On Sun, Apr 21, 2013 at 09:37:05PM +0200, Borislav Petkov wrote:
> On Tue, Apr 16, 2013 at 09:41:30AM -0700, Paul E. McKenney wrote:
> > From: "Paul E. McKenney" <paul...@linux.vnet.ibm.com>
> >
> > The Linux kernel uses a number of per-CPU kthreads, any of which might
> > contribute to OS jitter at any time. The usual approach to normal
> > kthreads, namely to bind them to a "housekeeping" CPU, does not work
> > with these kthreads because they cannot operate correctly if moved to
> > some other CPU. This commit therefore lists ways of controlling OS
> > jitter from the Linux kernel's per-CPU kthreads.
> >
> > Signed-off-by: Paul E. McKenney <paul...@linux.vnet.ibm.com>
> > Cc: Frederic Weisbecker <fweis...@gmail.com>
> > Cc: Steven Rostedt <rost...@goodmis.org>
> > Cc: Borislav Petkov <b...@alien8.de>
> > Cc: Arjan van de Ven <ar...@linux.intel.com>
> > Cc: Kevin Hilman <khil...@linaro.org>
> > Cc: Christoph Lameter <c...@linux.com>
> > Cc: Thomas Gleixner <t...@linutronix.de>
> > Cc: Olivier Baetz <olivier.ba...@novasparks.com>
> > Reviewed-by: Randy Dunlap <rdun...@infradead.org>
> > ---
> > Documentation/kernel-per-CPU-kthreads.txt | 186
> > ++++++++++++++++++++++++++++++
> > 1 file changed, 186 insertions(+)
> > create mode 100644 Documentation/kernel-per-CPU-kthreads.txt
> >
> > diff --git a/Documentation/kernel-per-CPU-kthreads.txt
> > b/Documentation/kernel-per-CPU-kthreads.txt
> > new file mode 100644
> > index 0000000..bfecc1c
> > --- /dev/null
> > +++ b/Documentation/kernel-per-CPU-kthreads.txt
> > @@ -0,0 +1,186 @@
> > +REDUCING OS JITTER DUE TO PER-CPU KTHREADS
> > +
> > +This document lists per-CPU kthreads in the Linux kernel and presents
> > +options to control OS jitter due to these kthreads. Note that kthreads
>
> s/due to/which can be caused by/
Same meaning, but "due to" is probably a bit more arcane. But how
about "and presents options to control these kthreads' OS jitter"?
> > +that are not per-CPU are not listed here -- to reduce OS jitter from
>
> one too many "that"s:
>
> s/that/which/
Fair point, but I can shorten it as follows:
Note that non-per-CPU kthreads CPU are not listed here --
to reduce OS jitter from non-per-CPU kthreads, bind them to a
"housekeeping" CPU that is dedicated to such work.
> > +non-per-CPU kthreads, bind them to a "housekeeping" CPU that is dedicated
>
> s/that/which/
Good catch -- I chose s/that is//.
> > +to such work.
> > +
> > +
> > +REFERENCES
> > +
> > +o Documentation/IRQ-affinity.txt: Binding interrupts to sets of CPUs.
> > +
> > +o Documentation/cgroups: Using cgroups to bind tasks to sets of CPUs.
> > +
> > +o man taskset: Using the taskset command to bind tasks to sets
> > + of CPUs.
> > +
> > +o man sched_setaffinity: Using the sched_setaffinity() system
> > + call to bind tasks to sets of CPUs.
> > +
> > +
> > +KTHREADS
> > +
> > +Name: ehca_comp/%u
> > +Purpose: Periodically process Infiniband-related work.
> > +To reduce corresponding OS jitter, do any of the following:
> > +1. Don't use EHCA Infiniband hardware. This will prevent these
>
> Sounds like this particular hardware is slow and its IRQ handler/softirq
> needs a lot of time. Yes, no?
>
> Can we have a reason why people shouldn't use that hw.
Because it has per-CPU kthreads that can cause OS jitter. ;-)
> > + kthreads from being created in the first place. (This will
> > + work for most people, as this hardware, though important,
> > + is relatively old and is produced in relatively low unit
> > + volumes.)
> > +2. Do all EHCA-Infiniband-related work on other CPUs, including
> > + interrupts.
> > +
> > +
> > +Name: irq/%d-%s
> > +Purpose: Handle threaded interrupts.
> > +To reduce corresponding OS jitter, do the following:
>
> This sentence keeps repeating; maybe explain the purpose of this doc in
> the beginning once and drop this sentence in the later sections.
There are "any of" and "all of" qualifiers. Also, I cannot count on
someone reading the document beginning to end. I would instead expect
many of them to search for the name of the kthread that is bothering
them and read only that part.
> > +1. Use irq affinity to force the irq threads to execute on
> > + some other CPU.
> > +
> > +Name: kcmtpd_ctr_%d
> > +Purpose: Handle Bluetooth work.
> > +To reduce corresponding OS jitter, do one of the following:
> > +1. Don't use Bluetooth, in which case these kthreads won't be
> > + created in the first place.
> > +2. Use irq affinity to force Bluetooth-related interrupts to
> > + occur on some other CPU and furthermore initiate all
> > + Bluetooth activity on some other CPU.
> > +
> > +Name: ksoftirqd/%u
> > +Purpose: Execute softirq handlers when threaded or when under heavy load.
> > +To reduce corresponding OS jitter, each softirq vector must be handled
> > +separately as follows:
> > +TIMER_SOFTIRQ: Do all of the following:
> > +1. To the extent possible, keep the CPU out of the kernel when it
> > + is non-idle, for example, by avoiding system calls and by forcing
> > + both kernel threads and interrupts to execute elsewhere.
> > +2. Build with CONFIG_HOTPLUG_CPU=y. After boot completes, force
> > + the CPU offline, then bring it back online. This forces
> > + recurring timers to migrate elsewhere. If you are concerned
>
> We don't migrate them back to that CPU when we online it again, do we?
Not unless the CPU it migrated to later is taken offline. Good point,
added words to that effect.
> > + with multiple CPUs, force them all offline before bringing the
> > + first one back online.
> > +NET_TX_SOFTIRQ and NET_RX_SOFTIRQ: Do all of the following:
> > +1. Force networking interrupts onto other CPUs.
> > +2. Initiate any network I/O on other CPUs.
> > +3. Once your application has started, prevent CPU-hotplug operations
> > + from being initiated from tasks that might run on the CPU to
> > + be de-jittered. (It is OK to force this CPU offline and then
> > + bring it back online before you start your application.)
> > +BLOCK_SOFTIRQ: Do all of the following:
> > +1. Force block-device interrupts onto some other CPU.
> > +2. Initiate any block I/O on other CPUs.
> > +3. Once your application has started, prevent CPU-hotplug operations
> > + from being initiated from tasks that might run on the CPU to
> > + be de-jittered. (It is OK to force this CPU offline and then
> > + bring it back online before you start your application.)
> > +BLOCK_IOPOLL_SOFTIRQ: Do all of the following:
> > +1. Force block-device interrupts onto some other CPU.
> > +2. Initiate any block I/O and block-I/O polling on other CPUs.
> > +3. Once your application has started, prevent CPU-hotplug operations
> > + from being initiated from tasks that might run on the CPU to
> > + be de-jittered. (It is OK to force this CPU offline and then
> > + bring it back online before you start your application.)
>
> more repeated text in brackets, maybe a footnote somewhere instead...
Indeed, it is a bit repetitive, but I expect that people will tend
to look just at the part that seems relevant rather than reading the
whole thing.
> > +TASKLET_SOFTIRQ: Do one or more of the following:
> > +1. Avoid use of drivers that use tasklets.
> > +2. Convert all drivers that you must use from tasklets to workqueues.
> > +3. Force interrupts for drivers using tasklets onto other CPUs,
> > + and also do I/O involving these drivers on other CPUs.
>
> How do I check which drivers use tasklets?
Good point -- I added "(Such drivers will contain calls to things like
tasklet_schedule().)"
> > +SCHED_SOFTIRQ: Do all of the following:
> > +1. Avoid sending scheduler IPIs to the CPU to be de-jittered,
> > + for example, ensure that at most one runnable kthread is
>
> To which sentence does "for example" belong to? Depending on the answer,
> you can split that sentence.
It belongs with the first sentence.
> > + present on that CPU. If a thread awakens that expects
> > + to run on the de-jittered CPU, the scheduler will send
>
> "If a thread expecting to run ont the de-jittered CPU awakens, the
> scheduler..."
Sold!
> > + an IPI that can result in a subsequent SCHED_SOFTIRQ.
> > +2. Build with CONFIG_RCU_NOCB_CPU=y, CONFIG_RCU_NOCB_CPU_ALL=y,
> > + CONFIG_NO_HZ_FULL=y, and in addition ensure that the CPU
>
> commas:
>
> , and, in addition, ensure...
Good catch, fixed.
> > + to be de-jittered is marked as an adaptive-ticks CPU using the
> > + "nohz_full=" boot parameter. This reduces the number of
> > + scheduler-clock interrupts that the de-jittered CPU receives,
> > + minimizing its chances of being selected to do load balancing,
>
> I don't think there's a "," if the "which... " part refers to the
> previous "load balancing" and not to the whole sentence.
Good point -- I can reword to:
This reduces the number of scheduler-clock interrupts that the
de-jittered CPU receives, minimizing its chances of being selected
to do the load balancing work that runs in SCHED_SOFTIRQ context.
> > + which happens in SCHED_SOFTIRQ context.
> > +3. To the extent possible, keep the CPU out of the kernel when it
> > + is non-idle, for example, by avoiding system calls and by
> > + forcing both kernel threads and interrupts to execute elsewhere.
>
> This time "for example" reads ok.
Glad you like it. ;-)
> > + This further reduces the number of scheduler-clock interrupts
> > + that the de-jittered CPU receives.
>
> s/that/which/ would suit better here IMHO.
Fair point, but how about this?
This further reduces the number of scheduler-clock interrupts
received by the de-jittered CPU.
> > +HRTIMER_SOFTIRQ: Do all of the following:
> > +1. To the extent possible, keep the CPU out of the kernel when it
> > + is non-idle, for example, by avoiding system calls and by forcing
> > + both kernel threads and interrupts to execute elsewhere.
>
> Ok, I think I get your "for example" usage pattern.
>
> "blabablabla. For example, do blabalbal."
>
> I think that would be a bit more readable.
In this case, agreed:
To the extent possible, keep the CPU out of the kernel when it
is non-idle. For example, avoid system calls and force both
kernel threads and interrupts to execute elsewhere.
> > +2. Build with CONFIG_HOTPLUG_CPU=y. Once boot completes, force the
> > + CPU offline, then bring it back online. This forces recurring
> > + timers to migrate elsewhere. If you are concerned with multiple
> > + CPUs, force them all offline before bringing the first one
> > + back online.
>
> Same question: do the timers get migrated back when the CPU reappears
> online?
Good point, applied the same change here.
> > +RCU_SOFTIRQ: Do at least one of the following:
> > +1. Offload callbacks and keep the CPU in either dyntick-idle or
> > + adaptive-ticks state by doing all of the following:
> > + a. Build with CONFIG_RCU_NOCB_CPU=y, CONFIG_RCU_NOCB_CPU_ALL=y,
> > + CONFIG_NO_HZ_FULL=y, and in addition ensure that the CPU
>
> , and, in addition,
>
> > + to be de-jittered is marked as an adaptive-ticks CPU
> > + using the "nohz_full=" boot parameter. Bind the rcuo
> > + kthreads to housekeeping CPUs that can tolerate OS jitter.
>
> which
Good point, took both.
> > + b. To the extent possible, keep the CPU out of the kernel
> > + when it is non-idle, for example, by avoiding system
> > + calls and by forcing both kernel threads and interrupts
> > + to execute elsewhere.
> > +2. Enable RCU to do its processing remotely via dyntick-idle by
> > + doing all of the following:
> > + a. Build with CONFIG_NO_HZ=y and CONFIG_RCU_FAST_NO_HZ=y.
> > + b. Ensure that the CPU goes idle frequently, allowing other
>
> I'm ensuring that by selecting the proper workload which has idle
> breathers?
Yep! Or, equivalently, by adding enough CPUs so that the workload
has idle breathers.
> > + CPUs to detect that it has passed through an RCU quiescent
> > + state. If the kernel is built with CONFIG_NO_HZ_FULL=y,
> > + userspace execution also allows other CPUs to detect that
> > + the CPU in question has passed through a quiescent state.
> > + c. To the extent possible, keep the CPU out of the kernel
> > + when it is non-idle, for example, by avoiding system
> > + calls and by forcing both kernel threads and interrupts
> > + to execute elsewhere.
> > +
> > +Name: rcuc/%u
> > +Purpose: Execute RCU callbacks in CONFIG_RCU_BOOST=y kernels.
> > +To reduce corresponding OS jitter, do at least one of the following:
> > +1. Build the kernel with CONFIG_PREEMPT=n. This prevents these
> > + kthreads from being created in the first place, and also prevents
> > + RCU priority boosting from ever being required. This approach
>
> "... this obviates the need for RCU priority boosting."
Sold!
> > + is feasible for workloads that do not require high degrees of
> > + responsiveness.
> > +2. Build the kernel with CONFIG_RCU_BOOST=n. This prevents these
> > + kthreads from being created in the first place. This approach
> > + is feasible only if your workload never requires RCU priority
> > + boosting, for example, if you ensure frequent idle time on all
> > + CPUs that might execute within the kernel.
> > +3. Build with CONFIG_RCU_NOCB_CPU=y and CONFIG_RCU_NOCB_CPU_ALL=y,
> > + which offloads all RCU callbacks to kthreads that can be moved
> > + off of CPUs susceptible to OS jitter. This approach prevents the
> > + rcuc/%u kthreads from having any work to do, so that they are
> > + never awakened.
> > +4. Ensure that the CPU never enters the kernel and in particular
>
> , and, in particular,
Good, fixed.
> > + avoid initiating any CPU hotplug operations on this CPU. This is
> > + another way of preventing any callbacks from being queued on the
> > + CPU, again preventing the rcuc/%u kthreads from having any work
> > + to do.
> > +
> > +Name: rcuob/%d, rcuop/%d, and rcuos/%d
> > +Purpose: Offload RCU callbacks from the corresponding CPU.
> > +To reduce corresponding OS jitter, do at least one of the following:
> > +1. Use affinity, cgroups, or other mechanism to force these kthreads
> > + to execute on some other CPU.
> > +2. Build with CONFIG_RCU_NOCB_CPUS=n, which will prevent these
> > + kthreads from being created in the first place. However,
> > + please note that this will not eliminate the corresponding
>
> can we drop "corresponding" here?
Yep! Dropped the preceding "the" as well, just to be on the safe side.
> > + OS jitter, but will instead shift it to RCU_SOFTIRQ.
> > +
> > +Name: watchdog/%u
> > +Purpose: Detect software lockups on each CPU.
> > +To reduce corresponding OS jitter, do at least one of the following:
>
> ditto.
I changed "corresponding" to "its" globally for this lead-in sentence.
> > +1. Build with CONFIG_LOCKUP_DETECTOR=n, which will prevent these
> > + kthreads from being created in the first place.
> > +2. Echo a zero to /proc/sys/kernel/watchdog to disable the
> > + watchdog timer.
> > +3. Echo a large number of /proc/sys/kernel/watchdog_thresh in
> > + order to reduce the frequency of OS jitter due to the watchdog
> > + timer down to a level that is acceptable for your workload.
Thank you for the thorough review and comments! Please see below for
an update.
Thanx, Paul
------------------------------------------------------------------------
REDUCING OS JITTER DUE TO PER-CPU KTHREADS
This document lists per-CPU kthreads in the Linux kernel and presents
options to control these kthreads' OS jitter. Note that non-per-CPU
kthreads CPU are not listed here. To reduce OS jitter from non-per-CPU
kthreads, bind them to a "housekeeping" CPU dedicated to such work.
REFERENCES
o Documentation/IRQ-affinity.txt: Binding interrupts to sets of CPUs.
o Documentation/cgroups: Using cgroups to bind tasks to sets of CPUs.
o man taskset: Using the taskset command to bind tasks to sets
of CPUs.
o man sched_setaffinity: Using the sched_setaffinity() system
call to bind tasks to sets of CPUs.
KTHREADS
Name: ehca_comp/%u
Purpose: Periodically process Infiniband-related work.
To reduce its OS jitter, do any of the following:
1. Don't use eHCA Infiniband hardware. This will prevent these
kthreads from being created in the first place. (This will
work for most people, as this hardware, though important,
is relatively old and is produced in relatively low unit
volumes.)
2. Do all eHCA-Infiniband-related work on other CPUs, including
interrupts.
Name: irq/%d-%s
Purpose: Handle threaded interrupts.
To reduce its OS jitter, do the following:
1. Use irq affinity to force the irq threads to execute on
some other CPU.
Name: kcmtpd_ctr_%d
Purpose: Handle Bluetooth work.
To reduce its OS jitter, do one of the following:
1. Don't use Bluetooth, in which case these kthreads won't be
created in the first place.
2. Use irq affinity to force Bluetooth-related interrupts to
occur on some other CPU and furthermore initiate all
Bluetooth activity on some other CPU.
Name: ksoftirqd/%u
Purpose: Execute softirq handlers when threaded or when under heavy load.
To reduce its OS jitter, each softirq vector must be handled
separately as follows:
TIMER_SOFTIRQ: Do all of the following:
1. To the extent possible, keep the CPU out of the kernel when it
is non-idle, for example, by avoiding system calls and by forcing
both kernel threads and interrupts to execute elsewhere.
2. Build with CONFIG_HOTPLUG_CPU=y. After boot completes, force
the CPU offline, then bring it back online. This forces
recurring timers to migrate elsewhere. If you are concerned
with multiple CPUs, force them all offline before bringing the
first one back online. Once you have onlined the CPUs in question,
do not offline any other CPUs, because doing so could force the
timer back onto one of the CPUs in question.
NET_TX_SOFTIRQ and NET_RX_SOFTIRQ: Do all of the following:
1. Force networking interrupts onto other CPUs.
2. Initiate any network I/O on other CPUs.
3. Once your application has started, prevent CPU-hotplug operations
from being initiated from tasks that might run on the CPU to
be de-jittered. (It is OK to force this CPU offline and then
bring it back online before you start your application.)
BLOCK_SOFTIRQ: Do all of the following:
1. Force block-device interrupts onto some other CPU.
2. Initiate any block I/O on other CPUs.
3. Once your application has started, prevent CPU-hotplug operations
from being initiated from tasks that might run on the CPU to
be de-jittered. (It is OK to force this CPU offline and then
bring it back online before you start your application.)
BLOCK_IOPOLL_SOFTIRQ: Do all of the following:
1. Force block-device interrupts onto some other CPU.
2. Initiate any block I/O and block-I/O polling on other CPUs.
3. Once your application has started, prevent CPU-hotplug operations
from being initiated from tasks that might run on the CPU to
be de-jittered. (It is OK to force this CPU offline and then
bring it back online before you start your application.)
TASKLET_SOFTIRQ: Do one or more of the following:
1. Avoid use of drivers that use tasklets. (Such drivers will contain
calls to things like tasklet_schedule().)
2. Convert all drivers that you must use from tasklets to workqueues.
3. Force interrupts for drivers using tasklets onto other CPUs,
and also do I/O involving these drivers on other CPUs.
SCHED_SOFTIRQ: Do all of the following:
1. Avoid sending scheduler IPIs to the CPU to be de-jittered,
for example, ensure that at most one runnable kthread is present
on that CPU. If a thread that expects to run on the de-jittered
CPU awakens, the scheduler will send an IPI that can result in
a subsequent SCHED_SOFTIRQ.
2. Build with CONFIG_RCU_NOCB_CPU=y, CONFIG_RCU_NOCB_CPU_ALL=y,
CONFIG_NO_HZ_FULL=y, and, in addition, ensure that the CPU
to be de-jittered is marked as an adaptive-ticks CPU using the
"nohz_full=" boot parameter. This reduces the number of
scheduler-clock interrupts that the de-jittered CPU receives,
minimizing its chances of being selected to do the load balancing
work that runs in SCHED_SOFTIRQ context.
3. To the extent possible, keep the CPU out of the kernel when it
is non-idle, for example, by avoiding system calls and by
forcing both kernel threads and interrupts to execute elsewhere.
This further reduces the number of scheduler-clock interrupts
received by the de-jittered CPU.
HRTIMER_SOFTIRQ: Do all of the following:
1. To the extent possible, keep the CPU out of the kernel when it
is non-idle. For example, avoid system calls and force both
kernel threads and interrupts to execute elsewhere.
2. Build with CONFIG_HOTPLUG_CPU=y. Once boot completes, force the
CPU offline, then bring it back online. This forces recurring
timers to migrate elsewhere. If you are concerned with multiple
CPUs, force them all offline before bringing the first one
back online. Once you have onlined the CPUs in question, do not
offline any other CPUs, because doing so could force the timer
back onto one of the CPUs in question.
RCU_SOFTIRQ: Do at least one of the following:
1. Offload callbacks and keep the CPU in either dyntick-idle or
adaptive-ticks state by doing all of the following:
a. Build with CONFIG_RCU_NOCB_CPU=y, CONFIG_RCU_NOCB_CPU_ALL=y,
CONFIG_NO_HZ_FULL=y, and, in addition ensure that the CPU
to be de-jittered is marked as an adaptive-ticks CPU using
the "nohz_full=" boot parameter. Bind the rcuo kthreads
to housekeeping CPUs, which can tolerate OS jitter.
b. To the extent possible, keep the CPU out of the kernel
when it is non-idle, for example, by avoiding system
calls and by forcing both kernel threads and interrupts
to execute elsewhere.
2. Enable RCU to do its processing remotely via dyntick-idle by
doing all of the following:
a. Build with CONFIG_NO_HZ=y and CONFIG_RCU_FAST_NO_HZ=y.
b. Ensure that the CPU goes idle frequently, allowing other
CPUs to detect that it has passed through an RCU quiescent
state. If the kernel is built with CONFIG_NO_HZ_FULL=y,
userspace execution also allows other CPUs to detect that
the CPU in question has passed through a quiescent state.
c. To the extent possible, keep the CPU out of the kernel
when it is non-idle, for example, by avoiding system
calls and by forcing both kernel threads and interrupts
to execute elsewhere.
Name: rcuc/%u
Purpose: Execute RCU callbacks in CONFIG_RCU_BOOST=y kernels.
To reduce its OS jitter, do at least one of the following:
1. Build the kernel with CONFIG_PREEMPT=n. This prevents these
kthreads from being created in the first place, and also obviates
the need for RCU priority boosting. This approach is feasible
for workloads that do not require high degrees of responsiveness.
2. Build the kernel with CONFIG_RCU_BOOST=n. This prevents these
kthreads from being created in the first place. This approach
is feasible only if your workload never requires RCU priority
boosting, for example, if you ensure frequent idle time on all
CPUs that might execute within the kernel.
3. Build with CONFIG_RCU_NOCB_CPU=y and CONFIG_RCU_NOCB_CPU_ALL=y,
which offloads all RCU callbacks to kthreads that can be moved
off of CPUs susceptible to OS jitter. This approach prevents the
rcuc/%u kthreads from having any work to do, so that they are
never awakened.
4. Ensure that the CPU never enters the kernel, and, in particular,
avoid initiating any CPU hotplug operations on this CPU. This is
another way of preventing any callbacks from being queued on the
CPU, again preventing the rcuc/%u kthreads from having any work
to do.
Name: rcuob/%d, rcuop/%d, and rcuos/%d
Purpose: Offload RCU callbacks from the corresponding CPU.
To reduce its OS jitter, do at least one of the following:
1. Use affinity, cgroups, or other mechanism to force these kthreads
to execute on some other CPU.
2. Build with CONFIG_RCU_NOCB_CPUS=n, which will prevent these
kthreads from being created in the first place. However, please
note that this will not eliminate OS jitter, but will instead
shift it to RCU_SOFTIRQ.
Name: watchdog/%u
Purpose: Detect software lockups on each CPU.
To reduce its OS jitter, do at least one of the following:
1. Build with CONFIG_LOCKUP_DETECTOR=n, which will prevent these
kthreads from being created in the first place.
2. Echo a zero to /proc/sys/kernel/watchdog to disable the
watchdog timer.
3. Echo a large number of /proc/sys/kernel/watchdog_thresh in
order to reduce the frequency of OS jitter due to the watchdog
timer down to a level that is acceptable for your workload.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
