… | |
… | |
66 | CORO THREAD LIFE CYCLE |
66 | CORO THREAD LIFE CYCLE |
67 | During the long and exciting (or not) life of a coro thread, it goes |
67 | During the long and exciting (or not) life of a coro thread, it goes |
68 | through a number of states: |
68 | through a number of states: |
69 | |
69 | |
70 | 1. Creation |
70 | 1. Creation |
71 | The first thing in the life of a coro thread is it's creation - |
71 | The first thing in the life of a coro thread is its creation - |
72 | obviously. The typical way to create a thread is to call the "async |
72 | obviously. The typical way to create a thread is to call the "async |
73 | BLOCK" function: |
73 | BLOCK" function: |
74 | |
74 | |
75 | async { |
75 | async { |
76 | # thread code goes here |
76 | # thread code goes here |
… | |
… | |
85 | This creates a new coro thread and puts it into the ready queue, |
85 | This creates a new coro thread and puts it into the ready queue, |
86 | meaning it will run as soon as the CPU is free for it. |
86 | meaning it will run as soon as the CPU is free for it. |
87 | |
87 | |
88 | "async" will return a Coro object - you can store this for future |
88 | "async" will return a Coro object - you can store this for future |
89 | reference or ignore it - a thread that is running, ready to run or |
89 | reference or ignore it - a thread that is running, ready to run or |
90 | waiting for some event is alive on it's own. |
90 | waiting for some event is alive on its own. |
91 | |
91 | |
92 | Another way to create a thread is to call the "new" constructor with |
92 | Another way to create a thread is to call the "new" constructor with |
93 | a code-reference: |
93 | a code-reference: |
94 | |
94 | |
95 | new Coro sub { |
95 | new Coro sub { |
… | |
… | |
239 | implements an endless loop, the $guard will not be cleaned up. |
239 | implements an endless loop, the $guard will not be cleaned up. |
240 | However, since the thread object returned by "async" is not stored |
240 | However, since the thread object returned by "async" is not stored |
241 | anywhere, the thread is initially referenced because it is in the |
241 | anywhere, the thread is initially referenced because it is in the |
242 | ready queue, when it runs it is referenced by $Coro::current, but |
242 | ready queue, when it runs it is referenced by $Coro::current, but |
243 | when it calls "schedule", it gets "cancel"ed causing the guard |
243 | when it calls "schedule", it gets "cancel"ed causing the guard |
244 | object to be destroyed (see the next section), and printing it's |
244 | object to be destroyed (see the next section), and printing its |
245 | message. |
245 | message. |
246 | |
246 | |
247 | If this seems a bit drastic, remember that this only happens when |
247 | If this seems a bit drastic, remember that this only happens when |
248 | nothing references the thread anymore, which means there is no way |
248 | nothing references the thread anymore, which means there is no way |
249 | to further execute it, ever. The only options at this point are |
249 | to further execute it, ever. The only options at this point are |
… | |
… | |
252 | 5. Cleanup |
252 | 5. Cleanup |
253 | Threads will allocate various resources. Most but not all will be |
253 | Threads will allocate various resources. Most but not all will be |
254 | returned when a thread terminates, during clean-up. |
254 | returned when a thread terminates, during clean-up. |
255 | |
255 | |
256 | Cleanup is quite similar to throwing an uncaught exception: perl |
256 | Cleanup is quite similar to throwing an uncaught exception: perl |
257 | will work it's way up through all subroutine calls and blocks. On |
257 | will work its way up through all subroutine calls and blocks. On its |
258 | it's way, it will release all "my" variables, undo all "local"'s and |
258 | way, it will release all "my" variables, undo all "local"'s and free |
259 | free any other resources truly local to the thread. |
259 | any other resources truly local to the thread. |
260 | |
260 | |
261 | So, a common way to free resources is to keep them referenced only |
261 | So, a common way to free resources is to keep them referenced only |
262 | by my variables: |
262 | by my variables: |
263 | |
263 | |
264 | async { |
264 | async { |
… | |
… | |
284 | those code blocks): |
284 | those code blocks): |
285 | |
285 | |
286 | async { |
286 | async { |
287 | my $window = new Gtk2::Window "toplevel"; |
287 | my $window = new Gtk2::Window "toplevel"; |
288 | # The window will not be cleaned up automatically, even when $window |
288 | # The window will not be cleaned up automatically, even when $window |
289 | # gets freed, so use a guard to ensure it's destruction |
289 | # gets freed, so use a guard to ensure its destruction |
290 | # in case of an error: |
290 | # in case of an error: |
291 | my $window_guard = Guard::guard { $window->destroy }; |
291 | my $window_guard = Guard::guard { $window->destroy }; |
292 | |
292 | |
293 | # we are safe here |
293 | # we are safe here |
294 | }; |
294 | }; |
… | |
… | |
534 | # at this place, the timezone is Antarctica/South_Pole, |
534 | # at this place, the timezone is Antarctica/South_Pole, |
535 | # without disturbing the TZ of any other coro. |
535 | # without disturbing the TZ of any other coro. |
536 | }; |
536 | }; |
537 | |
537 | |
538 | This can be used to localise about any resource (locale, uid, |
538 | This can be used to localise about any resource (locale, uid, |
539 | current working directory etc.) to a block, despite the existance of |
539 | current working directory etc.) to a block, despite the existence of |
540 | other coros. |
540 | other coros. |
541 | |
541 | |
542 | Another interesting example implements time-sliced multitasking |
542 | Another interesting example implements time-sliced multitasking |
543 | using interval timers (this could obviously be optimised, but does |
543 | using interval timers (this could obviously be optimised, but does |
544 | the job): |
544 | the job): |
… | |
… | |
626 | $state->is_new |
626 | $state->is_new |
627 | Returns true iff this Coro object is "new", i.e. has never been run |
627 | Returns true iff this Coro object is "new", i.e. has never been run |
628 | yet. Those states basically consist of only the code reference to |
628 | yet. Those states basically consist of only the code reference to |
629 | call and the arguments, but consumes very little other resources. |
629 | call and the arguments, but consumes very little other resources. |
630 | New states will automatically get assigned a perl interpreter when |
630 | New states will automatically get assigned a perl interpreter when |
631 | they are transfered to. |
631 | they are transferred to. |
632 | |
632 | |
633 | $state->is_zombie |
633 | $state->is_zombie |
634 | Returns true iff the Coro object has been cancelled, i.e. it's |
634 | Returns true iff the Coro object has been cancelled, i.e. its |
635 | resources freed because they were "cancel"'ed, "terminate"'d, |
635 | resources freed because they were "cancel"'ed, "terminate"'d, |
636 | "safe_cancel"'ed or simply went out of scope. |
636 | "safe_cancel"'ed or simply went out of scope. |
637 | |
637 | |
638 | The name "zombie" stems from UNIX culture, where a process that has |
638 | The name "zombie" stems from UNIX culture, where a process that has |
639 | exited and only stores and exit status and no other resources is |
639 | exited and only stores and exit status and no other resources is |
… | |
… | |
651 | |
651 | |
652 | $is_suspended = $coro->is_suspended |
652 | $is_suspended = $coro->is_suspended |
653 | Returns true iff this Coro object has been suspended. Suspended |
653 | Returns true iff this Coro object has been suspended. Suspended |
654 | Coros will not ever be scheduled. |
654 | Coros will not ever be scheduled. |
655 | |
655 | |
656 | $coro->cancel (arg...) |
656 | $coro->cancel ($arg...) |
657 | Terminates the given Coro thread and makes it return the given |
657 | Terminate the given Coro thread and make it return the given |
658 | arguments as status (default: an empty list). Never returns if the |
658 | arguments as status (default: an empty list). Never returns if the |
659 | Coro is the current Coro. |
659 | Coro is the current Coro. |
660 | |
660 | |
661 | This is a rather brutal way to free a coro, with some limitations - |
661 | This is a rather brutal way to free a coro, with some limitations - |
662 | if the thread is inside a C callback that doesn't expect to be |
662 | if the thread is inside a C callback that doesn't expect to be |
… | |
… | |
697 | context and can block if they wish. The downside is that there is no |
697 | context and can block if they wish. The downside is that there is no |
698 | guarantee that the thread can be cancelled when you call this |
698 | guarantee that the thread can be cancelled when you call this |
699 | method, and therefore, it might fail. It is also considerably slower |
699 | method, and therefore, it might fail. It is also considerably slower |
700 | than "cancel" or "terminate". |
700 | than "cancel" or "terminate". |
701 | |
701 | |
702 | A thread is in a safe-cancellable state if it either hasn't been run |
702 | A thread is in a safe-cancellable state if it either has never been |
|
|
703 | run yet, has already been canceled/terminated or otherwise |
703 | yet, or it has no C context attached and is inside an SLF function. |
704 | destroyed, or has no C context attached and is inside an SLF |
|
|
705 | function. |
704 | |
706 | |
|
|
707 | The first two states are trivial - a thread that hasnot started or |
|
|
708 | has already finished is safe to cancel. |
|
|
709 | |
705 | The latter two basically mean that the thread isn't currently inside |
710 | The last state basically means that the thread isn't currently |
706 | a perl callback called from some C function (usually via some XS |
711 | inside a perl callback called from some C function (usually via some |
707 | modules) and isn't currently executing inside some C function itself |
712 | XS modules) and isn't currently executing inside some C function |
708 | (via Coro's XS API). |
713 | itself (via Coro's XS API). |
709 | |
714 | |
710 | This call returns true when it could cancel the thread, or croaks |
715 | This call returns true when it could cancel the thread, or croaks |
711 | with an error otherwise (i.e. it either returns true or doesn't |
716 | with an error otherwise (i.e. it either returns true or doesn't |
712 | return at all). |
717 | return at all). |
713 | |
718 | |
… | |
… | |
778 | from multiple threads, and all will be resumed and given the status |
783 | from multiple threads, and all will be resumed and given the status |
779 | return once the $coro terminates. |
784 | return once the $coro terminates. |
780 | |
785 | |
781 | $coro->on_destroy (\&cb) |
786 | $coro->on_destroy (\&cb) |
782 | Registers a callback that is called when this coro thread gets |
787 | Registers a callback that is called when this coro thread gets |
783 | destroyed, that is, after it's resources have been freed but before |
788 | destroyed, that is, after its resources have been freed but before |
784 | it is joined. The callback gets passed the terminate/cancel |
789 | it is joined. The callback gets passed the terminate/cancel |
785 | arguments, if any, and *must not* die, under any circumstances. |
790 | arguments, if any, and *must not* die, under any circumstances. |
786 | |
791 | |
787 | There can be any number of "on_destroy" callbacks per coro, and |
792 | There can be any number of "on_destroy" callbacks per coro, and |
788 | there is currently no way to remove a callback once added. |
793 | there is currently no way to remove a callback once added. |
… | |
… | |
886 | $cb = rouse_cb |
891 | $cb = rouse_cb |
887 | Create and return a "rouse callback". That's a code reference that, |
892 | Create and return a "rouse callback". That's a code reference that, |
888 | when called, will remember a copy of its arguments and notify the |
893 | when called, will remember a copy of its arguments and notify the |
889 | owner coro of the callback. |
894 | owner coro of the callback. |
890 | |
895 | |
|
|
896 | Only the first invocation will store agruments and signal any waiter |
|
|
897 | - further calls will effectively be ignored, but it is ok to try. |
|
|
898 | |
891 | See the next function. |
899 | Also see the next function. |
892 | |
900 | |
893 | @args = rouse_wait [$cb] |
901 | @args = rouse_wait [$cb] |
894 | Wait for the specified rouse callback (or the last one that was |
902 | Wait for the specified rouse callback to be invoked (or if the |
895 | created in this coro). |
903 | argument is missing, use the most recently created callback in the |
|
|
904 | current coro). |
896 | |
905 | |
897 | As soon as the callback is invoked (or when the callback was invoked |
906 | As soon as the callback is invoked (or when the callback was invoked |
898 | before "rouse_wait"), it will return the arguments originally passed |
907 | before "rouse_wait"), it will return the arguments originally passed |
899 | to the rouse callback. In scalar context, that means you get the |
908 | to the rouse callback. In scalar context, that means you get the |
900 | *last* argument, just as if "rouse_wait" had a "return ($a1, $a2, |
909 | *last* argument, just as if "rouse_wait" had a "return ($a1, $a2, |
901 | $a3...)" statement at the end. |
910 | $a3...)" statement at the end. |
902 | |
911 | |
|
|
912 | You are only allowed to wait once for a given rouse callback. |
|
|
913 | |
903 | See the section HOW TO WAIT FOR A CALLBACK for an actual usage |
914 | See the section HOW TO WAIT FOR A CALLBACK for an actual usage |
904 | example. |
915 | example. |
|
|
916 | |
|
|
917 | As of Coro 6.57, you can reliably wait for a rouse callback in a |
|
|
918 | different thread than from where it was created. |
905 | |
919 | |
906 | HOW TO WAIT FOR A CALLBACK |
920 | HOW TO WAIT FOR A CALLBACK |
907 | It is very common for a coro to wait for some callback to be called. |
921 | It is very common for a coro to wait for some callback to be called. |
908 | This occurs naturally when you use coro in an otherwise event-based |
922 | This occurs naturally when you use coro in an otherwise event-based |
909 | program, or when you use event-based libraries. |
923 | program, or when you use event-based libraries. |
910 | |
924 | |
911 | These typically register a callback for some event, and call that |
925 | These typically register a callback for some event, and call that |
912 | callback when the event occured. In a coro, however, you typically want |
926 | callback when the event occurred. In a coro, however, you typically want |
913 | to just wait for the event, simplyifying things. |
927 | to just wait for the event, simplyifying things. |
914 | |
928 | |
915 | For example "AnyEvent->child" registers a callback to be called when a |
929 | For example "AnyEvent->child" registers a callback to be called when a |
916 | specific child has exited: |
930 | specific child has exited: |
917 | |
931 | |
… | |
… | |
1036 | processes. What makes it so bad is that on non-windows platforms, you |
1050 | processes. What makes it so bad is that on non-windows platforms, you |
1037 | can actually take advantage of custom hardware for this purpose (as |
1051 | can actually take advantage of custom hardware for this purpose (as |
1038 | evidenced by the forks module, which gives you the (i-) threads API, |
1052 | evidenced by the forks module, which gives you the (i-) threads API, |
1039 | just much faster). |
1053 | just much faster). |
1040 | |
1054 | |
1041 | Sharing data is in the i-threads model is done by transfering data |
1055 | Sharing data is in the i-threads model is done by transferring data |
1042 | structures between threads using copying semantics, which is very slow - |
1056 | structures between threads using copying semantics, which is very slow - |
1043 | shared data simply does not exist. Benchmarks using i-threads which are |
1057 | shared data simply does not exist. Benchmarks using i-threads which are |
1044 | communication-intensive show extremely bad behaviour with i-threads (in |
1058 | communication-intensive show extremely bad behaviour with i-threads (in |
1045 | fact, so bad that Coro, which cannot take direct advantage of multiple |
1059 | fact, so bad that Coro, which cannot take direct advantage of multiple |
1046 | CPUs, is often orders of magnitude faster because it shares data using |
1060 | CPUs, is often orders of magnitude faster because it shares data using |