8 years agoMFD: twl6040: Remove enum for PLL tracking
Peter Ujfalusi [Mon, 4 Jul 2011 07:28:28 +0000]
MFD: twl6040: Remove enum for PLL tracking

There is no need to have two different types for
tracking the selected PLL.
Use only the defines, when dealing with the PLLs.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Samuel Ortiz <sameo@linux.intel.com>

8 years agoASoC: twl6040: Configure PLL only once
Peter Ujfalusi [Sat, 2 Jul 2011 23:06:07 +0000]
ASoC: twl6040: Configure PLL only once

Avoid configuring the PLL several times during audio startup.
We can configure the PLL at prepare time with parameters collected
earlier hw_param, and set_dai_sysclk calls.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

8 years agoASoC: twl6040: Simplify sample rate constraint handling
Peter Ujfalusi [Wed, 29 Jun 2011 10:28:18 +0000]
ASoC: twl6040: Simplify sample rate constraint handling

We can manage the sample rate constraints without the need
to maintain a variable and a pointer.
This simplifies the handling of the constraint, and makes it
more robust.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

8 years agoASoC: twl6040: Move PLL selection to codec driver
Peter Ujfalusi [Mon, 27 Jun 2011 14:03:14 +0000]
ASoC: twl6040: Move PLL selection to codec driver

It is better if the selection between the Low power,
and High performance PLL is handled within the codec
driver, not in machine driver(s) to avoid duplicated
code, and also to have consistent tracking of the selected
PLL, and the resulting differences in supported sample
rates.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

8 years agoASoC: twl6040: Use neutral name for power mode text/enum
Peter Ujfalusi [Mon, 27 Jun 2011 10:33:14 +0000]
ASoC: twl6040: Use neutral name for power mode text/enum

Change the variable names to be neutral (not refering to HS).
This will ease up the introduction of PLL selection, which
going to use the same enum strings.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

8 years agoMFD: twl6040: Remove wrapper for threaded irq request
Peter Ujfalusi [Mon, 4 Jul 2011 17:16:23 +0000]
MFD: twl6040: Remove wrapper for threaded irq request

Remove the twl6040_request_irq/free_irq inline functions,
and use direct calls instead in the core driver to
register the threaded irq.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Reviewed-by: Felipe Balbi <balbi@ti.com>
Acked-by: Samuel Ortiz <sameo@linux.intel.com>

8 years agoMFD: twl6040: Demand valid interrupt configuration
Peter Ujfalusi [Mon, 4 Jul 2011 17:15:19 +0000]
MFD: twl6040: Demand valid interrupt configuration

In order to operate correctly twl6040 needs correct
interrupt configuration.
The slave drivers (vibra, and ASoC codec) will refuse
to probe, if the interrupt configuration is not correct.
In this way some checks can be removed from the code.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Reviewed-by: Felipe Balbi <balbi@ti.com>
Acked-by: Samuel Ortiz <sameo@linux.intel.com>

8 years agoASoC: twl6040: Do not use wrapper for irq request
Peter Ujfalusi [Mon, 4 Jul 2011 16:52:26 +0000]
ASoC: twl6040: Do not use wrapper for irq request

The twl6040_request_irq/free_irq inline functions are going
to be removed, so replace them with direct calls.
The irq number is provided by the core driver via resource.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Reviewed-by: Felipe Balbi <balbi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

8 years agoinput: twl6040-vibra: Do not use wrapper for irq request
Peter Ujfalusi [Mon, 4 Jul 2011 16:50:02 +0000]
input: twl6040-vibra: Do not use wrapper for irq request

The twl6040_request_irq/free_irq inline functions are going
to be removed, so replace them with direct calls.
The irq number is provided by the core driver via resource.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Reviewed-by: Felipe Balbi <balbi@ti.com>
Acked-by: Dmitry Torokhov <dtor@mail.ru>

8 years agoMFD: twl6040: Use resource to provide irq number for slaves
Peter Ujfalusi [Tue, 5 Jul 2011 08:40:33 +0000]
MFD: twl6040: Use resource to provide irq number for slaves

Provide the irq number for slaves via resource on the
platform device.
The irq number configuration is done in the twl6040-core
at probe time, so machine drivers do not need to be
modified.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Reviewed-by: Felipe Balbi <balbi@ti.com>
Acked-by: Samuel Ortiz <sameo@linux.intel.com>

8 years agoOMAP4: SDP4430: Add twl6040 codec platform support
Misael Lopez Cruz [Sat, 12 Feb 2011 01:50:37 +0000]
OMAP4: SDP4430: Add twl6040 codec platform support

Configure, and enable the twl6040 codec on SDP4430.

Signed-off-by: Misael Lopez Cruz <misael.lopez@ti.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>

8 years agoASoC: twl6040: Configure ramp step based on platform
Axel Castaneda Gonzalez [Thu, 24 Feb 2011 02:08:28 +0000]
ASoC: twl6040: Configure ramp step based on platform

Enable ramp down/up step to be configured based on
platform.

Signed-off-by: Axel Castaneda Gonzalez <x0055901@ti.com>
Signed-off-by: Misael Lopez Cruz <misael.lopez@ti.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

8 years agoASoC: twl6040: set default constraints.
Liam Girdwood [Mon, 7 Mar 2011 15:25:23 +0000]
ASoC: twl6040: set default constraints.

Set default sysclk constraints to high performance mode.

Signed-off-by: Liam Girdwood <lrg@ti.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

8 years agoASoC: twl6040: Remove pll and headset mode dependency
Misael Lopez Cruz [Fri, 11 Feb 2011 23:51:05 +0000]
ASoC: twl6040: Remove pll and headset mode dependency

Remove dependency between pll (hppll, lppll) and headset power
mode (low-power, high-performance), as headset power mode can
be used with any pll.

A new control is created to allow headset power mode configuration
from userspace. Changing headset power mode during earpiece related
usecases is not propagated down to the codec as earpiece requires
HS DAC in HP mode.

Signed-off-by: Misael Lopez Cruz <misael.lopez@ti.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

8 years agoASoC: twl6040: Support other sample rates in constraints.
Liam Girdwood [Fri, 18 Mar 2011 19:00:50 +0000]
ASoC: twl6040: Support other sample rates in constraints.

Add other supported sample rates to LP and HP modes.

Signed-off-by: Liam Girdwood <lrg@ti.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

8 years agoASoC: twl6040: add all ABE DAIs
Liam Girdwood [Fri, 11 Feb 2011 17:37:51 +0000]
ASoC: twl6040: add all ABE DAIs

Add all DAIs to fully support OMAP4 ABE.

Signed-off-by: Liam Girdwood <lrg@ti.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

8 years agoOMAP4: SDP4430: Add twl6040 vibrator platform support
Peter Ujfalusi [Wed, 1 Jun 2011 08:30:38 +0000]
OMAP4: SDP4430: Add twl6040 vibrator platform support

Add twl4030_vibra platform data, and the needed regulators
for twl6040 vibrator.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>

8 years agoinput: Add initial support for TWL6040 vibrator
Misael Lopez Cruz [Sun, 1 May 2011 08:51:24 +0000]
input: Add initial support for TWL6040 vibrator

Add twl6040_vibra as a child of MFD device twl6040_codec. This
implementation covers the PCM-to-PWM mode of TWL6040 vibrator
module.

Signed-off-by: Misael Lopez Cruz <misael.lopez@ti.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
CC: Tejun Heo <tj@kernel.org>
Acked-by: Dmitry Torokhov <dtor@mail.ru>

8 years agoMFD: twl6040: Change platform data for soc codec driver
Peter Ujfalusi [Wed, 1 Jun 2011 10:05:10 +0000]
MFD: twl6040: Change platform data for soc codec driver

Pass twl4030_codec_data instead of the twl4030_audio_data
for the ASoC codec driver.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Samuel Ortiz <sameo@linux.intel.com>

8 years agoASoC: twl6040: Convert into TWL6040 MFD child
Misael Lopez Cruz [Mon, 2 May 2011 02:27:00 +0000]
ASoC: twl6040: Convert into TWL6040 MFD child

Convert TWL6040 CODEC driver into a TWL6040 MFD child, it implies
that MFD-level operations like register accesses, clock setting
and power management are done through MFD APIs, not directly by
CODEC driver anymore. To avoid conflicts with the other MFD child,
vibrator registers are skipped in CODEC driver.

Signed-off-by: Misael Lopez Cruz <misael.lopez@ti.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>

8 years agomfd: twl6040: Add initial support
Misael Lopez Cruz [Wed, 27 Apr 2011 07:14:07 +0000]
mfd: twl6040: Add initial support

TWL6040 IC provides analog high-end audio codec functions for
handset applications. It contains several audio analog inputs
and outputs as well as vibrator support. It's connected to the
host processor via PDM interface for audio data communication.
The audio modules are controlled by internal registers that
can be accessed by I2C and PDM interface.

TWL6040 MFD will be registered as a child of TWL-CORE, and will
have two children of its own: twl6040-codec and twl6040-vibra.

This driver is based on TWL4030 and WM8350 MFD drivers.

Signed-off-by: Misael Lopez Cruz <misael.lopez@ti.com>
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Samuel Ortiz <sameo@linux.intel.com>

8 years agoMFD: twl4030-audio: Rename platform data
Peter Ujfalusi [Tue, 31 May 2011 12:21:13 +0000]
MFD: twl4030-audio: Rename platform data

Allign the platform data names for twl4030 audio submodule:
twl4030_audio_data: for the core MFD driver
twl4030_codec_data: for ASoC codec driver
twl4030_vibra_data: for the input/ForceFeedback driver

To avoid breakage, change all depending drivers, files
to use the new types.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Samuel Ortiz <sameo@linux.intel.com>

8 years agoMFD: twl4030-codec -> twl4030-audio: Rename the driver
Peter Ujfalusi [Tue, 31 May 2011 09:02:49 +0000]
MFD: twl4030-codec -> twl4030-audio: Rename the driver

Rename the driver, and header file from twl4030-codec to
twl4030-audio.
To avoid breakage change depending drivers at the same time.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
CC: Misael Lopez Cruz <misael.lopez@ti.com>
Acked-by: Samuel Ortiz <sameo@linux.intel.com>

8 years agoMFD: twl4030-codec: Rename internals from codec to audio
Peter Ujfalusi [Tue, 3 May 2011 15:03:43 +0000]
MFD: twl4030-codec: Rename internals from codec to audio

In preparation of renaming the driver from twl4030-codec
to twl4030-audio, first do some clean ups in the driver,
which does not cause any problems outside.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Samuel Ortiz <sameo@linux.intel.com>

8 years agoOMAP3: Move common regulator configuration to twl-common
Peter Ujfalusi [Tue, 7 Jun 2011 08:38:24 +0000]
OMAP3: Move common regulator configuration to twl-common

Some regulator config can be moved out from board files,
since they are close to identical.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>

8 years agoOMAP3: Move common twl configuration to twl-common
Peter Ujfalusi [Tue, 7 Jun 2011 07:28:54 +0000]
OMAP3: Move common twl configuration to twl-common

Reduce the amount of duplicated code by moving the common
configuration for twl4030/5030/tpsxx to the twl-common file.
Use the omap3_pmic_get_config function from board files to
properly configure the PMIC with the common fields.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>

8 years agoOMAP4: Move common twl6030 configuration to twl-common
Peter Ujfalusi [Tue, 7 Jun 2011 07:26:46 +0000]
OMAP4: Move common twl6030 configuration to twl-common

Reduce the amount of duplicated code by moving the common
configuration for TWL6030 (on OMAP4 platform) to the
twl-common file.
Use the omap4_pmic_get_config function from board files to
properly configure the PMIC with the common fields.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>

8 years agoOMAP: New twl-common for common TWL configuration
Peter Ujfalusi [Sat, 4 Jun 2011 05:16:41 +0000]
OMAP: New twl-common for common TWL configuration

Introduce a new file, which will be used to configure
common pmic (TWL) devices, regulators, and TWL audio.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Acked-by: Tony Lindgren <tony@atomide.com>

8 years agoMerge branch 'for_3.1/pm-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/khilm...
Tony Lindgren [Wed, 29 Jun 2011 11:45:16 +0000]
Merge branch 'for_3.1/pm-misc' of git://git./linux/kernel/git/khilman/linux-omap-pm into devel-cleanup

Conflicts:
arch/arm/mach-omap2/pm-debug.c
arch/arm/mach-omap2/pm.h

8 years agoMerge branch 'devel-timer' into devel-cleanup
Tony Lindgren [Wed, 29 Jun 2011 10:35:18 +0000]
Merge branch 'devel-timer' into devel-cleanup

8 years agoomap: mcbsp: Drop in-driver transfer support
Jarkko Nikula [Tue, 14 Jun 2011 11:23:52 +0000]
omap: mcbsp: Drop in-driver transfer support

We haven't seen either use for in-driver transfer API in McBSP driver
over the years so it looks they can be removed too.

Signed-off-by: Jarkko Nikula <jhnikula@gmail.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

8 years agoomap: mcbsp: Drop SPI mode support
Jarkko Nikula [Tue, 14 Jun 2011 11:23:51 +0000]
omap: mcbsp: Drop SPI mode support

We haven't seen any use for the SPI API in McBSP driver over the years. More
over, Peter Ujfalusi <peter.ujfalusi@ti.com> noticed that SPI mode is not
even supported since OMAP2430 so it's very unlikely that we'll see any use
for it in the future either.

Signed-off-by: Jarkko Nikula <jhnikula@gmail.com>
Acked-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

8 years agoomap: board-omap3evm: Fix compilation error
Peter Ujfalusi [Tue, 28 Jun 2011 10:16:55 +0000]
omap: board-omap3evm: Fix compilation error

Fix compilation error introduced with 786b01a8c1db0c0decca55d660a2a3ebd7cfb26b
(cleanup regulator supply definitions in mach-omap2).

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
[tony@atomide.com: updated comments]
Signed-off-by: Tony Lindgren <tony@atomide.com>

8 years agoomap: cleanup NAND platform data
Grazvydas Ignotas [Fri, 3 Jun 2011 19:56:33 +0000]
omap: cleanup NAND platform data

omap_nand_platform_data fields 'options', 'gpio_irq', 'nand_setup' and
'dma_channel' are never referenced by the NAND driver, yet various
board files are initializing those fields. This is both incorrect and
confusing, so remove them. This allows to get rid of a global
variable in gpmc-nand.c.

This also corrects an issue where some boards are trying to pass NAND
16bit flag through .options, but the driver is using .devsize instead
and ignoring .options.

Finally, .dev_ready is treated as a flag by the driver, so make it bool
instead of a function pointer.

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

8 years agoomap2+: Rename timer-gp.c into timer.c to combine timer init functions
Tony Lindgren [Tue, 29 Mar 2011 22:54:50 +0000]
omap2+: Rename timer-gp.c into timer.c to combine timer init functions

We can keep everything sys_timer and gptimer.c related code in
timer.c as the code will be very minimal.

Later on we can also remove timer-mpu.c, as it can be called from
omap4_timer_init function.

This allows us to get rid of confusing existing files. We currently
have timer-gp.c, timer-mpu.c, and patches have been posted to add
dmtimer.c. There's no need to have these multiple files, we can
put everything into timer.c.

Signed-off-by: Tony Lindgren <tony@atomide.com>

8 years agoomap2+: Remove omap2_gp_clockevent_set_gptimer
Tony Lindgren [Tue, 29 Mar 2011 22:54:49 +0000]
omap2+: Remove omap2_gp_clockevent_set_gptimer

This is no longer needed as we now just set the desired
.timer in MACHINE_START. We can now also remove timer-gp.h.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>

8 years agoomap2+: Use dmtimer macros for clocksource
Tony Lindgren [Tue, 29 Mar 2011 22:54:49 +0000]
omap2+: Use dmtimer macros for clocksource

Use dmtimer macros for clocksource. As with the clockevent,
this allows us to initialize the rest of dmtimer code later on.

Note that eventually we will be initializing the timesource
from init_early so sched_clock will work properly for
CONFIG_PRINTK_TIME.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>

8 years agoomap2+: Reserve clocksource and timesource and initialize dmtimer later
Tony Lindgren [Tue, 29 Mar 2011 22:54:49 +0000]
omap2+: Reserve clocksource and timesource and initialize dmtimer later

There's no need to initialize the dmtimer framework early.
Just mark the clocksource and timesource as reserved, and
initialize dmtimer with an arch_initcall.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>

8 years agoOMAP2+: PM: fix section mismatch in pm_dbg_init()
Sanjeev Premi [Thu, 16 Jun 2011 20:31:00 +0000]
OMAP2+: PM: fix section mismatch in pm_dbg_init()

Fix the section mismatch warning:

  WARNING: vmlinux.o(.text+0x21118): Section mismatch
  in reference from the function pm_dbg_init() to the
  function .init.text:pwrdms_setup()
  The function pm_dbg_init() references
  the function __init pwrdms_setup().
  This is often because pm_dbg_init lacks a __init
  annotation or the annotation of pwrdms_setup is wrong.

Signed-off-by: Sanjeev Premi <premi@ti.com>
Signed-off-by: Kevin Hilman <khilman@ti.com>

8 years agoOMAP3: PM debug: remove register dumping
Kevin Hilman [Thu, 26 May 2011 22:34:39 +0000]
OMAP3: PM debug: remove register dumping

Remove OMAP3-specific register dumping feature from PM debug layer.
This is removed because:

- it's ugly
- it's OMAP3-specific, and will obviously not scale to OMAP4+
- userspace /dev/mem-based tools (like omapconf) can do this much better

Tested-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Acked-by: Jean Pihet <j-pihet@ti.com>
Signed-off-by: Kevin Hilman <khilman@ti.com>

8 years agoOMAP2: PM debug: remove register dumping
Kevin Hilman [Thu, 26 May 2011 21:07:41 +0000]
OMAP2: PM debug: remove register dumping

Tested-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Signed-off-by: Kevin Hilman <khilman@ti.com>

8 years agoOMAP3: PM debug: remove sleep_while_idle feature
Kevin Hilman [Thu, 26 May 2011 21:48:19 +0000]
OMAP3: PM debug: remove sleep_while_idle feature

Remove the OMAP-specific PM debug 'sleep_while_idle' feature which is
currently available as an OMAP-specific debugfs entry.

This duplicates existing ARM-generic functionality available as a
boot-time option using the boot cmdline option 'hohlt'.

If runtime configuration of this is needed, then adding a debugfs
entry for the ARM-generic hlt/nohlt interface should be added.

Tested-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Acked-by: Jean Pihet <j-pihet@ti.com>
Signed-off-by: Kevin Hilman <khilman@ti.com>

8 years agoOMAP3+: SR: fix cosmetic indentation
Nishanth Menon [Mon, 14 Feb 2011 07:03:13 +0000]
OMAP3+: SR: fix cosmetic indentation

Error label case seems to have a 2 tab indentation when just 1 is
necessary.

Signed-off-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Kevin Hilman <khilman@ti.com>

8 years agoOMAP3+: SR: enable/disable SR only on need
Nishanth Menon [Mon, 14 Feb 2011 15:44:17 +0000]
OMAP3+: SR: enable/disable SR only on need

Since we already know the state of the autocomp enablement, we can
see if the requested state is different from the current state and
enable/disable SR only on the need basis.

Signed-off-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Kevin Hilman <khilman@ti.com>

8 years agoOMAP3+: SR: disable interrupt by default
Nishanth Menon [Mon, 14 Feb 2011 07:11:10 +0000]
OMAP3+: SR: disable interrupt by default

We will enable and disable interrupt on a need basis in the class
driver. We need to keep the IRQ disabled by default else the
forceupdate or vcbypass events could trigger events that we don't
need/expect to handle.

This is a preparation for SmartReflex AVS class drivers such as
class 2 and class 1.5 which would need to use interrupts. Existing
SmartReflex AVS class 3 driver does not require to use interrupts
and is not impacted by this change.

Signed-off-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Kevin Hilman <khilman@ti.com>

8 years agoOMAP3+: SR: make notify independent of class
Nishanth Menon [Mon, 14 Feb 2011 06:46:36 +0000]
OMAP3+: SR: make notify independent of class

Interrupt notification mechanism of SmartReflex can be used by the
choice of implementation of the class driver. For example, Class 2 and
Class 1.5 of SmartReflex can both use the interrupt notification to
identify the transition of voltage or other events.

Hence, the actual class does not matter for notifier. Let the class
driver's handling decide how it should be used. SmartReflex driver
should provide just the primitives.

Signed-off-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Kevin Hilman <khilman@ti.com>

8 years agoomap2+: Remove gptimer_wakeup for now
Tony Lindgren [Tue, 29 Mar 2011 22:54:49 +0000]
omap2+: Remove gptimer_wakeup for now

This removes the support for setting the wake-up timer for debugging.

Later on we can reserve gptimer1 for PM code only and have similar
functionality.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>

8 years agoomap2+: Use dmtimer macros for clockevent
Tony Lindgren [Tue, 29 Mar 2011 22:54:48 +0000]
omap2+: Use dmtimer macros for clockevent

This patch makes timer-gp.c to use only a subset of dmtimer
functions without the need to initialize dmtimer code early.

Also note that now with the inline functions, timer_set_next_event
becomes more efficient in the lines of assembly code.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>

8 years agoomap: Make a subset of dmtimer functions into inline functions
Tony Lindgren [Tue, 29 Mar 2011 22:54:48 +0000]
omap: Make a subset of dmtimer functions into inline functions

This will allow us to share the code between system timer and
dmtimer device driver code without having to initialize all the
dmtimers early. This change will also make the timer_set_next_event
more efficient as the inline functions will optimize the code
better for the timer reprogramming.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>

8 years agoomap: Move dmtimer defines to dmtimer.h
Tony Lindgren [Tue, 29 Mar 2011 22:54:48 +0000]
omap: Move dmtimer defines to dmtimer.h

These will be needed when dmtimer platform init code gets split
for omap1 and omap2+. These will also be needed for separate
sys_timer init and driver init for the rest of the hardware timers
in the following patches. No functional changes.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>

8 years agoomap: Set separate timer init functions to avoid cpu_is_omap tests
Tony Lindgren [Tue, 29 Mar 2011 22:54:48 +0000]
omap: Set separate timer init functions to avoid cpu_is_omap tests

This is needed for the following patches so we can initialize the
rest of the hardware timers later on.

As with the init_irq calls, there's no need to do cpu_is_omap calls
during the timer init as we only care about the major omap generation.
This means that we can initialize the sys_timer with the .timer
entries alone.

Note that for now we just set stubs for the various sys_timer entries
that will get populated in a later patch. The following patches will
also remove the omap_dm_timer_init calls and change the init for the
rest of the hardware timers to happen with an arch_initcall.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>

8 years agoomap: Use separate init_irq functions to avoid cpu_is_omap tests early
Tony Lindgren [Tue, 17 May 2011 10:51:26 +0000]
omap: Use separate init_irq functions to avoid cpu_is_omap tests early

This allows us to remove cpu_is_omap calls from init_irq functions.
There should not be any need for cpu_is_omap calls as at this point.
During the timer init we only care about SoC generation, and not about
subrevisions.

The main reason for the patch is that we want to initialize only
minimal omap specific code from the init_early call.

Signed-off-by: Tony Lindgren <tony@atomide.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>

8 years agoMerge branch 'fixes-v3.0-rc3' into devel-fixes
Tony Lindgren [Tue, 14 Jun 2011 12:53:42 +0000]
Merge branch 'fixes-v3.0-rc3' into devel-fixes

8 years agoOMAP1: PM: register notifiers with generic clock ops even when !PM_RUNTIME
Kevin Hilman [Tue, 14 Jun 2011 12:53:18 +0000]
OMAP1: PM: register notifiers with generic clock ops even when !PM_RUNTIME

When runtime PM is disabled, device clocks need to be enabled on
device add and disabled on device remove.  This currently is not
happening because in the !PM_RUNTIME case, no notifiers are registered
for OMAP1 devices.

Fix this by ensuring notifiers are registered, even in the !PM_RUNTIME case.

Reported-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Tested-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Signed-off-by: Kevin Hilman <khilman@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

8 years agoRemove old-style supply.dev assignments common in hsmmc init
Oleg Drokin [Mon, 6 Jun 2011 18:57:08 +0000]
Remove old-style supply.dev assignments common in hsmmc init

CC: Mark Brown <broonie@opensource.wolfsonmicro.com>
CC: Mike Rapoport <mike@compulab.co.il>
CC: Nishant Kamat <nskamat@ti.com>
CC: Steve Sakoman <steve@sakoman.com>
CC: Felipe Balbi <balbi@ti.com>
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Acked-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

8 years agocleanup regulator supply definitions in mach-omap2
Oleg Drokin [Mon, 6 Jun 2011 18:57:07 +0000]
cleanup regulator supply definitions in mach-omap2

to use REGULATOR_SUPPLY arrays.

CC: Mark Brown <broonie@opensource.wolfsonmicro.com>
CC: Mike Rapoport <mike@compulab.co.il>
CC: Nishant Kamat <nskamat@ti.com>
CC: Steve Sakoman <steve@sakoman.com>
CC: Felipe Balbi <balbi@ti.com>
CC: Santosh Shilimkar <santosh.shilimkar@ti.com>
CC: peter.barada@logicpd.com
Signed-off-by: Oleg Drokin <green@linuxhacker.ru>
Acked-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>

8 years agoLinux 3.0-rc3
Linus Torvalds [Mon, 13 Jun 2011 22:29:59 +0000]
Linux 3.0-rc3

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/penberg...
Linus Torvalds [Mon, 13 Jun 2011 20:00:53 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/penberg/slab-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/penberg/slab-2.6:
  SLAB: Record actual last user of freed objects.
  slub: always align cpu_slab to honor cmpxchg_double requirement

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph...
Linus Torvalds [Mon, 13 Jun 2011 18:21:50 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/sage/ceph-client

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client:
  ceph: unwind canceled flock state
  ceph: fix ENOENT logic in striped_read
  ceph: fix short sync reads from the OSD
  ceph: fix sync vs canceled write
  ceph: use ihold when we already have an inode ref

8 years agoMerge branch 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 13 Jun 2011 17:47:04 +0000]
Merge branch 'timers-urgent-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'timers-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  rtc: Staticize non-exported __rtc_set_alarm()
  rtc: Fix ioctl error path return
  ptp: Fix some locking bugs in ptp_read()
  ptp: Return -EFAULT on copy_to_user() errors

8 years agoMerge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 13 Jun 2011 17:45:49 +0000]
Merge branch 'perf-urgent-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  ftrace: Revert 8ab2b7efd ftrace: Remove unnecessary disabling of irqs
  kprobes/trace: Fix kprobe selftest for gcc 4.6
  ftrace: Fix possible undefined return code
  oprofile, dcookies: Fix possible circular locking dependency
  oprofile: Fix locking dependency in sync_start()
  oprofile: Free potentially owned tasks in case of errors
  oprofile, x86: Add comments to IBS LVT offset initialization

8 years agoMerge branches 'x86-urgent-for-linus' and 'irq-urgent-for-linus' of git://git.kernel...
Linus Torvalds [Mon, 13 Jun 2011 17:45:10 +0000]
Merge branches 'x86-urgent-for-linus' and 'irq-urgent-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86: devicetree: Add missing early_init_dt_setup_initrd_arch stub
  x86: cpu-hotplug: Prevent softirq wakeup on wrong CPU

* 'irq-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  genirq: Prevent potential NULL dereference in irq_set_irq_wake()

8 years agoMerge branch 'spi/merge' of git://git.secretlab.ca/git/linux-2.6
Linus Torvalds [Mon, 13 Jun 2011 17:33:31 +0000]
Merge branch 'spi/merge' of git://git.secretlab.ca/git/linux-2.6

* 'spi/merge' of git://git.secretlab.ca/git/linux-2.6:
  spi-pl022: Add missing return value update

8 years agoMerge branch 's5p-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Mon, 13 Jun 2011 17:32:53 +0000]
Merge branch 's5p-fixes-for-linus' of git://git./linux/kernel/git/kgene/linux-samsung

* 's5p-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/kgene/linux-samsung:
  ARM: S3C2410: remove the now empty mach-s3c2410/irq.c
  ARM: S3C24XX: Move s3c24xx_irq_syscore_ops to plat-s3c24xx/irq.c
  ARM: S3C24xx: Fix missing struct for s3c2410_dma_chan
  ARM: SAMSUNG: Remove unused onenand plat functions
  ARM: EXYNOS4: Remove compiler warning on exynos4_pwm4_resume
  ARM: S5P: Fix compilation error for exynos4_defconfig
  ARM: S5P: Should be S3C_VA_USB_HSPHY instead of S5P_VA_XX
  ARM: S5P64X0: Fix SPI platform device name
  ARM: S5PV210: Fix possible null pointer dereference

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ryusuke...
Linus Torvalds [Mon, 13 Jun 2011 17:32:24 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/ryusuke/nilfs2

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ryusuke/nilfs2:
  nilfs2: fix problem in setting checkpoint interval
  nilfs2: fix missing block address termination in btree node shrinking
  nilfs2: fix incorrect block address termination in node concatenation

8 years agospi-pl022: Add missing return value update
Virupax Sadashivpetimath [Mon, 13 Jun 2011 10:53:46 +0000]
spi-pl022: Add missing return value update

Return error on out of range cpsdvsr value.

Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Virupax Sadashivpetimath <virupax.sadashivpetimath@stericsson.com>
Signed-off-by: Grant Likely <grant.likely@secretlab.ca>

8 years agoMerge branch 'for_3.0/pm-fixes' of ssh://master.kernel.org/pub/scm/linux/kernel/git...
Tony Lindgren [Mon, 13 Jun 2011 14:40:25 +0000]
Merge branch 'for_3.0/pm-fixes' of ssh:///linux/kernel/git/khilman/linux-omap-pm into fixes

8 years agoomap: pandora: fix NAND support
Grazvydas Ignotas [Fri, 3 Jun 2011 20:24:03 +0000]
omap: pandora: fix NAND support

Commit d5ce2b65 "omap3630: nand: fix device size to work in polled mode"
changed values for .devsize in nand platform data, now we have to pass
NAND_BUSWIDTH_16 instead of '1' to select 16bit NAND.

Update pandora's platform data accordingly, also specify appropriate
transfer type.

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

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs...
Linus Torvalds [Sun, 12 Jun 2011 18:06:36 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/mason/btrfs-unstable

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable:
  Btrfs: use join_transaction in btrfs_evict_inode()
  Btrfs - use %pU to print fsid
  Btrfs: fix extent state leak on failed nodatasum reads
  btrfs: fix unlocked access of delalloc_inodes
  Btrfs: avoid stack bloat in btrfs_ioctl_fs_info()
  btrfs: remove 64bit alignment padding to allow extent_buffer to fit into one fewer cacheline
  Btrfs: clear current->journal_info on async transaction commit
  Btrfs: make sure to recheck for bitmaps in clusters
  btrfs: remove unneeded includes from scrub.c
  btrfs: reinitialize scrub workers
  btrfs: scrub: errors in tree enumeration
  Btrfs: don't map extent buffer if path->skip_locking is set
  Btrfs: unlock the trans lock properly
  Btrfs: don't map extent buffer if path->skip_locking is set
  Btrfs: fix duplicate checking logic
  Btrfs: fix the allocator loop logic
  Btrfs: fix bitmap regression
  Btrfs: don't commit the transaction if we dont have enough pinned bytes
  Btrfs: noinline the cluster searching functions
  Btrfs: cache bitmaps when searching for a cluster

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
Linus Torvalds [Sun, 12 Jun 2011 18:04:25 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/tiwai/sound-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
  ALSA: hda: Fix inaudible internal speakers on CyberpowerPC Gamer Xplorer N57001 laptop
  ALSA: Use %pV for snd_printk()
  ALSA: hda - Fix initialization of hp pins with master_mute in Realtek
  ALSA: hda - Fix invalid unsol tag for some alc262 model quirks
  ASoC: SAMSUNG: Fix the incorrect referencing of I2SCON register
  ASoC: snd_soc_new_{mixer,mux,pga} make sure to use right DAPM context
  ASoC: fsl: fix initialization of DMA buffers
  ASoC: WM8804 does not support sample rates below 32kHz
  ASoC: Fix WM8962 headphone volume update for use of advanced caches
  ASoC: Blackfin: bf5xx-ad1836: Fix codec device name
  ALSA: hda: Fix quirk for Dell Inspiron 910
  ASoC: AD1836: Fix setting the PCM format
  ASoC: Check for NULL register bank in snd_soc_get_cache_val()
  ASoC: Add missing break in WM8915 FLL source selection
  ASoC: Only update SYSCLK_ENA when pausing WM8915 SYSCLK
  ASoC: atmel_ssc: Don't try to free ssc if request failed

8 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/ide-2.6
Linus Torvalds [Sun, 12 Jun 2011 18:04:11 +0000]
Merge git://git./linux/kernel/git/davem/ide-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/ide-2.6:
  ide-cd: signedness warning fix again

8 years agoMerge branch 'gpio/merge' of git://git.secretlab.ca/git/linux-2.6
Linus Torvalds [Sun, 12 Jun 2011 18:03:29 +0000]
Merge branch 'gpio/merge' of git://git.secretlab.ca/git/linux-2.6

* 'gpio/merge' of git://git.secretlab.ca/git/linux-2.6:
  gpio/basic_mmio: add missing include of spinlock_types.h
  gpio/nomadik: fix sleepmode for elder Nomadik

8 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
Linus Torvalds [Sun, 12 Jun 2011 18:03:08 +0000]
Merge git://git./linux/kernel/git/davem/net-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (55 commits)
  ISDN, hfcsusb: Don't leak in hfcsusb_ph_info()
  netpoll: call dev_put() on error in netpoll_setup()
  net: ep93xx_eth: fix DMA API violations
  net: ep93xx_eth: drop GFP_DMA from call to dma_alloc_coherent()
  net: ep93xx_eth: allocate buffers using kmalloc()
  net: ep93xx_eth: pass struct device to DMA API functions
  ep93xx: set DMA masks for the ep93xx_eth
  vlan: Fix the ingress VLAN_FLAG_REORDER_HDR check
  dl2k: EEPROM CRC calculation wrong endianess on bigendian machine
  NET: am79c961: fix assembler warnings
  NET: am79c961: ensure multicast filter is correctly set at open
  NET: am79c961: ensure asm() statements are marked volatile
  ethtool.h: fix typos
  ep93xx_eth: Update MAINTAINERS
  ipv4: Fix packet size calculation for raw IPsec packets in __ip_append_data
  netpoll: prevent netpoll setup on slave devices
  net: pmtu_expires fixes
  gianfar:localized filer table
  iwlegacy: fix channel switch locking
  mac80211: fix IBSS teardown race
  ...

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris...
Linus Torvalds [Sun, 12 Jun 2011 02:56:25 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/jmorris/security-testing-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/security-testing-2.6:
  AppArmor: Fix sleep in invalid context from task_setrlimit

8 years agoISDN, hfcsusb: Don't leak in hfcsusb_ph_info()
Jesper Juhl [Sat, 11 Jun 2011 06:36:42 +0000]
ISDN, hfcsusb: Don't leak in hfcsusb_ph_info()

We leak the memory allocated to 'phi' when the variable goes out of scope
in hfcsusb_ph_info().

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agonetpoll: call dev_put() on error in netpoll_setup()
Dan Carpenter [Sun, 12 Jun 2011 01:55:22 +0000]
netpoll: call dev_put() on error in netpoll_setup()

There is a dev_put(ndev) missing on an error path.  This was
introduced in 0c1ad04aecb "netpoll: prevent netpoll setup on slave
devices".

Signed-off-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agonet: ep93xx_eth: fix DMA API violations
Mika Westerberg [Sat, 11 Jun 2011 08:39:58 +0000]
net: ep93xx_eth: fix DMA API violations

Russell King said:
>
> So, to summarize what its doing:
>
> 1. It allocates buffers for rx and tx.
> 2. It maps them with dma_map_single().
>       This transfers ownership of the buffer to the DMA device.
> 3. In ep93xx_xmit,
> 3a. It copies the data into the buffer with skb_copy_and_csum_dev()
>       This violates the DMA buffer ownership rules - the CPU should
>       not be writing to this buffer while it is (in principle) owned
>       by the DMA device.
> 3b. It then calls dma_sync_single_for_cpu() for the buffer.
>       This transfers ownership of the buffer to the CPU, which surely
>       is the wrong direction.
> 4. In ep93xx_rx,
> 4a. It calls dma_sync_single_for_cpu() for the buffer.
>       This at least transfers the DMA buffer ownership to the CPU
>       before the CPU reads the buffer
> 4b. It then uses skb_copy_to_linear_data() to copy the data out.
>       At no point does it transfer ownership back to the DMA device.
> 5. When the driver is removed, it dma_unmap_single()'s the buffer.
>       This transfers ownership of the buffer to the CPU.
> 6. It frees the buffer.
>
> While it may work on ep93xx, it's not respecting the DMA API rules,
> and with DMA debugging enabled it will probably encounter quite a few
> warnings.

This patch fixes these violations.

Signed-off-by: Mika Westerberg <mika.westerberg@iki.fi>
Acked-by: Russell King <rmk+kernel@arm.linux.org.uk>
Acked-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Tested-by: Petr Stetiar <ynezz@true.cz>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agonet: ep93xx_eth: drop GFP_DMA from call to dma_alloc_coherent()
Mika Westerberg [Sat, 11 Jun 2011 08:39:57 +0000]
net: ep93xx_eth: drop GFP_DMA from call to dma_alloc_coherent()

Commit a197b59ae6e8 (mm: fail GFP_DMA allocations when ZONE_DMA is not
configured) made page allocator to return NULL if GFP_DMA is set but
CONFIG_ZONE_DMA is disabled.

This causes ep93xx_eth to fail:

 WARNING: at mm/page_alloc.c:2251 __alloc_pages_nodemask+0x11c/0x638()
 Modules linked in:
 [<c0035498>] (unwind_backtrace+0x0/0xf4) from [<c0043da4>] (warn_slowpath_common+0x48/0x60)
 [<c0043da4>] (warn_slowpath_common+0x48/0x60) from [<c0043dd8>] (warn_slowpath_null+0x1c/0x24)
 [<c0043dd8>] (warn_slowpath_null+0x1c/0x24) from [<c0083b6c>] (__alloc_pages_nodemask+0x11c/0x638)
 [<c0083b6c>] (__alloc_pages_nodemask+0x11c/0x638) from [<c00366fc>] (__dma_alloc+0x8c/0x3ec)
 [<c00366fc>] (__dma_alloc+0x8c/0x3ec) from [<c0036adc>] (dma_alloc_coherent+0x54/0x60)
 [<c0036adc>] (dma_alloc_coherent+0x54/0x60) from [<c0227808>] (ep93xx_open+0x20/0x864)
 [<c0227808>] (ep93xx_open+0x20/0x864) from [<c0283144>] (__dev_open+0xb8/0x108)
 [<c0283144>] (__dev_open+0xb8/0x108) from [<c0280528>] (__dev_change_flags+0x70/0x128)
 [<c0280528>] (__dev_change_flags+0x70/0x128) from [<c0283054>] (dev_change_flags+0x10/0x48)
 [<c0283054>] (dev_change_flags+0x10/0x48) from [<c001a720>] (ip_auto_config+0x190/0xf68)
 [<c001a720>] (ip_auto_config+0x190/0xf68) from [<c00233b0>] (do_one_initcall+0x34/0x18c)
 [<c00233b0>] (do_one_initcall+0x34/0x18c) from [<c0008400>] (kernel_init+0x94/0x134)
 [<c0008400>] (kernel_init+0x94/0x134) from [<c0030858>] (kernel_thread_exit+0x0/0x8)

Since there is no restrictions for DMA on ep93xx, we can fix this by just
removing the GFP_DMA flag from the call.

Signed-off-by: Mika Westerberg <mika.westerberg@iki.fi>
Acked-by: Russell King <rmk+kernel@arm.linux.org.uk>
Acked-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Tested-by: Petr Stetiar <ynezz@true.cz>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agonet: ep93xx_eth: allocate buffers using kmalloc()
Mika Westerberg [Sat, 11 Jun 2011 08:39:56 +0000]
net: ep93xx_eth: allocate buffers using kmalloc()

We can use simply kmalloc() to allocate the buffers. This also simplifies the
code and allows us to perform DMA sync operations more easily.

Memory is allocated with only GFP_KERNEL since there are no DMA allocation
restrictions on this platform.

Signed-off-by: Mika Westerberg <mika.westerberg@iki.fi>
Acked-by: Russell King <rmk+kernel@arm.linux.org.uk>
Acked-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Tested-by: Petr Stetiar <ynezz@true.cz>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agonet: ep93xx_eth: pass struct device to DMA API functions
Mika Westerberg [Sat, 11 Jun 2011 08:39:55 +0000]
net: ep93xx_eth: pass struct device to DMA API functions

We shouldn't use NULL for any DMA API functions, unless we are dealing with
ISA or EISA device. So pass correct struct dev pointer to these functions.

Signed-off-by: Mika Westerberg <mika.westerberg@iki.fi>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoep93xx: set DMA masks for the ep93xx_eth
Mika Westerberg [Sat, 11 Jun 2011 08:39:54 +0000]
ep93xx: set DMA masks for the ep93xx_eth

Since the driver uses the DMA API, we should pass it valid DMA masks.

Signed-off-by: Mika Westerberg <mika.westerberg@iki.fi>
Acked-by: Russell King <rmk+kernel@arm.linux.org.uk>
Acked-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Tested-by: Petr Stetiar <ynezz@true.cz>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agovlan: Fix the ingress VLAN_FLAG_REORDER_HDR check
Jiri Pirko [Fri, 10 Jun 2011 06:56:58 +0000]
vlan: Fix the ingress VLAN_FLAG_REORDER_HDR check

Testing of VLAN_FLAG_REORDER_HDR does not belong in vlan_untag
but rather in vlan_do_receive.  Otherwise the vlan header
will not be properly put on the packet in the case of
vlan header accelleration.

As we remove the check from vlan_check_reorder_header
rename it vlan_reorder_header to keep the naming clean.

Fix up the skb->pkt_type early so we don't look at the packet
after adding the vlan tag, which guarantees we don't goof
and look at the wrong field.

Use a simple if statement instead of a complicated switch
statement to decided that we need to increment rx_stats
for a multicast packet.

Hopefully at somepoint we will just declare the case where
VLAN_FLAG_REORDER_HDR is cleared as unsupported and remove
the code.  Until then this keeps it working correctly.

Signed-off-by: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: Jiri Pirko <jpirko@redhat.com>
Acked-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agodl2k: EEPROM CRC calculation wrong endianess on bigendian machine
Daniel Hellstrom [Fri, 10 Jun 2011 04:55:16 +0000]
dl2k: EEPROM CRC calculation wrong endianess on bigendian machine

Signed-off-by: Daniel Hellstrom <daniel@gaisler.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoNET: am79c961: fix assembler warnings
Russell King - ARM Linux [Fri, 10 Jun 2011 00:52:35 +0000]
NET: am79c961: fix assembler warnings

Fix:
/tmp/ccvoZ6h8.s: Assembler messages:
/tmp/ccvoZ6h8.s:284: Warning: register range not in ascending order
/tmp/ccvoZ6h8.s:881: Warning: register range not in ascending order
/tmp/ccvoZ6h8.s:1087: Warning: register range not in ascending order

by ensuring that we have temporary variables placed into specific
registers.  Reorder the code a bit to allow the resulting assembly
to be slightly more optimal.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoNET: am79c961: ensure multicast filter is correctly set at open
Russell King - ARM Linux [Fri, 10 Jun 2011 00:52:14 +0000]
NET: am79c961: ensure multicast filter is correctly set at open

We were clearing out the multicast filter whenever the interface was
upped, and not setting the mode bits correctly.  This can cause
problems if there are any multicast addresses already set at this
point, or if ALLMULTI was set.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoNET: am79c961: ensure asm() statements are marked volatile
Russell King - ARM Linux [Fri, 10 Jun 2011 00:51:54 +0000]
NET: am79c961: ensure asm() statements are marked volatile

Without this the compiler can (and does) optimize register reads away
from within loops, and other such optimizations.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoide-cd: signedness warning fix again
Connor Hansen [Sat, 11 Jun 2011 22:06:48 +0000]
ide-cd: signedness warning fix again

One of the legit warnings 'make W=3 drivers/ide/ide-cd.c'
generates is:
drivers/ide/ide-cd.c: In function ide_cd_do_request
drivers/ide/ide-cd.c:828:2: warning: conversion to int from \
unsigned int may change the sign of the result
drivers/ide/ide-cd.c:833:2: warning: conversion to int from \
unsigned int may change the sign of the result

nsectors is declared int, should be unsigned int.

blk_rq_sectors() returns unsigned int, and ide_complete_rq
expects unsigned int as well.  Fixes both warnings.

Signed-off-by: Connor Hansen <cmdkhh@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agolinux/seqlock.h should #include asm/processor.h for cpu_relax()
David Howells [Sat, 11 Jun 2011 11:29:58 +0000]
linux/seqlock.h should #include asm/processor.h for cpu_relax()

It uses cpu_relax(), and so needs <asm/processor.h>

Without this patch, I see:

CC arch/mn10300/kernel/asm-offsets.s
  In file included from include/linux/time.h:8,
                   from include/linux/timex.h:56,
                   from include/linux/sched.h:57,
                   from arch/mn10300/kernel/asm-offsets.c:7:
  include/linux/seqlock.h: In function 'read_seqbegin':
  include/linux/seqlock.h:91: error: implicit declaration of function 'cpu_relax'

whilst building asb2364_defconfig on MN10300.

Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agoBtrfs: use join_transaction in btrfs_evict_inode()
Li Zefan [Wed, 8 Jun 2011 03:56:44 +0000]
Btrfs: use join_transaction in btrfs_evict_inode()

The WARN_ON() in start_transaction() was triggered while balancing.

The cause is btrfs_relocate_chunk() started a transaction and
then called iput() on the inode that stores free space cache,
and iput() called btrfs_start_transaction() again.

Reported-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Reviewed-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agonilfs2: fix problem in setting checkpoint interval
Ryusuke Konishi [Thu, 9 Jun 2011 15:33:06 +0000]
nilfs2: fix problem in setting checkpoint interval

Checkpoint generation interval of nilfs goes wrong after user has
changed the interval parameter with nilfs-tune tool.

 segctord starting. Construction interval = 5 seconds,
 CP frequency < 30 seconds
 segctord starting. Construction interval = 0 seconds,
 CP frequency < 30 seconds

This turned out to be caused by a trivial bug in initialization code
of log writer.  This will fix it.

Reported-by: Andrea Gelmini <andrea.gelmini@gmail.com>
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>

8 years agonilfs2: fix missing block address termination in btree node shrinking
Ryusuke Konishi [Wed, 25 May 2011 14:00:27 +0000]
nilfs2: fix missing block address termination in btree node shrinking

nilfs_btree_delete function does not terminate part of virtual block
addresses when shrinking the last remaining child node into the root
node.  The missing address termination causes that dead btree node
blocks persist and chip away free disk space.

This fixes the leak bug on the btree node deletion.

Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>

8 years agonilfs2: fix incorrect block address termination in node concatenation
Ryusuke Konishi [Wed, 25 May 2011 14:00:27 +0000]
nilfs2: fix incorrect block address termination in node concatenation

nilfs_btree_delete function wrongly terminates virtual block address
of the btree node held by its parent at index 0.  When concatenating
the index-0 node with its right sibling node, nilfs_btree_delete
terminates the block address of index-0 node instead of the right
sibling node which should be deleted.

This bug not only wears disk space in the long run, but also causes
file system corruption.  This will fix it.

Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>

8 years agoBtrfs - use %pU to print fsid
Ilya Dryomov [Wed, 9 Feb 2011 14:05:31 +0000]
Btrfs - use %pU to print fsid

Get rid of FIXME comment.  Uuids from dmesg are now the same as uuids
given by btrfs-progs.

Signed-off-by: Ilya Dryomov <idryomov@gmail.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agoBtrfs: fix extent state leak on failed nodatasum reads
Jan Schmidt [Wed, 4 May 2011 14:18:50 +0000]
Btrfs: fix extent state leak on failed nodatasum reads

When encountering an EIO while reading from a nodatasum extent, we
insert an error record into the inode's failure tree.
btrfs_readpage_end_io_hook returns early for nodatasum inodes. We'd
better clear the failure tree in that case, otherwise the kernel
complains about

BUG extent_state: Objects remaining on kmem_cache_close()

on rmmod.

Signed-off-by: Jan Schmidt <list.btrfs@jan-o-sch.net>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agoMerge branch 'for-chris' of git://git.kernel.org/pub/scm/linux/kernel/git/arne/btrfs...
Chris Mason [Fri, 10 Jun 2011 22:58:08 +0000]
Merge branch 'for-chris' of git://git./linux/kernel/git/arne/btrfs-unstable-arne into for-linus

8 years agobtrfs: fix unlocked access of delalloc_inodes
David Sterba [Thu, 9 Jun 2011 10:02:51 +0000]
btrfs: fix unlocked access of delalloc_inodes

list_splice_init will make delalloc_inodes empty, but without a spinlock
around, this may produce corrupted list head, accessed in many placess,
The race window is very tight and nobody seems to have hit it so far.

Signed-off-by: David Sterba <dsterba@suse.cz>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agoBtrfs: avoid stack bloat in btrfs_ioctl_fs_info()
Li Zefan [Wed, 8 Jun 2011 08:27:56 +0000]
Btrfs: avoid stack bloat in btrfs_ioctl_fs_info()

The size of struct btrfs_ioctl_fs_info_args is as big as 1KB, so
don't declare the variable on stack.

Signed-off-by: Li Zefan <lizf@cn.fujitsu.com>
Reviewed-by: Josef Bacik <josef@redhat.com>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agobtrfs: remove 64bit alignment padding to allow extent_buffer to fit into one fewer...
richard kennedy [Tue, 7 Jun 2011 10:46:32 +0000]
btrfs: remove 64bit alignment padding to allow extent_buffer to fit into one fewer cacheline

Reorder extent_buffer to remove 8 bytes of alignment padding on 64 bit
builds. This shrinks its size to 128 bytes allowing it to fit into one
fewer cache lines and allows more objects per slab in its kmem_cache.

slabinfo extent_buffer reports :-

 before:-
    Sizes (bytes)     Slabs
    ----------------------------------
    Object :     136  Total  :     123
    SlabObj:     136  Full   :     121
    SlabSiz:    4096  Partial:       0
    Loss   :       0  CpuSlab:       2
    Align  :       8  Objects:      30

 after :-
    Object :     128  Total  :       4
    SlabObj:     128  Full   :       2
    SlabSiz:    4096  Partial:       0
    Loss   :       0  CpuSlab:       2
    Align  :       8  Objects:      32

Signed-off-by: Richard Kennedy <richard@rsk.demon.co.uk>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agoBtrfs: clear current->journal_info on async transaction commit
Sage Weil [Fri, 10 Jun 2011 18:43:13 +0000]
Btrfs: clear current->journal_info on async transaction commit

Normally current->jouranl_info is cleared by commit_transaction.  For an
async snap or subvol creation, though, it runs in a work queue.  Clear
it in btrfs_commit_transaction_async() to avoid leaking a non-NULL
journal_info when we return to userspace.  When the actual commit runs in
the other thread it won't care that it's current->journal_info is already
NULL.

Signed-off-by: Sage Weil <sage@newdream.net>
Tested-by: Jim Schutt <jaschut@sandia.gov>
Signed-off-by: Chris Mason <chris.mason@oracle.com>

8 years agoBtrfs: make sure to recheck for bitmaps in clusters
Chris Mason [Fri, 10 Jun 2011 20:36:57 +0000]
Btrfs: make sure to recheck for bitmaps in clusters

Josef recently changed the free extent cache to look in
the block group cluster for any bitmaps before trying to
add a new bitmap for the same offset.  This avoids BUG_ON()s due
covering duplicate ranges.

But it didn't go quite far enough.  A given free range might span
between one or more bitmaps or free space entries.  The code has
looping to cover this, but it doesn't check for clustered bitmaps
every time.

This shuffles our gotos to check for a bitmap in the cluster
for every new bitmap entry we try to add.

Signed-off-by: Chris Mason <chris.mason@oracle.com>