… | |
… | |
53 | The newest version of this document is also available as a html-formatted |
53 | The newest version of this document is also available as a html-formatted |
54 | web page you might find easier to navigate when reading it for the first |
54 | web page you might find easier to navigate when reading it for the first |
55 | time: L<http://cvs.schmorp.de/libev/ev.html>. |
55 | time: L<http://cvs.schmorp.de/libev/ev.html>. |
56 | |
56 | |
57 | Libev is an event loop: you register interest in certain events (such as a |
57 | Libev is an event loop: you register interest in certain events (such as a |
58 | file descriptor being readable or a timeout occuring), and it will manage |
58 | file descriptor being readable or a timeout occurring), and it will manage |
59 | these event sources and provide your program with events. |
59 | these event sources and provide your program with events. |
60 | |
60 | |
61 | To do this, it must take more or less complete control over your process |
61 | To do this, it must take more or less complete control over your process |
62 | (or thread) by executing the I<event loop> handler, and will then |
62 | (or thread) by executing the I<event loop> handler, and will then |
63 | communicate events via a callback mechanism. |
63 | communicate events via a callback mechanism. |
… | |
… | |
313 | lot of inactive fds). It scales similarly to select, i.e. O(total_fds). |
313 | lot of inactive fds). It scales similarly to select, i.e. O(total_fds). |
314 | |
314 | |
315 | =item C<EVBACKEND_EPOLL> (value 4, Linux) |
315 | =item C<EVBACKEND_EPOLL> (value 4, Linux) |
316 | |
316 | |
317 | For few fds, this backend is a bit little slower than poll and select, |
317 | For few fds, this backend is a bit little slower than poll and select, |
318 | but it scales phenomenally better. While poll and select usually scale like |
318 | but it scales phenomenally better. While poll and select usually scale |
319 | O(total_fds) where n is the total number of fds (or the highest fd), epoll scales |
319 | like O(total_fds) where n is the total number of fds (or the highest fd), |
320 | either O(1) or O(active_fds). |
320 | epoll scales either O(1) or O(active_fds). The epoll design has a number |
|
|
321 | of shortcomings, such as silently dropping events in some hard-to-detect |
|
|
322 | cases and rewiring a syscall per fd change, no fork support and bad |
|
|
323 | support for dup: |
321 | |
324 | |
322 | While stopping and starting an I/O watcher in the same iteration will |
325 | While stopping, setting and starting an I/O watcher in the same iteration |
323 | result in some caching, there is still a syscall per such incident |
326 | will result in some caching, there is still a syscall per such incident |
324 | (because the fd could point to a different file description now), so its |
327 | (because the fd could point to a different file description now), so its |
325 | best to avoid that. Also, dup()ed file descriptors might not work very |
328 | best to avoid that. Also, C<dup ()>'ed file descriptors might not work |
326 | well if you register events for both fds. |
329 | very well if you register events for both fds. |
327 | |
330 | |
328 | Please note that epoll sometimes generates spurious notifications, so you |
331 | Please note that epoll sometimes generates spurious notifications, so you |
329 | need to use non-blocking I/O or other means to avoid blocking when no data |
332 | need to use non-blocking I/O or other means to avoid blocking when no data |
330 | (or space) is available. |
333 | (or space) is available. |
331 | |
334 | |
332 | =item C<EVBACKEND_KQUEUE> (value 8, most BSD clones) |
335 | =item C<EVBACKEND_KQUEUE> (value 8, most BSD clones) |
333 | |
336 | |
334 | Kqueue deserves special mention, as at the time of this writing, it |
337 | Kqueue deserves special mention, as at the time of this writing, it |
335 | was broken on all BSDs except NetBSD (usually it doesn't work with |
338 | was broken on I<all> BSDs (usually it doesn't work with anything but |
336 | anything but sockets and pipes, except on Darwin, where of course its |
339 | sockets and pipes, except on Darwin, where of course it's completely |
|
|
340 | useless. On NetBSD, it seems to work for all the FD types I tested, so it |
337 | completely useless). For this reason its not being "autodetected" |
341 | is used by default there). For this reason it's not being "autodetected" |
338 | unless you explicitly specify it explicitly in the flags (i.e. using |
342 | unless you explicitly specify it explicitly in the flags (i.e. using |
339 | C<EVBACKEND_KQUEUE>). |
343 | C<EVBACKEND_KQUEUE>) or libev was compiled on a known-to-be-good (-enough) |
|
|
344 | system like NetBSD. |
340 | |
345 | |
341 | It scales in the same way as the epoll backend, but the interface to the |
346 | It scales in the same way as the epoll backend, but the interface to the |
342 | kernel is more efficient (which says nothing about its actual speed, of |
347 | kernel is more efficient (which says nothing about its actual speed, |
343 | course). While starting and stopping an I/O watcher does not cause an |
348 | of course). While stopping, setting and starting an I/O watcher does |
344 | extra syscall as with epoll, it still adds up to four event changes per |
349 | never cause an extra syscall as with epoll, it still adds up to two event |
345 | incident, so its best to avoid that. |
350 | changes per incident, support for C<fork ()> is very bad and it drops fds |
|
|
351 | silently in similarly hard-to-detetc cases. |
346 | |
352 | |
347 | =item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8) |
353 | =item C<EVBACKEND_DEVPOLL> (value 16, Solaris 8) |
348 | |
354 | |
349 | This is not implemented yet (and might never be). |
355 | This is not implemented yet (and might never be). |
350 | |
356 | |
351 | =item C<EVBACKEND_PORT> (value 32, Solaris 10) |
357 | =item C<EVBACKEND_PORT> (value 32, Solaris 10) |
352 | |
358 | |
353 | This uses the Solaris 10 port mechanism. As with everything on Solaris, |
359 | This uses the Solaris 10 event port mechanism. As with everything on Solaris, |
354 | it's really slow, but it still scales very well (O(active_fds)). |
360 | it's really slow, but it still scales very well (O(active_fds)). |
355 | |
361 | |
356 | Please note that solaris ports can result in a lot of spurious |
362 | Please note that solaris event ports can deliver a lot of spurious |
357 | notifications, so you need to use non-blocking I/O or other means to avoid |
363 | notifications, so you need to use non-blocking I/O or other means to avoid |
358 | blocking when no data (or space) is available. |
364 | blocking when no data (or space) is available. |
359 | |
365 | |
360 | =item C<EVBACKEND_ALL> |
366 | =item C<EVBACKEND_ALL> |
361 | |
367 | |
… | |
… | |
407 | responsibility to either stop all watchers cleanly yoursef I<before> |
413 | responsibility to either stop all watchers cleanly yoursef I<before> |
408 | calling this function, or cope with the fact afterwards (which is usually |
414 | calling this function, or cope with the fact afterwards (which is usually |
409 | the easiest thing, you can just ignore the watchers and/or C<free ()> them |
415 | the easiest thing, you can just ignore the watchers and/or C<free ()> them |
410 | for example). |
416 | for example). |
411 | |
417 | |
412 | Not that certain global state, such as signal state, will not be freed by |
418 | Note that certain global state, such as signal state, will not be freed by |
413 | this function, and related watchers (such as signal and child watchers) |
419 | this function, and related watchers (such as signal and child watchers) |
414 | would need to be stopped manually. |
420 | would need to be stopped manually. |
415 | |
421 | |
416 | In general it is not advisable to call this function except in the |
422 | In general it is not advisable to call this function except in the |
417 | rare occasion where you really need to free e.g. the signal handling |
423 | rare occasion where you really need to free e.g. the signal handling |
… | |
… | |
469 | |
475 | |
470 | Returns the current "event loop time", which is the time the event loop |
476 | Returns the current "event loop time", which is the time the event loop |
471 | received events and started processing them. This timestamp does not |
477 | received events and started processing them. This timestamp does not |
472 | change as long as callbacks are being processed, and this is also the base |
478 | change as long as callbacks are being processed, and this is also the base |
473 | time used for relative timers. You can treat it as the timestamp of the |
479 | time used for relative timers. You can treat it as the timestamp of the |
474 | event occuring (or more correctly, libev finding out about it). |
480 | event occurring (or more correctly, libev finding out about it). |
475 | |
481 | |
476 | =item ev_loop (loop, int flags) |
482 | =item ev_loop (loop, int flags) |
477 | |
483 | |
478 | Finally, this is it, the event handler. This function usually is called |
484 | Finally, this is it, the event handler. This function usually is called |
479 | after you initialised all your watchers and you want to start handling |
485 | after you initialised all your watchers and you want to start handling |
… | |
… | |
924 | such as poll (fortunately in our Xlib example, Xlib already does this on |
930 | such as poll (fortunately in our Xlib example, Xlib already does this on |
925 | its own, so its quite safe to use). |
931 | its own, so its quite safe to use). |
926 | |
932 | |
927 | =head3 The special problem of disappearing file descriptors |
933 | =head3 The special problem of disappearing file descriptors |
928 | |
934 | |
929 | Some backends (e.g kqueue, epoll) need to be told about closing a file |
935 | Some backends (e.g. kqueue, epoll) need to be told about closing a file |
930 | descriptor (either by calling C<close> explicitly or by any other means, |
936 | descriptor (either by calling C<close> explicitly or by any other means, |
931 | such as C<dup>). The reason is that you register interest in some file |
937 | such as C<dup>). The reason is that you register interest in some file |
932 | descriptor, but when it goes away, the operating system will silently drop |
938 | descriptor, but when it goes away, the operating system will silently drop |
933 | this interest. If another file descriptor with the same number then is |
939 | this interest. If another file descriptor with the same number then is |
934 | registered with libev, there is no efficient way to see that this is, in |
940 | registered with libev, there is no efficient way to see that this is, in |
… | |
… | |
942 | descriptor even if the file descriptor number itself did not change. |
948 | descriptor even if the file descriptor number itself did not change. |
943 | |
949 | |
944 | This is how one would do it normally anyway, the important point is that |
950 | This is how one would do it normally anyway, the important point is that |
945 | the libev application should not optimise around libev but should leave |
951 | the libev application should not optimise around libev but should leave |
946 | optimisations to libev. |
952 | optimisations to libev. |
|
|
953 | |
|
|
954 | =head3 The special problem of dup'ed file descriptors |
|
|
955 | |
|
|
956 | Some backends (e.g. epoll), cannot register events for file descriptors, |
|
|
957 | but only events for the underlying file descriptions. That menas when you |
|
|
958 | have C<dup ()>'ed file descriptors and register events for them, only one |
|
|
959 | file descriptor might actually receive events. |
|
|
960 | |
|
|
961 | There is no workaorund possible except not registering events |
|
|
962 | for potentially C<dup ()>'ed file descriptors or to resort to |
|
|
963 | C<EVBACKEND_SELECT> or C<EVBACKEND_POLL>. |
|
|
964 | |
|
|
965 | =head3 The special problem of fork |
|
|
966 | |
|
|
967 | Some backends (epoll, kqueue) do not support C<fork ()> at all or exhibit |
|
|
968 | useless behaviour. Libev fully supports fork, but needs to be told about |
|
|
969 | it in the child. |
|
|
970 | |
|
|
971 | To support fork in your programs, you either have to call |
|
|
972 | C<ev_default_fork ()> or C<ev_loop_fork ()> after a fork in the child, |
|
|
973 | enable C<EVFLAG_FORKCHECK>, or resort to C<EVBACKEND_SELECT> or |
|
|
974 | C<EVBACKEND_POLL>. |
947 | |
975 | |
948 | |
976 | |
949 | =head3 Watcher-Specific Functions |
977 | =head3 Watcher-Specific Functions |
950 | |
978 | |
951 | =over 4 |
979 | =over 4 |
… | |
… | |
1706 | =head2 C<ev_embed> - when one backend isn't enough... |
1734 | =head2 C<ev_embed> - when one backend isn't enough... |
1707 | |
1735 | |
1708 | This is a rather advanced watcher type that lets you embed one event loop |
1736 | This is a rather advanced watcher type that lets you embed one event loop |
1709 | into another (currently only C<ev_io> events are supported in the embedded |
1737 | into another (currently only C<ev_io> events are supported in the embedded |
1710 | loop, other types of watchers might be handled in a delayed or incorrect |
1738 | loop, other types of watchers might be handled in a delayed or incorrect |
1711 | fashion and must not be used). |
1739 | fashion and must not be used). (See portability notes, below). |
1712 | |
1740 | |
1713 | There are primarily two reasons you would want that: work around bugs and |
1741 | There are primarily two reasons you would want that: work around bugs and |
1714 | prioritise I/O. |
1742 | prioritise I/O. |
1715 | |
1743 | |
1716 | As an example for a bug workaround, the kqueue backend might only support |
1744 | As an example for a bug workaround, the kqueue backend might only support |
… | |
… | |
1771 | ev_embed_start (loop_hi, &embed); |
1799 | ev_embed_start (loop_hi, &embed); |
1772 | } |
1800 | } |
1773 | else |
1801 | else |
1774 | loop_lo = loop_hi; |
1802 | loop_lo = loop_hi; |
1775 | |
1803 | |
|
|
1804 | =head2 Portability notes |
|
|
1805 | |
|
|
1806 | Kqueue is nominally embeddable, but this is broken on all BSDs that I |
|
|
1807 | tried, in various ways. Usually the embedded event loop will simply never |
|
|
1808 | receive events, sometimes it will only trigger a few times, sometimes in a |
|
|
1809 | loop. Epoll is also nominally embeddable, but many Linux kernel versions |
|
|
1810 | will always eport the epoll fd as ready, even when no events are pending. |
|
|
1811 | |
|
|
1812 | While libev allows embedding these backends (they are contained in |
|
|
1813 | C<ev_embeddable_backends ()>), take extreme care that it will actually |
|
|
1814 | work. |
|
|
1815 | |
|
|
1816 | When in doubt, create a dynamic event loop forced to use sockets (this |
|
|
1817 | usually works) and possibly another thread and a pipe or so to report to |
|
|
1818 | your main event loop. |
|
|
1819 | |
1776 | =head3 Watcher-Specific Functions and Data Members |
1820 | =head3 Watcher-Specific Functions and Data Members |
1777 | |
1821 | |
1778 | =over 4 |
1822 | =over 4 |
1779 | |
1823 | |
1780 | =item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop) |
1824 | =item ev_embed_init (ev_embed *, callback, struct ev_loop *embedded_loop) |
… | |
… | |
1791 | |
1835 | |
1792 | Make a single, non-blocking sweep over the embedded loop. This works |
1836 | Make a single, non-blocking sweep over the embedded loop. This works |
1793 | similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most |
1837 | similarly to C<ev_loop (embedded_loop, EVLOOP_NONBLOCK)>, but in the most |
1794 | apropriate way for embedded loops. |
1838 | apropriate way for embedded loops. |
1795 | |
1839 | |
1796 | =item struct ev_loop *loop [read-only] |
1840 | =item struct ev_loop *other [read-only] |
1797 | |
1841 | |
1798 | The embedded event loop. |
1842 | The embedded event loop. |
1799 | |
1843 | |
1800 | =back |
1844 | =back |
1801 | |
1845 | |
… | |
… | |
2140 | Libev can (and often is) directly embedded into host |
2184 | Libev can (and often is) directly embedded into host |
2141 | applications. Examples of applications that embed it include the Deliantra |
2185 | applications. Examples of applications that embed it include the Deliantra |
2142 | Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe) |
2186 | Game Server, the EV perl module, the GNU Virtual Private Ethernet (gvpe) |
2143 | and rxvt-unicode. |
2187 | and rxvt-unicode. |
2144 | |
2188 | |
2145 | The goal is to enable you to just copy the neecssary files into your |
2189 | The goal is to enable you to just copy the necessary files into your |
2146 | source directory without having to change even a single line in them, so |
2190 | source directory without having to change even a single line in them, so |
2147 | you can easily upgrade by simply copying (or having a checked-out copy of |
2191 | you can easily upgrade by simply copying (or having a checked-out copy of |
2148 | libev somewhere in your source tree). |
2192 | libev somewhere in your source tree). |
2149 | |
2193 | |
2150 | =head2 FILESETS |
2194 | =head2 FILESETS |
… | |
… | |
2240 | |
2284 | |
2241 | If defined to be C<1>, libev will try to detect the availability of the |
2285 | If defined to be C<1>, libev will try to detect the availability of the |
2242 | monotonic clock option at both compiletime and runtime. Otherwise no use |
2286 | monotonic clock option at both compiletime and runtime. Otherwise no use |
2243 | of the monotonic clock option will be attempted. If you enable this, you |
2287 | of the monotonic clock option will be attempted. If you enable this, you |
2244 | usually have to link against librt or something similar. Enabling it when |
2288 | usually have to link against librt or something similar. Enabling it when |
2245 | the functionality isn't available is safe, though, althoguh you have |
2289 | the functionality isn't available is safe, though, although you have |
2246 | to make sure you link against any libraries where the C<clock_gettime> |
2290 | to make sure you link against any libraries where the C<clock_gettime> |
2247 | function is hiding in (often F<-lrt>). |
2291 | function is hiding in (often F<-lrt>). |
2248 | |
2292 | |
2249 | =item EV_USE_REALTIME |
2293 | =item EV_USE_REALTIME |
2250 | |
2294 | |
2251 | If defined to be C<1>, libev will try to detect the availability of the |
2295 | If defined to be C<1>, libev will try to detect the availability of the |
2252 | realtime clock option at compiletime (and assume its availability at |
2296 | realtime clock option at compiletime (and assume its availability at |
2253 | runtime if successful). Otherwise no use of the realtime clock option will |
2297 | runtime if successful). Otherwise no use of the realtime clock option will |
2254 | be attempted. This effectively replaces C<gettimeofday> by C<clock_get |
2298 | be attempted. This effectively replaces C<gettimeofday> by C<clock_get |
2255 | (CLOCK_REALTIME, ...)> and will not normally affect correctness. See tzhe note about libraries |
2299 | (CLOCK_REALTIME, ...)> and will not normally affect correctness. See the |
2256 | in the description of C<EV_USE_MONOTONIC>, though. |
2300 | note about libraries in the description of C<EV_USE_MONOTONIC>, though. |
2257 | |
2301 | |
2258 | =item EV_USE_SELECT |
2302 | =item EV_USE_SELECT |
2259 | |
2303 | |
2260 | If undefined or defined to be C<1>, libev will compile in support for the |
2304 | If undefined or defined to be C<1>, libev will compile in support for the |
2261 | C<select>(2) backend. No attempt at autodetection will be done: if no |
2305 | C<select>(2) backend. No attempt at autodetection will be done: if no |
… | |
… | |
2440 | |
2484 | |
2441 | =item ev_set_cb (ev, cb) |
2485 | =item ev_set_cb (ev, cb) |
2442 | |
2486 | |
2443 | Can be used to change the callback member declaration in each watcher, |
2487 | Can be used to change the callback member declaration in each watcher, |
2444 | and the way callbacks are invoked and set. Must expand to a struct member |
2488 | and the way callbacks are invoked and set. Must expand to a struct member |
2445 | definition and a statement, respectively. See the F<ev.v> header file for |
2489 | definition and a statement, respectively. See the F<ev.h> header file for |
2446 | their default definitions. One possible use for overriding these is to |
2490 | their default definitions. One possible use for overriding these is to |
2447 | avoid the C<struct ev_loop *> as first argument in all cases, or to use |
2491 | avoid the C<struct ev_loop *> as first argument in all cases, or to use |
2448 | method calls instead of plain function calls in C++. |
2492 | method calls instead of plain function calls in C++. |
|
|
2493 | |
|
|
2494 | =head2 EXPORTED API SYMBOLS |
|
|
2495 | |
|
|
2496 | If you need to re-export the API (e.g. via a dll) and you need a list of |
|
|
2497 | exported symbols, you can use the provided F<Symbol.*> files which list |
|
|
2498 | all public symbols, one per line: |
|
|
2499 | |
|
|
2500 | Symbols.ev for libev proper |
|
|
2501 | Symbols.event for the libevent emulation |
|
|
2502 | |
|
|
2503 | This can also be used to rename all public symbols to avoid clashes with |
|
|
2504 | multiple versions of libev linked together (which is obviously bad in |
|
|
2505 | itself, but sometimes it is inconvinient to avoid this). |
|
|
2506 | |
|
|
2507 | A sed command like this will create wrapper C<#define>'s that you need to |
|
|
2508 | include before including F<ev.h>: |
|
|
2509 | |
|
|
2510 | <Symbols.ev sed -e "s/.*/#define & myprefix_&/" >wrap.h |
|
|
2511 | |
|
|
2512 | This would create a file F<wrap.h> which essentially looks like this: |
|
|
2513 | |
|
|
2514 | #define ev_backend myprefix_ev_backend |
|
|
2515 | #define ev_check_start myprefix_ev_check_start |
|
|
2516 | #define ev_check_stop myprefix_ev_check_stop |
|
|
2517 | ... |
2449 | |
2518 | |
2450 | =head2 EXAMPLES |
2519 | =head2 EXAMPLES |
2451 | |
2520 | |
2452 | For a real-world example of a program the includes libev |
2521 | For a real-world example of a program the includes libev |
2453 | verbatim, you can have a look at the EV perl module |
2522 | verbatim, you can have a look at the EV perl module |