Concurrency Managed Workqueue (cmwq)¶

Date:	September, 2010
Author:	Tejun Heo <tj@kernel.org>
Author:	Florian Mickler <florian@mickler.org>

Introduction¶

There are many cases where an asynchronous process execution context is needed and the workqueue (wq) API is the most commonly used mechanism for such cases.

When such an asynchronous execution context is needed, a work item describing which function to execute is put on a queue. An independent thread serves as the asynchronous execution context. The queue is called workqueue and the thread is called worker.

While there are work items on the workqueue the worker executes the functions associated with the work items one after the other. When there is no work item left on the workqueue the worker becomes idle. When a new work item gets queued, the worker begins executing again.

Why cmwq?¶

In the original wq implementation, a multi threaded (MT) wq had one worker thread per CPU and a single threaded (ST) wq had one worker thread system-wide. A single MT wq needed to keep around the same number of workers as the number of CPUs. The kernel grew a lot of MT wq users over the years and with the number of CPU cores continuously rising, some systems saturated the default 32k PID space just booting up.

Although MT wq wasted a lot of resource, the level of concurrency provided was unsatisfactory. The limitation was common to both ST and MT wq albeit less severe on MT. Each wq maintained its own separate worker pool. An MT wq could provide only one execution context per CPU while an ST wq one for the whole system. Work items had to compete for those very limited execution contexts leading to various problems including proneness to deadlocks around the single execution context.

The tension between the provided level of concurrency and resource usage also forced its users to make unnecessary tradeoffs like libata choosing to use ST wq for polling PIOs and accepting an unnecessary limitation that no two polling PIOs can progress at the same time. As MT wq don’t provide much better concurrency, users which require higher level of concurrency, like async or fscache, had to implement their own thread pool.

Concurrency Managed Workqueue (cmwq) is a reimplementation of wq with focus on the following goals.

Maintain compatibility with the original workqueue API.
Use per-CPU unified worker pools shared by all wq to provide flexible level of concurrency on demand without wasting a lot of resource.
Automatically regulate worker pool and level of concurrency so that the API users don’t need to worry about such details.

The Design¶

In order to ease the asynchronous execution of functions a new abstraction, the work item, is introduced.

A work item is a simple struct that holds a pointer to the function that is to be executed asynchronously. Whenever a driver or subsystem wants a function to be executed asynchronously it has to set up a work item pointing to that function and queue that work item on a workqueue.

Special purpose threads, called worker threads, execute the functions off of the queue, one after the other. If no work is queued, the worker threads become idle. These worker threads are managed in so called worker-pools.

The cmwq design differentiates between the user-facing workqueues that subsystems and drivers queue work items on and the backend mechanism which manages worker-pools and processes the queued work items.

There are two worker-pools, one for normal work items and the other for high priority ones, for each possible CPU and some extra worker-pools to serve work items queued on unbound workqueues - the number of these backing pools is dynamic.

Subsystems and drivers can create and queue work items through special workqueue API functions as they see fit. They can influence some aspects of the way the work items are executed by setting flags on the workqueue they are putting the work item on. These flags include things like CPU locality, concurrency limits, priority and more. To get a detailed overview refer to the API description of alloc_workqueue() below.

When a work item is queued to a workqueue, the target worker-pool is determined according to the queue parameters and workqueue attributes and appended on the shared worklist of the worker-pool. For example, unless specifically overridden, a work item of a bound workqueue will be queued on the worklist of either normal or highpri worker-pool that is associated to the CPU the issuer is running on.

For any worker pool implementation, managing the concurrency level (how many execution contexts are active) is an important issue. cmwq tries to keep the concurrency at a minimal but sufficient level. Minimal to save resources and sufficient in that the system is used at its full capacity.

Each worker-pool bound to an actual CPU implements concurrency management by hooking into the scheduler. The worker-pool is notified whenever an active worker wakes up or sleeps and keeps track of the number of the currently runnable workers. Generally, work items are not expected to hog a CPU and consume many cycles. That means maintaining just enough concurrency to prevent work processing from stalling should be optimal. As long as there are one or more runnable workers on the CPU, the worker-pool doesn’t start execution of a new work, but, when the last running worker goes to sleep, it immediately schedules a new worker so that the CPU doesn’t sit idle while there are pending work items. This allows using a minimal number of workers without losing execution bandwidth.

Keeping idle workers around doesn’t cost other than the memory space for kthreads, so cmwq holds onto idle ones for a while before killing them.

For unbound workqueues, the number of backing pools is dynamic. Unbound workqueue can be assigned custom attributes using apply_workqueue_attrs() and workqueue will automatically create backing worker pools matching the attributes. The responsibility of regulating concurrency level is on the users. There is also a flag to mark a bound wq to ignore the concurrency management. Please refer to the API section for details.

Forward progress guarantee relies on that workers can be created when more execution contexts are necessary, which in turn is guaranteed through the use of rescue workers. All work items which might be used on code paths that handle memory reclaim are required to be queued on wq’s that have a rescue-worker reserved for execution under memory pressure. Else it is possible that the worker-pool deadlocks waiting for execution contexts to free up.

Application Programming Interface (API)¶

alloc_workqueue() allocates a wq. The original create_*workqueue() functions are deprecated and scheduled for removal. alloc_workqueue() takes three arguments - @name, @flags and @max_active. @name is the name of the wq and also used as the name of the rescuer thread if there is one.

A wq no longer manages execution resources but serves as a domain for forward progress guarantee, flush and work item attributes. @flags and @max_active control how work items are assigned execution resources, scheduled and executed.

`flags`¶

WQ_UNBOUND

Work items queued to an unbound wq are served by the special worker-pools which host workers which are not bound to any specific CPU. This makes the wq behave as a simple execution context provider without concurrency management. The unbound worker-pools try to start execution of work items as soon as possible. Unbound wq sacrifices locality but is useful for the following cases.

Wide fluctuation in the concurrency level requirement is expected and using bound wq may end up creating large number of mostly unused workers across different CPUs as the issuer hops through different CPUs.
Long running CPU intensive workloads which can be better managed by the system scheduler.

WQ_FREEZABLE

A freezable wq participates in the freeze phase of the system suspend operations. Work items on the wq are drained and no new work item starts execution until thawed.

WQ_MEM_RECLAIM

All wq which might be used in the memory reclaim paths MUST have this flag set. The wq is guaranteed to have at least one execution context regardless of memory pressure.

WQ_HIGHPRI

Work items of a highpri wq are queued to the highpri worker-pool of the target cpu. Highpri worker-pools are served by worker threads with elevated nice level.

Note that normal and highpri worker-pools don’t interact with each other. Each maintains its separate pool of workers and implements concurrency management among its workers.

WQ_CPU_INTENSIVE

Work items of a CPU intensive wq do not contribute to the concurrency level. In other words, runnable CPU intensive work items will not prevent other work items in the same worker-pool from starting execution. This is useful for bound work items which are expected to hog CPU cycles so that their execution is regulated by the system scheduler.

Although CPU intensive work items don’t contribute to the concurrency level, start of their executions is still regulated by the concurrency management and runnable non-CPU-intensive work items can delay execution of CPU intensive work items.

This flag is meaningless for unbound wq.

Note that the flag WQ_NON_REENTRANT no longer exists as all workqueues are now non-reentrant - any work item is guaranteed to be executed by at most one worker system-wide at any given time.

`max_active`¶

@max_active determines the maximum number of execution contexts per CPU which can be assigned to the work items of a wq. For example, with @max_active of 16, at most 16 work items of the wq can be executing at the same time per CPU.

Currently, for a bound wq, the maximum limit for @max_active is 512 and the default value used when 0 is specified is 256. For an unbound wq, the limit is higher of 512 and 4 * num_possible_cpus(). These values are chosen sufficiently high such that they are not the limiting factor while providing protection in runaway cases.

The number of active work items of a wq is usually regulated by the users of the wq, more specifically, by how many work items the users may queue at the same time. Unless there is a specific need for throttling the number of active work items, specifying ‘0’ is recommended.

Some users depend on the strict execution ordering of ST wq. The combination of @max_active of 1 and WQ_UNBOUND used to achieve this behavior. Work items on such wq were always queued to the unbound worker-pools and only one work item could be active at any given time thus achieving the same ordering property as ST wq.

In the current implementation the above configuration only guarantees ST behavior within a given NUMA node. Instead alloc_ordered_queue() should be used to achieve system-wide ST behavior.

Example Execution Scenarios¶

The following example execution scenarios try to illustrate how cmwq behave under different configurations.

Work items w0, w1, w2 are queued to a bound wq q0 on the same CPU. w0 burns CPU for 5ms then sleeps for 10ms then burns CPU for 5ms again before finishing. w1 and w2 burn CPU for 5ms then sleep for 10ms.

Ignoring all other tasks, works and processing overhead, and assuming simple FIFO scheduling, the following is one highly simplified version of possible sequences of events with the original wq.

TIME IN MSECS  EVENT
            w0 starts and burns CPU
            w0 sleeps
           w0 wakes up and burns CPU
           w0 finishes
           w1 starts and burns CPU
           w1 sleeps
           w1 wakes up and finishes
           w2 starts and burns CPU
           w2 sleeps
           w2 wakes up and finishes

And with cmwq with @max_active >= 3,

TIME IN MSECS  EVENT
            w0 starts and burns CPU
            w0 sleeps
            w1 starts and burns CPU
           w1 sleeps
           w2 starts and burns CPU
           w2 sleeps
           w0 wakes up and burns CPU
           w0 finishes
           w1 wakes up and finishes
           w2 wakes up and finishes

If @max_active == 2,

TIME IN MSECS  EVENT
            w0 starts and burns CPU
            w0 sleeps
            w1 starts and burns CPU
           w1 sleeps
           w0 wakes up and burns CPU
           w0 finishes
           w1 wakes up and finishes
           w2 starts and burns CPU
           w2 sleeps
           w2 wakes up and finishes

Now, let’s assume w1 and w2 are queued to a different wq q1 which has WQ_CPU_INTENSIVE set,

TIME IN MSECS  EVENT
            w0 starts and burns CPU
            w0 sleeps
            w1 and w2 start and burn CPU
           w1 sleeps
           w2 sleeps
           w0 wakes up and burns CPU
           w0 finishes
           w1 wakes up and finishes
           w2 wakes up and finishes

Guidelines¶

Do not forget to use WQ_MEM_RECLAIM if a wq may process work items which are used during memory reclaim. Each wq with WQ_MEM_RECLAIM set has an execution context reserved for it. If there is dependency among multiple work items used during memory reclaim, they should be queued to separate wq each with WQ_MEM_RECLAIM.
Unless strict ordering is required, there is no need to use ST wq.
Unless there is a specific need, using 0 for @max_active is recommended. In most use cases, concurrency level usually stays well under the default limit.
A wq serves as a domain for forward progress guarantee (WQ_MEM_RECLAIM, flush and work item attributes. Work items which are not involved in memory reclaim and don’t need to be flushed as a part of a group of work items, and don’t require any special attribute, can use one of the system wq. There is no difference in execution characteristics between using a dedicated wq and a system wq.
Unless work items are expected to consume a huge amount of CPU cycles, using a bound wq is usually beneficial due to the increased level of locality in wq operations and work item execution.

Debugging¶

Because the work functions are executed by generic worker threads there are a few tricks needed to shed some light on misbehaving workqueue users.

Worker threads show up in the process list as:

root      5671  0.0  0.0      0     0 ?        S    12:07   0:00 [kworker/0:1]
root      5672  0.0  0.0      0     0 ?        S    12:07   0:00 [kworker/1:2]
root      5673  0.0  0.0      0     0 ?        S    12:12   0:00 [kworker/0:0]
root      5674  0.0  0.0      0     0 ?        S    12:13   0:00 [kworker/1:0]

If kworkers are going crazy (using too much cpu), there are two types of possible problems:

Something being scheduled in rapid succession

A single work item that consumes lots of cpu cycles

The first one can be tracked using tracing:

$ echo workqueue:workqueue_queue_work > /sys/kernel/debug/tracing/set_event
$ cat /sys/kernel/debug/tracing/trace_pipe > out.txt
(wait a few secs)
^C

If something is busy looping on work queueing, it would be dominating the output and the offender can be determined with the work item function.

For the second type of problems it should be possible to just check the stack trace of the offending worker thread.

$ cat /proc/THE_OFFENDING_KWORKER/stack

The work item’s function should be trivially visible in the stack trace.

Kernel Inline Documentations Reference¶

struct workqueue_attrs¶: A struct for workqueue attributes.

Definition

struct workqueue_attrs {
  int nice;
  cpumask_var_t cpumask;
  bool no_numa;
};

Members

nice

nice level

cpumask

allowed CPUs

no_numa

disable NUMA affinity

Unlike other fields, no_numa isn’t a property of a worker_pool. It only modifies how apply_workqueue_attrs() select pools and thus doesn’t participate in pool hash calculations or equality comparisons.

Description

This can be used to change attributes of an unbound workqueue.

work_pending(work)¶: Find out whether a work item is currently pending

Parameters

work: The work item in question

delayed_work_pending(w)¶: Find out whether a delayable work item is currently pending

Parameters

w: The work item in question

struct workqueue_struct * alloc_workqueue(const char * fmt, unsigned int flags, int max_active, ...)¶: allocate a workqueue

Parameters

const char * fmt: printf format for the name of the workqueue
unsigned int flags: WQ_* flags
int max_active: max in-flight work items, 0 for default remaining args: args for fmt
...: variable arguments

Description

Allocate a workqueue with the specified parameters. For detailed information on WQ_* flags, please refer to Documentation/core-api/workqueue.rst.

Return

Pointer to the allocated workqueue on success, NULL on failure.

alloc_ordered_workqueue(fmt, flags, args…)¶: allocate an ordered workqueue

Parameters

fmt: printf format for the name of the workqueue
flags: WQ_* flags (only WQ_FREEZABLE and WQ_MEM_RECLAIM are meaningful)
args...: args for fmt

Description

Allocate an ordered workqueue. An ordered workqueue executes at most one work item at any given time in the queued order. They are implemented as unbound workqueues with max_active of one.

Return

Pointer to the allocated workqueue on success, NULL on failure.

bool queue_work(struct workqueue_struct * wq, struct work_struct * work)¶: queue work on a workqueue

Parameters

struct workqueue_struct * wq: workqueue to use
struct work_struct * work: work to queue

Description

Returns false if work was already on a queue, true otherwise.

We queue the work to the CPU on which it was submitted, but if the CPU dies it can be processed by another CPU.

bool queue_delayed_work(struct workqueue_struct * wq, struct delayed_work * dwork, unsigned long delay)¶: queue work on a workqueue after delay

Parameters

struct workqueue_struct * wq: workqueue to use
struct delayed_work * dwork: delayable work to queue
unsigned long delay: number of jiffies to wait before queueing

Description

Equivalent to queue_delayed_work_on() but tries to use the local CPU.

bool mod_delayed_work(struct workqueue_struct * wq, struct delayed_work * dwork, unsigned long delay)¶: modify delay of or queue a delayed work

Parameters

struct workqueue_struct * wq: workqueue to use
struct delayed_work * dwork: work to queue
unsigned long delay: number of jiffies to wait before queueing

Description

mod_delayed_work_on() on local CPU.

bool schedule_work_on(int cpu, struct work_struct * work)¶: put work task on a specific cpu

Parameters

int cpu: cpu to put the work task on
struct work_struct * work: job to be done

Description

This puts a job on a specific cpu

bool schedule_work(struct work_struct * work)¶: put work task in global workqueue

Parameters

struct work_struct * work: job to be done

Description

Returns false if work was already on the kernel-global workqueue and true otherwise.

This puts a job in the kernel-global workqueue if it was not already queued and leaves it in the same position on the kernel-global workqueue otherwise.

void flush_scheduled_work(void)¶: ensure that any scheduled work has run to completion.

Parameters

void: no arguments

Description

Forces execution of the kernel-global workqueue and blocks until its completion.

Think twice before calling this function! It’s very easy to get into trouble if you don’t take great care. Either of the following situations will lead to deadlock:

One of the work items currently on the workqueue needs to acquire a lock held by your code or its caller.

Your code is running in the context of a work routine.

They will be detected by lockdep when they occur, but the first might not occur very often. It depends on what work items are on the workqueue and what locks they need, which you have no control over.

In most situations flushing the entire workqueue is overkill; you merely need to know that a particular work item isn’t queued and isn’t running. In such cases you should use cancel_delayed_work_sync() or cancel_work_sync() instead.

bool schedule_delayed_work_on(int cpu, struct delayed_work * dwork, unsigned long delay)¶: queue work in global workqueue on CPU after delay

Parameters

int cpu: cpu to use
struct delayed_work * dwork: job to be done
unsigned long delay: number of jiffies to wait

Description

After waiting for a given time this puts a job in the kernel-global workqueue on the specified CPU.

bool schedule_delayed_work(struct delayed_work * dwork, unsigned long delay)¶: put work task in global workqueue after delay

Parameters

struct delayed_work * dwork: job to be done
unsigned long delay: number of jiffies to wait or 0 for immediate execution

Description

After waiting for a given time this puts a job in the kernel-global workqueue.

Concurrency Managed Workqueue (cmwq)¶

Introduction¶

Why cmwq?¶

The Design¶

Application Programming Interface (API)¶

flags¶

max_active¶

Example Execution Scenarios¶

Guidelines¶

Debugging¶

Kernel Inline Documentations Reference¶

`flags`¶

`max_active`¶