Merge tag 'ux500-multiplatform-asoc' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-stericsson into next/multiplatform

From Linus Walleij <linus.walleij@linaro.org>:

Ux500 multiplatform support.  This tag builds upon the MFD-specific base
tag "ux500-multiplatform-mfd". This removes all <mach/*> dependencies
and makes the ux500 fully multi-platform.

* tag 'ux500-multiplatform-asoc' of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-stericsson:
  ARM: ux500: build hotplug.o for ARMv7-a
  ARM: ux500: move to multiplatform
  ARM: ux500: make remaining headers local
  ARM: ux500: make irqs.h local to platform
  ARM: ux500: get rid of <mach/[hardware|db8500-regs].h>
  staging: ste_rmi4: kill platform_data hack
  ARM: ux500: move mach/msp.h to <linux/platform_data/*>
  clk: ux500: pass clock base adresses in init call
  ARM: ux500: make debug macro stand-alone
  ARM: ux500: move debugmacro to debug includes
  ARM: ux500: split out prcmu initialization
  mfd: db8500-prcmu: drop unused includes
  ARM: ux500: move PM-related PRCMU functions to machine
  mfd: db8500-prcmu: get base address from resource
  mfd: prcmu: pass a base and size with the early initcall

Conflicts:
	arch/arm/Kconfig
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Linux 3.9-rc3

Conflicts:
	arch/arm/Kconfig
	arch/arm/mach-spear/spear3xx.c
	arch/arm/plat-spear/Kconfig

This is a dependency for ux500/multiplatform
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Merge tag 'bcm2835-for-3.10-multiplatform' of git://git.kernel.org/pub/scm/linux/kernel/git/swarren/linux-rpi into next/multiplatform

From Stephen Warren <swarren@wwwdotorg.org>:

ARM: bcm2835: convert to multi-platform

This branch (patch) converts BCM2835 to support being built into a
multi-platform single zImage. This mostly entails a few small Kconfig
tweaks, move the earlyprintk implementation to the standard multi-
platform location, and deleting some unnecessary files.

* tag 'bcm2835-for-3.10-multiplatform' of git://git.kernel.org/pub/scm/linux/kernel/git/swarren/linux-rpi:
  ARM: bcm2835: convert to multi-platform

Conflicts:
	arch/arm/Kconfig
	arch/arm/Kconfig.debug
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

If we try to build this file in a multiplatform configuration with
ARMv6 enabled, gas complains about the dsb operation being undefined.
Adding -march=armv7-a is safe because that code is only ever run
on ux500, which is Cortex-A9.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

Nothing is holding us up any more, and we can make ux500 coexist
with the rest of the platforms. The timex.h and uncompress.h
headers are no longer needed now.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

mach/setup.h and mach/devices.h are only needed from inside of mach-ux500
now, so we can simply move them out of the include/mach directory.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

With the PRCMU out of the way, nothing outside of mach-ux500
uses mach/irqs.h any more, so we can make it a local header file,
including the two sub-headers.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

This removes <mach/hardware.h> and <mach/db8500-regs.h>
from the Ux500, merging them into the local include
"db8500-regs.h" in mach-ux500. There is some impact
outside the ux500 machine, but most of it is dealt with
in earlier patches.

Contains portions of a clean-up patch from Arnd Bergmann.

Cc: Samuel Ortiz <sameo@linux.intel.com>
Cc: Ulf Hansson <ulf.hansson@linaro.org>
Acked-by: Mike Turquette <mturquette@linaro.org>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

There is only one instance of the platform data for synaptics_i2c_rmi4
in the mainline kernel, so there is no point of pretending its
variable here. The only member that has a dependency on the platform
is actually the interrupt number, and there is a field in the
i2c_client structure that gets initialized from the board info,
so we can trivially move the board_into into the platform without
knowledge of the platform_data structure.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

This header file only contains platform data structure definitions,
so it's straightforward to move.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
[Delete one include rather than move it]
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

The ux500 clock driver was including <mach/db8500-regs.h>
which will not work when building for multiplatform support
since <mach/*> is going away.

Pass the base adresses in the init call instead.

Cc: Ulf Hansson <ulf.hansson@linaro.org>
Cc: Mike Turquette <mturquette@linaro.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Ulf Hansson <ulf.hansson@linaro.org>
Acked-by: Mike Turquette <mturquette@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

This pushes the knowledge of physical addresses down into
the debug macro so we can get rid of the <mach/hardware.h>
dependency.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

This moves the Ux500 debug macro to the debug headers to
make way for multiplatform support.
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

This untangles the final bits of the prcmu code from the platform
code:

* The IRQ_PRCMU_* definitions move from irqs-db8500.h into prcmu.c
  because they are only of local significance.
* u8500_thsens_device goes into the prcmu, because it uses a PRCMU
  IRQ that the platform does not see.
* IRQ_DB8500_AB8500 and IRQ_PRCMU_BASE go into the platform data
  because the PRCMU does not see it.
Acked-by: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
[Fixed a oneliner bug]
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

These two <mach/*> includes are no longer used in the
PRCMU driver, so drop them.
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

We are trying to decompose and decentralize the code in
the DB8500 PRCMU out into subdrivers. The code moved in
this patch concerns a group of functions used for
decoupling and recoupling the IRQs from the GIC. During
sleep and idle the Ux500 system will transfer all IRQ
handling to the PRCMU using these functions.

Basically we are left with the two alternatives of code
placement as:

- arch/arm/mach-ux500/pm.c - this because the code is
  closely related to the GIC, and takes ownership of
  some of the registers from the PRCMU related to this
  PM functionality.

- drivers/mfd/db8500-prcmu-pm.c - because the code is
  affecting stuff in the PRCMU register range. But then
  this code needs to remap and handle GIC registers.

This patch implementation is taking the first approach.

Currently the cpuidle driver is the only piece of code
using this set of functions, but it will later also be
used by the suspend/resume code which is currently under
review.

The header file is moved to:
<linux/platform_data/arm-ux500-pm.h>
The function prototypes need to be placed in a globally
visible header since the CPUidle code is planned to move
out to drivers/cpuidle.
Acked-by: Samuel Ortiz <sameo@linux.intel.com>
Acked-by: Rickard Andersson <rickard.andersson@stericsson.com>
Acked-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

We cannot use a global variable stored in <mach/hardware.h> to
find the base address of the PRCMU. The real resource is already
there from the board, so use this to look up the base address
instead.

Currently the patch is kept minimal so as not to interfere with
other work being done on refactoring this driver, but at a later
point the defines using (prcmu_base + 0xnnn) need to be replaced
by pure offset defined for (0xnnn) and the base inlined with the
readl()/writel() and similar codepaths.
Acked-by: Samuel Ortiz <sameo@linux.intel.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

This patch will make an early remapping of the PRCMU, to be
used when setting up the clocks, that will call down into parts
of the PRCMU driver before it is probed.

Going forward this will be removed like this:

- The mailbox subsystem need to be merged.
  http://marc.info/?l=linux-kernel&m=136314559201983&w=2

- At this point the PRCMU clock code can be moved over to the
  ux500 clock driver in drivers/clk/ux500/* and maintained
  there in a decentralized manner.

- This early initcall and PRCMU base parameters become part of
  the ux500_clk_init() call instead.

Cc: Suman Anna <s-anna@ti.com>
Cc: Loic Pallardy <loic.pallardy@st.com>
Acked-by: Samuel Ortiz <sameo@linux.intel.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Ulf Hansson <ulf.hansson@linaro.org>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>

This patch finally adds the WM8850 SoC to the multi_v7_defconfig.
We don't have a seperate defconfig for this SoC as it is only
available as a multiplatform option.
Signed-off-by: Tony Prisk <linux@prisktech.co.nz>
Signed-off-by: Olof Johansson <olof@lixom.net>

This series enables multiplatform support on the SIRF prima2/marco/atlas6
platform. The code was already quite tidy, so this is a relatively simple
change, and it follows similar changes we made to other ARMv7 based
platforms recently.

* prima2/multiplatform:
  ARM: sirf: enable support in multi_v7_defconfig
  ARM: sirf: enable multiplatform support
  ARM: sirf: use clocksource_of infrastructure
  ARM: sirf: move debug-macro.S to include/debug/sirf.S
  ARM: sirf: enable sparse IRQ
  ARM: sirf: move irq driver to drivers/irqchip
  ARM: sirf: fix prima2 interrupt lookup
  pinctrl: sirf: convert to linear irq domain
  clocksource: make CLOCKSOURCE_OF_DECLARE type safe
  ARM/dts: prima2: add .dtsi for atlas6 and .dts for atla6-evb board
  arm: prima2: add new SiRFatlas6 machine in common board
  ARM: smp_twd: convert to use CLKSRC_OF init
  clocksource: tegra20: use the device_node pointer passed to init
  clocksource: pass DT node pointer to init functions
  clocksource: add empty version of clocksource_of_init

Conflicts:
	arch/arm/configs/multi_v7_defconfig
	arch/arm/mach-spear/spear13xx.c
Tested-by: Barry Song <Barry.Song@csr.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

This enables all sirf specific drivers in the defconfig.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Barry Song <Baohua.Song@csr.com>

All the prerequisites are there now, so we can move sirf into multiplatform.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

This moves the two sirf clocksource drivers to drivers/clocksource
and integrates them into the framework for locating the clock sources
automatically.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Tested-by: Barry Song <Baohua.Song@csr.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: Thomas Gleixner <tglx@linutronix.de>

The new style ll_debug implementation for multiplatform requires the
platform glue to be in include/debug, so let's move it there to
separate the debugging logic from the platform code.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Tested-by: Barry Song <Baohua.Song@csr.com>

Now that both irqchips for sirf are converted to not rely on
legacy domains, let's move all of the platform over to sparse
IRQ.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

This updates the irqchip drier for prima2 to the current practices by
moving it into drivers/irqchip and integrating it into the irqchip_init
infrastructure. We also now use a linear irq domain as a preparation
for sparse IRQ suport.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Thomas Gleixner <tglx@linutronix.de>

We must not read the interrupts property manually but instead
use irq_of_parse_and_map() to guarantee that we get the correct
interrupt number once we stop using the legacy IRQ domain.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Tested-by: Barry Song <Baohua.Song@csr.com>

The sirf platforms use no hardcoded IRQ numbers, so there is no reason to
use the legacy domain, and by converting to the linear domain, we get
a more efficient representation of sparse IRQs and remove the dependency
on the mach/irqs.h header file.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Linus Walleij <linus.walleij@linaro.org>

This ensures that a function pointer passed into CLOCKSOURCE_OF_DECLARE
takes the same arguments that we use for calling that function later.

Also fix the extraneous semicolon at end of the CLOCKSOURCE_OF_DECLARE
definition.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Rob Herring <rob.herring@calxeda.com>

The hotplug.c file uses assembly instructions that are only available
on ARMv7 but not on ARMv6. This is ok because we know that code will
only run on arm ARMv7 SPEARr13xx, but it produces build errors when
we also enable one of the ARMv6 targets in a multiplatform configuration.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

The nomadik multiplatform support made it possible to select
MACH_NOMADIK_8815NHK without selecting ARCH_NOMADIK, which leads
to build errors when we also select ARMv6/v7 targets. Adding the
ifdef here restores the intended behavior.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

atlas6.dtsi is basically a copy of prima2.dtsi as most components are
compatible with prima2 except that:
1. node of l2 cache is deleted
2. node multimedia engine is deleted
3. node of sata is deleted
4. node of sdmmc4 is deleted
5. powervr is moved to "powervr,sgx510"
6. pinctrl is moved to atlas6 as pinmux layout has big changes in
   atlas6
7. clock is moved to atlas6 as clock layout has changes in atlas6
Signed-off-by: Barry Song <Baohua.Song@csr.com>
Signed-off-by: Jiansong Chen <Jiansong.Chen@csr.com>

SiRFatlas6's machine definition is almost seem with SiRFprimaII
except that prima2 has a 256MB DMA zone.
This patch adds SiRFatlas6 machine in common board files, and
also adds atlas6 arch node in Kconfig.
Signed-off-by: Barry Song <Baohua.Song@csr.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Commit 1d9d8639 ("perf,x86: fix kernel crash with PEBS/BTS after
suspend/resume") introduces a link failure since
perf_restore_debug_store() is only defined for CONFIG_CPU_SUP_INTEL:

	arch/x86/power/built-in.o: In function `restore_processor_state':
	(.text+0x45c): undefined reference to `perf_restore_debug_store'

Fix it by defining the dummy function appropriately.
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Commit 1d9d8639 ("perf,x86: fix kernel crash with PEBS/BTS after
suspend/resume") fixed a crash when doing PEBS performance profiling
after resuming, but in using init_debug_store_on_cpu() to restore the
DS_AREA mtrr it also resulted in a new WARN_ON() triggering.

init_debug_store_on_cpu() uses "wrmsr_on_cpu()", which in turn uses CPU
cross-calls to do the MSR update.  Which is not really valid at the
early resume stage, and the warning is quite reasonable.  Now, it all
happens to _work_, for the simple reason that smp_call_function_single()
ends up just doing the call directly on the CPU when the CPU number
matches, but we really should just do the wrmsr() directly instead.

This duplicates the wrmsr() logic, but hopefully we can just remove the
wrmsr_on_cpu() version eventually.
Reported-and-tested-by: Parag Warudkar <parag.lkml@gmail.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Pull btrfs fixes from Chris Mason:
 "Eric's rcu barrier patch fixes a long standing problem with our
  unmount code hanging on to devices in workqueue helpers.  Liu Bo
  nailed down a difficult assertion for in-memory extent mappings."

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs:
  Btrfs: fix warning of free_extent_map
  Btrfs: fix warning when creating snapshots
  Btrfs: return as soon as possible when edquot happens
  Btrfs: return EIO if we have extent tree corruption
  btrfs: use rcu_barrier() to wait for bdev puts at unmount
  Btrfs: remove btrfs_try_spin_lock
  Btrfs: get better concurrency for snapshot-aware defrag work

Users report that an extent map's list is still linked when it's actually
going to be freed from cache.

The story is that

a) when we're going to drop an extent map and may split this large one into
smaller ems, and if this large one is flagged as EXTENT_FLAG_LOGGING which means
that it's on the list to be logged, then the smaller ems split from it will also
be flagged as EXTENT_FLAG_LOGGING, and this is _not_ expected.

b) we'll keep ems from unlinking the list and freeing when they are flagged with
EXTENT_FLAG_LOGGING, because the log code holds one reference.

The end result is the warning, but the truth is that we set the flag
EXTENT_FLAG_LOGGING only during fsync.

So clear flag EXTENT_FLAG_LOGGING for extent maps split from a large one.
Reported-by: Johannes Hirte <johannes.hirte@fem.tu-ilmenau.de>
Reported-by: Darrick J. Wong <darrick.wong@oracle.com>
Signed-off-by: Liu Bo <bo.li.liu@oracle.com>
Signed-off-by: Chris Mason <chris.mason@fusionio.com>

Pull kbuild fix from Michal Marek:
 "One fix for for make headers_install/headers_check to not require make
  3.81.  The requirement has been accidentally introduced in 3.7."

* 'kbuild' of git://git.kernel.org/pub/scm/linux/kernel/git/mmarek/kbuild:
  kbuild: fix make headers_check with make 3.80