--- libev/ev.pod 2007/11/12 08:29:11 1.9 +++ libev/ev.pod 2007/11/23 19:35:09 1.35 @@ -41,19 +41,28 @@ argument of name C (which is always of type C) will not have this argument. -=head1 TIME AND OTHER GLOBAL FUNCTIONS +=head1 TIME REPRESENTATION Libev represents time as a single floating point number, representing the (fractional) number of seconds since the (POSIX) epoch (somewhere near the beginning of 1970, details are complicated, don't ask). This type is called C, which is what you should use too. It usually aliases -to the double type in C. +to the C type in C, and when you need to do any calculations on +it, you should treat it as such. + + +=head1 GLOBAL FUNCTIONS + +These functions can be called anytime, even before initialising the +library in any way. =over 4 =item ev_tstamp ev_time () -Returns the current time as libev would use it. +Returns the current time as libev would use it. Please note that the +C function is usually faster and also often returns the timestamp +you actually want to know. =item int ev_version_major () @@ -70,6 +79,45 @@ compatible to older versions, so a larger minor version alone is usually not a problem. +Example: make sure we haven't accidentally been linked against the wrong +version: + + assert (("libev version mismatch", + ev_version_major () == EV_VERSION_MAJOR + && ev_version_minor () >= EV_VERSION_MINOR)); + +=item unsigned int ev_supported_backends () + +Return the set of all backends (i.e. their corresponding C +value) compiled into this binary of libev (independent of their +availability on the system you are running on). See C for +a description of the set values. + +Example: make sure we have the epoll method, because yeah this is cool and +a must have and can we have a torrent of it please!!!11 + + assert (("sorry, no epoll, no sex", + ev_supported_backends () & EVBACKEND_EPOLL)); + +=item unsigned int ev_recommended_backends () + +Return the set of all backends compiled into this binary of libev and also +recommended for this platform. This set is often smaller than the one +returned by C, as for example kqueue is broken on +most BSDs and will not be autodetected unless you explicitly request it +(assuming you know what you are doing). This is the set of backends that +libev will probe for if you specify no backends explicitly. + +=item unsigned int ev_embeddable_backends () + +Returns the set of backends that are embeddable in other event loops. This +is the theoretical, all-platform, value. To find which backends +might be supported on the current system, you would need to look at +C, likewise for +recommended ones. + +See the description of C watchers for more info. + =item ev_set_allocator (void *(*cb)(void *ptr, long size)) Sets the allocation function to use (the prototype is similar to the @@ -82,6 +130,26 @@ free some memory if it cannot allocate memory, to use a special allocator, or even to sleep a while and retry until some memory is available. +Example: replace the libev allocator with one that waits a bit and then +retries: better than mine). + + static void * + persistent_realloc (void *ptr, long size) + { + for (;;) + { + void *newptr = realloc (ptr, size); + + if (newptr) + return newptr; + + sleep (60); + } + } + + ... + ev_set_allocator (persistent_realloc); + =item ev_set_syserr_cb (void (*cb)(const char *msg)); Set the callback function to call on a retryable syscall error (such @@ -92,6 +160,18 @@ requested operation, or, if the condition doesn't go away, do bad stuff (such as abort). +Example: do the same thing as libev does internally: + + static void + fatal_error (const char *msg) + { + perror (msg); + abort (); + } + + ... + ev_set_syserr_cb (fatal_error); + =back =head1 FUNCTIONS CONTROLLING THE EVENT LOOP @@ -101,7 +181,7 @@ events, and dynamically created loops which do not. If you use threads, a common model is to run the default event loop -in your main thread (or in a separate thrad) and for each thread you +in your main thread (or in a separate thread) and for each thread you create, you also create another event loop. Libev itself does no locking whatsoever, so if you mix calls to the same event loop in different threads, make sure you lock (this is usually a bad idea, though, even if @@ -114,24 +194,24 @@ This will initialise the default event loop if it hasn't been initialised yet and return it. If the default loop could not be initialised, returns false. If it already was initialised it simply returns it (and ignores the -flags). +flags. If that is troubling you, check C afterwards). If you don't know what event loop to use, use the one returned from this function. The flags argument can be used to specify special behaviour or specific -backends to use, and is usually specified as 0 (or EVFLAG_AUTO). +backends to use, and is usually specified as C<0> (or C). -It supports the following flags: +The following flags are supported: =over 4 -=item EVFLAG_AUTO +=item C The default flags value. Use this if you have no clue (it's the right thing, believe me). -=item EVFLAG_NOENV +=item C If this flag bit is ored into the flag value (or the program runs setuid or setgid) then libev will I look at the environment variable @@ -140,24 +220,95 @@ useful to try out specific backends to test their performance, or to work around bugs. -=item EVMETHOD_SELECT (portable select backend) +=item C (value 1, portable select backend) -=item EVMETHOD_POLL (poll backend, available everywhere except on windows) +This is your standard select(2) backend. Not I standard, as +libev tries to roll its own fd_set with no limits on the number of fds, +but if that fails, expect a fairly low limit on the number of fds when +using this backend. It doesn't scale too well (O(highest_fd)), but its usually +the fastest backend for a low number of fds. -=item EVMETHOD_EPOLL (linux only) +=item C (value 2, poll backend, available everywhere except on windows) -=item EVMETHOD_KQUEUE (some bsds only) +And this is your standard poll(2) backend. It's more complicated than +select, but handles sparse fds better and has no artificial limit on the +number of fds you can use (except it will slow down considerably with a +lot of inactive fds). It scales similarly to select, i.e. O(total_fds). -=item EVMETHOD_DEVPOLL (solaris 8 only) +=item C (value 4, Linux) -=item EVMETHOD_PORT (solaris 10 only) +For few fds, this backend is a bit little slower than poll and select, +but it scales phenomenally better. While poll and select usually scale like +O(total_fds) where n is the total number of fds (or the highest fd), epoll scales +either O(1) or O(active_fds). -If one or more of these are ored into the flags value, then only these -backends will be tried (in the reverse order as given here). If one are -specified, any backend will do. +While stopping and starting an I/O watcher in the same iteration will +result in some caching, there is still a syscall per such incident +(because the fd could point to a different file description now), so its +best to avoid that. Also, dup()ed file descriptors might not work very +well if you register events for both fds. + +Please note that epoll sometimes generates spurious notifications, so you +need to use non-blocking I/O or other means to avoid blocking when no data +(or space) is available. + +=item C (value 8, most BSD clones) + +Kqueue deserves special mention, as at the time of this writing, it +was broken on all BSDs except NetBSD (usually it doesn't work with +anything but sockets and pipes, except on Darwin, where of course its +completely useless). For this reason its not being "autodetected" +unless you explicitly specify it explicitly in the flags (i.e. using +C). + +It scales in the same way as the epoll backend, but the interface to the +kernel is more efficient (which says nothing about its actual speed, of +course). While starting and stopping an I/O watcher does not cause an +extra syscall as with epoll, it still adds up to four event changes per +incident, so its best to avoid that. + +=item C (value 16, Solaris 8) + +This is not implemented yet (and might never be). + +=item C (value 32, Solaris 10) + +This uses the Solaris 10 port mechanism. As with everything on Solaris, +it's really slow, but it still scales very well (O(active_fds)). + +Please note that solaris ports can result in a lot of spurious +notifications, so you need to use non-blocking I/O or other means to avoid +blocking when no data (or space) is available. + +=item C + +Try all backends (even potentially broken ones that wouldn't be tried +with C). Since this is a mask, you can do stuff such as +C. =back +If one or more of these are ored into the flags value, then only these +backends will be tried (in the reverse order as given here). If none are +specified, most compiled-in backend will be tried, usually in reverse +order of their flag values :) + +The most typical usage is like this: + + if (!ev_default_loop (0)) + fatal ("could not initialise libev, bad $LIBEV_FLAGS in environment?"); + +Restrict libev to the select and poll backends, and do not allow +environment settings to be taken into account: + + ev_default_loop (EVBACKEND_POLL | EVBACKEND_SELECT | EVFLAG_NOENV); + +Use whatever libev has to offer, but make sure that kqueue is used if +available (warning, breaks stuff, best use only with your own private +event loop and only if you know the OS supports your types of fds): + + ev_default_loop (ev_recommended_backends () | EVBACKEND_KQUEUE); + =item struct ev_loop *ev_loop_new (unsigned int flags) Similar to C, but always creates a new event loop that is @@ -165,6 +316,12 @@ handle signal and child watchers, and attempts to do so will be greeted by undefined behaviour (or a failed assertion if assertions are enabled). +Example: try to create a event loop that uses epoll and nothing else. + + struct ev_loop *epoller = ev_loop_new (EVBACKEND_EPOLL | EVFLAG_NOENV); + if (!epoller) + fatal ("no epoll found here, maybe it hides under your chair"); + =item ev_default_destroy () Destroys the default loop again (frees all memory and kernel state @@ -183,9 +340,9 @@ after forking, in either the parent or child process (or both, but that again makes little sense). -You I call this function after forking if and only if you want to -use the event library in both processes. If you just fork+exec, you don't -have to call it. +You I call this function in the child process after forking if and +only if you want to use the event library in both processes. If you just +fork+exec, you don't have to call it. The function itself is quite fast and it's usually not a problem to call it just in case after a fork. To make this easy, the function will fit in @@ -193,24 +350,28 @@ pthread_atfork (0, 0, ev_default_fork); +At the moment, C and C are safe to use +without calling this function, so if you force one of those backends you +do not need to care. + =item ev_loop_fork (loop) Like C, but acts on an event loop created by C. Yes, you have to call this on every allocated event loop after fork, and how you do this is entirely your own problem. -=item unsigned int ev_method (loop) +=item unsigned int ev_backend (loop) -Returns one of the C flags indicating the event backend in +Returns one of the C flags indicating the event backend in use. =item ev_tstamp ev_now (loop) Returns the current "event loop time", which is the time the event loop -got events and started processing them. This timestamp does not change -as long as callbacks are being processed, and this is also the base time -used for relative timers. You can treat it as the timestamp of the event -occuring (or more correctly, the mainloop finding out about it). +received events and started processing them. This timestamp does not +change as long as callbacks are being processed, and this is also the base +time used for relative timers. You can treat it as the timestamp of the +event occuring (or more correctly, libev finding out about it). =item ev_loop (loop, int flags) @@ -218,8 +379,14 @@ after you initialised all your watchers and you want to start handling events. -If the flags argument is specified as 0, it will not return until either -no event watchers are active anymore or C was called. +If the flags argument is specified as C<0>, it will not return until +either no event watchers are active anymore or C was called. + +Please note that an explicit C is usually better than +relying on all watchers to be stopped when deciding when a program has +finished (especially in interactive programs), but having a program that +automatically loops as long as it has to and no longer by virtue of +relying on its watchers stopping correctly is a thing of beauty. A flags value of C will look for new events, will handle those events and any outstanding ones, but will not block your process in @@ -228,17 +395,45 @@ A flags value of C will look for new events (waiting if neccessary) and will handle those and any outstanding ones. It will block your process until at least one new event arrives, and will return after -one iteration of the loop. - -This flags value could be used to implement alternative looping -constructs, but the C and C watchers provide a better and -more generic mechanism. +one iteration of the loop. This is useful if you are waiting for some +external event in conjunction with something not expressible using other +libev watchers. However, a pair of C/C watchers is +usually a better approach for this kind of thing. + +Here are the gory details of what C does: + + * If there are no active watchers (reference count is zero), return. + - Queue prepare watchers and then call all outstanding watchers. + - If we have been forked, recreate the kernel state. + - Update the kernel state with all outstanding changes. + - Update the "event loop time". + - Calculate for how long to block. + - Block the process, waiting for any events. + - Queue all outstanding I/O (fd) events. + - Update the "event loop time" and do time jump handling. + - Queue all outstanding timers. + - Queue all outstanding periodics. + - If no events are pending now, queue all idle watchers. + - Queue all check watchers. + - Call all queued watchers in reverse order (i.e. check watchers first). + Signals and child watchers are implemented as I/O watchers, and will + be handled here by queueing them when their watcher gets executed. + - If ev_unloop has been called or EVLOOP_ONESHOT or EVLOOP_NONBLOCK + were used, return, otherwise continue with step *. + +Example: queue some jobs and then loop until no events are outsanding +anymore. + + ... queue jobs here, make sure they register event watchers as long + ... as they still have work to do (even an idle watcher will do..) + ev_loop (my_loop, 0); + ... jobs done. yeah! =item ev_unloop (loop, how) Can be used to make a call to C return early (but only after it has processed all outstanding events). The C argument must be either -C, which will make the innermost C call return, or +C, which will make the innermost C call return, or C, which will make all nested C calls return. =item ev_ref (loop) @@ -256,13 +451,26 @@ way to do this for generic recurring timers or from within third-party libraries. Just remember to I and I. +Example: create a signal watcher, but keep it from keeping C +running when nothing else is active. + + struct dv_signal exitsig; + ev_signal_init (&exitsig, sig_cb, SIGINT); + ev_signal_start (myloop, &exitsig); + evf_unref (myloop); + +Example: for some weird reason, unregister the above signal handler again. + + ev_ref (myloop); + ev_signal_stop (myloop, &exitsig); + =back =head1 ANATOMY OF A WATCHER A watcher is a structure that you create and register to record your interest in some event. For instance, if you want to wait for STDIN to -become readable, you would create an ev_io watcher for that: +become readable, you would create an C watcher for that: static void my_cb (struct ev_loop *loop, struct ev_io *w, int revents) { @@ -299,63 +507,63 @@ As long as your watcher is active (has been started but not stopped) you must not touch the values stored in it. Most specifically you must never -reinitialise it or call its set method. +reinitialise it or call its set macro. -You cna check whether an event is active by calling the C macro. To see whether an event is outstanding (but the -callback for it has not been called yet) you cna use the C macro. Each and every callback receives the event loop pointer as first, the registered watcher structure as second, and a bitset of received events as third argument. -The rceeived events usually include a single bit per event type received +The received events usually include a single bit per event type received (you can receive multiple events at the same time). The possible bit masks are: =over 4 -=item EV_READ +=item C -=item EV_WRITE +=item C -The file descriptor in the ev_io watcher has become readable and/or +The file descriptor in the C watcher has become readable and/or writable. -=item EV_TIMEOUT +=item C -The ev_timer watcher has timed out. +The C watcher has timed out. -=item EV_PERIODIC +=item C -The ev_periodic watcher has timed out. +The C watcher has timed out. -=item EV_SIGNAL +=item C -The signal specified in the ev_signal watcher has been received by a thread. +The signal specified in the C watcher has been received by a thread. -=item EV_CHILD +=item C -The pid specified in the ev_child watcher has received a status change. +The pid specified in the C watcher has received a status change. -=item EV_IDLE +=item C -The ev_idle watcher has determined that you have nothing better to do. +The C watcher has determined that you have nothing better to do. -=item EV_PREPARE +=item C -=item EV_CHECK +=item C -All ev_prepare watchers are invoked just I C starts -to gather new events, and all ev_check watchers are invoked just after +All C watchers are invoked just I C starts +to gather new events, and all C watchers are invoked just after C has gathered them, but before it invokes any callbacks for any received events. Callbacks of both watcher types can start and stop as many watchers as they want, and all of them will be taken into account -(for example, a ev_prepare watcher might start an idle watcher to keep +(for example, a C watcher might start an idle watcher to keep C from blocking). -=item EV_ERROR +=item C An unspecified error has occured, the watcher has been stopped. This might happen because the watcher could not be properly started because libev @@ -374,7 +582,7 @@ =head2 ASSOCIATING CUSTOM DATA WITH A WATCHER Each watcher has, by default, a member C that you can change -and read at any time, libev will completely ignore it. This cna be used +and read at any time, libev will completely ignore it. This can be used to associate arbitrary data with your watcher. If you need more data and don't want to allocate memory and store a pointer to it in that data member, you can also "subclass" the watcher type and provide your own @@ -406,15 +614,16 @@ This section describes each watcher in detail, but will not repeat information given in the last section. -=head2 struct ev_io - is my file descriptor readable or writable + +=head2 C - is this file descriptor readable or writable I/O watchers check whether a file descriptor is readable or writable in each iteration of the event loop (This behaviour is called level-triggering because you keep receiving events as long as the -condition persists. Remember you cna stop the watcher if you don't want to +condition persists. Remember you can stop the watcher if you don't want to act on the event and neither want to receive future events). -In general you can register as many read and/or write event watchers oer +In general you can register as many read and/or write event watchers per fd as you want (as long as you don't confuse yourself). Setting all file descriptors to non-blocking mode is also usually a good idea (but not required if you know what you are doing). @@ -422,11 +631,12 @@ You have to be careful with dup'ed file descriptors, though. Some backends (the linux epoll backend is a notable example) cannot handle dup'ed file descriptors correctly if you register interest in two or more fds pointing -to the same file/socket etc. description. +to the same underlying file/socket etc. description (that is, they share +the same underlying "file open"). If you must do this, then force the use of a known-to-be-good backend -(at the time of this writing, this includes only EVMETHOD_SELECT and -EVMETHOD_POLL). +(at the time of this writing, this includes only C and +C). =over 4 @@ -434,31 +644,65 @@ =item ev_io_set (ev_io *, int fd, int events) -Configures an ev_io watcher. The fd is the file descriptor to rceeive +Configures an C watcher. The fd is the file descriptor to rceeive events for and events is either C, C or C to receive the given events. +Please note that most of the more scalable backend mechanisms (for example +epoll and solaris ports) can result in spurious readyness notifications +for file descriptors, so you practically need to use non-blocking I/O (and +treat callback invocation as hint only), or retest separately with a safe +interface before doing I/O (XLib can do this), or force the use of either +C or C, which don't suffer from this +problem. Also note that it is quite easy to have your callback invoked +when the readyness condition is no longer valid even when employing +typical ways of handling events, so its a good idea to use non-blocking +I/O unconditionally. + =back -=head2 struct ev_timer - relative and optionally recurring timeouts +Example: call C when STDIN_FILENO has become, well +readable, but only once. Since it is likely line-buffered, you could +attempt to read a whole line in the callback: + + static void + stdin_readable_cb (struct ev_loop *loop, struct ev_io *w, int revents) + { + ev_io_stop (loop, w); + .. read from stdin here (or from w->fd) and haqndle any I/O errors + } + + ... + struct ev_loop *loop = ev_default_init (0); + struct ev_io stdin_readable; + ev_io_init (&stdin_readable, stdin_readable_cb, STDIN_FILENO, EV_READ); + ev_io_start (loop, &stdin_readable); + ev_loop (loop, 0); + + +=head2 C - relative and optionally recurring timeouts Timer watchers are simple relative timers that generate an event after a given time, and optionally repeating in regular intervals after that. The timers are based on real time, that is, if you register an event that -times out after an hour and youreset your system clock to last years +times out after an hour and you reset your system clock to last years time, it will still time out after (roughly) and hour. "Roughly" because -detecting time jumps is hard, and soem inaccuracies are unavoidable (the +detecting time jumps is hard, and some inaccuracies are unavoidable (the monotonic clock option helps a lot here). The relative timeouts are calculated relative to the C time. This is usually the right thing as this timestamp refers to the time -of the event triggering whatever timeout you are modifying/starting. If -you suspect event processing to be delayed and you *need* to base the timeout -ion the current time, use something like this to adjust for this: +of the event triggering whatever timeout you are modifying/starting. If +you suspect event processing to be delayed and you I to base the timeout +on the current time, use something like this to adjust for this: ev_timer_set (&timer, after + ev_now () - ev_time (), 0.); +The callback is guarenteed to be invoked only when its timeout has passed, +but if multiple timers become ready during the same loop iteration then +order of execution is undefined. + =over 4 =item ev_timer_init (ev_timer *, callback, ev_tstamp after, ev_tstamp repeat) @@ -473,7 +717,7 @@ The timer itself will do a best-effort at avoiding drift, that is, if you configure a timer to trigger every 10 seconds, then it will trigger at exactly 10 second intervals. If, however, your program cannot keep up with -the timer (ecause it takes longer than those 10 seconds to do stuff) the +the timer (because it takes longer than those 10 seconds to do stuff) the timer will not fire more than once per event loop iteration. =item ev_timer_again (loop) @@ -490,30 +734,65 @@ example: Imagine you have a tcp connection and you want a so-called idle timeout, that is, you want to be called when there have been, say, 60 seconds of inactivity on the socket. The easiest way to do this is to -configure an ev_timer with after=repeat=60 and calling ev_timer_again each +configure an C with after=repeat=60 and calling ev_timer_again each time you successfully read or write some data. If you go into an idle state where you do not expect data to travel on the socket, you can stop the timer, and again will automatically restart it if need be. =back -=head2 ev_periodic - to cron or not to cron it +Example: create a timer that fires after 60 seconds. + + static void + one_minute_cb (struct ev_loop *loop, struct ev_timer *w, int revents) + { + .. one minute over, w is actually stopped right here + } + + struct ev_timer mytimer; + ev_timer_init (&mytimer, one_minute_cb, 60., 0.); + ev_timer_start (loop, &mytimer); + +Example: create a timeout timer that times out after 10 seconds of +inactivity. + + static void + timeout_cb (struct ev_loop *loop, struct ev_timer *w, int revents) + { + .. ten seconds without any activity + } + + struct ev_timer mytimer; + ev_timer_init (&mytimer, timeout_cb, 0., 10.); /* note, only repeat used */ + ev_timer_again (&mytimer); /* start timer */ + ev_loop (loop, 0); + + // and in some piece of code that gets executed on any "activity": + // reset the timeout to start ticking again at 10 seconds + ev_timer_again (&mytimer); + + +=head2 C - to cron or not to cron Periodic watchers are also timers of a kind, but they are very versatile (and unfortunately a bit complex). -Unlike ev_timer's, they are not based on real time (or relative time) +Unlike C's, they are not based on real time (or relative time) but on wallclock time (absolute time). You can tell a periodic watcher to trigger "at" some specific point in time. For example, if you tell a periodic watcher to trigger in 10 seconds (by specifiying e.g. c) and then reset your system clock to the last year, then it will -take a year to trigger the event (unlike an ev_timer, which would trigger +take a year to trigger the event (unlike an C, which would trigger roughly 10 seconds later and of course not if you reset your system time again). They can also be used to implement vastly more complex timers, such as triggering an event on eahc midnight, local time. +As with timers, the callback is guarenteed to be invoked only when the +time (C) has been passed, but if multiple periodic timers become ready +during the same loop iteration then order of execution is undefined. + =over 4 =item ev_periodic_init (ev_periodic *, callback, ev_tstamp at, ev_tstamp interval, reschedule_cb) @@ -523,7 +802,6 @@ Lots of arguments, lets sort it out... There are basically three modes of operation, and we will explain them from simplest to complex: - =over 4 =item * absolute timer (interval = reschedule_cb = 0) @@ -546,11 +824,11 @@ This doesn't mean there will always be 3600 seconds in between triggers, but only that the the callback will be called when the system time shows a -full hour (UTC), or more correct, when the system time is evenly divisible +full hour (UTC), or more correctly, when the system time is evenly divisible by 3600. Another way to think about it (for the mathematically inclined) is that -ev_periodic will try to run the callback in this mode at the next possible +C will try to run the callback in this mode at the next possible time where C