struct call_single_data is used in IPIs to transfer information between
CPUs. Its size is bigger than sizeof(unsigned long) and less than
cache line size. Currently it is not allocated with any explicit alignment
requirements. This makes it possible for allocated call_single_data to
cross two cache lines, which results in double the number of the cache lines
that need to be transferred among CPUs.
This can be fixed by requiring call_single_data to be aligned with the
size of call_single_data. Currently the size of call_single_data is the
power of 2. If we add new fields to call_single_data, we may need to
add padding to make sure the size of new definition is the power of 2
as well.
Fortunately, this is enforced by GCC, which will report bad sizes.
To set alignment requirements of call_single_data to the size of
call_single_data, a struct definition and a typedef is used.
To test the effect of the patch, I used the vm-scalability multiple
thread swap test case (swap-w-seq-mt). The test will create multiple
threads and each thread will eat memory until all RAM and part of swap
is used, so that huge number of IPIs are triggered when unmapping
memory. In the test, the throughput of memory writing improves ~5%
compared with misaligned call_single_data, because of faster IPIs.
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Huang, Ying <ying.huang@intel.com>
[ Add call_single_data_t and align with size of call_single_data. ]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Aaron Lu <aaron.lu@intel.com>
Cc: Borislav Petkov <bp@suse.de>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Juergen Gross <jgross@suse.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/87bmnqd6lz.fsf@yhuang-mobile.sh.intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
This was found with the -RT patch enabled, but the fix should apply to
non-RT also.
Used multi_v7_defconfig+PREEMPT_RT_FULL=y and this caused a compilation
warning without this fix:
../drivers/cpuidle/coupled.c:122:21: warning: 'cpuidle_coupled_lock'
defined but not used [-Wunused-variable]
Signed-off-by: Anders Roxell <anders.roxell@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Since we are using cpuidle_driver::safe_state_index directly as the
target state index, it is better to add the sanity check at the point
of registering the driver.
Signed-off-by: Xunlei Pang <pang.xunlei@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
For cpuidle_state_is_coupled(), 'dev' is not used, so remove it.
Signed-off-by: Xunlei Pang <pang.xunlei@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
cpuidle_device::safe_state_index need to be initialized before
use, it should be the same as cpuidle_driver::safe_state_index.
We tackled this issue by removing the safe_state_index from the
cpuidle_device structure and use the one in the cpuidle_driver
structure instead.
Suggested-by: Daniel Lezcano <daniel.lezcano@linaro.org>
Signed-off-by: Xunlei Pang <pang.xunlei@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Thanks to spatch, plus manual removal of "&*". Then a sweep for
for_each_cpu_mask => for_each_cpu.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Acked-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Jason Cooper <jason@lakedaemon.net>
Cc: Chris Metcalf <cmetcalf@ezchip.com>
Cc: netdev@vger.kernel.org
The name __smp_call_function_single() doesn't tell much about the
properties of this function, especially when compared to
smp_call_function_single().
The comments above the implementation are also misleading. The main
point of this function is actually not to be able to embed the csd
in an object. This is actually a requirement that result from the
purpose of this function which is to raise an IPI asynchronously.
As such it can be called with interrupts disabled. And this feature
comes at the cost of the caller who then needs to serialize the
IPIs on this csd.
Lets rename the function and enhance the comments so that they reflect
these properties.
Suggested-by: Christoph Hellwig <hch@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@fb.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
The main point of calling __smp_call_function_single() is to send
an IPI in a pure asynchronous way. By embedding a csd in an object,
a caller can send the IPI without waiting for a previous one to complete
as is required by smp_call_function_single() for example. As such,
sending this kind of IPI can be safe even when irqs are disabled.
This flexibility comes at the expense of the caller who then needs to
synchronize the csd lifecycle by himself and make sure that IPIs on a
single csd are serialized.
This is how __smp_call_function_single() works when wait = 0 and this
usecase is relevant.
Now there don't seem to be any usecase with wait = 1 that can't be
covered by smp_call_function_single() instead, which is safer. Lets look
at the two possible scenario:
1) The user calls __smp_call_function_single(wait = 1) on a csd embedded
in an object. It looks like a nice and convenient pattern at the first
sight because we can then retrieve the object from the IPI handler easily.
But actually it is a waste of memory space in the object since the csd
can be allocated from the stack by smp_call_function_single(wait = 1)
and the object can be passed an the IPI argument.
Besides that, embedding the csd in an object is more error prone
because the caller must take care of the serialization of the IPIs
for this csd.
2) The user calls __smp_call_function_single(wait = 1) on a csd that
is allocated on the stack. It's ok but smp_call_function_single()
can do it as well and it already takes care of the allocation on the
stack. Again it's more simple and less error prone.
Therefore, using the underscore prepend API version with wait = 1
is a bad pattern and a sign that the caller can do safer and more
simple.
There was a single user of that which has just been converted.
So lets remove this option to discourage further users.
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Jens Axboe <axboe@fb.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Some comments in cpuidle core files contain trivial mistakes.
This patch fixes them.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The coupled cpuidle waiting loop clears pending pokes before
entering the safe state. If a poke arrives just before the
pokes are cleared, but after the while loop condition checks,
the poke will be lost and the cpu will stay in the safe state
until another interrupt arrives. This may cause the cpu that
sent the poke to spin in the ready loop with interrupts off
until another cpu receives an interrupt, and if no other cpus
have interrupts routed to them it can spin forever.
Change the return value of cpuidle_coupled_clear_pokes to
return if a poke was cleared, and move the need_resched()
checks into the callers. In the waiting loop, if
a poke was cleared restart the loop to repeat the while
condition checks.
Reported-by: Neil Zhang <zhangwm@marvell.com>
Signed-off-by: Colin Cross <ccross@android.com>
Cc: 3.6+ <stable@vger.kernel.org> # 3.6+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Joseph Lo <josephl@nvidia.com> reported a lockup on Tegra20 caused
by a race condition in coupled cpuidle. When two or more cpus
enter idle at the same time, the first cpus to arrive may go to the
ready loop without processing pending pokes from the last cpu to
arrive.
This patch adds a check for pending pokes once all cpus have been
synchronized in the ready loop and resets the coupled state and
retries if any cpus failed to handle their pending poke.
Retrying on all cpus may trigger the same issue again, so this patch
also adds a check to ensure that each cpu has received at least one
poke between when it enters the waiting loop and when it moves on to
the ready loop.
Reported-and-tested-by: Joseph Lo <josephl@nvidia.com>
Tested-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Colin Cross <ccross@android.com>
Cc: 3.6+ <stable@vger.kernel.org> # 3.6+
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Calling cpuidle_enter_state is expected to return with interrupts
enabled, but interrupts must be disabled before starting the
ready loop synchronization stage. Call local_irq_disable after
each call to cpuidle_enter_state for the safe state.
Tested-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Colin Cross <ccross@android.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
The ready_waiting_counts atomic variable is compared against the wrong
online cpu count. The latter is computed incorrectly using logical-OR
instead of bit-OR. This patch fixes that.
Signed-off-by: Sivaram Nair <sivaramn@nvidia.com>
Acked-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Acked-by: Colin Cross <ccross@android.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
When a kernel is built to support multiple hardware types it's possible
that CONFIG_ARCH_NEEDS_CPU_IDLE_COUPLED is set but the hardware the
kernel is run on doesn't support cpuidle and therefore doesn't load a
driver for it. In this case, when the system is shut down,
cpuidle_coupled_cpu_notify() gets called with cpuidle_devices set to
NULL. There are quite possibly other circumstances where this
situation can also occur and we should check for it.
Signed-off-by: Jon Medhurst <tixy@linaro.org>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
The cpu hotplug notifier gets called in both atomic and non-atomic
contexts, it is not always safe to lock a mutex. Filter out all events
except the six necessary ones, which are all sleepable, before taking
the mutex.
Signed-off-by: Colin Cross <ccross@android.com>
Reviewed-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Adds cpuidle_coupled_parallel_barrier, which can be used by coupled
cpuidle state enter functions to handle resynchronization after
determining if any cpu needs to abort. The normal use case will
be:
static bool abort_flag;
static atomic_t abort_barrier;
int arch_cpuidle_enter(struct cpuidle_device *dev, ...)
{
if (arch_turn_off_irq_controller()) {
/* returns an error if an irq is pending and would be lost
if idle continued and turned off power */
abort_flag = true;
}
cpuidle_coupled_parallel_barrier(dev, &abort_barrier);
if (abort_flag) {
/* One of the cpus didn't turn off it's irq controller */
arch_turn_on_irq_controller();
return -EINTR;
}
/* continue with idle */
...
}
This will cause all cpus to abort idle together if one of them needs
to abort.
Reviewed-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Tested-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>
Tested-by: Kevin Hilman <khilman@ti.com>
Signed-off-by: Colin Cross <ccross@android.com>
Signed-off-by: Len Brown <len.brown@intel.com>
On some ARM SMP SoCs (OMAP4460, Tegra 2, and probably more), the
cpus cannot be independently powered down, either due to
sequencing restrictions (on Tegra 2, cpu 0 must be the last to
power down), or due to HW bugs (on OMAP4460, a cpu powering up
will corrupt the gic state unless the other cpu runs a work
around). Each cpu has a power state that it can enter without
coordinating with the other cpu (usually Wait For Interrupt, or
WFI), and one or more "coupled" power states that affect blocks
shared between the cpus (L2 cache, interrupt controller, and
sometimes the whole SoC). Entering a coupled power state must
be tightly controlled on both cpus.
The easiest solution to implementing coupled cpu power states is
to hotplug all but one cpu whenever possible, usually using a
cpufreq governor that looks at cpu load to determine when to
enable the secondary cpus. This causes problems, as hotplug is an
expensive operation, so the number of hotplug transitions must be
minimized, leading to very slow response to loads, often on the
order of seconds.
This file implements an alternative solution, where each cpu will
wait in the WFI state until all cpus are ready to enter a coupled
state, at which point the coupled state function will be called
on all cpus at approximately the same time.
Once all cpus are ready to enter idle, they are woken by an smp
cross call. At this point, there is a chance that one of the
cpus will find work to do, and choose not to enter idle. A
final pass is needed to guarantee that all cpus will call the
power state enter function at the same time. During this pass,
each cpu will increment the ready counter, and continue once the
ready counter matches the number of online coupled cpus. If any
cpu exits idle, the other cpus will decrement their counter and
retry.
To use coupled cpuidle states, a cpuidle driver must:
Set struct cpuidle_device.coupled_cpus to the mask of all
coupled cpus, usually the same as cpu_possible_mask if all cpus
are part of the same cluster. The coupled_cpus mask must be
set in the struct cpuidle_device for each cpu.
Set struct cpuidle_device.safe_state to a state that is not a
coupled state. This is usually WFI.
Set CPUIDLE_FLAG_COUPLED in struct cpuidle_state.flags for each
state that affects multiple cpus.
Provide a struct cpuidle_state.enter function for each state
that affects multiple cpus. This function is guaranteed to be
called on all cpus at approximately the same time. The driver
should ensure that the cpus all abort together if any cpu tries
to abort once the function is called.
update1:
cpuidle: coupled: fix count of online cpus
online_count was never incremented on boot, and was also counting
cpus that were not part of the coupled set. Fix both issues by
introducting a new function that counts online coupled cpus, and
call it from register as well as the hotplug notifier.
update2:
cpuidle: coupled: fix decrementing ready count
cpuidle_coupled_set_not_ready sometimes refuses to decrement the
ready count in order to prevent a race condition. This makes it
unsuitable for use when finished with idle. Add a new function
cpuidle_coupled_set_done that decrements both the ready count and
waiting count, and call it after idle is complete.
Cc: Amit Kucheria <amit.kucheria@linaro.org>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Trinabh Gupta <g.trinabh@gmail.com>
Cc: Deepthi Dharwar <deepthi@linux.vnet.ibm.com>
Reviewed-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Tested-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Reviewed-by: Kevin Hilman <khilman@ti.com>
Tested-by: Kevin Hilman <khilman@ti.com>
Signed-off-by: Colin Cross <ccross@android.com>
Acked-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Len Brown <len.brown@intel.com>
Ying Huang