| 22 |
|
* subtasks may be hosted by a small number of actual threads in a |
| 23 |
|
* ForkJoinPool, at the price of some usage limitations. |
| 24 |
|
* |
| 25 |
< |
* <p> A "main" ForkJoinTask begins execution when submitted to a |
| 26 |
< |
* {@link ForkJoinPool}. Once started, it will usually in turn start |
| 27 |
< |
* other subtasks. As indicated by the name of this class, many |
| 28 |
< |
* programs using ForkJoinTasks employ only methods {@code fork} and |
| 29 |
< |
* {@code join}, or derivatives such as {@code invokeAll}. However, |
| 30 |
< |
* this class also provides a number of other methods that can come |
| 31 |
< |
* into play in advanced usages, as well as extension mechanics that |
| 32 |
< |
* allow support of new forms of fork/join processing. |
| 25 |
> |
* <p>A "main" {@code ForkJoinTask} begins execution when submitted |
| 26 |
> |
* to a {@link ForkJoinPool}. Once started, it will usually in turn |
| 27 |
> |
* start other subtasks. As indicated by the name of this class, |
| 28 |
> |
* many programs using {@code ForkJoinTask} employ only methods |
| 29 |
> |
* {@link #fork} and {@link #join}, or derivatives such as {@link |
| 30 |
> |
* #invokeAll}. However, this class also provides a number of other |
| 31 |
> |
* methods that can come into play in advanced usages, as well as |
| 32 |
> |
* extension mechanics that allow support of new forms of fork/join |
| 33 |
> |
* processing. |
| 34 |
|
* |
| 35 |
< |
* <p>A ForkJoinTask is a lightweight form of {@link Future}. The |
| 36 |
< |
* efficiency of ForkJoinTasks stems from a set of restrictions (that |
| 37 |
< |
* are only partially statically enforceable) reflecting their |
| 38 |
< |
* intended use as computational tasks calculating pure functions or |
| 39 |
< |
* operating on purely isolated objects. The primary coordination |
| 40 |
< |
* mechanisms are {@link #fork}, that arranges asynchronous execution, |
| 41 |
< |
* and {@link #join}, that doesn't proceed until the task's result has |
| 42 |
< |
* been computed. Computations should avoid {@code synchronized} |
| 43 |
< |
* methods or blocks, and should minimize other blocking |
| 44 |
< |
* synchronization apart from joining other tasks or using |
| 45 |
< |
* synchronizers such as Phasers that are advertised to cooperate with |
| 46 |
< |
* fork/join scheduling. Tasks should also not perform blocking IO, |
| 47 |
< |
* and should ideally access variables that are completely independent |
| 48 |
< |
* of those accessed by other running tasks. Minor breaches of these |
| 49 |
< |
* restrictions, for example using shared output streams, may be |
| 50 |
< |
* tolerable in practice, but frequent use may result in poor |
| 51 |
< |
* performance, and the potential to indefinitely stall if the number |
| 52 |
< |
* of threads not waiting for IO or other external synchronization |
| 53 |
< |
* becomes exhausted. This usage restriction is in part enforced by |
| 54 |
< |
* not permitting checked exceptions such as {@code IOExceptions} |
| 55 |
< |
* to be thrown. However, computations may still encounter unchecked |
| 56 |
< |
* exceptions, that are rethrown to callers attempting join |
| 57 |
< |
* them. These exceptions may additionally include |
| 58 |
< |
* RejectedExecutionExceptions stemming from internal resource |
| 59 |
< |
* exhaustion such as failure to allocate internal task queues. |
| 35 |
> |
* <p>A {@code ForkJoinTask} is a lightweight form of {@link Future}. |
| 36 |
> |
* The efficiency of {@code ForkJoinTask}s stems from a set of |
| 37 |
> |
* restrictions (that are only partially statically enforceable) |
| 38 |
> |
* reflecting their intended use as computational tasks calculating |
| 39 |
> |
* pure functions or operating on purely isolated objects. The |
| 40 |
> |
* primary coordination mechanisms are {@link #fork}, that arranges |
| 41 |
> |
* asynchronous execution, and {@link #join}, that doesn't proceed |
| 42 |
> |
* until the task's result has been computed. Computations should |
| 43 |
> |
* avoid {@code synchronized} methods or blocks, and should minimize |
| 44 |
> |
* other blocking synchronization apart from joining other tasks or |
| 45 |
> |
* using synchronizers such as Phasers that are advertised to |
| 46 |
> |
* cooperate with fork/join scheduling. Tasks should also not perform |
| 47 |
> |
* blocking IO, and should ideally access variables that are |
| 48 |
> |
* completely independent of those accessed by other running |
| 49 |
> |
* tasks. Minor breaches of these restrictions, for example using |
| 50 |
> |
* shared output streams, may be tolerable in practice, but frequent |
| 51 |
> |
* use may result in poor performance, and the potential to |
| 52 |
> |
* indefinitely stall if the number of threads not waiting for IO or |
| 53 |
> |
* other external synchronization becomes exhausted. This usage |
| 54 |
> |
* restriction is in part enforced by not permitting checked |
| 55 |
> |
* exceptions such as {@code IOExceptions} to be thrown. However, |
| 56 |
> |
* computations may still encounter unchecked exceptions, that are |
| 57 |
> |
* rethrown to callers attempting join them. These exceptions may |
| 58 |
> |
* additionally include RejectedExecutionExceptions stemming from |
| 59 |
> |
* internal resource exhaustion such as failure to allocate internal |
| 60 |
> |
* task queues. |
| 61 |
|
* |
| 62 |
|
* <p>The primary method for awaiting completion and extracting |
| 63 |
|
* results of a task is {@link #join}, but there are several variants: |
| 76 |
|
* performs the most common form of parallel invocation: forking a set |
| 77 |
|
* of tasks and joining them all. |
| 78 |
|
* |
| 79 |
< |
* <p> The ForkJoinTask class is not usually directly subclassed. |
| 79 |
> |
* <p>The ForkJoinTask class is not usually directly subclassed. |
| 80 |
|
* Instead, you subclass one of the abstract classes that support a |
| 81 |
|
* particular style of fork/join processing, typically {@link |
| 82 |
|
* RecursiveAction} for computations that do not return results, or |
| 700 |
|
* still ensure that these minimal properties hold. In particular, |
| 701 |
|
* the cancel method itself must not throw exceptions. |
| 702 |
|
* |
| 703 |
< |
* <p> This method is designed to be invoked by <em>other</em> |
| 703 |
> |
* <p>This method is designed to be invoked by <em>other</em> |
| 704 |
|
* tasks. To terminate the current task, you can just return or |
| 705 |
|
* throw an unchecked exception from its computation method, or |
| 706 |
|
* invoke {@link #completeExceptionally}. |
