2 Concurrency Managed Workqueue (cmwq)
4 September, 2010 Tejun Heo <tj@kernel.org>
5 Florian Mickler <florian@mickler.org>
12 4. Application Programming Interface (API)
13 5. Example Execution Scenarios
20 There are many cases where an asynchronous process execution context
21 is needed and the workqueue (wq) API is the most commonly used
22 mechanism for such cases.
24 When such an asynchronous execution context is needed, a work item
25 describing which function to execute is put on a queue. An
26 independent thread serves as the asynchronous execution context. The
27 queue is called workqueue and the thread is called worker.
29 While there are work items on the workqueue the worker executes the
30 functions associated with the work items one after the other. When
31 there is no work item left on the workqueue the worker becomes idle.
32 When a new work item gets queued, the worker begins executing again.
37 In the original wq implementation, a multi threaded (MT) wq had one
38 worker thread per CPU and a single threaded (ST) wq had one worker
39 thread system-wide. A single MT wq needed to keep around the same
40 number of workers as the number of CPUs. The kernel grew a lot of MT
41 wq users over the years and with the number of CPU cores continuously
42 rising, some systems saturated the default 32k PID space just booting
45 Although MT wq wasted a lot of resource, the level of concurrency
46 provided was unsatisfactory. The limitation was common to both ST and
47 MT wq albeit less severe on MT. Each wq maintained its own separate
48 worker pool. A MT wq could provide only one execution context per CPU
49 while a ST wq one for the whole system. Work items had to compete for
50 those very limited execution contexts leading to various problems
51 including proneness to deadlocks around the single execution context.
53 The tension between the provided level of concurrency and resource
54 usage also forced its users to make unnecessary tradeoffs like libata
55 choosing to use ST wq for polling PIOs and accepting an unnecessary
56 limitation that no two polling PIOs can progress at the same time. As
57 MT wq don't provide much better concurrency, users which require
58 higher level of concurrency, like async or fscache, had to implement
59 their own thread pool.
61 Concurrency Managed Workqueue (cmwq) is a reimplementation of wq with
62 focus on the following goals.
64 * Maintain compatibility with the original workqueue API.
66 * Use per-CPU unified worker pools shared by all wq to provide
67 flexible level of concurrency on demand without wasting a lot of
70 * Automatically regulate worker pool and level of concurrency so that
71 the API users don't need to worry about such details.
76 In order to ease the asynchronous execution of functions a new
77 abstraction, the work item, is introduced.
79 A work item is a simple struct that holds a pointer to the function
80 that is to be executed asynchronously. Whenever a driver or subsystem
81 wants a function to be executed asynchronously it has to set up a work
82 item pointing to that function and queue that work item on a
85 Special purpose threads, called worker threads, execute the functions
86 off of the queue, one after the other. If no work is queued, the
87 worker threads become idle. These worker threads are managed in so
90 The cmwq design differentiates between the user-facing workqueues that
91 subsystems and drivers queue work items on and the backend mechanism
92 which manages worker-pools and processes the queued work items.
94 There are two worker-pools, one for normal work items and the other
95 for high priority ones, for each possible CPU and some extra
96 worker-pools to serve work items queued on unbound workqueues - the
97 number of these backing pools is dynamic.
99 Subsystems and drivers can create and queue work items through special
100 workqueue API functions as they see fit. They can influence some
101 aspects of the way the work items are executed by setting flags on the
102 workqueue they are putting the work item on. These flags include
103 things like CPU locality, concurrency limits, priority and more. To
104 get a detailed overview refer to the API description of
105 alloc_workqueue() below.
107 When a work item is queued to a workqueue, the target worker-pool is
108 determined according to the queue parameters and workqueue attributes
109 and appended on the shared worklist of the worker-pool. For example,
110 unless specifically overridden, a work item of a bound workqueue will
111 be queued on the worklist of either normal or highpri worker-pool that
112 is associated to the CPU the issuer is running on.
114 For any worker pool implementation, managing the concurrency level
115 (how many execution contexts are active) is an important issue. cmwq
116 tries to keep the concurrency at a minimal but sufficient level.
117 Minimal to save resources and sufficient in that the system is used at
120 Each worker-pool bound to an actual CPU implements concurrency
121 management by hooking into the scheduler. The worker-pool is notified
122 whenever an active worker wakes up or sleeps and keeps track of the
123 number of the currently runnable workers. Generally, work items are
124 not expected to hog a CPU and consume many cycles. That means
125 maintaining just enough concurrency to prevent work processing from
126 stalling should be optimal. As long as there are one or more runnable
127 workers on the CPU, the worker-pool doesn't start execution of a new
128 work, but, when the last running worker goes to sleep, it immediately
129 schedules a new worker so that the CPU doesn't sit idle while there
130 are pending work items. This allows using a minimal number of workers
131 without losing execution bandwidth.
133 Keeping idle workers around doesn't cost other than the memory space
134 for kthreads, so cmwq holds onto idle ones for a while before killing
137 For unbound workqueues, the number of backing pools is dynamic.
138 Unbound workqueue can be assigned custom attributes using
139 apply_workqueue_attrs() and workqueue will automatically create
140 backing worker pools matching the attributes. The responsibility of
141 regulating concurrency level is on the users. There is also a flag to
142 mark a bound wq to ignore the concurrency management. Please refer to
143 the API section for details.
145 Forward progress guarantee relies on that workers can be created when
146 more execution contexts are necessary, which in turn is guaranteed
147 through the use of rescue workers. All work items which might be used
148 on code paths that handle memory reclaim are required to be queued on
149 wq's that have a rescue-worker reserved for execution under memory
150 pressure. Else it is possible that the worker-pool deadlocks waiting
151 for execution contexts to free up.
154 4. Application Programming Interface (API)
156 alloc_workqueue() allocates a wq. The original create_*workqueue()
157 functions are deprecated and scheduled for removal. alloc_workqueue()
158 takes three arguments - @name, @flags and @max_active. @name is the
159 name of the wq and also used as the name of the rescuer thread if
162 A wq no longer manages execution resources but serves as a domain for
163 forward progress guarantee, flush and work item attributes. @flags
164 and @max_active control how work items are assigned execution
165 resources, scheduled and executed.
171 Work items queued to an unbound wq are served by the special
172 woker-pools which host workers which are not bound to any
173 specific CPU. This makes the wq behave as a simple execution
174 context provider without concurrency management. The unbound
175 worker-pools try to start execution of work items as soon as
176 possible. Unbound wq sacrifices locality but is useful for
179 * Wide fluctuation in the concurrency level requirement is
180 expected and using bound wq may end up creating large number
181 of mostly unused workers across different CPUs as the issuer
182 hops through different CPUs.
184 * Long running CPU intensive workloads which can be better
185 managed by the system scheduler.
189 A freezable wq participates in the freeze phase of the system
190 suspend operations. Work items on the wq are drained and no
191 new work item starts execution until thawed.
195 All wq which might be used in the memory reclaim paths _MUST_
196 have this flag set. The wq is guaranteed to have at least one
197 execution context regardless of memory pressure.
201 Work items of a highpri wq are queued to the highpri
202 worker-pool of the target cpu. Highpri worker-pools are
203 served by worker threads with elevated nice level.
205 Note that normal and highpri worker-pools don't interact with
206 each other. Each maintain its separate pool of workers and
207 implements concurrency management among its workers.
211 Work items of a CPU intensive wq do not contribute to the
212 concurrency level. In other words, runnable CPU intensive
213 work items will not prevent other work items in the same
214 worker-pool from starting execution. This is useful for bound
215 work items which are expected to hog CPU cycles so that their
216 execution is regulated by the system scheduler.
218 Although CPU intensive work items don't contribute to the
219 concurrency level, start of their executions is still
220 regulated by the concurrency management and runnable
221 non-CPU-intensive work items can delay execution of CPU
222 intensive work items.
224 This flag is meaningless for unbound wq.
226 Note that the flag WQ_NON_REENTRANT no longer exists as all workqueues
227 are now non-reentrant - any work item is guaranteed to be executed by
228 at most one worker system-wide at any given time.
232 @max_active determines the maximum number of execution contexts per
233 CPU which can be assigned to the work items of a wq. For example,
234 with @max_active of 16, at most 16 work items of the wq can be
235 executing at the same time per CPU.
237 Currently, for a bound wq, the maximum limit for @max_active is 512
238 and the default value used when 0 is specified is 256. For an unbound
239 wq, the limit is higher of 512 and 4 * num_possible_cpus(). These
240 values are chosen sufficiently high such that they are not the
241 limiting factor while providing protection in runaway cases.
243 The number of active work items of a wq is usually regulated by the
244 users of the wq, more specifically, by how many work items the users
245 may queue at the same time. Unless there is a specific need for
246 throttling the number of active work items, specifying '0' is
249 Some users depend on the strict execution ordering of ST wq. The
250 combination of @max_active of 1 and WQ_UNBOUND is used to achieve this
251 behavior. Work items on such wq are always queued to the unbound
252 worker-pools and only one work item can be active at any given time thus
253 achieving the same ordering property as ST wq.
256 5. Example Execution Scenarios
258 The following example execution scenarios try to illustrate how cmwq
259 behave under different configurations.
261 Work items w0, w1, w2 are queued to a bound wq q0 on the same CPU.
262 w0 burns CPU for 5ms then sleeps for 10ms then burns CPU for 5ms
263 again before finishing. w1 and w2 burn CPU for 5ms then sleep for
266 Ignoring all other tasks, works and processing overhead, and assuming
267 simple FIFO scheduling, the following is one highly simplified version
268 of possible sequences of events with the original wq.
271 0 w0 starts and burns CPU
273 15 w0 wakes up and burns CPU
275 20 w1 starts and burns CPU
277 35 w1 wakes up and finishes
278 35 w2 starts and burns CPU
280 50 w2 wakes up and finishes
282 And with cmwq with @max_active >= 3,
285 0 w0 starts and burns CPU
287 5 w1 starts and burns CPU
289 10 w2 starts and burns CPU
291 15 w0 wakes up and burns CPU
293 20 w1 wakes up and finishes
294 25 w2 wakes up and finishes
299 0 w0 starts and burns CPU
301 5 w1 starts and burns CPU
303 15 w0 wakes up and burns CPU
305 20 w1 wakes up and finishes
306 20 w2 starts and burns CPU
308 35 w2 wakes up and finishes
310 Now, let's assume w1 and w2 are queued to a different wq q1 which has
311 WQ_CPU_INTENSIVE set,
314 0 w0 starts and burns CPU
316 5 w1 and w2 start and burn CPU
319 15 w0 wakes up and burns CPU
321 20 w1 wakes up and finishes
322 25 w2 wakes up and finishes
327 * Do not forget to use WQ_MEM_RECLAIM if a wq may process work items
328 which are used during memory reclaim. Each wq with WQ_MEM_RECLAIM
329 set has an execution context reserved for it. If there is
330 dependency among multiple work items used during memory reclaim,
331 they should be queued to separate wq each with WQ_MEM_RECLAIM.
333 * Unless strict ordering is required, there is no need to use ST wq.
335 * Unless there is a specific need, using 0 for @max_active is
336 recommended. In most use cases, concurrency level usually stays
337 well under the default limit.
339 * A wq serves as a domain for forward progress guarantee
340 (WQ_MEM_RECLAIM, flush and work item attributes. Work items which
341 are not involved in memory reclaim and don't need to be flushed as a
342 part of a group of work items, and don't require any special
343 attribute, can use one of the system wq. There is no difference in
344 execution characteristics between using a dedicated wq and a system
347 * Unless work items are expected to consume a huge amount of CPU
348 cycles, using a bound wq is usually beneficial due to the increased
349 level of locality in wq operations and work item execution.
354 Because the work functions are executed by generic worker threads
355 there are a few tricks needed to shed some light on misbehaving
358 Worker threads show up in the process list as:
360 root 5671 0.0 0.0 0 0 ? S 12:07 0:00 [kworker/0:1]
361 root 5672 0.0 0.0 0 0 ? S 12:07 0:00 [kworker/1:2]
362 root 5673 0.0 0.0 0 0 ? S 12:12 0:00 [kworker/0:0]
363 root 5674 0.0 0.0 0 0 ? S 12:13 0:00 [kworker/1:0]
365 If kworkers are going crazy (using too much cpu), there are two types
366 of possible problems:
368 1. Something beeing scheduled in rapid succession
369 2. A single work item that consumes lots of cpu cycles
371 The first one can be tracked using tracing:
373 $ echo workqueue:workqueue_queue_work > /sys/kernel/debug/tracing/set_event
374 $ cat /sys/kernel/debug/tracing/trace_pipe > out.txt
378 If something is busy looping on work queueing, it would be dominating
379 the output and the offender can be determined with the work item
382 For the second type of problems it should be possible to just check
383 the stack trace of the offending worker thread.
385 $ cat /proc/THE_OFFENDING_KWORKER/stack
387 The work item's function should be trivially visible in the stack