[ViewVC] Diff of: cvs/libeio/eio.pod

Comparing libeio/eio.pod (file contents):
Revision 1.7 by root, Sun Jun 5 22:44:30 2011 UTC vs.
Revision 1.35 by root, Mon Aug 18 08:11:54 2014 UTC

…		…
25	similar functions, as well as less rarely ones such as C<mknod>, C<futime>	25	similar functions, as well as less rarely ones such as C<mknod>, C<futime>
26	or C<readlink>.	26	or C<readlink>.
27		27
28	It also offers wrappers around C<sendfile> (Solaris, Linux, HP-UX and	28	It also offers wrappers around C<sendfile> (Solaris, Linux, HP-UX and
29	FreeBSD, with emulation on other platforms) and C<readahead> (Linux, with	29	FreeBSD, with emulation on other platforms) and C<readahead> (Linux, with
30	emulation elsewhere>).	30	emulation elsewhere).
31		31
32	The goal is to enable you to write fully non-blocking programs. For	32	The goal is to enable you to write fully non-blocking programs. For
33	example, in a game server, you would not want to freeze for a few seconds	33	example, in a game server, you would not want to freeze for a few seconds
34	just because the server is running a backup and you happen to call	34	just because the server is running a backup and you happen to call
35	C<readdir>.	35	C<readdir>.
…		…
45	Unlike the name component C<stamp> might indicate, it is also used for	45	Unlike the name component C<stamp> might indicate, it is also used for
46	time differences throughout libeio.	46	time differences throughout libeio.
47		47
48	=head2 FORK SUPPORT	48	=head2 FORK SUPPORT
49		49
50	Calling C<fork ()> is fully supported by this module. It is implemented in these steps:	50	Usage of pthreads in a program changes the semantics of fork
		51	considerably. Specifically, only async-safe functions can be called after
		52	fork. Libeio uses pthreads, so this applies, and makes using fork hard for
		53	anything but relatively fork + exec uses.
51		54
52	1. wait till all requests in "execute" state have been handled	55	This library only works in the process that initialised it: Forking is
53	(basically requests that are already handed over to the kernel).	56	fully supported, but using libeio in any other process than the one that
54	2. fork	57	called C<eio_init> is not.
55	3. in the parent, continue business as usual, done
56	4. in the child, destroy all ready and pending requests and free the
57	memory used by the worker threads. This gives you a fully empty
58	libeio queue.
59		58
60	Note, however, since libeio does use threads, thr above guarantee doesn't	59	You might get around by not I<using> libeio before (or after) forking in
61	cover your libc, for example, malloc and other libc functions are not	60	the parent, and using it in the child afterwards. You could also try to
62	fork-safe, so there is very little you can do after a fork, and in fatc,	61	call the L<eio_init> function again in the child, which will brutally
63	the above might crash, and thus change.	62	reinitialise all data structures, which isn't POSIX conformant, but
		63	typically works.
		64
		65	Otherwise, the only recommendation you should follow is: treat fork code
		66	the same way you treat signal handlers, and only ever call C<eio_init> in
		67	the process that uses it, and only once ever.
64		68
65	=head1 INITIALISATION/INTEGRATION	69	=head1 INITIALISATION/INTEGRATION
66		70
67	Before you can call any eio functions you first have to initialise the	71	Before you can call any eio functions you first have to initialise the
68	library. The library integrates into any event loop, but can also be used	72	library. The library integrates into any event loop, but can also be used
…		…
77	This function initialises the library. On success it returns C<0>, on	81	This function initialises the library. On success it returns C<0>, on
78	failure it returns C<-1> and sets C<errno> appropriately.	82	failure it returns C<-1> and sets C<errno> appropriately.
79		83
80	It accepts two function pointers specifying callbacks as argument, both of	84	It accepts two function pointers specifying callbacks as argument, both of
81	which can be C<0>, in which case the callback isn't called.	85	which can be C<0>, in which case the callback isn't called.
		86
		87	There is currently no way to change these callbacks later, or to
		88	"uninitialise" the library again.
82		89
83	=item want_poll callback	90	=item want_poll callback
84		91
85	The C<want_poll> callback is invoked whenever libeio wants attention (i.e.	92	The C<want_poll> callback is invoked whenever libeio wants attention (i.e.
86	it wants to be polled by calling C<eio_poll>). It is "edge-triggered",	93	it wants to be polled by calling C<eio_poll>). It is "edge-triggered",
…		…
124	=back	131	=back
125		132
126	For libev, you would typically use an C<ev_async> watcher: the	133	For libev, you would typically use an C<ev_async> watcher: the
127	C<want_poll> callback would invoke C<ev_async_send> to wake up the event	134	C<want_poll> callback would invoke C<ev_async_send> to wake up the event
128	loop. Inside the callback set for the watcher, one would call C<eio_poll	135	loop. Inside the callback set for the watcher, one would call C<eio_poll
129	()> (followed by C<ev_async_send> again if C<eio_poll> indicates that not	136	()>.
130	all requests have been handled yet). The race is taken care of because	137
131	libev resets/rearms the async watcher before calling your callback,	138	If C<eio_poll ()> is configured to not handle all results in one go
132	and therefore, before calling C<eio_poll>. This might result in (some)	139	(i.e. it returns C<-1>) then you should start an idle watcher that calls
133	spurious wake-ups, but is generally harmless.	140	C<eio_poll> until it returns something C<!= -1>.
		141
		142	A full-featured connector between libeio and libev would look as follows
		143	(if C<eio_poll> is handling all requests, it can of course be simplified a
		144	lot by removing the idle watcher logic):
		145
		146	static struct ev_loop *loop;
		147	static ev_idle repeat_watcher;
		148	static ev_async ready_watcher;
		149
		150	/* idle watcher callback, only used when eio_poll */
		151	/* didn't handle all results in one call */
		152	static void
		153	repeat (EV_P_ ev_idle *w, int revents)
		154	{
		155	if (eio_poll () != -1)
		156	ev_idle_stop (EV_A_ w);
		157	}
		158
		159	/* eio has some results, process them */
		160	static void
		161	ready (EV_P_ ev_async *w, int revents)
		162	{
		163	if (eio_poll () == -1)
		164	ev_idle_start (EV_A_ &repeat_watcher);
		165	}
		166
		167	/* wake up the event loop */
		168	static void
		169	want_poll (void)
		170	{
		171	ev_async_send (loop, &ready_watcher)
		172	}
		173
		174	void
		175	my_init_eio ()
		176	{
		177	loop = EV_DEFAULT;
		178
		179	ev_idle_init (&repeat_watcher, repeat);
		180	ev_async_init (&ready_watcher, ready);
		181	ev_async_start (loop, &watcher);
		182
		183	eio_init (want_poll, 0);
		184	}
134		185
135	For most other event loops, you would typically use a pipe - the event	186	For most other event loops, you would typically use a pipe - the event
136	loop should be told to wait for read readiness on the read end. In	187	loop should be told to wait for read readiness on the read end. In
137	C<want_poll> you would write a single byte, in C<done_poll> you would try	188	C<want_poll> you would write a single byte, in C<done_poll> you would try
138	to read that byte, and in the callback for the read end, you would call	189	to read that byte, and in the callback for the read end, you would call
139	C<eio_poll>. The race is avoided here because the event loop should invoke
140	your callback again and again until the byte has been read (as the pipe
141	read callback does not read it, only C<done_poll>).
142
143	=head2 CONFIGURATION
144
145	The functions in this section can sometimes be useful, but the default
146	configuration will do in most case, so you should skip this section on
147	first reading.
148
149	=over 4
150
151	=item eio_set_max_poll_time (eio_tstamp nseconds)
152
153	This causes C<eio_poll ()> to return after it has detected that it was
154	running for C<nsecond> seconds or longer (this number can be fractional).
155
156	This can be used to limit the amount of time spent handling eio requests,
157	for example, in interactive programs, you might want to limit this time to
158	C<0.01> seconds or so.
159
160	Note that:
161
162	a) libeio doesn't know how long your request callbacks take, so the time
163	spent in C<eio_poll> is up to one callback invocation longer then this
164	interval.
165
166	b) this is implemented by calling C<gettimeofday> after each request,
167	which can be costly.
168
169	c) at least one request will be handled.
170
171	=item eio_set_max_poll_reqs (unsigned int nreqs)
172
173	When C<nreqs> is non-zero, then C<eio_poll> will not handle more than
174	C<nreqs> requests per invocation. This is a less costly way to limit the
175	amount of work done by C<eio_poll> then setting a time limit.
176
177	If you know your callbacks are generally fast, you could use this to
178	encourage interactiveness in your programs by setting it to C<10>, C<100>
179	or even C<1000>.
180
181	=item eio_set_min_parallel (unsigned int nthreads)
182
183	Make sure libeio can handle at least this many requests in parallel. It
184	might be able handle more.
185
186	=item eio_set_max_parallel (unsigned int nthreads)
187
188	Set the maximum number of threads that libeio will spawn.
189
190	=item eio_set_max_idle (unsigned int nthreads)
191
192	Libeio uses threads internally to handle most requests, and will start and stop threads on demand.
193
194	This call can be used to limit the number of idle threads (threads without
195	work to do): libeio will keep some threads idle in preparation for more
196	requests, but never longer than C<nthreads> threads.
197
198	In addition to this, libeio will also stop threads when they are idle for
199	a few seconds, regardless of this setting.
200
201	=item unsigned int eio_nthreads ()
202
203	Return the number of worker threads currently running.
204
205	=item unsigned int eio_nreqs ()
206
207	Return the number of requests currently handled by libeio. This is the
208	total number of requests that have been submitted to libeio, but not yet
209	destroyed.
210
211	=item unsigned int eio_nready ()
212
213	Returns the number of ready requests, i.e. requests that have been
214	submitted but have not yet entered the execution phase.
215
216	=item unsigned int eio_npending ()
217
218	Returns the number of pending requests, i.e. requests that have been
219	executed and have results, but have not been finished yet by a call to
220	C<eio_poll>).	190	C<eio_poll>.
221		191
222	=back	192	You don't have to take special care in the case C<eio_poll> doesn't handle
		193	all requests, as the done callback will not be invoked, so the event loop
		194	will still signal readiness for the pipe until I<all> results have been
		195	processed.
223		196
224		197
225	=head1 HIGH LEVEL REQUEST API	198	=head1 HIGH LEVEL REQUEST API
226		199
227	Libeio has both a high-level API, which consists of calling a request	200	Libeio has both a high-level API, which consists of calling a request
…		…
234		207
235	You submit a request by calling the relevant C<eio_TYPE> function with the	208	You submit a request by calling the relevant C<eio_TYPE> function with the
236	required parameters, a callback of type C<int (eio_cb)(eio_req req)>	209	required parameters, a callback of type C<int (eio_cb)(eio_req req)>
237	(called C<eio_cb> below) and a freely usable C<void *data> argument.	210	(called C<eio_cb> below) and a freely usable C<void *data> argument.
238		211
239	The return value will either be 0	212	The return value will either be 0, in case something went really wrong
		213	(which can basically only happen on very fatal errors, such as C<malloc>
		214	returning 0, which is rather unlikely), or a pointer to the newly-created
		215	and submitted C<eio_req *>.
240		216
241	The callback will be called with an C<eio_req *> which contains the	217	The callback will be called with an C<eio_req *> which contains the
242	results of the request. The members you can access inside that structure	218	results of the request. The members you can access inside that structure
243	vary from request to request, except for:	219	vary from request to request, except for:
244		220
…		…
256	=item C<void *data>	232	=item C<void *data>
257		233
258	The C<void *data> member simply stores the value of the C<data> argument.	234	The C<void *data> member simply stores the value of the C<data> argument.
259		235
260	=back	236	=back
		237
		238	Members not explicitly described as accessible must not be
		239	accessed. Specifically, there is no guarantee that any members will still
		240	have the value they had when the request was submitted.
261		241
262	The return value of the callback is normally C<0>, which tells libeio to	242	The return value of the callback is normally C<0>, which tells libeio to
263	continue normally. If a callback returns a nonzero value, libeio will	243	continue normally. If a callback returns a nonzero value, libeio will
264	stop processing results (in C<eio_poll>) and will return the value to its	244	stop processing results (in C<eio_poll>) and will return the value to its
265	caller.	245	caller.
266		246
267	Memory areas passed to libeio must stay valid as long as a request	247	Memory areas passed to libeio wrappers must stay valid as long as a
268	executes, with the exception of paths, which are being copied	248	request executes, with the exception of paths, which are being copied
269	internally. Any memory libeio itself allocates will be freed after the	249	internally. Any memory libeio itself allocates will be freed after the
270	finish callback has been called. If you want to manage all memory passed	250	finish callback has been called. If you want to manage all memory passed
271	to libeio yourself you can use the low-level API.	251	to libeio yourself you can use the low-level API.
272		252
273	For example, to open a file, you could do this:	253	For example, to open a file, you could do this:
…		…
291	}	271	}
292		272
293	/* the first three arguments are passed to open(2) */	273	/* the first three arguments are passed to open(2) */
294	/* the remaining are priority, callback and data */	274	/* the remaining are priority, callback and data */
295	if (!eio_open ("/etc/passwd", O_RDONLY, 0, 0, file_open_done, 0))	275	if (!eio_open ("/etc/passwd", O_RDONLY, 0, 0, file_open_done, 0))
296	abort (); /* something ent wrong, we will all die!!! */	276	abort (); /* something went wrong, we will all die!!! */
297		277
298	Note that you additionally need to call C<eio_poll> when the C<want_cb>	278	Note that you additionally need to call C<eio_poll> when the C<want_cb>
299	indicates that requests are ready to be processed.	279	indicates that requests are ready to be processed.
		280
		281	=head2 CANCELLING REQUESTS
		282
		283	Sometimes the need for a request goes away before the request is
		284	finished. In that case, one can cancel the request by a call to
		285	C<eio_cancel>:
		286
		287	=over 4
		288
		289	=item eio_cancel (eio_req *req)
		290
		291	Cancel the request (and all its subrequests). If the request is currently
		292	executing it might still continue to execute, and in other cases it might
		293	still take a while till the request is cancelled.
		294
		295	When cancelled, the finish callback will not be invoked.
		296
		297	C<EIO_CANCELLED> is still true for requests that have successfully
		298	executed, as long as C<eio_cancel> was called on them at some point.
		299
		300	=back
300		301
301	=head2 AVAILABLE REQUESTS	302	=head2 AVAILABLE REQUESTS
302		303
303	The following request functions are available. I<All> of them return the	304	The following request functions are available. I<All> of them return the
304	C<eio_req *> on success and C<0> on failure, and I<all> of them have the	305	C<eio_req *> on success and C<0> on failure, and I<all> of them have the
…		…
307	custom data value as C<data>.	308	custom data value as C<data>.
308		309
309	=head3 POSIX API WRAPPERS	310	=head3 POSIX API WRAPPERS
310		311
311	These requests simply wrap the POSIX call of the same name, with the same	312	These requests simply wrap the POSIX call of the same name, with the same
312	arguments:	313	arguments. If a function is not implemented by the OS and cannot be emulated
		314	in some way, then all of these return C<-1> and set C<errorno> to C<ENOSYS>.
313		315
314	=over 4	316	=over 4
315		317
316	=item eio_open (const char path, int flags, mode_t mode, int pri, eio_cb cb, void data)	318	=item eio_open (const char path, int flags, mode_t mode, int pri, eio_cb cb, void data)
317		319
		320	=item eio_truncate (const char path, off_t offset, int pri, eio_cb cb, void data)
		321
		322	=item eio_chown (const char path, uid_t uid, gid_t gid, int pri, eio_cb cb, void data)
		323
		324	=item eio_chmod (const char path, mode_t mode, int pri, eio_cb cb, void data)
		325
		326	=item eio_mkdir (const char path, mode_t mode, int pri, eio_cb cb, void data)
		327
		328	=item eio_rmdir (const char path, int pri, eio_cb cb, void data)
		329
		330	=item eio_unlink (const char path, int pri, eio_cb cb, void data)
		331
318	=item eio_utime (const char path, eio_tstamp atime, eio_tstamp mtime, int pri, eio_cb cb, void data)	332	=item eio_utime (const char path, eio_tstamp atime, eio_tstamp mtime, int pri, eio_cb cb, void data)
319		333
320	=item eio_truncate (const char path, off_t offset, int pri, eio_cb cb, void data)
321
322	=item eio_chown (const char path, uid_t uid, gid_t gid, int pri, eio_cb cb, void data)
323
324	=item eio_chmod (const char path, mode_t mode, int pri, eio_cb cb, void data)
325
326	=item eio_mkdir (const char path, mode_t mode, int pri, eio_cb cb, void data)
327
328	=item eio_rmdir (const char path, int pri, eio_cb cb, void data)
329
330	=item eio_unlink (const char path, int pri, eio_cb cb, void data)
331
332	=item eio_readlink (const char path, int pri, eio_cb cb, void data) /* result=ptr2 allocated dynamically */
333
334	=item eio_stat (const char path, int pri, eio_cb cb, void data) /* stat buffer=ptr2 allocated dynamically */
335
336	=item eio_lstat (const char path, int pri, eio_cb cb, void data) /* stat buffer=ptr2 allocated dynamically */
337
338	=item eio_statvfs (const char path, int pri, eio_cb cb, void data) /* stat buffer=ptr2 allocated dynamically */
339
340	=item eio_mknod (const char path, mode_t mode, dev_t dev, int pri, eio_cb cb, void data)	334	=item eio_mknod (const char path, mode_t mode, dev_t dev, int pri, eio_cb cb, void data)
341		335
342	=item eio_link (const char path, const char new_path, int pri, eio_cb cb, void *data)	336	=item eio_link (const char path, const char new_path, int pri, eio_cb cb, void *data)
343		337
344	=item eio_symlink (const char path, const char new_path, int pri, eio_cb cb, void *data)	338	=item eio_symlink (const char path, const char new_path, int pri, eio_cb cb, void *data)
345		339
346	=item eio_rename (const char path, const char new_path, int pri, eio_cb cb, void *data)	340	=item eio_rename (const char path, const char new_path, int pri, eio_cb cb, void *data)
347		341
348	=item eio_msync (void addr, size_t length, int flags, int pri, eio_cb cb, void data)
349
350	=item eio_mlock (void addr, size_t length, int pri, eio_cb cb, void data)	342	=item eio_mlock (void addr, size_t length, int pri, eio_cb cb, void data)
351
352	=item eio_mlockall (int flags, int pri, eio_cb cb, void *data)
353		343
354	=item eio_close (int fd, int pri, eio_cb cb, void *data)	344	=item eio_close (int fd, int pri, eio_cb cb, void *data)
355		345
356	=item eio_sync (int pri, eio_cb cb, void *data)	346	=item eio_sync (int pri, eio_cb cb, void *data)
357		347
…		…
386		376
387	Not surprisingly, pread and pwrite are not thread-safe on Darwin (OS/X),	377	Not surprisingly, pread and pwrite are not thread-safe on Darwin (OS/X),
388	so it is advised not to submit multiple requests on the same fd on this	378	so it is advised not to submit multiple requests on the same fd on this
389	horrible pile of garbage.	379	horrible pile of garbage.
390		380
		381	=item eio_mlockall (int flags, int pri, eio_cb cb, void *data)
		382
		383	Like C<mlockall>, but the flag value constants are called
		384	C<EIO_MCL_CURRENT> and C<EIO_MCL_FUTURE>.
		385
		386	=item eio_msync (void addr, size_t length, int flags, int pri, eio_cb cb, void data)
		387
		388	Just like msync, except that the flag values are called C<EIO_MS_ASYNC>,
		389	C<EIO_MS_INVALIDATE> and C<EIO_MS_SYNC>.
		390
		391	=item eio_readlink (const char path, int pri, eio_cb cb, void data)
		392
		393	If successful, the path read by C<readlink(2)> can be accessed via C<<
		394	req->ptr2 >> and is I<NOT> null-terminated, with the length specified as
		395	C<< req->result >>.
		396
		397	if (req->result >= 0)
		398	{
		399	char target = strndup ((char )req->ptr2, req->result);
		400
		401	free (target);
		402	}
		403
		404	=item eio_realpath (const char path, int pri, eio_cb cb, void data)
		405
		406	Similar to the realpath libc function, but unlike that one, C<<
		407	req->result >> is C<-1> on failure. On success, the result is the length
		408	of the returned path in C<ptr2> (which is I<NOT> 0-terminated) - this is
		409	similar to readlink.
		410
		411	=item eio_stat (const char path, int pri, eio_cb cb, void data)
		412
		413	=item eio_lstat (const char path, int pri, eio_cb cb, void data)
		414
391	=item eio_fstat (int fd, int pri, eio_cb cb, void *data)	415	=item eio_fstat (int fd, int pri, eio_cb cb, void *data)
392		416
393	Stats a file - if C<< req->result >> indicates success, then you can	417	Stats a file - if C<< req->result >> indicates success, then you can
394	access the C<struct stat>-like structure via C<< req->ptr2 >>:	418	access the C<struct stat>-like structure via C<< req->ptr2 >>:
395		419
396	EIO_STRUCT_STAT statdata = (EIO_STRUCT_STAT )req->ptr2;	420	EIO_STRUCT_STAT statdata = (EIO_STRUCT_STAT )req->ptr2;
397		421
398	=item eio_fstatvfs (int fd, int pri, eio_cb cb, void data) / stat buffer=ptr2 allocated dynamically */	422	=item eio_statvfs (const char path, int pri, eio_cb cb, void data)
		423
		424	=item eio_fstatvfs (int fd, int pri, eio_cb cb, void *data)
399		425
400	Stats a filesystem - if C<< req->result >> indicates success, then you can	426	Stats a filesystem - if C<< req->result >> indicates success, then you can
401	access the C<struct statvfs>-like structure via C<< req->ptr2 >>:	427	access the C<struct statvfs>-like structure via C<< req->ptr2 >>:
402		428
403	EIO_STRUCT_STATVFS statdata = (EIO_STRUCT_STATVFS )req->ptr2;	429	EIO_STRUCT_STATVFS statdata = (EIO_STRUCT_STATVFS )req->ptr2;
404		430
405	=back	431	=back
406		432
407	=head3 READING DIRECTORIES	433	=head3 READING DIRECTORIES
408		434
409	Reading directories sounds simple, but can be rather demanding, especially	435	Reading directories sounds simple, but can be rather demanding, especially
410	if you want to do stuff such as traversing a diretcory hierarchy or	436	if you want to do stuff such as traversing a directory hierarchy or
411	processing all files in a directory. Libeio can assist thess complex tasks	437	processing all files in a directory. Libeio can assist these complex tasks
412	with it's C<eio_readdir> call.	438	with it's C<eio_readdir> call.
413		439
414	=over 4	440	=over 4
415		441
416	=item eio_readdir (const char path, int flags, int pri, eio_cb cb, void data)	442	=item eio_readdir (const char path, int flags, int pri, eio_cb cb, void data)
…		…
419	(via the C<opendir>, C<readdir> and C<closedir> calls) and returns either	445	(via the C<opendir>, C<readdir> and C<closedir> calls) and returns either
420	the names or an array of C<struct eio_dirent>, depending on the C<flags>	446	the names or an array of C<struct eio_dirent>, depending on the C<flags>
421	argument.	447	argument.
422		448
423	The C<< req->result >> indicates either the number of files found, or	449	The C<< req->result >> indicates either the number of files found, or
424	C<-1> on error. On success, zero-terminated names can be found as C<< req->ptr2 >>,	450	C<-1> on error. On success, null-terminated names can be found as C<< req->ptr2 >>,
425	and C<struct eio_dirents>, if requested by C<flags>, can be found via C<<	451	and C<struct eio_dirents>, if requested by C<flags>, can be found via C<<
426	req->ptr1 >>.	452	req->ptr1 >>.
427		453
428	Here is an example that prints all the names:	454	Here is an example that prints all the names:
429		455
…		…
448		474
449	If this flag is specified, then, in addition to the names in C<ptr2>,	475	If this flag is specified, then, in addition to the names in C<ptr2>,
450	also an array of C<struct eio_dirent> is returned, in C<ptr1>. A C<struct	476	also an array of C<struct eio_dirent> is returned, in C<ptr1>. A C<struct
451	eio_dirent> looks like this:	477	eio_dirent> looks like this:
452		478
453	struct eio_dirent	479	struct eio_dirent
454	{	480	{
455	int nameofs; /* offset of null-terminated name string in (char )req->ptr2 /	481	int nameofs; /* offset of null-terminated name string in (char )req->ptr2 /
456	unsigned short namelen; /* size of filename without trailing 0 */	482	unsigned short namelen; /* size of filename without trailing 0 */
457	unsigned char type; /* one of EIO_DT_* */	483	unsigned char type; /* one of EIO_DT_* */
458	signed char score; /* internal use */	484	signed char score; /* internal use */
459	ino_t inode; /* the inode number, if available, otherwise unspecified */	485	ino_t inode; /* the inode number, if available, otherwise unspecified */
460	};	486	};
461		487
462	The only members you normally would access are C<nameofs>, which is the	488	The only members you normally would access are C<nameofs>, which is the
463	byte-offset from C<ptr2> to the start of the name, C<namelen> and C<type>.	489	byte-offset from C<ptr2> to the start of the name, C<namelen> and C<type>.
464		490
465	C<type> can be one of:	491	C<type> can be one of:
…		…
508	When this flag is specified, then the names will be returned in an order	534	When this flag is specified, then the names will be returned in an order
509	suitable for stat()'ing each one. That is, when you plan to stat()	535	suitable for stat()'ing each one. That is, when you plan to stat()
510	all files in the given directory, then the returned order will likely	536	all files in the given directory, then the returned order will likely
511	be fastest.	537	be fastest.
512		538
513	If both this flag and C<EIO_READDIR_DIRS_FIRST> are specified, then	539	If both this flag and C<EIO_READDIR_DIRS_FIRST> are specified, then the
514	the likely dirs come first, resulting in a less optimal stat order.	540	likely directories come first, resulting in a less optimal stat order.
515		541
516	=item EIO_READDIR_FOUND_UNKNOWN	542	=item EIO_READDIR_FOUND_UNKNOWN
517		543
518	This flag should not be specified when calling C<eio_readdir>. Instead,	544	This flag should not be specified when calling C<eio_readdir>. Instead,
519	it is being set by C<eio_readdir> (you can access the C<flags> via C<<	545	it is being set by C<eio_readdir> (you can access the C<flags> via C<<
520	req->int1 >>, when any of the C<type>'s found were C<EIO_DT_UNKNOWN>. The	546	req->int1 >>, when any of the C<type>'s found were C<EIO_DT_UNKNOWN>. The
521	absense of this flag therefore indicates that all C<type>'s are known,	547	absence of this flag therefore indicates that all C<type>'s are known,
522	which can be used to speed up some algorithms.	548	which can be used to speed up some algorithms.
523		549
524	A typical use case would be to identify all subdirectories within a	550	A typical use case would be to identify all subdirectories within a
525	directory - you would ask C<eio_readdir> for C<EIO_READDIR_DIRS_FIRST>. If	551	directory - you would ask C<eio_readdir> for C<EIO_READDIR_DIRS_FIRST>. If
526	then this flag is I<NOT> set, then all the entries at the beginning of the	552	then this flag is I<NOT> set, then all the entries at the beginning of the
…		…
556	=item eio_readahead (int fd, off_t offset, size_t length, int pri, eio_cb cb, void *data)	582	=item eio_readahead (int fd, off_t offset, size_t length, int pri, eio_cb cb, void *data)
557		583
558	Calls C<readahead(2)>. If the syscall is missing, then the call is	584	Calls C<readahead(2)>. If the syscall is missing, then the call is
559	emulated by simply reading the data (currently in 64kiB chunks).	585	emulated by simply reading the data (currently in 64kiB chunks).
560		586
		587	=item eio_syncfs (int fd, int pri, eio_cb cb, void *data)
		588
		589	Calls Linux' C<syncfs> syscall, if available. Returns C<-1> and sets
		590	C<errno> to C<ENOSYS> if the call is missing I<but still calls sync()>,
		591	if the C<fd> is C<< >= 0 >>, so you can probe for the availability of the
		592	syscall with a negative C<fd> argument and checking for C<-1/ENOSYS>.
		593
561	=item eio_sync_file_range (int fd, off_t offset, size_t nbytes, unsigned int flags, int pri, eio_cb cb, void *data)	594	=item eio_sync_file_range (int fd, off_t offset, size_t nbytes, unsigned int flags, int pri, eio_cb cb, void *data)
562		595
563	Calls C<sync_file_range>. If the syscall is missing, then this is the same	596	Calls C<sync_file_range>. If the syscall is missing, then this is the same
564	as calling C<fdatasync>.	597	as calling C<fdatasync>.
565		598
		599	Flags can be any combination of C<EIO_SYNC_FILE_RANGE_WAIT_BEFORE>,
		600	C<EIO_SYNC_FILE_RANGE_WRITE> and C<EIO_SYNC_FILE_RANGE_WAIT_AFTER>.
		601
		602	=item eio_fallocate (int fd, int mode, off_t offset, off_t len, int pri, eio_cb cb, void *data)
		603
		604	Calls C<fallocate> (note: I<NOT> C<posix_fallocate>!). If the syscall is
		605	missing, then it returns failure and sets C<errno> to C<ENOSYS>.
		606
		607	The C<mode> argument can be C<0> (for behaviour similar to
		608	C<posix_fallocate>), or C<EIO_FALLOC_FL_KEEP_SIZE>, which keeps the size
		609	of the file unchanged (but still preallocates space beyond end of file).
		610
566	=back	611	=back
567		612
568	=head3 LIBEIO-SPECIFIC REQUESTS	613	=head3 LIBEIO-SPECIFIC REQUESTS
569		614
570	These requests are specific to libeio and do not correspond to any OS call.	615	These requests are specific to libeio and do not correspond to any OS call.
571		616
572	=over 4	617	=over 4
573		618
574	=item eio_mtouch (void addr, size_t length, int flags, int pri, eio_cb cb, void data)	619	=item eio_mtouch (void addr, size_t length, int flags, int pri, eio_cb cb, void data)
575		620
		621	Reads (C<flags == 0>) or modifies (C<flags == EIO_MT_MODIFY>) the given
		622	memory area, page-wise, that is, it reads (or reads and writes back) the
		623	first octet of every page that spans the memory area.
		624
		625	This can be used to page in some mmapped file, or dirty some pages. Note
		626	that dirtying is an unlocked read-write access, so races can ensue when
		627	the some other thread modifies the data stored in that memory area.
		628
576	=item eio_custom (void ()(eio_req ) execute, int pri, eio_cb cb, void *data)	629	=item eio_custom (void ()(eio_req ) execute, int pri, eio_cb cb, void *data)
577		630
578	Executes a custom request, i.e., a user-specified callback.	631	Executes a custom request, i.e., a user-specified callback.
579		632
580	The callback gets the C<eio_req *> as parameter and is expected to read	633	The callback gets the C<eio_req *> as parameter and is expected to read
581	and modify any request-specific members. Specifically, it should set C<<	634	and modify any request-specific members. Specifically, it should set C<<
…		…
601	req->result = open (req->data, O_RDONLY);	654	req->result = open (req->data, O_RDONLY);
602	}	655	}
603		656
604	eio_custom (my_open, 0, my_open_done, "/etc/passwd");	657	eio_custom (my_open, 0, my_open_done, "/etc/passwd");
605		658
606	=item eio_busy (eio_tstamp delay, int pri, eio_cb cb, void *data)	659	=item eio_busy (eio_tstamp delay, int pri, eio_cb cb, void *data)
607		660
608	This is a a request that takes C<delay> seconds to execute, but otherwise	661	This is a request that takes C<delay> seconds to execute, but otherwise
609	does nothing - it simply puts one of the worker threads to sleep for this	662	does nothing - it simply puts one of the worker threads to sleep for this
610	long.	663	long.
611		664
612	This request can be used to artificially increase load, e.g. for debugging	665	This request can be used to artificially increase load, e.g. for debugging
613	or benchmarking reasons.	666	or benchmarking reasons.
614		667
615	=item eio_nop (int pri, eio_cb cb, void *data)	668	=item eio_nop (int pri, eio_cb cb, void *data)
616		669
617	This request does nothing, except go through the whole request cycle. This	670	This request does nothing, except go through the whole request cycle. This
618	can be used to measure latency or in some cases to simplify code, but is	671	can be used to measure latency or in some cases to simplify code, but is
619	not really of much use.	672	not really of much use.
620		673
621	=back	674	=back
622		675
623	=head3 GROUPING AND LIMITING REQUESTS	676	=head3 GROUPING AND LIMITING REQUESTS
624		677
		678	There is one more rather special request, C<eio_grp>. It is a very special
		679	aio request: Instead of doing something, it is a container for other eio
		680	requests.
		681
		682	There are two primary use cases for this: a) bundle many requests into a
		683	single, composite, request with a definite callback and the ability to
		684	cancel the whole request with its subrequests and b) limiting the number
		685	of "active" requests.
		686
		687	Further below you will find more discussion of these topics - first
		688	follows the reference section detailing the request generator and other
		689	methods.
		690
		691	=over 4
		692
		693	=item eio_req grp = eio_grp (eio_cb cb, void data)
		694
		695	Creates, submits and returns a group request. Note that it doesn't have a
		696	priority, unlike all other requests.
		697
		698	=item eio_grp_add (eio_req grp, eio_req req)
		699
		700	Adds a request to the request group.
		701
		702	=item eio_grp_cancel (eio_req *grp)
		703
		704	Cancels all requests I<in> the group, but I<not> the group request
		705	itself. You can cancel the group request I<and> all subrequests via a
		706	normal C<eio_cancel> call.
		707
		708	=back
		709
		710	=head4 GROUP REQUEST LIFETIME
		711
		712	Left alone, a group request will instantly move to the pending state and
		713	will be finished at the next call of C<eio_poll>.
		714
		715	The usefulness stems from the fact that, if a subrequest is added to a
		716	group I<before> a call to C<eio_poll>, via C<eio_grp_add>, then the group
		717	will not finish until all the subrequests have finished.
		718
		719	So the usage cycle of a group request is like this: after it is created,
		720	you normally instantly add a subrequest. If none is added, the group
		721	request will finish on it's own. As long as subrequests are added before
		722	the group request is finished it will be kept from finishing, that is the
		723	callbacks of any subrequests can, in turn, add more requests to the group,
		724	and as long as any requests are active, the group request itself will not
		725	finish.
		726
		727	=head4 CREATING COMPOSITE REQUESTS
		728
		729	Imagine you wanted to create an C<eio_load> request that opens a file,
		730	reads it and closes it. This means it has to execute at least three eio
		731	requests, but for various reasons it might be nice if that request looked
		732	like any other eio request.
		733
		734	This can be done with groups:
		735
		736	=over 4
		737
		738	=item 1) create the request object
		739
		740	Create a group that contains all further requests. This is the request you
		741	can return as "the load request".
		742
		743	=item 2) open the file, maybe
		744
		745	Next, open the file with C<eio_open> and add the request to the group
		746	request and you are finished setting up the request.
		747
		748	If, for some reason, you cannot C<eio_open> (path is a null ptr?) you
		749	can set C<< grp->result >> to C<-1> to signal an error and let the group
		750	request finish on its own.
		751
		752	=item 3) open callback adds more requests
		753
		754	In the open callback, if the open was not successful, copy C<<
		755	req->errorno >> to C<< grp->errorno >> and set C<< grp->result >> to
		756	C<-1> to signal an error.
		757
		758	Otherwise, malloc some memory or so and issue a read request, adding the
		759	read request to the group.
		760
		761	=item 4) continue issuing requests till finished
		762
		763	In the read callback, check for errors and possibly continue with
		764	C<eio_close> or any other eio request in the same way.
		765
		766	As soon as no new requests are added, the group request will finish. Make
		767	sure you I<always> set C<< grp->result >> to some sensible value.
		768
		769	=back
		770
		771	=head4 REQUEST LIMITING
		772
		773
625	#TODO	774	#TODO
626		775
627	/*****************************************************************************/
628	/* groups */
629
630	eio_req eio_grp (eio_cb cb, void data);
631	void eio_grp_feed (eio_req grp, void (feed)(eio_req *req), int limit);
632	void eio_grp_limit (eio_req *grp, int limit);	776	void eio_grp_limit (eio_req *grp, int limit);
633	void eio_grp_add (eio_req grp, eio_req req);
634	void eio_grp_cancel (eio_req grp); / cancels all sub requests but not the group */
635		777
636
637	=back
638		778
639		779
640	=head1 LOW LEVEL REQUEST API	780	=head1 LOW LEVEL REQUEST API
641		781
642	#TODO	782	#TODO
…		…
645	=head1 ANATOMY AND LIFETIME OF AN EIO REQUEST	785	=head1 ANATOMY AND LIFETIME OF AN EIO REQUEST
646		786
647	A request is represented by a structure of type C<eio_req>. To initialise	787	A request is represented by a structure of type C<eio_req>. To initialise
648	it, clear it to all zero bytes:	788	it, clear it to all zero bytes:
649		789
650	eio_req req;	790	eio_req req;
651		791
652	memset (&req, 0, sizeof (req));	792	memset (&req, 0, sizeof (req));
653		793
654	A more common way to initialise a new C<eio_req> is to use C<calloc>:	794	A more common way to initialise a new C<eio_req> is to use C<calloc>:
655		795
656	eio_req req = calloc (1, sizeof (req));	796	eio_req req = calloc (1, sizeof (req));
657		797
658	In either case, libeio neither allocates, initialises or frees the	798	In either case, libeio neither allocates, initialises or frees the
659	C<eio_req> structure for you - it merely uses it.	799	C<eio_req> structure for you - it merely uses it.
660		800
661	zero	801	zero
662		802
663	#TODO	803	#TODO
		804
		805	=head2 CONFIGURATION
		806
		807	The functions in this section can sometimes be useful, but the default
		808	configuration will do in most case, so you should skip this section on
		809	first reading.
		810
		811	=over 4
		812
		813	=item eio_set_max_poll_time (eio_tstamp nseconds)
		814
		815	This causes C<eio_poll ()> to return after it has detected that it was
		816	running for C<nsecond> seconds or longer (this number can be fractional).
		817
		818	This can be used to limit the amount of time spent handling eio requests,
		819	for example, in interactive programs, you might want to limit this time to
		820	C<0.01> seconds or so.
		821
		822	Note that:
		823
		824	=over 4
		825
		826	=item a) libeio doesn't know how long your request callbacks take, so the
		827	time spent in C<eio_poll> is up to one callback invocation longer then
		828	this interval.
		829
		830	=item b) this is implemented by calling C<gettimeofday> after each
		831	request, which can be costly.
		832
		833	=item c) at least one request will be handled.
		834
		835	=back
		836
		837	=item eio_set_max_poll_reqs (unsigned int nreqs)
		838
		839	When C<nreqs> is non-zero, then C<eio_poll> will not handle more than
		840	C<nreqs> requests per invocation. This is a less costly way to limit the
		841	amount of work done by C<eio_poll> then setting a time limit.
		842
		843	If you know your callbacks are generally fast, you could use this to
		844	encourage interactiveness in your programs by setting it to C<10>, C<100>
		845	or even C<1000>.
		846
		847	=item eio_set_min_parallel (unsigned int nthreads)
		848
		849	Make sure libeio can handle at least this many requests in parallel. It
		850	might be able handle more.
		851
		852	=item eio_set_max_parallel (unsigned int nthreads)
		853
		854	Set the maximum number of threads that libeio will spawn.
		855
		856	=item eio_set_max_idle (unsigned int nthreads)
		857
		858	Libeio uses threads internally to handle most requests, and will start and stop threads on demand.
		859
		860	This call can be used to limit the number of idle threads (threads without
		861	work to do): libeio will keep some threads idle in preparation for more
		862	requests, but never longer than C<nthreads> threads.
		863
		864	In addition to this, libeio will also stop threads when they are idle for
		865	a few seconds, regardless of this setting.
		866
		867	=item unsigned int eio_nthreads ()
		868
		869	Return the number of worker threads currently running.
		870
		871	=item unsigned int eio_nreqs ()
		872
		873	Return the number of requests currently handled by libeio. This is the
		874	total number of requests that have been submitted to libeio, but not yet
		875	destroyed.
		876
		877	=item unsigned int eio_nready ()
		878
		879	Returns the number of ready requests, i.e. requests that have been
		880	submitted but have not yet entered the execution phase.
		881
		882	=item unsigned int eio_npending ()
		883
		884	Returns the number of pending requests, i.e. requests that have been
		885	executed and have results, but have not been finished yet by a call to
		886	C<eio_poll>).
		887
		888	=back
664		889
665	=head1 EMBEDDING	890	=head1 EMBEDDING
666		891
667	Libeio can be embedded directly into programs. This functionality is not	892	Libeio can be embedded directly into programs. This functionality is not
668	documented and not (yet) officially supported.	893	documented and not (yet) officially supported.
…		…
685	This symbol governs the stack size for each eio thread. Libeio itself	910	This symbol governs the stack size for each eio thread. Libeio itself
686	was written to use very little stackspace, but when using C<EIO_CUSTOM>	911	was written to use very little stackspace, but when using C<EIO_CUSTOM>
687	requests, you might want to increase this.	912	requests, you might want to increase this.
688		913
689	If this symbol is undefined (the default) then libeio will use its default	914	If this symbol is undefined (the default) then libeio will use its default
690	stack size (C<sizeof (long) * 4096> currently). If it is defined, but	915	stack size (C<sizeof (void ) 4096> currently). In all other cases, the
691	C<0>, then the default operating system stack size will be used. In all
692	other cases, the value must be an expression that evaluates to the desired	916	value must be an expression that evaluates to the desired stack size.
693	stack size.
694		917
695	=back	918	=back
696		919
697		920
698	=head1 PORTABILITY REQUIREMENTS	921	=head1 PORTABILITY REQUIREMENTS

Diff Legend

-–
+Removed lines
-+
+Added lines
-<
+Changed lines
->
+Changed lines

Comparing libeio/eio.pod (file contents): Revision 1.7 by root, Sun Jun 5 22:44:30 2011 UTC vs. Revision 1.35 by root, Mon Aug 18 08:11:54 2014 UTC

Diff Legend

Comparing libeio/eio.pod (file contents):
Revision 1.7 by root, Sun Jun 5 22:44:30 2011 UTC vs.
Revision 1.35 by root, Mon Aug 18 08:11:54 2014 UTC