 Feb 2013 20:26:19 +0000 (14:26 -0600)]
mfd: twl-core: Fix kernel panic on boot

Commit 8a6aaa3 (mfd: twl-core: Collect global variables behind one
private structure (global)) removed the variable "inuse" that is used
to determine if the device has been initialised and now use the
twl_priv structure instead. This is causing the kernel to panic on
OMAP3+ devices using the twl driver, because we try to access the
twl_priv->ready member before checking if twl_priv is initialised. Fix
this and move this test to the beginning of the twl_i2c_read/write
function because twl_get_last_module() also uses the twl_priv structure.

Signed-off-by: Jon Hunter <jon-hunter@ti.com>
Acked-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agoMerge branch 'abx500-pinctrl-for-mfd' of git://git.kernel.org/pub/scm/linux/kernel...
Samuel Ortiz [Wed, 13 Feb 2013 23:24:41 +0000 (00:24 +0100)]
Merge branch 'abx500-pinctrl-for-mfd' of git://git./linux/kernel/git/linusw/linux-pinctrl into for-next

Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: rtsx: Fix issue that booting OS with SD card inserted
Wei WANG [Fri, 8 Feb 2013 07:24:27 +0000 (15:24 +0800)]
mfd: rtsx: Fix issue that booting OS with SD card inserted

Realtek card reader supports both SD and MS card. According to the
settings of rtsx MFD driver, SD host will be probed before MS host.
If we boot/reboot Linux with SD card inserted, the resetting flow of SD
card will succeed, and the following resetting flow of MS is sure to fail.
Then MS upper-level driver will ask rtsx driver to turn power off. This
request leads to the result that the following SD commands fail and SD card
can't be accessed again.

In this commit, Realtek's SD and MS host driver will check whether the card
that upper driver requesting is the one existing in the slot. If not, Realtek's
host driver will refuse the operation to make sure the exlusive accessing
at the same time.

Signed-off-by: Wei WANG <wei_wang@realsil.com.cn>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: ab8500: Fix compile error
Linus Walleij [Wed, 6 Feb 2013 22:23:01 +0000 (23:23 +0100)]
mfd: ab8500: Fix compile error

When compiling the AB8500 core driver in the latest
MFD tree the following happens:

  CC      drivers/mfd/ab8500-debugfs.o
/home/elinwal/linux-next/drivers/mfd/ab8500-debugfs.c:157:3: error: 'AB8500_SYS_CTRL1_BLOCK' undeclared here (not in a function)
/home/elinwal/linux-next/drivers/mfd/ab8500-debugfs.c:157:2: error: array index in initializer not of integer type
/home/elinwal/linux-next/drivers/mfd/ab8500-debugfs.c:157:2: error: (near initialization for 'debug_ranges')
(...)

This is due to a missing include statement, so fix
it up.

Cc: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: Add missing GENERIC_HARDIRQS dependecies
Heiko Carstens [Wed, 6 Feb 2013 16:23:51 +0000 (17:23 +0100)]
mfd: Add missing GENERIC_HARDIRQS dependecies

A lot of mfd drivers select MFD_CORE which however depends on
GENERIC_HARDIRQS support.
So add the missing dependency to all drivers to get rid of
this link error:

ERROR: "irq_create_mapping" [drivers/mfd/mfd-core.ko] undefined!

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agoMerge branch 'for-mfd' of git://git.linaro.org/people/ljones/linux-3.0-ux500 into...
Samuel Ortiz [Wed, 13 Feb 2013 23:23:17 +0000 (00:23 +0100)]
Merge branch 'for-mfd' of git://git.linaro.org/people/ljones/linux-3.0-ux500 into for-next

Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agoDocumentation: Add docs for max8925 dt
Qing Xu [Mon, 4 Feb 2013 15:40:47 +0000 (23:40 +0800)]
Documentation: Add docs for max8925 dt

add docs for dt of max8925-mfd, max8925-backlight, and
max8925-battery

Signed-off-by: Qing Xu <qingx@marvell.com>
Signed-off-by: Haojian Zhuang <haojian.zhuang@gmail.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: max8925: Add dts
Qing Xu [Mon, 4 Feb 2013 15:40:46 +0000 (23:40 +0800)]
mfd: max8925: Add dts

Add max8925 dts support into mmp2 brownstone platform

Signed-off-by: Qing Xu <qingx@marvell.com>
Signed-off-by: Haojian Zhuang <haojian.zhuang@gmail.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: max8925: Support dt for backlight
Qing Xu [Mon, 4 Feb 2013 15:40:45 +0000 (23:40 +0800)]
mfd: max8925: Support dt for backlight

Add device tree support in max8925 backlight.

Signed-off-by: Qing Xu <qingx@marvell.com>
Signed-off-by: Haojian Zhuang <haojian.zhuang@gmail.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: max8925: Fix onkey driver irq base
Qing Xu [Mon, 4 Feb 2013 15:40:44 +0000 (23:40 +0800)]
mfd: max8925: Fix onkey driver irq base

update onkey driver's irq base, it should get from max8925,
but not save in a private value

Signed-off-by: Qing Xu <qingx@marvell.com>
Acked-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: max8925: Fix mfd device register failure
Qing Xu [Mon, 4 Feb 2013 15:40:43 +0000 (23:40 +0800)]
mfd: max8925: Fix mfd device register failure

we encounter rtc/power/touch driver registry failure, root cause
it is resources confilict in insert_resouce, solved by changing
mfd_add_devices 5th parameter to NULL

Signed-off-by: Qing Xu <qingx@marvell.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: max8925: Add irqdomain for dt
Qing Xu [Mon, 4 Feb 2013 15:40:42 +0000 (23:40 +0800)]
mfd: max8925: Add irqdomain for dt

Add irqdomains for max8925's main irq, wrap irq register operations
into irqdomain's map func. it is necessary for dt support.

Also, add dt support for max8925 driver.

Signed-off-by: Qing Xu <qingx@marvell.com>
Signed-off-by: Haojian Zhuang <haojian.zhuang@gmail.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: vexpress: Allow vexpress-sysreg to self-initialise
Catalin Marinas [Mon, 4 Feb 2013 18:08:02 +0000 (18:08 +0000)]
mfd: vexpress: Allow vexpress-sysreg to self-initialise

The vexpress_sysreg_init() is a core_initcall() already and it can
trigger the early initialisation if a matching node is found. This patch
allows the SoC code to avoid calling vexpress_sysreg_of_early_init()
explicitly.

Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Pawel Moll <pawel.moll@arm.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: rtsx: Support RTS5227
Roger Tseng [Mon, 4 Feb 2013 07:45:59 +0000 (15:45 +0800)]
mfd: rtsx: Support RTS5227

Support new model RTS5227.

Signed-off-by: Roger Tseng <rogerable@realtek.com>
Reviewed-by: Wei WANG <wei_wang@realsil.com.cn>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: rtsx: Implement driving adjustment to device-dependent callbacks
Roger Tseng [Mon, 4 Feb 2013 07:45:58 +0000 (15:45 +0800)]
mfd: rtsx: Implement driving adjustment to device-dependent callbacks

Implement different ways of selecting driving capability(a necessary adjustment
along with voltage change). It was origionally in device-independent
mmc/host/rtsx_pci_sdmmc.c. Moving it here to support devices which may have a
different way of adjustment.

Signed-off-by: Roger Tseng <rogerable@realtek.com>
Reviewed-by: Wei WANG <wei_wang@realsil.com.cn>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: vexpress: Add pseudo-GPIO based LEDs
Pawel Moll [Wed, 30 Jan 2013 10:33:16 +0000 (10:33 +0000)]
mfd: vexpress: Add pseudo-GPIO based LEDs

The LEDs on the Versatile Express motherboard are controlled
through simple memory-mapped register. This patch extends
the pseudo-GPIO controller definition for these lines and
creates generic "leds-gpio" device using them

Signed-off-by: Pawel Moll <pawel.moll@arm.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: ab8500: Rename ab8500 to abx500 for hwmon driver
Hongbo Zhang [Wed, 30 Jan 2013 10:21:27 +0000 (18:21 +0800)]
mfd: ab8500: Rename ab8500 to abx500 for hwmon driver

We are using a generic abx500 hwmon layer, so rename specific ab8500 to generic
abx500 for hwmon device and driver matching.

Signed-off-by: Hongbo Zhang <hongbo.zhang@linaro.org>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: rtsx: Optimize card detect flow
Wei WANG [Tue, 29 Jan 2013 07:21:37 +0000 (15:21 +0800)]
mfd: rtsx: Optimize card detect flow

1. Schedule card detect work at the end of the ISR
2. Callback function ops->cd_deglitch may delay for a period of time.
It is not proper to call this callback when local irq disabled.
3. Card detect flow can't be executed in parallel with other card reader
operations, so it's better to be protected by mutex.

Signed-off-by: Wei WANG <wei_wang@realsil.com.cn>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: rtsx: Use macros to replace some variables
Wei WANG [Tue, 29 Jan 2013 07:21:36 +0000 (15:21 +0800)]
mfd: rtsx: Use macros to replace some variables

In function rtsx_pci_switch_clock, some variables, such as min_n, max_n,
and max_div, are not necessary. And those assigned values look very
obscure for others. It's more proper to use macro definitions here to
replace these variables.

Signed-off-by: Wei WANG <wei_wang@realsil.com.cn>
Acked-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: rtsx: Fix checkpatch warning
Wei WANG [Tue, 29 Jan 2013 07:21:35 +0000 (15:21 +0800)]
mfd: rtsx: Fix checkpatch warning

WARNING: Avoid CamelCase: <min_N>
+ u8 N, min_N, max_N, clk_divider;

WARNING: Avoid CamelCase: <max_N>
+ u8 N, min_N, max_N, clk_divider;

Signed-off-by: Wei WANG <wei_wang@realsil.com.cn>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: rtsx: Declare that the DMA address limitation is 32bit explicitly
Wei WANG [Tue, 29 Jan 2013 07:21:34 +0000 (15:21 +0800)]
mfd: rtsx: Declare that the DMA address limitation is 32bit explicitly

Realtek PCIe card reader only supports 32bit DMA.
This declaration can improve the readability.

Signed-off-by: Wei WANG <wei_wang@realsil.com.cn>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: rtsx: Remove redundant code
Wei WANG [Tue, 29 Jan 2013 07:21:33 +0000 (15:21 +0800)]
mfd: rtsx: Remove redundant code

In function rtsx_pci_add_sg_tbl, the statement "ptr++" is useless.

Signed-off-by: Wei WANG <wei_wang@realsil.com.cn>
Acked-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: rtsx: Fix typo in comment
Wei WANG [Tue, 29 Jan 2013 07:21:32 +0000 (15:21 +0800)]
mfd: rtsx: Fix typo in comment

Fix a misspelling word in comment

Signed-off-by: Wei WANG <wei_wang@realsil.com.cn>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: tps65090: remove suspend/resume callbacks
Laxman Dewangan [Tue, 29 Jan 2013 09:05:18 +0000 (14:35 +0530)]
mfd: tps65090: remove suspend/resume callbacks

The tps65090 mfd driver implement the suspend/resume callbacks
which just disable and enable irqs in suspend/resume respectively.

This operation is already done in irq suspend and irq_resume and
hence it is not require to implement the same in the driver.

Remove this non-require code.

Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: tps65090: Pass irq domain when adding mfd sub devices
Laxman Dewangan [Tue, 29 Jan 2013 09:05:17 +0000 (14:35 +0530)]
mfd: tps65090: Pass irq domain when adding mfd sub devices

When device is get added through DT then irq_base is 0 (zero)
and in this case regmap_irq_chip_get_base() generates warning.
The interrupt of this device get added through irq_domain_add_linear()
when irq_base is 0.

Hence pass the irq domain in place of base_irq when calling
mfd_add_devices().

Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: tps65090: add DT support for tps65090
Laxman Dewangan [Tue, 29 Jan 2013 09:05:15 +0000 (14:35 +0530)]
mfd: tps65090: add DT support for tps65090

Add device tree support for the TI PMIC TPS65090.
The device can be registered through platform or DT.

Add device tree binding document for this device.

Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: wm5102: Make DSP scratch registers readable
Mark Brown [Fri, 25 Jan 2013 09:09:23 +0000 (17:09 +0800)]
mfd: wm5102: Make DSP scratch registers readable

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: tps6507x: Add DT support
Vishwanathrao Badarkhe, Manish [Thu, 24 Jan 2013 10:55:17 +0000 (16:25 +0530)]
mfd: tps6507x: Add DT support

Add device tree based initialization support for TI's
tps6507x mfd device.

Signed-off-by: Vishwanathrao Badarkhe, Manish <manishv.b@ti.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: 88pm80x: Remove redundant devm_* calls
Yi Zhang [Tue, 22 Jan 2013 02:43:45 +0000 (10:43 +0800)]
mfd: 88pm80x: Remove redundant devm_* calls

devm_* functions are device managed and make error handling
and code simpler; it also fix error exit paths

Signed-off-by: Yi Zhang <yizhang@marvell.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: db8500-prcmu: Add ux500_wdt mfd_cell
Fabio Baltieri [Fri, 18 Jan 2013 11:40:14 +0000 (12:40 +0100)]
mfd: db8500-prcmu: Add ux500_wdt mfd_cell

This patch adds the necessary structures to use the watchdog
functionality of PRCMU.

The watchdog driver is named ux500_wdt.

Acked-by: Lee Jones <lee.jones@linaro.org>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Fabio Baltieri <fabio.baltieri@linaro.org>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agowatchdog: Add support for ux500_wdt watchdog
Fabio Baltieri [Tue, 29 Jan 2013 08:57:19 +0000 (09:57 +0100)]
watchdog: Add support for ux500_wdt watchdog

This patch adds support for the ux500_wdt watchdog that is found in
ST-Ericsson Ux500 platform.  The driver is based on PRCMU APIs.

Acked-by: Linus Walleij <linus.walleij@linaro.org>
Acked-by: Lee Jones <lee.jones@linaro.org>
Acked-by: Wim Van Sebroeck <wim@iguana.be>
Signed-off-by: Fabio Baltieri <fabio.baltieri@linaro.org>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: dbx500-prcmu: Export a9wdog functions
Fabio Baltieri [Fri, 18 Jan 2013 11:40:12 +0000 (12:40 +0100)]
mfd: dbx500-prcmu: Export a9wdog functions

Add EXPORT_SYMBOL to db500_prcmu_*_a9wdog functions to allow usage from
module.

Acked-by: Lee Jones <lee.jones@linaro.org>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Fabio Baltieri <fabio.baltieri@linaro.org>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: dbx500-prcmu: Add watchdog ID definitions
Fabio Baltieri [Fri, 18 Jan 2013 11:40:11 +0000 (12:40 +0100)]
mfd: dbx500-prcmu: Add watchdog ID definitions

Add definition of watchdog IDs to be used by ux500_wdt driver.

Acked-by: Lee Jones <lee.jones@linaro.org>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Fabio Baltieri <fabio.baltieri@linaro.org>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: lpc_ich: Fix gpio base and control offsets
Aaron Sierra [Thu, 24 Jan 2013 20:52:39 +0000 (14:52 -0600)]
mfd: lpc_ich: Fix gpio base and control offsets

In ICH5 and earlier the GPIOBASE and GPIOCTRL registers are found at
offsets 0x58 and 0x5C, respectively. This patch allows GPIO access to
properly be enabled (and disabled) for these chipsets.

Signed-off-by: Agócs Pál <agocs.pal.86@gmail.com>
Signed-off-by: Aaron Sierra <asierra@xes-inc.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: twl-core: Move twl_i2c_read/write_u8 to header as inline function
Peter Ujfalusi [Wed, 16 Jan 2013 13:53:59 +0000 (14:53 +0100)]
mfd: twl-core: Move twl_i2c_read/write_u8 to header as inline function

twl_i2c_read/write_u8 become as a simple wrapper over the twl_i2c_read/write.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: twl-core: Remove no longer valid comment regarding to write buffer size
Peter Ujfalusi [Wed, 16 Jan 2013 13:53:58 +0000 (14:53 +0100)]
mfd: twl-core: Remove no longer valid comment regarding to write buffer size

With the regmap conversion there is no longeer a need to allocate bigger
buffer for writes

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: twl-core: Collect global variables behind one private structure (global)
Peter Ujfalusi [Wed, 16 Jan 2013 13:53:57 +0000 (14:53 +0100)]
mfd: twl-core: Collect global variables behind one private structure (global)

Gather the global variables under a single structure and allocate it with
devm_kzalloc(). It is easier to see them and if in the future we try to add
support for multiple instance of twl in the system it is going to be much
simpler.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: twl-core: Move 'inuse' check early at probe time
Peter Ujfalusi [Wed, 16 Jan 2013 13:53:56 +0000 (14:53 +0100)]
mfd: twl-core: Move 'inuse' check early at probe time

We can fail earlier in case multiple instance of the twl-core is tried to
be loaded.
The twl-core by design only supports one instance.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: twl-core: Do not create dummy pdata when booted with DT
Peter Ujfalusi [Wed, 16 Jan 2013 13:53:55 +0000 (14:53 +0100)]
mfd: twl-core: Do not create dummy pdata when booted with DT

When booted with DT we can manage without the dummy pdata.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: twl-core: Do not try to call legacy mfd add_children() when booted with DT
Peter Ujfalusi [Wed, 16 Jan 2013 13:53:54 +0000 (14:53 +0100)]
mfd: twl-core: Do not try to call legacy mfd add_children() when booted with DT

There is really no point to retry to add children devices in case the
of_platform_populate() fails.
We do not have any information provided via pdata in this case anyways.
Depending on the boot type (legacy or DT) only execute either one.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: twl-core: Allocate twl_modules dynamically
Peter Ujfalusi [Wed, 16 Jan 2013 13:53:53 +0000 (14:53 +0100)]
mfd: twl-core: Allocate twl_modules dynamically

At boot time we can allocate the twl_modules array dynamically based on the
twl class we are using with devm_kzalloc() instead of the static
twl_modules[] array.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: twl-core: Use the lookup table to find the correct subchip for the modules
Peter Ujfalusi [Wed, 16 Jan 2013 13:53:52 +0000 (14:53 +0100)]
mfd: twl-core: Use the lookup table to find the correct subchip for the modules

Instead of using SUB_CHIP_ID* or magic numbers use the twl_mapping table to
look for the subchip ID.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: twl-core: No need to check for invalid subchip ID
Peter Ujfalusi [Wed, 16 Jan 2013 13:53:51 +0000 (14:53 +0100)]
mfd: twl-core: No need to check for invalid subchip ID

The module id table no longer can have invalid/unused entries.
No need for checking the ID for validity.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: twl-core: Clean up module id lookup and definitions
Peter Ujfalusi [Wed, 16 Jan 2013 13:53:50 +0000 (14:53 +0100)]
mfd: twl-core: Clean up module id lookup and definitions

Use enums for all module definitions:
twl_module_ids for common functionality among twl4030/twl6030
twl4030_module_ids for twl4030 specific ids
twl6030_module_ids for twl6030 specific ids

In this way the list can be managed easier when new functionality going to
be implemented.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agoARM: OMAP: zoom-display: Remove the use of TWL4030_MODULE_PWM1
Peter Ujfalusi [Wed, 16 Jan 2013 13:53:49 +0000 (14:53 +0100)]
ARM: OMAP: zoom-display: Remove the use of TWL4030_MODULE_PWM1

Use the future proof TWL_MODULE_PWM module id instead to aim the twl-core
cleanup planed for 3.9 kernel cycle.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: arizona: Provide platform data for MICBIAS configuration
Mark Brown [Mon, 28 Jan 2013 16:47:37 +0000 (00:47 +0800)]
mfd: arizona: Provide platform data for MICBIAS configuration

Allow the MICBIAS voltages and other attributes to be configured by the
platform.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: arizona: Clarify mixer underclocking error
Mark Brown [Sun, 27 Jan 2013 16:32:53 +0000 (00:32 +0800)]
mfd: arizona: Clarify mixer underclocking error

If the mixer is underclocked it will drop a sample so log that error
more directly.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: arizona: Disable interrupts during resume
Mark Brown [Sun, 27 Jan 2013 04:07:32 +0000 (12:07 +0800)]
mfd: arizona: Disable interrupts during resume

Runtime power management does not function during system suspend but the
Arizona devices need to use runtime power management to power up the device
in order to handle interrupts. Try to avoid interrupts firing during
resume by disabling the primary IRQ before interrupts are reenabled on
resume and only reenabling it again during main resume.

The goal is to avoid issues in the situation where an interrupt is asserted
during resume (eg, due to it being the wake source) and the interrupt
handling gets scheduled prior to the device being able to handle runtime
PM.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: wm5102: Refresh register defaults
Mark Brown [Sun, 27 Jan 2013 04:07:30 +0000 (12:07 +0800)]
mfd: wm5102: Refresh register defaults

The WM5102 register defaults are not up to date with the current register
map, synchronise them with those for current devices.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: wm5102: Mark DSP memory regions as volatile and readable
Mark Brown [Sun, 27 Jan 2013 04:07:23 +0000 (12:07 +0800)]
mfd: wm5102: Mark DSP memory regions as volatile and readable

We can cache some of them but this is simpler for now.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: wm8994: Use devm_regulator_bulk_get API
Sachin Kamat [Tue, 8 Jan 2013 08:31:22 +0000 (14:01 +0530)]
mfd: wm8994: Use devm_regulator_bulk_get API

devm_regulator_bulk_get is device managed and saves some cleanup
and exit code.

Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: wm5102: Update rev B patch for latest evaluation
Mark Brown [Wed, 2 Jan 2013 14:30:18 +0000 (14:30 +0000)]
mfd: wm5102: Update rev B patch for latest evaluation

The latest evaluation of the revision B silicon suggests some changes to
the tuning applied for optimal performance.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: arizona: Register MICVDD supply first to ensure no retries
Mark Brown [Thu, 20 Dec 2012 15:38:03 +0000 (15:38 +0000)]
mfd: arizona: Register MICVDD supply first to ensure no retries

Not strictly required as probe deferral will take care of everything but
it makes boot a little smoother.

Reported-by: Ryo Tsutsui <Ryo.Tsutsui@wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: wm5102: Mark only extant DSP registers volatile
Mark Brown [Tue, 11 Dec 2012 07:51:39 +0000 (16:51 +0900)]
mfd: wm5102: Mark only extant DSP registers volatile

Since regmap sometimes uses volatile as a proxy for readable simply
having a blanket condition can mark too many registers as readable.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agortc: palmas: Add RTC driver Palmas series PMIC
Laxman Dewangan [Thu, 3 Jan 2013 10:47:00 +0000 (16:17 +0530)]
rtc: palmas: Add RTC driver Palmas series PMIC

TI Palmas series PMIC support the RTC and alarm functionality.
Add RTC driver with alarm support for this device.

Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agogpio: palmas: Add support for Palmas GPIO
Laxman Dewangan [Thu, 3 Jan 2013 10:46:59 +0000 (16:16 +0530)]
gpio: palmas: Add support for Palmas GPIO

Add gpio driver for TI Palmas series PMIC. This has 8 gpio which can
work as input/output.

Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: palmas: Add APIs to access the Palmas' registers
Laxman Dewangan [Thu, 3 Jan 2013 10:46:58 +0000 (16:16 +0530)]
mfd: palmas: Add APIs to access the Palmas' registers

Palmas register set is divided into different blocks (base and offset)
and hence different i2c addresses. The i2c address offsets are derived
from base address of block of registers.

Add inline APIs to access the Palma's registers which takes the base of
register block and register offset. The i2c address offset is  derived
from the base address of register blocks.

Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: palmas: Add rtc irq number as irq resource for palmas-rtc
Laxman Dewangan [Thu, 3 Jan 2013 10:46:57 +0000 (16:16 +0530)]
mfd: palmas: Add rtc irq number as irq resource for palmas-rtc

Palma RTC is capable of generating alarm interrupt. Pass the alarm interrupt
as IRQ_RESOURCE for palmas-rtc sub device driver so that rtc driver can get
irq as platform_get_irq().

Also pass the irq domain in mfd_add_devices() to properly offset the irqs for
sub devices. This is needed when adding device through DT.

Signed-off-by: Laxman Dewangan <ldewangan@nvidia.com>
Signed-off-by: Samuel Ortiz <sameo@linux.intel.com>
11 years agomfd: ab8500: allow AB9540 based devices to use ABX500 pinctrl
Lee Jones [Wed, 16 Jan 2013 15:09:36 +0000 (15:09 +0000)]
mfd: ab8500: allow AB9540 based devices to use ABX500 pinctrl

The old AB8500 GPIO driver has been un-BROKEN and converted into a
multi-platform pinctrl driver. If any AB9540 based devices wish to
request any GPIO pins that it offers, they can after this patch.

Cc: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
11 years agomfd: ab8500: ensure new AB8500 pinctrl driver is probed correctly
Lee Jones [Wed, 16 Jan 2013 14:49:45 +0000 (14:49 +0000)]
mfd: ab8500: ensure new AB8500 pinctrl driver is probed correctly

The old, BROKEN AB8500 GPIO driver has been revamped as a shiny
new pinctrl driver and has been renamed as such. So, if we would
like to make use of it, we need to register it via its new name.

Cc: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
11 years agomfd: ab8500: provide a irq_set_type() function
Lee Jones [Mon, 7 Jan 2013 12:23:48 +0000 (12:23 +0000)]
mfd: ab8500: provide a irq_set_type() function

In the AB8500 IRQ mask and unmask functions, we rely on testing for
IRQ_TYPE_EDGE_RISING and IRQ_TYPE_EDGE_FALLING interrupts to
physically mask and unmask the correct interrupt lines. In order
for us to do that, the trigger needs to be set in the associated
flags. However, unless a irq_set_type() function pointer is passed
when registering the IRQ chip, the IRQ subsystem will refuse to do
it. For that reason, we're providing one.

Cc: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
11 years agomfd: ab8500: actually handle the AB8500 GPIO IRQs correctly
Linus Walleij [Wed, 6 Feb 2013 20:54:34 +0000 (21:54 +0100)]
mfd: ab8500: actually handle the AB8500 GPIO IRQs correctly

The patch:
"mfd: ab8500: prepare to handle AB8500 GPIO's IRQs correctly"
altered the AB8500 IRQ mask/unmask functions such that they
would handle masking on/off the falling edge IRQ if this was
requested by the consumer. However the bit mask for hwirqs
43 and 44 was shifting the bit mask incorrectly, resulting in
the wrong IRQ being mased/unmasked.

Further while the patch would mask/unmask the correct line,
when the interrupt actually came in, it would still be treated
as a valid hwirq. The offsetting applied when masking/unmasking
was not applied when handling the IRQ, i.e. the falling edge
lines were not routed back to the rising edge lines.

This fixes both cases. The end result has been tested with
the SIM detect IRQ, GPIO12, hwirq 46 and 62.

Cc: Samuel Ortiz <sameo@linux.intel.com>
Cc: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@stericsson.com>
11 years agomfd: ab8500: prepare to handle AB8500 GPIO's IRQs correctly
Lee Jones [Thu, 20 Dec 2012 11:23:42 +0000 (11:23 +0000)]
mfd: ab8500: prepare to handle AB8500 GPIO's IRQs correctly

In an upcoming patch, the gpio-ab8500 driver will relinquish all
IRQ handling capability and pass it back into the AB8500 core
driver. This will aid in reducing massive code duplication within
the kernel. Also, most of the functionality is already in the
AB8500 core driver, as the GPIO IRQs are actually sandwiched
between lots of other IRQs which the core driver already handles.

All we're doing here is providing the core driver with knowledge
that each GPIO has two IRQs assigned to it; one for rising and
a separate one for falling.

Cc: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
11 years agomfd: ab8500-gpadc: Use new ab8500_gpadc_get() with name parameter
Philippe Langlais [Wed, 18 Apr 2012 13:52:59 +0000 (15:52 +0200)]
mfd: ab8500-gpadc: Use new ab8500_gpadc_get() with name parameter

The new format of ab8500_gpadc_get() accepts a device name as a
parameter to specify which device to retrieve. This patch
enforces the use of that new format.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Philippe Langlais <philippe.langlais@linaro.org>
11 years agomfd: ab8500-gpadc: Remove unused 'struct ab8500_gpadc'
Lee Jones [Mon, 28 Jan 2013 09:21:58 +0000 (09:21 +0000)]
mfd: ab8500-gpadc: Remove unused 'struct ab8500_gpadc'

The ab8500_gpadc structure carries lots of important information which
most functions make good use of. The initial expectation was that
ab8500_gpadc_runtime_idle() would be no exception; however, this hasn't
been the case to date. Let's remove it for now and add it back in only
when we have a use for it.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
11 years agomfd: ab8500-gpadc: Reduce conversion timeout
Lee Jones [Mon, 28 Jan 2013 09:20:45 +0000 (09:20 +0000)]
mfd: ab8500-gpadc: Reduce conversion timeout

Reduce the conversion timeout from 2s to 0.5s

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Jonas Aaberg <jonas.aberg@stericsson.com>
Reviewed-by: Karl KOMIEROWSKI <karl.komierowski@stericsson.com>
11 years agomfd: ab8500-debugfs: sizeof() mismatch bugfix
Ashok G [Tue, 28 Feb 2012 04:51:00 +0000 (10:21 +0530)]
mfd: ab8500-debugfs: sizeof() mismatch bugfix

Simple pointer error fix to obtain the expected sizeof() result.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Ashok G <ashok.g@stericsson.com>
Reviewed-by: Mattias WALLIN <mattias.wallin@stericsson.com>
11 years agomfd: ab8500-gpadc: Add runtime pm support
Lee Jones [Mon, 4 Feb 2013 08:33:13 +0000 (08:33 +0000)]
mfd: ab8500-gpadc: Add runtime pm support

Add runtime pm support to speed up multiple ADC reads in a row.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Jonas Aaberg <jonas.aberg@stericsson.com>
Reviewed-by: Ulf HANSSON <ulf.hansson@stericsson.com>
11 years agomfd: ab8500-debugfs: Add interrupt debug
Bengt Jonsson [Tue, 14 Feb 2012 12:01:00 +0000 (13:01 +0100)]
mfd: ab8500-debugfs: Add interrupt debug

This patch adds an entry in debugfs to check number of interrupts
from the AB.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Bengt Jonsson <bengt.g.jonsson@stericsson.com>
Reviewed-by: Rabin VINCENT <rabin.vincent@stericsson.com>
11 years agomfd ab8500-gpadc: Introduce new AB version detection
Michel JAOUEN [Thu, 9 Feb 2012 11:06:47 +0000 (12:06 +0100)]
mfd ab8500-gpadc: Introduce new AB version detection

Add support for AB8505 and AB9540

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Maxime Coquelin <maxime.coquelin@stericsson.com>
Signed-off-by: Bengt Jonsson <bengt.g.jonsson@stericsson.com>
Reviewed-by: Rabin VINCENT <rabin.vincent@stericsson.com>
11 years agomfd: ab8500-debugfs: Allow number of IRQs to be provided more dynamically
Linus Walleij [Fri, 3 Feb 2012 10:19:05 +0000 (11:19 +0100)]
mfd: ab8500-debugfs: Allow number of IRQs to be provided more dynamically

With the introduction of new AB* platforms, it's important to allow
as much code reuse as possible. By allowing a system's number of IRQs
to be dynamically passed, we can reuse almost all of the -debugfs
driver.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@stericsson.com>
11 years agomfd: ab8500-core: Allow the possibility to dump all AB8500 registers
Mian Yousaf Kaukab [Fri, 27 Jan 2012 10:35:41 +0000 (11:35 +0100)]
mfd: ab8500-core: Allow the possibility to dump all AB8500 registers

Implement an API so that a user may dump all AB8500 registers
via debugfs file access.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Mian Yousaf Kaukab <mian.yousaf.kaukab@stericsson.com>
Reviewed-by: Linus WALLEIJ <linus.walleij@stericsson.com>
Reviewed-by: Jonas ABERG <jonas.aberg@stericsson.com>
11 years agomfd: ab8500-debugfs: add debugfs node to read all registers
Mian Yousaf Kaukab [Thu, 26 Jan 2012 14:39:20 +0000 (15:39 +0100)]
mfd: ab8500-debugfs: add debugfs node to read all registers

Update the ab8500_registers_print() to reuse it from multiple places.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Mian Yousaf Kaukab <mian.yousaf.kaukab@stericsson.com>
Reviewed-by: Linus WALLEIJ <linus.walleij@stericsson.com>
Reviewed-by: Jonas ABERG <jonas.aberg@stericsson.com>
11 years agomfd: ab8500-debugfs: Set the USB charging current to 300mA for ABV3
Philippe Langlais [Mon, 17 Oct 2011 07:48:55 +0000 (09:48 +0200)]
mfd: ab8500-debugfs: Set the USB charging current to 300mA for ABV3

In case of AB-V3, the eye diagram related issues are resolved.
So, set the device charging current to 300mA when connected to
standard host. Also, add the USB PHY tuning values to improve
the USB eye diagram

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: <thirupathi.chippakurthy@stericsson.com>
11 years agomfd: ab8500-debugfs: Export all AB8500 ADCs as debugfs nodes
John Beckett [Tue, 31 May 2011 12:54:27 +0000 (13:54 +0100)]
mfd: ab8500-debugfs: Export all AB8500 ADCs as debugfs nodes

Allow a user to take a glimpse into the inner workings of the
AB8500 Analogue-to-Digital Converters, via debugfs.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: John Beckett <john.beckett@stericsson.com>
Reviewed-by: Mattias WALLIN <mattias.wallin@stericsson.com>
11 years agomfd: ab8500-debugfs: Formated access AB8500 registers from debugfs entry
carriere etienne [Fri, 8 Apr 2011 14:26:36 +0000 (16:26 +0200)]
mfd: ab8500-debugfs: Formated access AB8500 registers from debugfs entry

Add debugfs entry ab8500/hwreg to read/write bit-field in AB8500 registers.
Check the debugfs entries usage from heading comments in ab8500-debugfs.c

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: carriere etienne <etienne.carriere@stericsson.com>
Reviewed-by: Mattias WALLIN <mattias.wallin@stericsson.com>
11 years agomfd: ab8500-debugfs: Use NULL to initialise remaining NULL pointer
Lee Jones [Mon, 14 Jan 2013 17:17:34 +0000 (17:17 +0000)]
mfd: ab8500-debugfs: Use NULL to initialise remaining NULL pointer

Partly for coding style reasons, but mostly because sparse warns on it.

This patch is a completion of a previous patch by Mark Brown.

Reviewed-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Lee Jones <lee.jones@linaro.org>
11 years agomfd: ab8500-debugfs: Keep count of IRQs in debugfs
Mattias Wallin [Fri, 19 Nov 2010 16:55:11 +0000 (17:55 +0100)]
mfd: ab8500-debugfs: Keep count of IRQs in debugfs

This patch adds a counter to the sysfs file dynamically created
by debugfs. It also fixes an array index error.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Mattias Wallin <mattias.wallin@stericsson.com>
Reviewed-by: Mattias NILSSON <mattias.i.nilsson@stericsson.com>
11 years agomfd: ab8500-debugfs: Provide a means for a user subscribe to IRQs
Lee Jones [Mon, 14 Jan 2013 16:10:36 +0000 (16:10 +0000)]
mfd: ab8500-debugfs: Provide a means for a user subscribe to IRQs

Allow users to subscribe to and view IRQ events live from debugfs.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
11 years agomfd: abx500-core: Provide an API to dump all ABx500 registers
Mian Yousaf Kaukab [Fri, 27 Jan 2012 10:22:16 +0000 (11:22 +0100)]
mfd: abx500-core: Provide an API to dump all ABx500 registers

Some drivers can detect subsystem failures e.g. shared memory driver
can detect modem sub system failures. It would be helpful in analyzing
these failures if AB register dump is available at that point. This
patch adds the API for the drivers to dump AB registers in the kernel
log.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Mian Yousaf Kaukab <mian.yousaf.kaukab@stericsson.com>
Reviewed-by: Linus WALLEIJ <linus.walleij@stericsson.com>
Reviewed-by: Jonas ABERG <jonas.aberg@stericsson.com>
11 years agomfd: ab8500-sysctrl: export read/write symbols
Jonas Aaberg [Fri, 11 Nov 2011 06:52:10 +0000 (07:52 +0100)]
mfd: ab8500-sysctrl: export read/write symbols

Export ab8500_sysctrl_read() and ab8500_sysctrl_write() symobols.

They will be used by LTP test cases.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Jonas Aaberg <jonas.aberg@stericsson.com>
Reviewed-by: Bengt JONSSON <bengt.g.jonsson@stericsson.com>
11 years agomfd: ab8500-sysctrl: Provide configuration for SysClkReqRfClkBuf registers
Kennet Wallden [Tue, 27 Sep 2011 07:23:56 +0000 (09:23 +0200)]
mfd: ab8500-sysctrl: Provide configuration for SysClkReqRfClkBuf registers

Add the possibility to pass configuration settings for
SysCl1kReqRfClkBuf to SysClk8ReqRfClkBuf via platform data.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Kennet Wallden <kennet.wallden@stericsson.com>
Reviewed-by: Karl-Johan PERNTZ <karl-johan.perntz@stericsson.com>
Reviewed-by: Bengt JONSSON <bengt.g.jonsson@stericsson.com>
11 years agomfd: ab8500-gpadc: Change to usleep_range() for greater resolution
Lee Jones [Mon, 29 Aug 2011 06:32:36 +0000 (08:32 +0200)]
mfd: ab8500-gpadc: Change to usleep_range() for greater resolution

The resolution of msleep is related to HZ, so with HZ set to
100 any msleep of less than 10ms will become ~10ms.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
11 years agomfd: ab8500-sysctrl.c: Provide charging as reset reason
Jonas Aaberg [Thu, 18 Aug 2011 08:14:38 +0000 (10:14 +0200)]
mfd: ab8500-sysctrl.c: Provide charging as reset reason

It's possible to supply a string to provide a reason for
triggering a restart. In this case our reason is to enter
charging-only mode, as a charger was found to be present.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Jonas Aaberg <jonas.aberg@stericsson.com>
Reviewed-by: Karl KOMIEROWSKI <karl.komierowski@stericsson.com>
11 years agomfd: ab8500-sysctrl: Only reboot into charging mode if battery type is known
Jonas Aaberg [Wed, 17 Aug 2011 13:58:52 +0000 (15:58 +0200)]
mfd: ab8500-sysctrl: Only reboot into charging mode if battery type is known

When a charger is connected, we usually want AB8500 based systems to
reboot into charging-only mode. However, if the battery type cannot
be identified this would be futile, so we'll just shut the system
down instead.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Jonas Aaberg <jonas.aberg@stericsson.com>
Reviewed-by: Karl KOMIEROWSKI <karl.komierowski@stericsson.com>
11 years agomfd: ab8500-sysctrl: If a charger is present, reboot instead
Jonas Aaberg [Wed, 17 Aug 2011 11:20:21 +0000 (13:20 +0200)]
mfd: ab8500-sysctrl: If a charger is present, reboot instead

If a charger is attached on power off, reboot the system
into charging mode instead of powering it off.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Jonas Aaberg <jonas.aberg@stericsson.com>
Reviewed-by: Karl KOMIEROWSKI <karl.komierowski@stericsson.com>
11 years agomfd: ab8500-sysctrl: Provide a platform specific pm_power_off() call-back
Lee Jones [Mon, 14 Jan 2013 13:26:15 +0000 (13:26 +0000)]
mfd: ab8500-sysctrl: Provide a platform specific pm_power_off() call-back

The kernel allows us to specify a function call-back which will be
invoked when a system power-off request has been received. Here we
provide one which is to be used when shutting down AB8500 based
platforms.

Signed-off-by: Lee Jones <lee.jones@linaro.org>
11 years agoLinux 3.8-rc6
Linus Torvalds [Fri, 1 Feb 2013 01:08:14 +0000 (12:08 +1100)]
Linux 3.8-rc6

11 years agoMerge tag 'dm-3.8-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/agk/linux-dm
Linus Torvalds [Fri, 1 Feb 2013 01:04:22 +0000 (12:04 +1100)]
Merge tag 'dm-3.8-fixes-2' of git://git./linux/kernel/git/agk/linux-dm

Pull more device-mapper fixes from Alasdair G Kergon:
 "A fix for stacked dm thin devices and a fix for the new dm WRITE SAME
  support."

* tag 'dm-3.8-fixes-2' of git://git.kernel.org/pub/scm/linux/kernel/git/agk/linux-dm:
  dm: fix write same requests counting
  dm thin: fix queue limits stacking

11 years agoMerge branch 'for-3.8/upstream-fixes' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Thu, 31 Jan 2013 21:44:59 +0000 (08:44 +1100)]
Merge branch 'for-3.8/upstream-fixes' of git://git./linux/kernel/git/jikos/hid

PullHID fixes from Jiri Kosina:

 - fix i2c-hid and hidraw interaction, by Benjamin Tissoires

 - a quirk to make a particular device (Formosa IR receiver) work
   properly, by Nicholas Santos

* 'for-3.8/upstream-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/hid:
  HID: i2c-hid: fix i2c_hid_output_raw_report
  HID: usbhid: quirk for Formosa IR receiver
  HID: remove x bit from sensor doc

11 years agoMerge tag 'nfs-for-3.8-4' of git://git.linux-nfs.org/projects/trondmy/linux-nfs
Linus Torvalds [Thu, 31 Jan 2013 21:43:52 +0000 (08:43 +1100)]
Merge tag 'nfs-for-3.8-4' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

Pull NFS client bugfixes from Trond Myklebust:

 - Error reporting in nfs_xdev_mount incorrectly maps all errors to
   ENOMEM

 - Fix an NFSv4 refcounting issue

 - Fix a mount failure when the server reboots during NFSv4 trunking
   discovery

 - NFSv4.1 mounts may need to run the lease recovery thread.

 - Don't silently fail setattr() requests on mountpoints

 - Fix a SUNRPC socket/transport livelock and priority queue issue

 - We must handle NFS4ERR_DELAY when resetting the NFSv4.1 session.

* tag 'nfs-for-3.8-4' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
  NFSv4.1: Handle NFS4ERR_DELAY when resetting the NFSv4.1 session
  SUNRPC: When changing the queue priority, ensure that we change the owner
  NFS: Don't silently fail setattr() requests on mountpoints
  NFSv4.1: Ensure that nfs41_walk_client_list() does start lease recovery
  NFSv4: Fix NFSv4 trunking discovery
  NFSv4: Fix NFSv4 reference counting for trunked sessions
  NFS: Fix error reporting in nfs_xdev_mount

11 years agoMerge branch 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus
Linus Torvalds [Thu, 31 Jan 2013 21:43:04 +0000 (08:43 +1100)]
Merge branch 'upstream' of git://git.linux-mips.org/ralf/upstream-linus

Pull MIPS updates from Ralf Baechle:
 "A number of fixes all across the MIPS tree.  No area is particularly
  standing out and things have cooled down quite nicely for a release."

* 'upstream' of git://git.linux-mips.org/pub/scm/ralf/upstream-linus:
  MIPS: Function tracer: Fix broken function tracing
  mips: Move __virt_addr_valid() to a place for MIPS 64
  MIPS: Netlogic: Fix UP compilation on XLR
  MIPS: AR71xx: Fix AR71XX_PCI_MEM_SIZE
  MIPS: AR724x: Fix AR724X_PCI_MEM_SIZE
  MIPS: Lantiq: Fix cp0_perfcount_irq mapping
  MIPS: DSP: Fix DSP mask for registers.
  MIPS: Fix build failure by adding definition of pfn_pmd().
  MIPS: Octeon: Fix warning.
  MIPS: delay.c: Check BITS_PER_LONG instead of __SIZEOF_LONG__
  MIPS: PNX833x: Fix comment.
  MIPS: Add struct p_format to union mips_instruction.
  MIPS: Export <asm/break.h>.
  MIPS: BCM47xx: Enable SSB prerequisite SSB_DRIVER_PCICORE.
  MIPS: BCM47xx: Select GPIOLIB for BCMA on bcm47xx platform
  MIPS: vpe.c: Fix null pointer dereference in print arguments.

11 years agoHID: i2c-hid: fix i2c_hid_output_raw_report
Benjamin Tissoires [Thu, 31 Jan 2013 16:50:02 +0000 (17:50 +0100)]
HID: i2c-hid: fix i2c_hid_output_raw_report

i2c_hid_output_raw_report is used by hidraw to forward set_report requests.
The current implementation of i2c_hid_set_report needs to take the
report_id as an argument. The report_id is stored in the first byte
of the buffer in argument of i2c_hid_output_raw_report.

Not removing the report_id from the given buffer adds this byte 2 times
in the command, leading to a non working command.

Reported-by: Andrew Duggan <aduggan@synaptics.com>
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@gmail.com>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
11 years agoMIPS: Function tracer: Fix broken function tracing
Al Cooper [Wed, 16 Jan 2013 22:43:28 +0000 (22:43 +0000)]
MIPS: Function tracer: Fix broken function tracing

Function tracing is currently broken for all 32 bit MIPS platforms.
When tracing is enabled, the kernel immediately hangs on boot.
This is a result of commit b732d439cb43336cd6d7e804ecb2c81193ef63b0
that changes the kernel/trace/Kconfig file so that is no longer
forces FRAME_POINTER when FUNCTION_TRACING is enabled.

MIPS frame pointers are generally considered to be useless because
they cannot be used to unwind the stack. Unfortunately the MIPS
function tracing code has bugs that are masked by the use of frame
pointers. This commit fixes the bugs so that MIPS frame pointers
don't need to be enabled.

The bugs are a result of the odd calling sequence used to call the trace
routine. This calling sequence is inserted into every traceable function
when the tracing CONFIG option is enabled. This sequence is generated
for 32bit MIPS platforms by the compiler via the "-pg" flag.

Part of the sequence is "addiu sp,sp,-8" in the delay slot after every
call to the trace routine "_mcount" (some legacy thing where 2 arguments
used to be pushed on the stack). The _mcount routine is expected to
adjust the sp by +8 before returning.  So when not disabled, the original
jalr and addiu will be there, so _mcount has to adjust sp.

The problem is that when tracing is disabled for a function, the
"jalr _mcount" instruction is replaced with a nop, but the
"addiu sp,sp,-8" is still executed and the stack pointer is left
trashed. When frame pointers are enabled the problem is masked
because any access to the stack is done through the frame
pointer and the stack pointer is restored from the frame pointer when
the function returns.

This patch writes two nops starting at the address of the "jalr _mcount"
instruction whenever tracing is disabled. This means that the
"addiu sp,sp.-8" will be converted to a nop along with the "jalr".  When
disabled, there will be two nops.

This is SMP safe because the first time this happens is during
ftrace_init() which is before any other processor has been started.
Subsequent calls to enable/disable tracing when other CPUs ARE running
will still be safe because the enable will only change the first nop
to a "jalr" and the disable, while writing 2 nops, will only be changing
the "jalr". This patch also stops using stop_machine() to call the
tracer enable/disable routines and calls them directly because the
routines are SMP safe.

When the kernel first boots we have to be able to handle the gcc
generated jalr, addui sequence until ftrace_init gets a chance to run
and change the sequence. At this point mcount just adjusts the stack
and returns. When ftrace_init runs, we convert the jalr/addui to nops.
Then whenever tracing is enabled we convert the first nop to a "jalr
mcount+8". The mcount+8 entry point skips the stack adjust.

[ralf@linux-mips.org: Folded in  Steven Rostedt's build fix.]

Signed-off-by: Al Cooper <alcooperx@gmail.com>
Cc: rostedt@goodmis.org
Cc: ddaney.cavm@gmail.com
Cc: linux-mips@linux-mips.org
Cc: linux-kernel@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/4806/
Patchwork: https://patchwork.linux-mips.org/patch/4841/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
11 years agodm: fix write same requests counting
Alasdair G Kergon [Thu, 31 Jan 2013 14:23:36 +0000 (14:23 +0000)]
dm: fix write same requests counting

When processing write same requests, fix dm to send the configured
number of WRITE SAME requests to the target rather than the number of
discards, which is not always the same.

Device-mapper WRITE SAME support was introduced by commit
23508a96cd2e857d57044a2ed7d305f2d9daf441 ("dm: add WRITE SAME support").

Signed-off-by: Alasdair G Kergon <agk@redhat.com>
Acked-by: Mike Snitzer <snitzer@redhat.com>
11 years agomips: Move __virt_addr_valid() to a place for MIPS 64
Steven Rostedt [Fri, 25 Jan 2013 17:13:15 +0000 (17:13 +0000)]
mips: Move __virt_addr_valid() to a place for MIPS 64

Commit d3ce88431892 "MIPS: Fix modpost error in modules attepting to use
virt_addr_valid()" moved __virt_addr_valid() from a macro in a header
file to a function in ioremap.c. But ioremap.c is only compiled for MIPS
32, and not for MIPS 64.

When compiling for my yeeloong2, which supposedly supports hibernation,
which compiles kernel/power/snapshot.c which calls virt_addr_valid(), I
got this error:

  LD      init/built-in.o
kernel/built-in.o: In function `memory_bm_free':
snapshot.c:(.text+0x4c9c4): undefined reference to `__virt_addr_valid'
snapshot.c:(.text+0x4ca58): undefined reference to `__virt_addr_valid'
kernel/built-in.o: In function `snapshot_write_next':
(.text+0x4e44c): undefined reference to `__virt_addr_valid'
kernel/built-in.o: In function `snapshot_write_next':
(.text+0x4e890): undefined reference to `__virt_addr_valid'
make[1]: *** [vmlinux] Error 1
make: *** [sub-make] Error 2

I suspect that __virt_addr_valid() is fine for mips 64. I moved it to
mmap.c such that it gets compiled for mips 64 and 32.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Cc: linux-kernel@vger.kernel.org
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/4842/
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
11 years agodm thin: fix queue limits stacking
Mike Snitzer [Thu, 31 Jan 2013 14:11:14 +0000 (14:11 +0000)]
dm thin: fix queue limits stacking

thin_io_hints() is blindly copying the queue limits from the thin-pool
which can lead to incorrect limits being set.  The fix here simply
deletes the thin_io_hints() hook which leaves the existing stacking
infrastructure to set the limits correctly.

When a thin-pool uses an MD device for the data device a thin device
from the thin-pool must respect MD's constraints about disallowing a bio
from spanning multiple chunks.  Otherwise we can see problems.  If the raid0
chunksize is 1152K and thin-pool chunksize is 256K I see the following
md/raid0 error (with extra debug tracing added to thin_endio) when
mkfs.xfs is executed against the thin device:

md/raid0:md99: make_request bug: can't convert block across chunks or bigger than 1152k 6688 127
device-mapper: thin: bio sector=2080 err=-5 bi_size=130560 bi_rw=17 bi_vcnt=32 bi_idx=0

This extra DM debugging shows that the failing bio is spanning across
the first and second logical 1152K chunk (sector 2080 + 255 takes the
bio beyond the first chunk's boundary of sector 2304).  So the bio
splitting that DM is doing clearly isn't respecting the MD limits.

max_hw_sectors_kb is 127 for both the thin-pool and thin device
(queue_max_hw_sectors returns 255 so we'll excuse sysfs's lack of
precision).  So this explains why bi_size is 130560.

But the thin device's max_hw_sectors_kb should be 4 (PAGE_SIZE) given
that it doesn't have a .merge function (for bio_add_page to consult
indirectly via dm_merge_bvec) yet the thin-pool does sit above an MD
device that has a compulsory merge_bvec_fn.  This scenario is exactly
why DM must resort to sending single PAGE_SIZE bios to the underlying
layer. Some additional context for this is available in the header for
commit 8cbeb67a ("dm: avoid unsupported spanning of md stripe boundaries").

Long story short, the reason a thin device doesn't properly get
configured to have a max_hw_sectors_kb of 4 (PAGE_SIZE) is that
thin_io_hints() is blindly copying the queue limits from the thin-pool
device directly to the thin device's queue limits.

Fix this by eliminating thin_io_hints.  Doing so is safe because the
block layer's queue limits stacking already enables the upper level thin
device to inherit the thin-pool device's discard and minimum_io_size and
optimal_io_size limits that get set in pool_io_hints.  But avoiding the
queue limits copy allows the thin and thin-pool limits to be different
where it is important, namely max_hw_sectors_kb.

Reported-by: Daniel Browning <db@kavod.com>
Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Cc: stable@vger.kernel.org
Signed-off-by: Alasdair G Kergon <agk@redhat.com>