| Age | Commit message (Collapse) | Author |
|---|
|
The hisi SMP code always uses the hi3xxx_set_cpu() function
defined in the hotplug.c file, so we cannot build without
this when CONFIG_SMP is enabled. This patch slightly restructures
the code so we always build the parts of hotplug.c that we need
but just leave out the CPU disable logic if CONFIG_HOTPLUG_CPU
is turned off.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Haojian Zhuang <haojian.zhuang@gmail.com>
|
|
ep93xx_chip_revision is used by the pata_ep93xx driver,
which can be a loadable module. Exporting the symbol
avoids a link error in this case.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Hartley Sweeten <hsweeten@visionengravers.com>
Cc: Ryan Mallon <rmallon@gmail.com>
|
|
The efm32 platform does not provide a zreladdr-y line its Makefile.boot,
so we always have to use CONFIG_AUTO_ZRELADDR in order to successfully
build and link a zImage.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Uwe Kleine-König <kernel@pengutronix.de>
|
|
The DAVINCI_DA850_EVM board uses an unusual method to
enable the GPIO_PCA953X and KEYBOARD_GPIO_POLLED symbols,
which leads to the dependencies on these symbols being
ignored. As GPIO_PCA953X actually requires I2C, that
can lead to build failures when I2C is disabled.
This patch removes the duplicate symbol definitions
and instead enables them from the davinci_all_defconfig
file.
A different question whether we actually want to automatically
enable them at all or rather put them into defconfig,
but that should be a separate patch.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Sekhar Nori <nsekhar@ti.com>
Cc: Kevin Hilman <khilman@deeprootsystems.com>
Cc: davinci-linux-open-source@linux.davincidsp.com
|
|
We cannot call phy_register_fixup_for_uid() if CONFIG_PHYLIB
is not built into the kernel, and we should not enforce that
to be built into vmlinux either, because one might want to
disable the entire network stack.
This change uses a compile-time condition on CONFIG_PHYLIB
to remove the call in the cases where it cannot work.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Sekhar Nori <nsekhar@ti.com>
Cc: Kevin Hilman <khilman@deeprootsystems.com>
Cc: davinci-linux-open-source@linux.davincidsp.com
|
|
It makes no sense for sama5 support to be enabled if we don't
also enable USE_OF. Making this automatic in Kconfig avoids
a possible randconfig conflict between the old and new clock
support code.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Acked-by: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com>
|
|
If CONFIG_PATA_AT91 is disabled, the code in at91_add_device_cf
is turned into invalid C statements due to the lack of an
expression before the 'else' clause.
This moves the first half of the condition inside of the #ifdef,
which seems to be what the author intended.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Cc: Andrew Victor <linux@maxim.org.za>
Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com>
|
|
The pata_at91 driver uses interfaces defined in the sam9_smc
platform code. Since the pata driver can be a loadable module,
we have to export those symbols in order to link cleanly.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Cc: Andrew Victor <linux@maxim.org.za>
Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com>
|
|
at91x40 has no support for device tree, but Kconfig allows
us to enable CONFIG_OF anyway, causing a link error in the
at91 reset controller initialization.
The easiest fix is to adapt the existing #ifdef to omit
the broken code on at91x40 where it is never called anyway.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Cc: Andrew Victor <linux@maxim.org.za>
Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com>
|
|
at91x40 is different from all the other at91 machines, and it is
impossible to build a kernel that works on both this SoC and
any of the others, even though it is possible to build a noMMU
kernel for any at91 machine.
By turning at91x40 into a separate top-level option, we explicitly
forbid enabling invalid configurations that include mutually exclusive
machines.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Cc: Andrew Victor <linux@maxim.org.za>
Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com>
|
|
The big series refactoring Exynos suspend to RAM handling missed the
cpuidle driver that is disabled in exynos_defconfig, leaving it
including old mach/pm_core.h header and using old s3c_cpu_resume symbol
instead of new exynos_cpu_resume, resulting in compilation failures with
CONFIG_ARCH_EXYNOS and CONFIG_CPU_IDLE enabled.
This patch fixes that silly mistake and performs necessary modification
to arhc/arm/exynos/cpuidle.c to make it compile again.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
This patch fixes compilation failure due to missing explicit inclusion
of linux/serial_s3c.h in mach/pm-core.h, which contains definitions
required for further code in this header.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Remove some entries from Dalmore's device tree that attempt to set some
options which aren't supported for the drive_gma pin group.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Olof Johansson <olof@lixom.net>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux into next/boards
Merge "ARM: tegra: defconfig updates for 3.15" from Stephen Warren:
The following options are enabled:
- ams AS3722 is used as system PMIC for Tegra124 based Venice2.
- NXP PCA9546 is used in Cardhu Tegra30 board to connect to 3 cameras.
- GSPCA to support the webcam on Venice2.
- Allocate 64 MiB for CMA by default; the default 16MiB is not enough
for the majority of use-cases. This can still be overridden by the cma
command-line option.
* tag 'tegra-for-3.15-defconfig' of git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux:
ARM: tegra: defconfig updates
Signed-off-by: Olof Johansson <olof@lixom.net>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux into next/dt
Merge "ARM: tegra: device tree changes for 3.15" from Stephen Warren:
This enables:
- host1x and eDP support on Tegra124.
- LCD panel support for a few Tegra20 devices and Venice2.
- Enables power down, SPI flash, and USB on Venice2.
- Documents which Dalmore revision is supported.
- Adds an I2C bus mux to Cardhu.
Additionally, Tegra124 is converted to use #address-cells=<2> since the
HW suports more than 32-bits of address space, and various cleanups are
included.
* tag 'tegra-for-3.15-dt' of git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux: (21 commits)
ARM: dts: tegra: add PCIe interrupt mapping properties
ARM: tegra: use 2 address cells for Tegra124 DT
ARM: tegra: Rename as3722 node to pmic
ARM: tegra: Fix whitespace around '='
ARM: tegra: Enable USB on Venice2
ARM: tegra: Add Tegra124 USB support
ARM: tegra: Enable eDP for Venice2
ARM: tegra: Add Tegra124 eDP support
ARM: tegra: Add Tegra124 host1x support
ARM: tegra: Hook up SDMMC3 power-supply on Venice2
ARM: tegra: Overhaul Venice2 regulators
ARM: tegra: Combine VBUS enable pins into one node
ARM: tegra: Use "disabled" for status property
ARM: tegra: add SPI flash to Venice2 DT
ARM: tegra: enable PCA9546 on Cardhu
ARM: tegra: enable LCD panel on Ventana
ARM: tegra: enable LCD panel on Seaboard
ARM: tegra: add system-power-controller property for PMIC node
ARM: tegra: document which Dalmore revisions are supported
ARM: tegra: Properly sort clocks property
...
Signed-off-by: Olof Johansson <olof@lixom.net>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux into next/soc
Merge "ARM: tegra: core SoC changes for 3.15" from Stephen Warren:
A couple of minor fixes, plus the removal of the EMC scaling driver,
which hasn't been active since we converted clocks to device tree.
* tag 'tegra-for-3.15-soc' of git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux:
ARM: tegra: Export I/O rail functions
ARM: tegra: remove tegra EMC scaling driver
ARM: tegra: don't timeout if CPU is powergated
Signed-off-by: Olof Johansson <olof@lixom.net>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux into next/soc
Merge "ARM: tegra: Trusted Foundations work for 3.15" from Stephen Warren:
This pull request contains a number of cleanups and enhancements for the
Trusted Foundations firmware used on production Tegra SoCs. The changes
allow kernels without TF support to run on HW that uses TF, albeit with
reduced functionality, and also fix the cpuidle feature.
* tag 'tegra-for-3.15-tf' of git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux:
ARM: tegra: cpuidle: use firmware for power down
ARM: trusted_foundations: implement prepare_idle()
ARM: firmware: add prepare_idle() operation
ARM: firmware: enable Trusted Foundations by default
ARM: trusted_foundations: fallback when TF support is missing
ARM: trusted_foundations: fix vendor prefix typos
Signed-off-by: Olof Johansson <olof@lixom.net>
|
|
As suggested by Olof Johansson at
http://www.spinics.net/lists/arm-kernel/msg314009.html.
It be better just keeping a "ti,keystone" top-level compatible and
just using that to probe. If so we don't have to touch the file
for new boards in the future.
So use common "ti,keystone" compatible in keystone.c for all boards.
Cc: Santosh Shilimkar <santosh.shilimkar@ti.com>
Cc: Olof Johansson <olof@lixom.net>
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@ti.com>
Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
|
|
The UBIFS partition size was incorrectly defined as for 128M NAND,
but on k2hk-evm installed NAND flash size is 512M.
Hence, correct it.
Signed-off-by: Murali Karicheri <m-karicheri2@ti.com>
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@ti.com>
Signed-off-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
|
|
This is a dummy placeholder file. Delete it.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
The contents of this header file are not referenced anywhere in the
included .c files except in devs.c. Remove its inclusion. For devs.c,
explicitly include sizes.h header for SZ_* macros.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Remove the code that is not referenced anywhere. While at it also
remove incorrect file path.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Most of the clock related dt-binding header files are located in
dt-bindings/clock folder. It would be good to keep all the similar
header files at a single location.
Signed-off-by: Tushar Behera <tushar.behera@linaro.org>
Reviewed-by: Sachin Kamat <sachin.kamat@linaro.org>
Acked-by: Tomasz Figa <t.figa@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
LDO3 and LDO23 need to be enabled in order for soft-reset to work.
Additionally LDO9 needs to be enabled for USB operations.
Signed-off-by: Tushar Behera <tushar.behera@linaro.org>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
MDMA1 can support both secure and non-secure AXI transactions. When this
is enabled in the kernel for boards that run in secure mode, we get
imprecise external aborts causing the kernel to oops.
Unhandled fault: imprecise external abort (0x1406) at 0x00000000
Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000007
Suggested-by: Javi Merino <javi.merino@arm.com>
Signed-off-by: Tushar Behera <tushar.behera@linaro.org>
Tested-by: Javi Merino <javi.merino@arm.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Some very recent change appears to have removed an implicit inclusion of
serial_s3c.h causing build failures due to references to UART registers
in the serial port restore code in next-20140318. Include it explicitly
to fix the build.
Signed-off-by: Mark Brown <broonie@linaro.org>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
This patch restores the ability to receive wake-up events from internal
GIC interrupts, e.g. RTC tick or alarm interrupts.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Since Exynos SoCs does not follow most of the semantics of older SoCs
when configuring the system to enter sleep, there is no reason to rely
on the legacy Samsung PM core anymore.
This patch adds local Exynos suspend ops and removes all the code left
unnecessary. As a side effect, suspend support on Exynos becomes
multiplatform-friendly.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
This patch simplifies Exynos PM initialization and makes it
multiplatform friendly by replacing initcalls used originally to invoke
all the initialization code with explicit function calls.
In addition, an useless subsys_interface is removed, as all its .add_dev
callback did was setting two function pointers.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
CONFIG_PM means that at least one of CONFIG_PM_SLEEP and
CONFIG_PM_RUNTIME is enabled, while multiple entries in
mach-exynos/Kconfig abused it to enable sleep- and runtime-specific
functionality.
This patch fixes this abuse by replacing dependencies on CONFIG_PM with
appropriate dependencies on either CONFIG_PM_SLEEP or CONFIG_PM_RUNTIME,
whichever is appropriate.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
To allow using Samsung PM memory check helpers on platforms that do not
use the legacy Samsung PM core, this patch moves prototypes of relevant
functions to plat/pm-common.h header.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
To separate legacy PM code from generic helpers, this patch moves the
generic register save/restore helpers to a new file called pm-common.c
that is compiled always when CONFIG_PM_SLEEP is enabled, to allow
platforms that do not want to use the legacy PM code use the generic
helpers.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Not all Samsung SoC platforms are going to use the legacy Samsung PM
code enabled by CONFIG_SAMSUNG_PM_DEBUG. To allow using Samsung PM debug
helpers on such platforms, related code is moved to separate file and
a plat/pm-common.h header is added to separate legacy and generic code.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
This patch removes one-line functions that was used just to pass
constant arguments to lower level functions. After previous patches the
need for those constants has been eliminated, so the main functions can
be called directly.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
This patch modifies Samsung PM debug helpers to use a multiplatform
friendly way of getting base address of debug UART port, so instead
of using a per-mach static macro, a generic debug_ll_addr() helper
is used.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
The only SoC that does not have DIVSLOT register is S3C2410, so instead
of exporting a variable for platforms to set if DIVSLOT register should
be preserved, it's enough to simply check whether we are running on
a S3C2410 instead.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Due to the S3C2410 SoC being quite different from other S3C24xx SoCs
in some aspects, such as availability of DIVSLOT register in its UART
blocks, there is a need sometimes to check whether we are running on
this SoC, not just the S3C24xx series. This patch adds soc_is_s3c2410()
helper function for this purpose.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Trying to resume l2x0 if it was not enabled before suspend leads to
system crash. This patch prevents this by checking if l2x0_regs_phys is
a valid pointer to l2x0 context data saved on initialization.
Signed-off-by: Tomasz Figa <t.figa@samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
|
|
Add generic compatible strings to the respective board DT files.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Add generic compatible strings for Exynos4 and 5 platforms so that
future SoCs can use them if there is nothing extra/specific to be
differentiated.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Since there is very little difference between these two files,
they can be easily combined into one with necessary SoC checks.
While at it also merge the common.c file into this as it does
not have any other users.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
cpu_table was used to distinguish between different Exynos4 and 5
SoCs and based on the type do the initialization and io mapping.
exynos_init is dummy and no longer needed as we do a DT based booting.
By having a common io mapping function we can get rid of the whole
table and avoid populating it for every SoC.
Tested on Exynos4210, 5250 and 5420 based boards.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Tested-by: Tomasz Figa <t.figa@samsung.com>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Add helpers to check for Exynos4 and 5 family of SoCs.
This will eliminate comparing long list of SoCs and make
code simpler.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Reserve memory for MFC-v7 IP.
Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
Reorganize code so that "plat/mfc.h" is no more referred
from mach-exynos directory.
Signed-off-by: Tushar Behera <tushar.behera@linaro.org>
Signed-off-by: Kukjin Kim <kgene.kim@samsung.com>
|
|
The syscall.h headers were including linux/audit.h but really only
needed the uapi/linux/audit.h to get the requisite defines. Switch to
the uapi headers.
Signed-off-by: Eric Paris <eparis@redhat.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-mips@linux-mips.org
Cc: linux-s390@vger.kernel.org
Cc: x86@kernel.org
|
|
Every caller of syscall_get_arch() uses current for the task and no
implementors of the function need args. So just get rid of both of
those things. Admittedly, since these are inline functions we aren't
wasting stack space, but it just makes the prototypes better.
Signed-off-by: Eric Paris <eparis@redhat.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-mips@linux-mips.org
Cc: linux390@de.ibm.com
Cc: x86@kernel.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-s390@vger.kernel.org
Cc: linux-arch@vger.kernel.org
|
|
Currently AUDITSYSCALL has a long list of architecture depencency:
depends on AUDIT && (X86 || PARISC || PPC || S390 || IA64 || UML ||
SPARC64 || SUPERH || (ARM && AEABI && !OABI_COMPAT) || ALPHA)
The purpose of this patch is to replace it with HAVE_ARCH_AUDITSYSCALL
for simplicity.
Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Acked-by: Will Deacon <will.deacon@arm.com> (arm)
Acked-by: Richard Guy Briggs <rgb@redhat.com> (audit)
Acked-by: Matt Turner <mattst88@gmail.com> (alpha)
Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
Signed-off-by: Eric Paris <eparis@redhat.com>
|
|
On 03/15/2014 12:40 AM, Christoffer Dall wrote:
> On Fri, Mar 14, 2014 at 11:13:29AM +0530, Srivatsa S. Bhat wrote:
>> On 03/13/2014 04:51 AM, Christoffer Dall wrote:
>>> On Tue, Mar 11, 2014 at 02:05:38AM +0530, Srivatsa S. Bhat wrote:
>>>> Subsystems that want to register CPU hotplug callbacks, as well as perform
>>>> initialization for the CPUs that are already online, often do it as shown
>>>> below:
>>>>
[...]
>>> Just so we're clear, the existing code was simply racy as not prone to
>>> deadlocks, right?
>>>
>>> This makes it clear that the test above for compatible CPUs can be quite
>>> easily evaded by using CPU hotplug, but we don't really have a good
>>> solution for handling that yet... Hmmm, grumble grumble, I guess if you
>>> hotplug unsupported CPUs on a KVM/ARM system for now, stuff will break.
>>>
>>
>> In this particular case, there was no deadlock possibility, rather the
>> existing code had insufficient synchronization against CPU hotplug.
>>
>> init_hyp_mode() would invoke cpu_init_hyp_mode() on currently online CPUs
>> using on_each_cpu(). If a CPU came online after this point and before calling
>> register_cpu_notifier(), that CPU would remain uninitialized because this
>> subsystem would miss the hot-online event. This patch fixes this bug and
>> also uses the new synchronization method (instead of get/put_online_cpus())
>> to ensure that we don't deadlock with CPU hotplug.
>>
>
> Yes, that was my conclusion as well. Thanks for clarifying. (It could
> be noted in the commit message as well if you should feel so inclined).
>
Please find the patch with updated changelog (and your Ack) below.
(No changes in code).
From: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Subject: [PATCH] arm, kvm: Fix CPU hotplug callback registration
Subsystems that want to register CPU hotplug callbacks, as well as perform
initialization for the CPUs that are already online, often do it as shown
below:
get_online_cpus();
for_each_online_cpu(cpu)
init_cpu(cpu);
register_cpu_notifier(&foobar_cpu_notifier);
put_online_cpus();
This is wrong, since it is prone to ABBA deadlocks involving the
cpu_add_remove_lock and the cpu_hotplug.lock (when running concurrently
with CPU hotplug operations).
Instead, the correct and race-free way of performing the callback
registration is:
cpu_notifier_register_begin();
for_each_online_cpu(cpu)
init_cpu(cpu);
/* Note the use of the double underscored version of the API */
__register_cpu_notifier(&foobar_cpu_notifier);
cpu_notifier_register_done();
In the existing arm kvm code, there is no synchronization with CPU hotplug
to avoid missing the hotplug events that might occur after invoking
init_hyp_mode() and before calling register_cpu_notifier(). Fix this bug
and also use the new synchronization method (instead of get/put_online_cpus())
to ensure that we don't deadlock with CPU hotplug.
Cc: Gleb Natapov <gleb@kernel.org>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Ingo Molnar <mingo@kernel.org>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Acked-by: Christoffer Dall <christoffer.dall@linaro.org>
Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
|
