diff options
Diffstat (limited to 'Documentation')
-rw-r--r-- | Documentation/kernel-per-CPU-kthreads.txt | 202 | ||||
-rw-r--r-- | Documentation/power/devices.txt | 15 | ||||
-rw-r--r-- | Documentation/power/interface.txt | 4 | ||||
-rw-r--r-- | Documentation/power/notifiers.txt | 6 | ||||
-rw-r--r-- | Documentation/power/states.txt | 30 |
5 files changed, 233 insertions, 24 deletions
diff --git a/Documentation/kernel-per-CPU-kthreads.txt b/Documentation/kernel-per-CPU-kthreads.txt new file mode 100644 index 0000000..cbf7ae4 --- /dev/null +++ b/Documentation/kernel-per-CPU-kthreads.txt @@ -0,0 +1,202 @@ +REDUCING OS JITTER DUE TO PER-CPU KTHREADS + +This document lists per-CPU kthreads in the Linux kernel and presents +options to control their OS jitter. Note that non-per-CPU kthreads 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. + +o /sys/devices/system/cpu/cpuN/online: Control CPU N's hotplug state, + writing "0" to offline and "1" to online. + +o In order to locate kernel-generated OS jitter on CPU N: + + cd /sys/kernel/debug/tracing + echo 1 > max_graph_depth # Increase the "1" for more detail + echo function_graph > current_tracer + # run workload + cat per_cpu/cpuN/trace + + +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, instead choosing hardware + that does not require per-CPU kthreads. 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. +3. Rework the eHCA driver so that its per-CPU kthreads are + provisioned only on selected CPUs. + + +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. diff --git a/Documentation/power/devices.txt b/Documentation/power/devices.txt index 504dfe4..a66c982 100644 --- a/Documentation/power/devices.txt +++ b/Documentation/power/devices.txt @@ -268,7 +268,7 @@ situations. System Power Management Phases ------------------------------ Suspending or resuming the system is done in several phases. Different phases -are used for standby or memory sleep states ("suspend-to-RAM") and the +are used for freeze, standby, and memory sleep states ("suspend-to-RAM") and the hibernation state ("suspend-to-disk"). Each phase involves executing callbacks for every device before the next phase begins. Not all busses or classes support all these callbacks and not all drivers use all the callbacks. The @@ -309,7 +309,8 @@ execute the corresponding method from dev->driver->pm instead if there is one. Entering System Suspend ----------------------- -When the system goes into the standby or memory sleep state, the phases are: +When the system goes into the freeze, standby or memory sleep state, +the phases are: prepare, suspend, suspend_late, suspend_noirq. @@ -368,7 +369,7 @@ the devices that were suspended. Leaving System Suspend ---------------------- -When resuming from standby or memory sleep, the phases are: +When resuming from freeze, standby or memory sleep, the phases are: resume_noirq, resume_early, resume, complete. @@ -433,8 +434,8 @@ the system log. Entering Hibernation -------------------- -Hibernating the system is more complicated than putting it into the standby or -memory sleep state, because it involves creating and saving a system image. +Hibernating the system is more complicated than putting it into the other +sleep states, because it involves creating and saving a system image. Therefore there are more phases for hibernation, with a different set of callbacks. These phases always run after tasks have been frozen and memory has been freed. @@ -485,8 +486,8 @@ image forms an atomic snapshot of the system state. At this point the system image is saved, and the devices then need to be prepared for the upcoming system shutdown. This is much like suspending them -before putting the system into the standby or memory sleep state, and the phases -are similar. +before putting the system into the freeze, standby or memory sleep state, +and the phases are similar. 9. The prepare phase is discussed above. diff --git a/Documentation/power/interface.txt b/Documentation/power/interface.txt index c537834..f1f0f59a 100644 --- a/Documentation/power/interface.txt +++ b/Documentation/power/interface.txt @@ -7,8 +7,8 @@ running. The interface exists in /sys/power/ directory (assuming sysfs is mounted at /sys). /sys/power/state controls system power state. Reading from this file -returns what states are supported, which is hard-coded to 'standby' -(Power-On Suspend), 'mem' (Suspend-to-RAM), and 'disk' +returns what states are supported, which is hard-coded to 'freeze', +'standby' (Power-On Suspend), 'mem' (Suspend-to-RAM), and 'disk' (Suspend-to-Disk). Writing to this file one of those strings causes the system to diff --git a/Documentation/power/notifiers.txt b/Documentation/power/notifiers.txt index c2a4a34..a81fa25 100644 --- a/Documentation/power/notifiers.txt +++ b/Documentation/power/notifiers.txt @@ -15,8 +15,10 @@ A suspend/hibernation notifier may be used for this purpose. The subsystems or drivers having such needs can register suspend notifiers that will be called upon the following events by the PM core: -PM_HIBERNATION_PREPARE The system is going to hibernate or suspend, tasks will - be frozen immediately. +PM_HIBERNATION_PREPARE The system is going to hibernate, tasks will be frozen + immediately. This is different from PM_SUSPEND_PREPARE + below because here we do additional work between notifiers + and drivers freezing. PM_POST_HIBERNATION The system memory state has been restored from a hibernation image or an error occurred during diff --git a/Documentation/power/states.txt b/Documentation/power/states.txt index 4416b28..442d43d 100644 --- a/Documentation/power/states.txt +++ b/Documentation/power/states.txt @@ -2,12 +2,26 @@ System Power Management States -The kernel supports three power management states generically, though -each is dependent on platform support code to implement the low-level -details for each state. This file describes each state, what they are +The kernel supports four power management states generically, though +one is generic and the other three are dependent on platform support +code to implement the low-level details for each state. +This file describes each state, what they are commonly called, what ACPI state they map to, and what string to write to /sys/power/state to enter that state +state: Freeze / Low-Power Idle +ACPI state: S0 +String: "freeze" + +This state is a generic, pure software, light-weight, low-power state. +It allows more energy to be saved relative to idle by freezing user +space and putting all I/O devices into low-power states (possibly +lower-power than available at run time), such that the processors can +spend more time in their idle states. +This state can be used for platforms without Standby/Suspend-to-RAM +support, or it can be used in addition to Suspend-to-RAM (memory sleep) +to provide reduced resume latency. + State: Standby / Power-On Suspend ACPI State: S1 @@ -22,9 +36,6 @@ We try to put devices in a low-power state equivalent to D1, which also offers low power savings, but low resume latency. Not all devices support D1, and those that don't are left on. -A transition from Standby to the On state should take about 1-2 -seconds. - State: Suspend-to-RAM ACPI State: S3 @@ -42,9 +53,6 @@ transition back to the On state. For at least ACPI, STR requires some minimal boot-strapping code to resume the system from STR. This may be true on other platforms. -A transition from Suspend-to-RAM to the On state should take about -3-5 seconds. - State: Suspend-to-disk ACPI State: S4 @@ -74,7 +82,3 @@ low-power state (like ACPI S4), or it may simply power down. Powering down offers greater savings, and allows this mechanism to work on any system. However, entering a real low-power state allows the user to trigger wake up events (e.g. pressing a key or opening a laptop lid). - -A transition from Suspend-to-Disk to the On state should take about 30 -seconds, though it's typically a bit more with the current -implementation. |