Checkpoint synchronization
You are encouraged to solve this task according to the task description, using any language you may know.
The checkpoint synchronization is a problem of synchronizing multiple tasks. Consider a workshop where several workers (tasks) assembly details of some mechanism. When each of them completes his work they put the details together. There is no store, so a worker who finished its part first must wait for others before starting another one. Putting details together is the checkpoint at which tasks synchronize themselves before going their paths apart.
The task
Implement checkpoint synchronization in your language.
Make sure that the solution is race condition-free. Note that a straightforward solution based on events is exposed to race condition. Let two tasks A and B need to be synchronized at a checkpoint. Each signals its event (EA and EB correspondingly), then waits for the AND-combination of the events (EA&EB) and resets its event. Consider the following scenario: A signals EA first and gets blocked waiting for EA&EB. Then B signals EB and loses the processor. Then A is released (both events are signaled) and resets EA. Now if B returns and enters waiting for EA&EB, it gets lost.
When a worker is ready it shall not continue before others finish. A typical implementation bug is when a worker is counted twice within one working cycle causing its premature completion. This happens when the quickest worker serves its cycle two times while the laziest one is lagging behind.
If you can, implement workers joining and leaving.
Ada
<lang Ada>with Ada.Calendar; use Ada.Calendar; with Ada.Numerics.Float_Random; with Ada.Text_IO; use Ada.Text_IO;
procedure Test_Checkpoint is
package FR renames Ada.Numerics.Float_Random;
No_Of_Cubicles: constant Positive := 3;
-- That many workers can work in parallel
No_Of_Workers: constant Positive := 6;
-- That many workers are potentially available
-- some will join the team when others quit the job
type Activity_Array is array(Character) of Boolean;
-- we want to know who is currently working
protected Checkpoint is
entry Deliver;
entry Join (Label : out Character; Tolerance: out Float);
entry Leave(Label : in Character);
private
Signaling : Boolean := False;
Ready_Count : Natural := 0;
Worker_Count : Natural := 0;
Unused_Label : Character := 'A';
Likelyhood_To_Quit: Float := 1.0;
Active : Activity_Array := (others => false);
entry Lodge;
end Checkpoint;
protected body Checkpoint is
entry Join (Label : out Character; Tolerance: out Float)
when not Signaling and Worker_Count < No_Of_Cubicles is
begin
Label := Unused_Label;
Active(Label):= True;
Unused_Label := Character'Succ (Unused_Label);
Worker_Count := Worker_Count + 1;
Likelyhood_To_Quit := Likelyhood_To_Quit / 2.0;
Tolerance := Likelyhood_To_Quit;
end Join;
entry Leave(Label: in Character) when not Signaling is
begin
Worker_Count := Worker_Count - 1;
Active(Label) := False;
end Leave;
entry Deliver when not Signaling is
begin
Ready_Count := Ready_Count + 1;
requeue Lodge;
end Deliver;
entry Lodge when Ready_Count = Worker_Count or Signaling is
begin
if Ready_Count = Worker_Count then
Put("---Sync Point [");
for C in Character loop
if Active(C) then
Put(C);
end if;
end loop;
Put_Line("]---");
end if;
Ready_Count := Ready_Count - 1;
Signaling := Ready_Count /= 0;
end Lodge;
end Checkpoint;
task type Worker;
task body Worker is
Dice : FR.Generator;
Label : Character;
Tolerance : Float;
Shift_End : Time := Clock + 2.0;
-- Trade unions are hard!
begin
FR.Reset (Dice);
Checkpoint.Join (Label, Tolerance);
Put_Line(Label & " joins the team");
loop
Put_Line (Label & " is working");
delay Duration (FR.Random (Dice) * 0.500);
Put_Line (Label & " is ready");
Checkpoint.Deliver;
if FR.Random(Dice) < Tolerance then
Put_Line(Label & " leaves the team");
exit;
elsif Clock >= Shift_End then
Put_Line(Label & " ends shift");
exit;
end if;
end loop;
Checkpoint.Leave(Label);
end Worker;
Set : array (1..No_Of_Workers) of Worker;
begin
null; -- Nothing to do here
end Test_Checkpoint;
</lang> Sample output:
A joins the team A is working B joins the team B is working C joins the team C is working B is ready C is ready A is ready ---Sync Point [ABC]--- A is working C is working B is working C is ready B is ready A is ready ---Sync Point [ABC]--- A is working B is working C is working B is ready C is ready A is ready ---Sync Point [ABC]--- A leaves the team D joins the team D is working C is working B is working C is ready B is ready D is ready ---Sync Point [BCD]--- D is working C is working B is working C is ready B is ready D is ready ---Sync Point [BCD]--- D is working C is working B is working D is ready B is ready C is ready ---Sync Point [BCD]--- C leaves the team E joins the team E is working D is working B leaves the team F joins the team F is working D is ready E is ready F is ready ---Sync Point [DEF]--- D is working F is working E is working D is ready F is ready E is ready ---Sync Point [DEF]--- E ends shift F ends shift D ends shift
BBC BASIC
<lang bbcbasic> INSTALL @lib$+"TIMERLIB"
nWorkers% = 3
DIM tID%(nWorkers%)
tID%(1) = FN_ontimer(10, PROCworker1, 1)
tID%(2) = FN_ontimer(11, PROCworker2, 1)
tID%(3) = FN_ontimer(12, PROCworker3, 1)
DEF PROCworker1 : PROCtask(1) : ENDPROC
DEF PROCworker2 : PROCtask(2) : ENDPROC
DEF PROCworker3 : PROCtask(3) : ENDPROC
ON ERROR PROCcleanup : REPORT : PRINT : END
ON CLOSE PROCcleanup : QUIT
REPEAT
WAIT 0
UNTIL FALSE
END
DEF PROCtask(worker%)
PRIVATE cnt%()
DIM cnt%(nWorkers%)
CASE cnt%(worker%) OF
WHEN 0:
cnt%(worker%) = RND(30)
PRINT "Worker "; worker% " starting (" ;cnt%(worker%) " ticks)"
WHEN -1:
OTHERWISE:
cnt%(worker%) -= 1
IF cnt%(worker%) = 0 THEN
PRINT "Worker "; worker% " ready and waiting"
cnt%(worker%) = -1
PROCcheckpoint
cnt%(worker%) = 0
ENDIF
ENDCASE
ENDPROC
DEF PROCcheckpoint
PRIVATE checked%, sync%
IF checked% = 0 sync% = FALSE
checked% += 1
WHILE NOT sync%
WAIT 0
IF checked% = nWorkers% THEN
sync% = TRUE
PRINT "--Sync Point--"
ENDIF
ENDWHILE
checked% -= 1
ENDPROC
DEF PROCcleanup
LOCAL I%
FOR I% = 1 TO nWorkers%
PROC_killtimer(tID%(I%))
NEXT
ENDPROC</lang>
Output:
Worker 1 starting (23 ticks) Worker 3 starting (26 ticks) Worker 2 starting (13 ticks) Worker 2 ready and waiting Worker 1 ready and waiting Worker 3 ready and waiting --Sync Point-- Worker 3 starting (2 ticks) Worker 1 starting (23 ticks) Worker 2 starting (2 ticks) Worker 3 ready and waiting Worker 2 ready and waiting Worker 1 ready and waiting --Sync Point-- Worker 3 starting (8 ticks) Worker 1 starting (28 ticks) Worker 2 starting (5 ticks)
C
Using OpenMP. Compiled with gcc -Wall -fopenmp.
<lang C>#include <stdio.h>
- include <stdlib.h>
- include <unistd.h>
- include <omp.h>
int main() {
int jobs = 41, tid;
omp_set_num_threads(5);
#pragma omp parallel shared(jobs) private(tid)
{
tid = omp_get_thread_num();
while (jobs > 0) {
/* this is the checkpoint */
#pragma omp barrier
if (!jobs) break;
printf("%d: taking job %d\n", tid, jobs--);
usleep(100000 + rand() / (double) RAND_MAX * 3000000);
printf("%d: done job\n", tid);
}
printf("[%d] leaving\n", tid);
/* this stops jobless thread from exiting early and killing workers */
#pragma omp barrier
}
return 0;
}</lang>
Clojure
With a fixed number of workers, this would be very straightforward in Clojure by using a CyclicBarrier from java.util.concurrent. So to make it interesting, this version supports workers dynamically joining and parting, and uses the new (2013) core.async library to use Go-like channels. Also, each worker passes a value to the checkpoint, so that some combine function could consume them once they're all received. <lang clojure>(ns checkpoint.core
(:gen-class)
(:require [clojure.core.async :as async :refer [go <! >! <!! >!! alts! close!]]
[clojure.string :as string]))
(defn coordinate [ctl-ch resp-ch combine]
(go
(<! (async/timeout 2000)) ;delay a bit to allow worker setup
(loop [members {}, received {}] ;maps by in-channel of out-channels & received data resp.
(let [rcvd-count (count received)
release #(doseq [outch (vals members)] (go (>! outch %)))
received (if (and (pos? rcvd-count) (= rcvd-count (count members)))
(do (-> received vals combine release) {})
received)
[v ch] (alts! (cons ctl-ch (keys members)))]
;receive a message on ctrl-ch or any member input channel
(if (= ch ctl-ch)
(let [[op inch outch] v] ;only a Checkpoint (see below) sends on ctl-ch
(condp = op
:join (do (>! resp-ch :ok)
(recur (assoc members inch outch) received))
:part (do (>! resp-ch :ok)
(close! inch) (close! outch)
(recur (dissoc members inch) (dissoc received inch)))
:exit :exit))
(if (nil? v) ;is the channel closed?
(do
(close! (get members ch))
(recur (dissoc members ch) (dissoc received ch)))
(recur members (assoc received ch v))))))))
(defprotocol ICheckpoint
(join [this]) (part [this inch outch]))
(deftype Checkpoint [ctl-ch resp-ch sync]
ICheckpoint
(join [this]
(let [inch (async/chan), outch (async/chan 1)]
(go
(>! ctl-ch [:join inch outch])
(<! resp-ch)
[inch outch])))
(part [this inch outch]
(go
(>! ctl-ch [:part inch outch]))))
(defn checkpoint [combine]
(let [ctl-ch (async/chan), resp-ch (async/chan 1)] (->Checkpoint ctl-ch resp-ch (coordinate ctl-ch resp-ch combine))))
(defn worker
([ckpt repeats] (worker ckpt repeats (fn [& args] nil)))
([ckpt repeats mon]
(go
(let [[send recv] (<! (join ckpt))]
(doseq [n (range repeats)]
(<! (async/timeout (rand-int 5000)))
(>! send n) (mon "sent" n)
(<! recv) (mon "recvd"))
(part ckpt send recv)))))
(defn -main
[& args]
(let [ckpt (checkpoint identity)
monitor (fn [id]
(fn [& args] (println (apply str "worker" id ":" (string/join " " args)))))]
(worker ckpt 10 (monitor 1))
(worker ckpt 10 (monitor 2))))
</lang>
D
<lang d>import std.stdio; import std.parallelism: taskPool, defaultPoolThreads, totalCPUs;
void buildMechanism(uint nparts) {
auto details = new uint[nparts];
foreach (i, ref detail; taskPool.parallel(details)) {
writeln("Build detail ", i);
detail = i;
}
// This could be written more concisely via std.parallelism.reduce,
// but we want to see the checkpoint explicitly.
writeln("Checkpoint reached. Assemble details ...");
uint sum = 0;
foreach (immutable detail; details)
sum += detail;
writeln("Mechanism with ", nparts, " parts finished: ", sum);
}
void main() {
defaultPoolThreads = totalCPUs + 1; // totalCPUs - 1 on default. buildMechanism(42); buildMechanism(11);
}</lang>
- Example output:
Build detail 0 Build detail 2 Build detail 6 Build detail 3 Build detail 8 Build detail 10 Build detail 4 Build detail 5 Build detail 7 Build detail 9 Build detail 1 Checkpoint reached. Assemble details ... Mechanism with 11 parts finished: 55
E
The problem as stated is somewhat unnatural in E. We would prefer to define the control flow in association with the data flow; for example, such that the workers return values that are combined at the checkpoint; the availability of that result value naturally defines when the workers should proceed with the next round.
That said, here is an implementation of the task as stated. We start by defining a 'flag set' data structure (which is hopefully also useful for other problems), which allows us to express the checkpoint algorithm straightforwardly while being protected against the possibility of a task calling deliver or leave too many times. Note also that each task gets its own reference denoting its membership in the checkpoint group; thus it can only speak for itself and not break any global invariants.
<lang e>/** A flagSet solves this problem: There are N things, each in a true or false
* state, and we want to know whether they are all true (or all false), and be * able to bulk-change all of them, and all this without allowing double- * counting -- setting a flag twice is idempotent. */
def makeFlagSet() {
# Each flag object is either in the true set or the false set.
def trues := [].asSet().diverge()
def falses := [].asSet().diverge()
return def flagSet {
/** Add a flag to the set. */
to join() {
def flag {
/** Get the value of this flag. */
to get() :boolean {
}
/** Set the value of this flag. */
to put(v :boolean) {
def [del,add] := if (v) { [falses,trues] } else { [trues,falses] }
if (del.contains(flag)) {
del.remove(flag)
add.addElement(flag)
}
}
/** Remove this flag from the set. */
to leave() :void {
trues.remove(flag)
falses.remove(flag)
}
}
falses.addElement(flag)
return flag
}
/** Are all the flags true (none false)? */
to allTrue() { return falses.size().isZero() }
/** Are all the flags false (none true)? */
to allFalse() { return trues.size().isZero() }
/** Set all the flags to the same value. */
to setAll(v :boolean) {
def [del,add] := if (v) { [falses,trues] } else { [trues,falses] }
add.addAll(del)
del.removeAll(del)
}
}
}
def makeCheckpoint() {
def [var continueSignal, var continueRes] := Ref.promise()
def readies := makeFlagSet()
/** Check whether all tasks have reached the checkpoint, and if so send the
* signal and go to the next round. */
def check() {
if (readies.allTrue()) {
readies.setAll(false)
continueRes.resolve(null) # send the continue signal
def [p, r] := Ref.promise() # prepare a new continue signal
continueSignal := p
continueRes := r
}
}
return def checkpoint {
to join() {
def &flag := readies.join()
return def membership {
to leave() {
(&flag).leave()
check <- ()
}
to deliver() {
flag := true
check <- ()
return continueSignal
}
}
}
}
}
def makeWorker(piece, checkpoint) {
def stops := timer.now() + 3000 + entropy.nextInt(2000)
var count := 0
def checkpointMember := checkpoint <- join()
def stopped
def run() {
# Pretend to do something lengthy; up to 1000 ms.
timer.whenPast(timer.now() + entropy.nextInt(1000), fn {
if (timer.now() >= stops) {
checkpointMember <- leave()
bind stopped := true
} else {
count += 1
println(`Delivering $piece#$count`)
when (checkpointMember <- deliver()) -> {
println(`Delivered $piece#$count`)
run()
}
}
})
}
run()
return stopped
}
def checkpoint := makeCheckpoint() var waits := [] for piece in 1..5 {
waits with= makeWorker(piece, checkpoint)
} interp.waitAtTop(promiseAllFulfilled(waits))</lang>
Erlang
A team of 5 workers assemble 3 items. The time it takes to assemble 1 item is 0 - 100 milliseconds. <lang Erlang> -module( checkpoint_synchronization ).
-export( [task/0] ).
task() ->
Pid = erlang:spawn( fun() -> checkpoint_loop([], []) end ),
[erlang:spawn(fun() -> random:seed(X, 1, 0), worker_loop(X, 3, Pid) end) || X <- lists:seq(1, 5)],
erlang:exit( Pid, normal ).
checkpoint_loop( Assemblings, Completes ) ->
receive
{starting, Worker} -> checkpoint_loop( [Worker | Assemblings], Completes );
{done, Worker} ->
New_assemblings = lists:delete( Worker, Assemblings ),
New_completes = checkpoint_loop_release( New_assemblings, [Worker | Completes] ),
checkpoint_loop( New_assemblings, New_completes )
end.
checkpoint_loop_release( [], Completes ) ->
[X ! all_complete || X <- Completes],
[];
checkpoint_loop_release( _Assemblings, Completes ) -> Completes.
worker_loop( _Worker, 0, _Checkpoint ) -> ok; worker_loop( Worker, N, Checkpoint ) ->
Checkpoint ! {starting, erlang:self()},
io:fwrite( "Worker ~p ~p~n", [Worker, N] ),
timer:sleep( random:uniform(100) ),
Checkpoint ! {done, erlang:self()},
receive
all_complete -> ok
end,
worker_loop( Worker, N - 1, Checkpoint ).
</lang>
- Output:
36> checkpoint_synchronization:task(). Worker 1 item 3 Worker 2 item 3 Worker 3 item 3 Worker 4 item 3 Worker 5 item 3 Worker 5 item 2 Worker 4 item 2 Worker 2 item 2 Worker 3 item 2 Worker 1 item 2 Worker 1 item 1 Worker 2 item 1 Worker 3 item 1 Worker 4 item 1 Worker 5 item 1
Go
As of February 2011, Go has checkpoint synchronization in the standard library, with a type called WaitGroup in the package sync. Code below uses this feature and completes the task with the workshop scenario, including workers joining and leaving.
Also see the Go solution(s) to concurrent computing. That is a much simpler task, and shown there are two different implementations of checkpoint synchronization. <lang go>package main
import (
"log" "math/rand" "os" "sync" "time"
)
const nMech = 5 const detailsPerMech = 4
var l = log.New(os.Stdout, "", 0)
func main() {
assemble := make(chan int) var complete sync.WaitGroup
go solicit(assemble, &complete, nMech*detailsPerMech)
for i := 1; i <= nMech; i++ {
complete.Add(detailsPerMech)
for j := 0; j < detailsPerMech; j++ {
assemble <- 0
}
// Go checkpoint feature
complete.Wait()
// checkpoint reached
l.Println("mechanism", i, "completed")
}
}
func solicit(a chan int, c *sync.WaitGroup, nDetails int) {
rand.Seed(time.Now().UnixNano())
var id int // worker id, for output
for nDetails > 0 {
// some random time to find a worker
time.Sleep(time.Duration(5e8 + rand.Int63n(5e8)))
id++
// contract to assemble a certain number of details
contract := rand.Intn(5) + 1
if contract > nDetails {
contract = nDetails
}
dword := "details"
if contract == 1 {
dword = "detail"
}
l.Println("worker", id, "contracted to assemble", contract, dword)
go worker(a, c, contract, id)
nDetails -= contract
}
}
func worker(a chan int, c *sync.WaitGroup, contract, id int) {
// some random time it takes for this worker to assemble a detail
assemblyTime := time.Duration(5e8 + rand.Int63n(5e8))
l.Println("worker", id, "enters shop")
for i := 0; i < contract; i++ {
<-a
l.Println("worker", id, "assembling")
time.Sleep(assemblyTime)
l.Println("worker", id, "completed detail")
c.Done()
}
l.Println("worker", id, "leaves shop")
}</lang> Output:
worker 1 contracted to assemble 2 details worker 1 enters shop worker 1 assembling worker 2 contracted to assemble 5 details worker 2 enters shop worker 2 assembling worker 1 completed detail worker 1 assembling worker 2 completed detail worker 2 assembling worker 3 contracted to assemble 1 detail worker 3 enters shop worker 1 completed detail worker 1 leaves shop worker 2 completed detail mechanism 1 completed worker 3 assembling worker 2 assembling ... worker 5 completed detail worker 7 completed detail worker 7 leaves shop mechanism 4 completed worker 6 assembling worker 5 assembling worker 6 completed detail worker 6 assembling worker 5 completed detail worker 5 leaves shop worker 6 completed detail worker 6 assembling worker 6 completed detail worker 6 leaves shop mechanism 5 completed
Haskell
Although not being sure if the approach might be right, this example shows several workers performing a series of tasks simultaneously and synchronizing themselves before starting the next task.
Each worker has several tasks in order to complete his work. As an example, they have to calculate big sums. A worker can be idle during one of the tasks. The tasks are arranged so that we get a list of the first task of all workers, then a list of the second task of all workers, and so on. Idle workers are skipped. The tasks are taken out of the Task data type and returned as plain values. The notion of a worker vanishes here. The definition of the worker's tasks allows us to keep track of what each worker actually performs.
The function "runTasks" gets one of those groups of tasks and executes them in parallel and gathers the result of each task. But this function doesn't return until all tasks are finished. This function doesn't know what each worker is doing.
Once the first group of tasks is finished, a function is applied to the results (in the example, the results are simply added together).
Then the second group of tasks is passed to "runTasks", and so on.
The workers can have any number of tasks, and they can contain idle phases. That way, a worker doesn't need to join at the first task, and may skip tasks. The function "groupTasks" takes care of idle states.
Caveats:
- A group of tasks must return values of the same type.
- If each group of tasks should return values a different type, you have to define groups of workers for each different task instead of defining workers with several tasks each.
- More flexibility can be achieved with the use of custom data types.
- Due to the use of parallel computation, only pure functions (without side effects) can be executed. Moreover, only a few Haskell compilers (GHC among them) support parallel computation. The program must be compiled with the -threaded and -rtsopts options enabled and run with the +RTS -N commandline option for computations to be actually performed in parallel.
- For effectful computations, you should use concurrent threads (forkIO and MVar from the module Control.Concurrent), software transactional memory (STM) or alternatives provided by other modules.
<lang Haskell>import Control.Parallel
data Task a = Idle | Make a type TaskList a = [a] type Results a = [a] type TaskGroups a = [TaskList a] type WorkerList a = [Worker a] type Worker a = [Task a]
-- run tasks in parallel and collect their results -- the function doesn't return until all tasks are done, therefore -- finished threads wait for the others to finish. runTasks :: TaskList a -> Results a runTasks [] = [] runTasks (x:[]) = x : [] runTasks (x:y:[]) = y `par` x : y : [] runTasks (x:y:ys) = y `par` x : y : runTasks ys
-- take a list of workers with different numbers of tasks and group -- them: first the first task of each worker, then the second one etc. groupTasks :: WorkerList a -> TaskGroups a groupTasks [] = [] groupTasks xs
| allWorkersIdle xs = []
| otherwise =
concatMap extractTask xs : groupTasks (map removeTask xs)
-- return a task as a plain value extractTask :: Worker a -> [a] extractTask [] = [] extractTask (Idle:_) = [] extractTask (Make a:_) = [a]
-- remove the foremost task of each worker removeTask :: Worker a -> Worker a removeTask = drop 1
-- checks whether all workers are idle in this task allWorkersIdle :: WorkerList a -> Bool allWorkersIdle = all null . map extractTask
-- the workers must calculate big sums. the first sum of each worker -- belongs to the first task, and so on. -- because of laziness, nothing is computed yet.
-- worker1 has 5 tasks to do worker1 :: Worker Integer worker1 = map Make [ sum [1..n*1000000] | n <- [1..5] ]
-- worker2 has 4 tasks to do worker2 :: Worker Integer worker2 = map Make [ sum [1..n*100000] | n <- [1..4] ]
-- worker3 has 3 tasks to do worker3 :: Worker Integer worker3 = map Make [ sum [1..n*1000000] | n <- [1..3] ]
-- worker4 has 5 tasks to do worker4 :: Worker Integer worker4 = map Make [ sum [1..n*300000] | n <- [1..5] ]
-- worker5 has 4 tasks to do, but starts at the second task. worker5 :: Worker Integer worker5 = [Idle] ++ map Make [ sum [1..n*400000] | n <- [1..4] ]
-- group the workers' tasks tasks :: TaskGroups Integer tasks = groupTasks [worker1, worker2, worker3, worker4, worker5]
-- a workshop: take a function to operate the results and a group of tasks, -- execute the tasks showing the process and process the results workshop :: (Show a, Num a, Show b, Num b) => ([a] -> b) -> a -> IO () workshop func a = mapM_ doWork $ zip [1..length a] a
where
doWork (x, y) = do
putStrLn $ "Doing task " ++ show x ++ "."
putStrLn $ "There are " ++ show (length y) ++ " workers for this task."
putStrLn "Waiting for all workers..."
print $ func $ runTasks y
putStrLn $ "Task " ++ show x ++ " done."
main = workshop sum tasks </lang>
The following version works with the concurrency model provided by the module Control.Concurrent
A workshop is an MVar that holds three values: the number of workers doing something, the number of workers ready for the next task and the total number of workers at the moment.
A worker takes a list of actions. Before executing the actions, he joins the workshop and the total number of workers is increased. Then, he reports that he has started an action and the number of active workers is increased. Next, the worker carries out an action. After that, he reports that he is ready for the next action. The number of active and ready workers is updated. Then he enters the check point loop, where he stays until all other workers have reported being ready. Then he goes into active state again and executes the next action, and so on. After the last action, he leaves the workshop and the total number of workers is decreased.
The checkPoint function keeps reading the values of the MVar and looping until there are 0 or less active workers and the number of ready workers is equal to the total number of workers. At this point, the MVar is reset. In order to avoid race conditions, if there are zero active and ready workers, the function returns immediately as to allow the worker to start an action.
The example code prints some useful messages to the screen, such as when a check point is reached and what each worker is currently doing (it also shows his thread ID).
The "shop" function forks the worker threads and returns their ID's, so the threads can be killed easily from the "main" function when the user hits a key.
The "main" function launches three workers first, and after 5 seconds it launches two workers more. It then waits for a key press and kills all threads, if they're still active.
Other than the parallel version above, this code runs in the IO Monad and makes it possible to perform IO actions such as accessing the hardware. However, all actions must have the return type IO (). If the workers must return some useful values, the MVar should be extended with the necessary fields and the workers should use those fields to store the results they produce.
Note: This code has been tested on GHC 7.6.1 and will most probably not run under other Haskell implementations due to the use of some functions from the module Control.Concurrent. It won't work if compiled with the -O2 compiler switch. Compile with the -threaded compiler switch if you want to run the threads in parallel.
<lang Haskell>import Control.Concurrent import Control.Monad -- needed for "forM", "forM_"
-- (workers working, workers done, workers total) type Workshop = MVar (Int, Int, Int) -- list of IO actions to be performed by one worker type Actions = [IO ()]
newWorkshop :: IO Workshop newWorkshop = newMVar (0, 0, 0)
-- check point: workers wait here for the other workers to -- finish, before resuming execution/restarting checkPoint :: Workshop -> IO () checkPoint w = do
(working, done, count) <- takeMVar w
-- all workers are done: reset counters and return (threads
-- resume execution or restart)
if working <= 0 && done == count
then do
putStrLn "---- Check Point"
putMVar w (0, 0, count)
-- mvar was just initialized: do nothing, just return.
-- otherwise, a race condition may arise
else if working == 0 && done == 0
then putMVar w (working, done, count)
-- workers are still working: wait for them (loop)
else do
putMVar w (working, done, count)
checkPoint w
-- join the workshop addWorker :: Workshop -> ThreadId -> IO () addWorker w i = do
(working, done, count) <- takeMVar w putStrLn $ "Worker " ++ show i ++ " has joined the group." putMVar w (working, done, count + 1)
-- leave the workshop removeWorker :: Workshop -> ThreadId -> IO () removeWorker w i = do
(working, done, count) <- takeMVar w putStrLn $ "Worker " ++ show i ++ " has left the group." putMVar w (working, done, count - 1)
-- increase the number of workers doing something. -- optionally, print a message using the thread's ID startWork :: Workshop -> ThreadId -> IO () startWork w i = do
(working, done, count) <- takeMVar w putStrLn $ "Worker " ++ show i ++ " has started." putMVar w (working + 1, done, count)
-- decrease the number of workers doing something and increase the -- number of workers done. optionally, print a message using -- the thread's ID finishWork :: Workshop -> ThreadId -> IO () finishWork w i = do
(working, done, count) <- takeMVar w putStrLn $ "Worker " ++ show i ++ " is ready." putMVar w (working - 1, done + 1, count)
-- put a worker to do his tasks. the steps are: -- 1. join the workshop "w" -- 2. report that the worker has started an action -- 3. perform one action -- 4. report that the worker is ready for the next action -- 5. wait for the other workers to finish -- 6. repeat from 2 until the worker has nothing more to do -- 7. leave the workshop worker :: Workshop -> Actions -> IO () worker w actions = do
i <- myThreadId
addWorker w i
forM_ actions $ \action -> do
startWork w i
action
finishWork w i
checkPoint w
removeWorker w i
-- launch several worker threads. their thread ID's are returned shop :: Workshop -> [Actions] -> IO [ThreadId] shop w actions = do
forM actions $ \x -> forkIO (worker w x)
main = do
-- make a workshop
w <- newWorkshop
-- the workers won't be doing anything special, just wait for n
-- regular intervals. pids gathers the ID's of the threads
-- this are the first workers joining the workshop
pids1 <- shop w
[replicate 5 $ threadDelay 1300000
,replicate 10 $ threadDelay 759191
,replicate 7 $ threadDelay 965300]
-- wait for 5 secs before the next workers join
threadDelay 5000000
-- these are other workers that join the workshop later
pids2 <- shop w
[replicate 6 $ threadDelay 380000
,replicate 4 $ threadDelay 250000]
-- wait for a key press
getChar
-- kill all worker threads before exit, if they're still running
forM_ (pids1 ++ pids2) killThread</lang>
Output:
main +RTS -N2 Worker ThreadId 30 has joined the group. Worker ThreadId 31 has joined the group. Worker ThreadId 32 has joined the group. Worker ThreadId 30 has started. Worker ThreadId 31 has started. Worker ThreadId 32 has started. Worker ThreadId 31 is ready. Worker ThreadId 32 is ready. Worker ThreadId 30 is ready. ---- Check Point Worker ThreadId 32 has started. Worker ThreadId 31 has started. Worker ThreadId 30 has started. Worker ThreadId 31 is ready. Worker ThreadId 32 is ready. Worker ThreadId 30 is ready. ---- Check Point Worker ThreadId 32 has started. Worker ThreadId 31 has started. Worker ThreadId 30 has started. Worker ThreadId 31 is ready. Worker ThreadId 32 is ready. Worker ThreadId 30 is ready. ---- Check Point Worker ThreadId 32 has started. Worker ThreadId 31 has started. Worker ThreadId 30 has started. Worker ThreadId 31 is ready. Worker ThreadId 32 is ready. Worker ThreadId 33 has joined the group. Worker ThreadId 34 has joined the group. Worker ThreadId 33 has started. Worker ThreadId 34 has started. Worker ThreadId 30 is ready. Worker ThreadId 34 is ready. Worker ThreadId 33 is ready. ---- Check Point Worker ThreadId 32 has started. Worker ThreadId 34 has started. Worker ThreadId 31 has started. Worker ThreadId 30 has started. Worker ThreadId 33 has started. Worker ThreadId 34 is ready. Worker ThreadId 33 is ready. Worker ThreadId 31 is ready. Worker ThreadId 32 is ready. Worker ThreadId 30 is ready. ---- Check Point Worker ThreadId 31 has started. Worker ThreadId 32 has started. Worker ThreadId 34 has started. Worker ThreadId 33 has started. Worker ThreadId 30 has left the group. Worker ThreadId 34 is ready. Worker ThreadId 33 is ready. Worker ThreadId 31 is ready. Worker ThreadId 32 is ready. ---- Check Point Worker ThreadId 34 has started. Worker ThreadId 33 has started. Worker ThreadId 31 has started. Worker ThreadId 32 has started. Worker ThreadId 34 is ready. Worker ThreadId 33 is ready. Worker ThreadId 31 is ready. Worker ThreadId 32 is ready. ---- Check Point Worker ThreadId 31 has started. Worker ThreadId 33 has started. Worker ThreadId 34 has left the group. Worker ThreadId 32 has left the group. Worker ThreadId 33 is ready. Worker ThreadId 31 is ready. ---- Check Point Worker ThreadId 33 has started. Worker ThreadId 31 has started. Worker ThreadId 33 is ready. Worker ThreadId 31 is ready. ---- Check Point Worker ThreadId 33 has left the group. Worker ThreadId 31 has started. Worker ThreadId 31 is ready. ---- Check Point Worker ThreadId 31 has left the group.
Icon and Unicon
The following only works in Unicon:
<lang unicon>global nWorkers, workers, cv
procedure main(A)
nWorkers := integer(A[1]) | 3 cv := condvar() every put(workers := [], worker(!nWorkers)) every wait(!workers)
end
procedure worker(n)
return thread every !3 do { # Union limits each worker to 3 pieces
write(n," is working")
delay(?3 * 1000)
write(n," is done")
countdown()
}
end
procedure countdown()
critical cv: {
if (nWorkers -:= 1) <= 0 then {
write("\t\tAll done")
nWorkers := *workers
return (unlock(cv),signal(cv, 0))
}
wait(cv)
}
end</lang>
Sample run:
->cps 1 is working 2 is working 3 is working 3 is done 1 is done 2 is done All done 2 is working 3 is working 1 is working 1 is done 3 is done 2 is done All done 2 is working 1 is working 3 is working 2 is done 1 is done 3 is done All done ->
J
The current implementations of J are all single threaded. However, the language definition offers a lot of parallelism which I imagine will eventually be supported, once performance gains significantly better than a factor of 2 on common problems become economically viable.
For example in 1 2 3 + 4 5 6, we have three addition operations which are specified to be carried out in parallel, and this kind of parallelism pervades the language definition.
Java
<lang Java>import java.util.Scanner; import java.util.Random;
public class CheckpointSync{ public static void main(String[] args){ System.out.print("Enter number of workers to use: "); Scanner in = new Scanner(System.in); Worker.nWorkers = in.nextInt(); System.out.print("Enter number of tasks to complete:"); runTasks(in.nextInt()); }
/* * Informs that workers started working on the task and * starts running threads. Prior to proceeding with next * task syncs using static Worker.checkpoint() method. */ private static void runTasks(int nTasks){ for(int i = 0; i < nTasks; i++){ System.out.println("Starting task number " + (i+1) + "."); runThreads(); Worker.checkpoint(); } }
/* * Creates a thread for each worker and runs it. */ private static void runThreads(){ for(int i = 0; i < Worker.nWorkers; i ++){ new Thread(new Worker(i+1)).start(); } }
/* * Worker inner static class. */ public static class Worker implements Runnable{ public Worker(int threadID){ this.threadID = threadID; } public void run(){ work(); }
/* * Notifies that thread started running for 100 to 1000 msec. * Once finished increments static counter 'nFinished' * that counts number of workers finished their work. */ private synchronized void work(){ try { int workTime = rgen.nextInt(900) + 100; System.out.println("Worker " + threadID + " will work for " + workTime + " msec."); Thread.sleep(workTime); //work for 'workTime' nFinished++; //increases work finished counter System.out.println("Worker " + threadID + " is ready"); } catch (InterruptedException e) { System.err.println("Error: thread execution interrupted"); e.printStackTrace(); } }
/* * Used to synchronize Worker threads using 'nFinished' static integer. * Waits (with step of 10 msec) until 'nFinished' equals to 'nWorkers'. * Once they are equal resets 'nFinished' counter. */ public static synchronized void checkpoint(){ while(nFinished != nWorkers){ try { Thread.sleep(10); } catch (InterruptedException e) { System.err.println("Error: thread execution interrupted"); e.printStackTrace(); } } nFinished = 0; }
/* inner class instance variables */ private int threadID;
/* static variables */ private static Random rgen = new Random(); private static int nFinished = 0; public static int nWorkers = 0; } }</lang> Output:
Enter number of workers to use: 5 Enter number of tasks to complete:3 Starting task number 1. Worker 1 will work for 882 msec. Worker 2 will work for 330 msec. Worker 3 will work for 618 msec. Worker 4 will work for 949 msec. Worker 5 will work for 805 msec. Worker 2 is ready Worker 3 is ready Worker 5 is ready Worker 1 is ready Worker 4 is ready Starting task number 2. Worker 1 will work for 942 msec. Worker 2 will work for 247 msec. Worker 3 will work for 545 msec. Worker 4 will work for 850 msec. Worker 5 will work for 888 msec. Worker 2 is ready Worker 3 is ready Worker 4 is ready Worker 5 is ready Worker 1 is ready Starting task number 3. Worker 2 will work for 976 msec. Worker 1 will work for 194 msec. Worker 4 will work for 532 msec. Worker 3 will work for 515 msec. Worker 5 will work for 326 msec. Worker 1 is ready Worker 5 is ready Worker 3 is ready Worker 4 is ready Worker 2 is ready
<lang java5>import java.util.Random; import java.util.concurrent.CountDownLatch;
public class Sync { static class Worker implements Runnable { private final CountDownLatch doneSignal; private int threadID;
public Worker(int id, CountDownLatch doneSignal) { this.doneSignal = doneSignal; threadID = id; }
public void run() { doWork(); doneSignal.countDown(); }
void doWork() { try { int workTime = new Random().nextInt(900) + 100; System.out.println("Worker " + threadID + " will work for " + workTime + " msec."); Thread.sleep(workTime); //work for 'workTime' System.out.println("Worker " + threadID + " is ready"); } catch (InterruptedException e) { System.err.println("Error: thread execution interrupted"); e.printStackTrace(); } } }
public static void main(String[] args) { int n = 3;//6 workers and 3 tasks for(int task = 1; task <= n; task++) { CountDownLatch latch = new CountDownLatch(n * 2); System.out.println("Starting task " + task); for(int worker = 0; worker < n * 2; worker++) { new Thread(new Worker(worker, latch)).start(); } try { latch.await();//wait for n*2 threads to signal the latch } catch (InterruptedException e) { e.printStackTrace(); } System.out.println("Task " + task + " complete"); } } }</lang> Output:
Starting task 1 Worker 0 will work for 959 msec. Worker 1 will work for 905 msec. Worker 3 will work for 622 msec. Worker 2 will work for 969 msec. Worker 4 will work for 577 msec. Worker 5 will work for 727 msec. Worker 4 is ready Worker 3 is ready Worker 5 is ready Worker 1 is ready Worker 0 is ready Worker 2 is ready Task 1 complete Starting task 2 Worker 0 will work for 305 msec. Worker 2 will work for 541 msec. Worker 4 will work for 663 msec. Worker 1 will work for 883 msec. Worker 3 will work for 324 msec. Worker 5 will work for 459 msec. Worker 0 is ready Worker 3 is ready Worker 5 is ready Worker 2 is ready Worker 4 is ready Worker 1 is ready Task 2 complete Starting task 3 Worker 0 will work for 554 msec. Worker 2 will work for 727 msec. Worker 1 will work for 203 msec. Worker 4 will work for 249 msec. Worker 3 will work for 612 msec. Worker 5 will work for 723 msec. Worker 1 is ready Worker 4 is ready Worker 0 is ready Worker 3 is ready Worker 5 is ready Worker 2 is ready Task 3 complete
Logtalk
The following example can be found in the Logtalk distribution and is used here with permission. It's based on the Erlang solution for this task. Works when using SWI-Prolog, XSB, or YAP as the backend compiler. <lang logtalk>
- - object(checkpoint).
:- threaded.
:- public(run/3).
:- mode(run(+integer,+integer,+float), one).
:- info(run/3, [
comment is 'Assemble items using a team of workers with a maximum time per item assembly.',
arguments is ['Workers'-'Number of workers', 'Items'-'Number of items to assemble', 'Time'-'Maximum time in seconds to assemble one item']
]).
:- public(run/0).
:- mode(run, one).
:- info(run/0, [
comment is 'Assemble three items using a team of five workers with a maximum of 0.1 seconds per item assembly.'
]).
:- uses(integer, [between/3]). :- uses(random, [random/3]).
run(Workers, Items, Time) :-
% start the workers
forall(
between(1, Workers, Worker),
threaded_ignore(worker(Worker, Items, Time))
),
% assemble the items
checkpoint_loop(Workers, Items).
run :-
% default values
run(5, 3, 0.100).
checkpoint_loop(_, 0) :-
!,
write('All assemblies done.'), nl.
checkpoint_loop(Workers, Item) :-
% wait for all threads to reach the checkpoint
forall(
between(1, Workers, Worker),
threaded_wait(done(Worker, Item))
),
write('Assembly of item '), write(Item), write(' done.'), nl,
% signal the workers to procede to the next assembly
NextItem is Item - 1,
forall(
between(1, Workers, Worker),
threaded_notify(next(Worker, NextItem))
),
checkpoint_loop(Workers, NextItem).
worker(_, 0, _) :-
!.
worker(Worker, Item, Time) :-
% the time necessary to assemble one item varies between 0.0 and Time seconds
random(0.0, Time, AssemblyTime), thread_sleep(AssemblyTime),
write('Worker '), write(Worker), write(' item '), write(Item), nl,
% notify checkpoint that the worker have done his/her part of this item
threaded_notify(done(Worker, Item)),
% wait for green light to move to the next item
NextItem is Item - 1,
threaded_wait(next(Worker, NextItem)),
worker(Worker, NextItem, Time).
- - end_object.
</lang> Output: <lang text> | ?- checkpoint::run. Worker 1 item 3 Worker 3 item 3 Worker 5 item 3 Worker 2 item 3 Worker 4 item 3 Assembly of item 3 done. Worker 4 item 2 Worker 1 item 2 Worker 5 item 2 Worker 3 item 2 Worker 2 item 2 Assembly of item 2 done. Worker 4 item 1 Worker 1 item 1 Worker 2 item 1 Worker 3 item 1 Worker 5 item 1 Assembly of item 1 done. All assemblies done. yes </lang>
Oforth
Checkpoint is implemented as a task. It :
- Receives n "job done" events from n tasks into a "jobs" channel.
- Then sends $done event to all tasks so they can work again.
Each task :
- Sleeps randomly between 1 and 100n milliseconds, simulating its job.
- Then sends its number to the checkpoint using "jobs" channel.
- And waits for checkpoint return on its personal channel.
<lang Oforth>func: task(n, jobs, myChannel) {
while(true) [
System.Out "TASK " << n << " : Beginning my work..." << cr
System sleep(n 100 * rand)
jobs send(n) drop
System.Out "TASK " << n << " : Finish, waiting for others..." << cr
myChannel receive drop
]
}
func: checkPoint(n, jobs, channels) { | ch |
while(true) [
#[ jobs receive drop ] times(n)
"CHECKPOINT : All jobs done, sending done to all tasks" println
channels forEach: ch [ $done ch send drop ]
]
}
func: testCheckPoint(n) { | jobs channels i |
n seq map(#[ drop Channel new]) ->channels Channel newSize(n) ->jobs
#[ checkPoint(n, jobs, channels) ] & n loop: i [ #[ task(i, jobs, channels at(i)) ] & ]
}</lang>
Perl
The perlipc man page details several approaches to interprocess communication. Here's one of my favourites: socketpair and fork. I've omitted some error-checking for brevity.
<lang perl>#!/usr/bin/perl use warnings; use strict; use v5.10;
use Socket;
my $nr_items = 3;
sub short_sleep($) {
(my $seconds) = @_; select undef, undef, undef, $seconds;
}
- This is run in a worker thread. It repeatedly waits for a character from
- the main thread, and sends a value back to the main thread. A short
- sleep introduces random timing, just to keep us honest.
sub be_worker($$) {
my ($socket, $value) = @_;
for (1 .. $nr_items) {
sysread $socket, my $dummy, 1;
short_sleep rand 0.5;
syswrite $socket, $value;
++$value;
}
exit;
}
- This function forks a worker and sends it a socket on which to talk to
- the main thread, as well as an initial value to work with. It returns
- (to the main thread) a socket on which to talk to the worker.
sub fork_worker($) {
(my $value) = @_;
socketpair my $kidsock, my $dadsock, AF_UNIX, SOCK_STREAM, PF_UNSPEC
or die "socketpair: $!";
if (fork // die "fork: $!") {
# We're the parent
close $dadsock;
return $kidsock;
}
else {
# We're the child
close $kidsock;
be_worker $dadsock, $value;
# Never returns
}
}
- Fork two workers, send them start signals, retrieve the values they send
- back, and print them
my $alpha_sock = fork_worker 'A'; my $digit_sock = fork_worker 1;
for (1 .. $nr_items) {
syswrite $_, 'x' for $alpha_sock, $digit_sock; sysread $alpha_sock, my $alpha, 1; sysread $digit_sock, my $digit, 1; say $alpha, $digit;
}
- If the main thread were planning to run for a long time after the
- workers had terminate, it would need to reap them to avoid zombies:
wait; wait;</lang>
A sample run:
msl@64Lucid:~/perl$ ./checkpoint A1 B2 C3 msl@64Lucid:~/perl$
PicoLisp
The following solution implements each worker as a coroutine. Therefore, it works only in the 64-bit version.
'checkpoints' takes a number of projects to do, and a number of workers. Each worker is started with a random number of steps to do (between 2 and 5), and is kept in a list of 'Staff' members. Whenever a worker finishes, he is removed from that list, until it is empty and the project is done.
'worker' takes a number of steps to perform. It "works" by printing each step, and returning NIL when done. <lang PicoLisp>(de checkpoints (Projects Workers)
(for P Projects
(prinl "Starting project number " P ":")
(for
(Staff
(mapcar
'((I) (worker (format I) (rand 2 5))) # Create staff of workers
(range 1 Workers) )
Staff # While still busy
(filter worker Staff) ) ) # Remove finished workers
(prinl "Project number " P " is done.") ) )
(de worker (ID Steps)
(co ID
(prinl "Worker " ID " has " Steps " steps to do")
(for N Steps
(yield ID)
(prinl "Worker " ID " step " N) )
NIL ) )</lang>
Output:
: (checkpoints 2 3) # Start two projects with 3 workers Starting project number 1: Worker 1 has 2 steps to do Worker 2 has 3 steps to do Worker 3 has 5 steps to do Worker 1 step 1 Worker 2 step 1 Worker 3 step 1 Worker 1 step 2 Worker 2 step 2 Worker 3 step 2 Worker 2 step 3 Worker 3 step 3 Worker 3 step 4 Worker 3 step 5 Project number 1 is done. Starting project number 2: Worker 1 has 4 steps to do Worker 2 has 3 steps to do Worker 3 has 2 steps to do Worker 1 step 1 Worker 2 step 1 Worker 3 step 1 Worker 1 step 2 Worker 2 step 2 Worker 3 step 2 Worker 1 step 3 Worker 2 step 3 Worker 1 step 4 Project number 2 is done.
PureBasic
PureBasic normally uses Semaphores and Mutex’s to synchronize parallel systems. This system only relies on semaphores between each thread and the controller (CheckPoint-procedure). For exchanging data a Mutex based message stack could easily be added, both synchronized according to this specific task or non-blocking if each worker could be allowed that freedom. <lang PureBasic>#MaxWorktime=8000 ; "Workday" in msec
- Structure that each thread uses
Structure MyIO
ThreadID.i Semaphore_Joining.i Semaphore_Release.i Semaphore_Deliver.i Semaphore_Leaving.i
EndStructure
- Array of used threads
Global Dim Comm.MyIO(0)
- Master loop synchronizing the threads via semaphores
Procedure CheckPoint()
Protected i, j, maxthreads=ArraySize(Comm())
Protected Worker_count, Deliver_count
Repeat
For i=1 To maxthreads
With Comm(i)
If TrySemaphore(\Semaphore_Leaving)
Worker_count-1
ElseIf TrySemaphore(\Semaphore_Deliver)
Deliver_count+1
If Deliver_count=Worker_count
PrintN("All Workers reported in, starting next task.")
Deliver_count=0
For j=1 To maxthreads
SignalSemaphore(Comm(j)\Semaphore_Release)
Next j
EndIf
ElseIf TrySemaphore(\Semaphore_Joining)
PrintN("A new Worker joined the force.")
Worker_count+1: SignalSemaphore(\Semaphore_Release)
ElseIf Worker_count=0
ProcedureReturn
EndIf
Next i
EndWith
ForEver
StartAll=0
EndProcedure
- A worker thread, all orchestrated by the Checkpoint() routine
Procedure Worker(ID)
Protected EndTime=ElapsedMilliseconds()+#MaxWorktime, n
With Comm(ID)
SignalSemaphore(\Semaphore_Joining)
Repeat
Repeat ; Use a non-blocking semaphore check to avoid dead-locking at shutdown.
If ElapsedMilliseconds()>EndTime
SignalSemaphore(\Semaphore_Leaving)
PrintN("Thread #"+Str(ID)+" is done.")
ProcedureReturn
EndIf
Delay(1)
Until TrySemaphore(\Semaphore_Release)
n=Random(1000)
PrintN("Thread #"+Str(ID)+" will work for "+Str(n)+" msec.")
Delay(n): PrintN("Thread #"+Str(ID)+" delivering")
SignalSemaphore(\Semaphore_Deliver)
ForEver
EndWith
EndProcedure
- User IO & init
If OpenConsole()
Define i, j
Repeat
Print("Enter number of workers to use [2-2000]: ")
j=Val(Input())
Until j>=2 And j<=2000
ReDim Comm(j)
For i=1 To j
With Comm(i)
\Semaphore_Release =CreateSemaphore()
\Semaphore_Joining =CreateSemaphore()
\Semaphore_Deliver =CreateSemaphore()
\Semaphore_Leaving =CreateSemaphore()
\ThreadID = CreateThread(@Worker(),i)
EndWith
Next
PrintN("Work started, "+Str(j)+" workers has been called.")
CheckPoint()
Print("Press ENTER to exit"): Input()
EndIf</lang>
Enter number of workers to use [2-2000]: 5 Work started, 5 workers has been called. A new Worker joined the force. A new Worker joined the force. A new Worker joined the force. A new Worker joined the force. A new Worker joined the force. Thread #5 will work for 908 msec. Thread #3 will work for 405 msec. Thread #1 will work for 536 msec. Thread #2 will work for 632 msec. Thread #4 will work for 202 msec. Thread #4 delivering Thread #3 delivering Thread #1 delivering Thread #2 delivering Thread #5 delivering All Workers reported in, starting next task. Thread #2 will work for 484 msec. Thread #4 will work for 836 msec. Thread #5 will work for 464 msec. Thread #3 will work for 251 msec. Thread #1 will work for 734 msec. Thread #3 delivering Thread #5 delivering Thread #2 delivering Thread #1 delivering Thread #4 delivering All Workers reported in, starting next task. Thread #3 will work for 864 msec. Thread #1 will work for 526 msec. Thread #5 will work for 145 msec. Thread #2 will work for 762 msec. Thread #4 will work for 283 msec. Thread #5 delivering Thread #4 delivering Thread #1 delivering Thread #2 delivering Thread #3 delivering All Workers reported in, starting next task. Thread #2 will work for 329 msec. Thread #4 will work for 452 msec. Thread #1 will work for 176 msec. Thread #5 will work for 702 msec. Thread #3 will work for 500 msec. Thread #1 delivering Thread #2 delivering Thread #4 delivering Thread #3 delivering Thread #5 delivering All Workers reported in, starting next task. Thread #5 will work for 681 msec. Thread #3 will work for 71 msec. Thread #2 will work for 267 msec. Thread #1 will work for 151 msec. Thread #4 will work for 252 msec. Thread #3 delivering Thread #1 delivering Thread #4 delivering Thread #2 delivering Thread #5 delivering All Workers reported in, starting next task. Thread #5 will work for 963 msec. Thread #3 will work for 378 msec. Thread #1 will work for 209 msec. Thread #4 will work for 897 msec. Thread #2 will work for 736 msec. Thread #1 delivering Thread #3 delivering Thread #2 delivering Thread #5 delivering Thread #4 delivering All Workers reported in, starting next task. Thread #2 will work for 44 msec. Thread #4 will work for 973 msec. Thread #1 will work for 700 msec. Thread #3 will work for 505 msec. Thread #5 will work for 256 msec. Thread #2 delivering Thread #5 delivering Thread #3 delivering Thread #1 delivering Thread #4 delivering All Workers reported in, starting next task. Thread #2 will work for 703 msec. Thread #4 will work for 296 msec. Thread #1 will work for 702 msec. Thread #3 will work for 99 msec. Thread #5 will work for 114 msec. Thread #3 delivering Thread #5 delivering Thread #4 delivering Thread #1 delivering Thread #2 delivering All Workers reported in, starting next task. Thread #3 will work for 97 msec. Thread #5 will work for 192 msec. Thread #2 will work for 762 msec. Thread #1 will work for 232 msec. Thread #4 will work for 484 msec. Thread #3 delivering Thread #5 delivering Thread #1 delivering Thread #4 delivering Thread #2 delivering All Workers reported in, starting next task. Thread #1 will work for 790 msec. Thread #5 will work for 602 msec. Thread #3 will work for 105 msec. Thread #2 will work for 449 msec. Thread #4 will work for 180 msec. Thread #3 delivering Thread #4 delivering Thread #2 delivering Thread #2 is done. Thread #4 is done. Thread #3 is done. Thread #5 delivering Thread #5 is done. Thread #1 delivering Thread #1 is done. Press ENTER to exit
Racket
This solution uses a double barrier to synchronize the five threads. The method can be found on page 41 of the delightful book "The Little Book of Semaphores" by Allen B. Downey. <lang racket>
- lang racket
(define t 5) ; total number of threads (define count 0) ; number of threads arrived at rendezvous (define mutex (make-semaphore 1)) ; exclusive access to count (define turnstile (make-semaphore 0)) (define turnstile2 (make-semaphore 1)) (define ch (make-channel))
(define (make-producer name start)
(λ ()
(let loop ([n start])
(sleep (* 0.01 (random 10))) ; "compute" something
;; rendezvous
(semaphore-wait mutex)
(set! count (+ count 1)) ; we have arrived
(when (= count t) ; are we the last to arrive?
(semaphore-wait turnstile2)
(semaphore-post turnstile))
(semaphore-post mutex)
; avoid deadlock problem:
(semaphore-wait turnstile)
(semaphore-post turnstile)
; critical point
(channel-put ch n) ; send result to controller
; leave properly
(semaphore-wait mutex)
(set! count (- count 1))
(when (= count 0) ; are we the last to leave?
(semaphore-wait turnstile)
(semaphore-post turnstile2))
(semaphore-post mutex)
(semaphore-wait turnstile2)
(semaphore-post turnstile2)
(loop (+ n t)))))
- start t workers
(map (λ(start) (thread (make-producer start start)))
(range 0 t))
(let loop ()
(displayln (for/list ([_ t]) (channel-get ch))) (loop))
</lang> Output: <lang racket> (1 4 2 0 3) (6 9 7 8 5) (11 10 14 12 13) (16 15 18 19 17) (24 21 20 23 22) (29 25 28 27 26) (30 33 34 32 31) (37 38 39 35 36) (44 43 41 40 42) (46 45 48 49 47) (50 53 51 54 52) (56 57 58 55 59) (60 63 62 61 64) (66 69 65 68 67) (73 70 74 71 72) (78 77 76 79 75) (82 80 81 84 83) (87 89 88 86 85) (92 93 90 91 94) (97 98 99 95 96) ... </lang>
Ruby
<lang ruby>require 'socket'
- A Workshop runs all of its workers, then collects their results. Use
- Workshop#add to add workers and Workshop#work to run them.
- This implementation forks some processes to run the workers in
- parallel. Ruby must provide Kernel#fork and 'socket' library must
- provide UNIXSocket.
- Why processes and not threads? C Ruby still has a Global VM Lock,
- where only one thread can hold the lock. One platform, OpenBSD, still
- has userspace threads, with all threads on one cpu core. Multiple
- processes will not compete for a single Global VM Lock and can run
- on multiple cpu cores.
class Workshop
# Creates a Workshop.
def initialize
@sockets = {}
end
# Adds a worker to this Workshop. Returns a worker id _wid_ for this # worker. The worker is a block that takes some _args_ and returns # some value. Workshop#work will run the block. # # This implementation forks a process for the worker. This process # will use Marshal with UNIXSocket to receive the _args_ and to send # the return value. The _wid_ is a process id. The worker also # inherits _IO_ objects, which might be a problem if the worker holds # open a pipe or socket, and the other end never reads EOF. def add child, parent = UNIXSocket.pair
wid = fork do
# I am the child.
child.close
@sockets.each_value { |sibling| sibling.close }
# Prevent that all the children print their backtraces (to a mess
# of mixed lines) when user presses Control-C.
Signal.trap("INT") { exit! }
loop do
# Wait for a command.
begin
command, args = Marshal.load(parent)
rescue EOFError
# Parent probably died.
break
end
case command
when :work
# Do work. Send result to parent.
result = yield *args
Marshal.dump(result, parent)
when :remove
break
else
fail "bad command from workshop"
end
end
end
# I am the parent. parent.close @sockets[wid] = child wid end
# Runs all of the workers, and collects the results in a Hash. Passes
# the same _args_ to each of the workers. Returns a Hash that pairs
# _wid_ => _result_, where _wid_ is the worker id and _result_ is the
# return value from the worker.
#
# This implementation runs the workers in parallel, and waits until
# _all_ of the workers finish their results. Workshop provides no way
# to start the work without waiting for the work to finish. If a
# worker dies (for example, by raising an Exception), then
# Workshop#work raises a RuntimeError.
def work(*args)
message = [:work, args]
@sockets.each_pair do |wid, child|
Marshal.dump(message, child)
end
# Checkpoint! Wait for all workers to finish.
result = {}
@sockets.each_pair do |wid, child|
begin
# This waits until the child finishes a result.
result[wid] = Marshal.load(child)
rescue EOFError
fail "Worker #{wid} died"
end
end
result
end
# Removes a worker from the Workshop, who has a worker id _wid_.
# If there is no such worker, raises ArgumentError.
#
# This implementation kills and reaps the process for the worker.
def remove(wid)
unless child = @sockets.delete(wid)
raise ArgumentError, "No worker #{wid}"
else
Marshal.dump([:remove, nil], child)
child.close
Process.wait(wid)
end
end
end
- First create a Workshop.
require 'pp' shop = Workshop.new wids = []
- Our workers must not use the same random numbers after the fork.
@fixed_rand = false def fix_rand
unless @fixed_rand; srand; @fixed_rand = true; end
end
- Start with some workers.
6.times do
wids << shop.add do |i|
# This worker slowly calculates a Fibonacci number.
fix_rand
f = proc { |n| if n < 2 then n else f[n - 1] + f[n - 2] end }
[i, f[25 + rand(10)]]
end
end
6.times do |i|
# Do one cycle of work, and print the result. pp shop.work(i)
# Remove a worker. victim = rand(wids.length) shop.remove wids[victim] wids.slice! victim
# Add another worker.
wids << shop.add do |j|
# This worker slowly calculates a number from
# the sequence 0, 1, 2, 3, 6, 11, 20, 37, 68, 125, ...
fix_rand
f = proc { |n| if n < 3 then n else f[n - 1] + f[n - 2] + f[n - 3] end }
[j, i, f[20 + rand(10)]]
end
end
- Remove all workers.
wids.each { |wid| shop.remove wid } pp shop.work(6)</lang>
Example of output:
{23187=>[0, 1346269],
17293=>[0, 1346269],
9974=>[0, 317811],
31730=>[0, 196418],
30156=>[0, 2178309],
25663=>[0, 832040]}
...
{23187=>[5, 5702887],
17293=>[5, 832040],
31730=>[5, 514229],
17459=>[5, 2, 24548655],
18683=>[5, 3, 187427],
4494=>[5, 4, 1166220]}
{}
Tcl
This implementation works by having a separate thread handle the synchronization (inter-thread message delivery already being serialized). The alternative, using a read-write mutex, is more complex and more likely to run into trouble with multi-core machines. <lang tcl>package require Tcl 8.5 package require Thread
namespace eval checkpoint {
namespace export {[a-z]*}
namespace ensemble create
variable members {}
variable waiting {}
variable event
# Back-end of join operation
proc Join {id} {
variable members variable counter if {$id ni $members} { lappend members $id } return $id
}
# Back-end of leave operation
proc Leave {id} {
variable members set idx [lsearch -exact $members $id] if {$idx > -1} { set members [lreplace $members $idx $idx] variable event if {![info exists event]} { set event [after idle ::checkpoint::Release] } } return
}
# Back-end of deliver operation
proc Deliver {id} {
variable waiting lappend waiting $id
variable event if {![info exists event]} { set event [after idle ::checkpoint::Release] } return
}
# Releasing is done as an "idle" action to prevent deadlocks
proc Release {} {
variable members variable waiting variable event unset event if {[llength $members] != [llength $waiting]} return set w $waiting set waiting {} foreach id $w { thread::send -async $id {incr ::checkpoint::Delivered} }
}
# Make a thread and attach it to the public API of the checkpoint proc makeThread Template:Script "" {
set id [thread::create thread::wait] thread::send $id { namespace eval checkpoint { namespace export {[a-z]*} namespace ensemble create
# Call to actually join the checkpoint group proc join {} { variable checkpoint thread::send $checkpoint [list \ ::checkpoint::Join [thread::id]] } # Call to actually leave the checkpoint group proc leave {} { variable checkpoint thread::send $checkpoint [list \ ::checkpoint::Leave [thread::id]] } # Call to wait for checkpoint synchronization proc deliver {} { variable checkpoint # Do this from within the [vwait] to ensure that we're already waiting after 0 [list thread::send $checkpoint [list \ ::checkpoint::Deliver [thread::id]]] vwait ::checkpoint::Delivered } } } thread::send $id [list set ::checkpoint::checkpoint [thread::id]] thread::send $id $script return $id
}
# Utility to help determine whether the checkpoint is in use
proc anyJoined {} {
variable members expr {[llength $members] > 0}
}
}</lang> Demonstration of how this works.
<lang tcl># Build the workers foreach worker {A B C D} {
dict set ids $worker [checkpoint makeThread {
proc task {name} { checkpoint join set deadline [expr {[clock seconds] + 2}] while {[clock seconds] <= $deadline} { puts "$name is working" after [expr {int(500 * rand())}] puts "$name is ready" checkpoint deliver } checkpoint leave thread::release; # Ask the thread to finish }
}]
}
- Set them all processing in the background
dict for {name id} $ids {
thread::send -async $id "task $name"
}
- Wait until all tasks are done (i.e., they have unregistered)
while 1 {
after 100 set s 1; vwait s; # Process events for 100ms
if {![checkpoint anyJoined]} {
break
}
}</lang> Output:
A is working C is working B is working D is working B is ready A is ready D is ready C is ready B is working A is working D is working C is working D is ready A is ready C is ready B is ready B is working D is working A is working C is working D is ready C is ready B is ready A is ready D is working C is working B is working A is working D is ready A is ready C is ready B is ready D is working C is working A is working B is working C is ready A is ready B is ready D is ready