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

Comparing libeio/eio.pod (file contents):
Revision 1.10 by root, Sun Jun 5 23:22:04 2011 UTC vs.
Revision 1.21 by root, Thu Jul 7 22:36:18 2011 UTC

…		…
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	Calling C<fork ()> is fully supported by this module - but you must not
		51	rely on this. It is currently implemented in these steps:
51		52
52	1. wait till all requests in "execute" state have been handled	53	1. wait till all requests in "execute" state have been handled
53	(basically requests that are already handed over to the kernel).	54	(basically requests that are already handed over to the kernel).
54	2. fork	55	2. fork
55	3. in the parent, continue business as usual, done	56	3. in the parent, continue business as usual, done
56	4. in the child, destroy all ready and pending requests and free the	57	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	memory used by the worker threads. This gives you a fully empty
58	libeio queue.	59	libeio queue.
59		60
60	Note, however, since libeio does use threads, thr above guarantee doesn't	61	Note, however, since libeio does use threads, the above guarantee doesn't
61	cover your libc, for example, malloc and other libc functions are not	62	cover your libc, for example, malloc and other libc functions are not
62	fork-safe, so there is very little you can do after a fork, and in fatc,	63	fork-safe, so there is very little you can do after a fork, and in fact,
63	the above might crash, and thus change.	64	the above might crash, and thus change.
64		65
65	=head1 INITIALISATION/INTEGRATION	66	=head1 INITIALISATION/INTEGRATION
66		67
67	Before you can call any eio functions you first have to initialise the	68	Before you can call any eio functions you first have to initialise the
…		…
124	=back	125	=back
125		126
126	For libev, you would typically use an C<ev_async> watcher: the	127	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	128	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	129	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	130	()>.
130	all requests have been handled yet). The race is taken care of because	131
131	libev resets/rearms the async watcher before calling your callback,	132	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)	133	(i.e. it returns C<-1>) then you should start an idle watcher that calls
133	spurious wake-ups, but is generally harmless.	134	C<eio_poll> until it returns something C<!= -1>.
		135
		136	A full-featured connector between libeio and libev would look as follows
		137	(if C<eio_poll> is handling all requests, it can of course be simplified a
		138	lot by removing the idle watcher logic):
		139
		140	static struct ev_loop *loop;
		141	static ev_idle repeat_watcher;
		142	static ev_async ready_watcher;
		143
		144	/* idle watcher callback, only used when eio_poll */
		145	/* didn't handle all results in one call */
		146	static void
		147	repeat (EV_P_ ev_idle *w, int revents)
		148	{
		149	if (eio_poll () != -1)
		150	ev_idle_stop (EV_A_ w);
		151	}
		152
		153	/* eio has some results, process them */
		154	static void
		155	ready (EV_P_ ev_async *w, int revents)
		156	{
		157	if (eio_poll () == -1)
		158	ev_idle_start (EV_A_ &repeat_watcher);
		159	}
		160
		161	/* wake up the event loop */
		162	static void
		163	want_poll (void)
		164	{
		165	ev_async_send (loop, &ready_watcher)
		166	}
		167
		168	void
		169	my_init_eio ()
		170	{
		171	loop = EV_DEFAULT;
		172
		173	ev_idle_init (&repeat_watcher, repeat);
		174	ev_async_init (&ready_watcher, ready);
		175	ev_async_start (loop &watcher);
		176
		177	eio_init (want_poll, 0);
		178	}
134		179
135	For most other event loops, you would typically use a pipe - the event	180	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	181	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	182	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	183	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	184	C<eio_poll>.
140	your callback again and again until the byte has been read (as the pipe	185
141	read callback does not read it, only C<done_poll>).	186	You don't have to take special care in the case C<eio_poll> doesn't handle
		187	all requests, as the done callback will not be invoked, so the event loop
		188	will still signal readiness for the pipe until I<all> results have been
		189	processed.
142		190
143		191
144	=head1 HIGH LEVEL REQUEST API	192	=head1 HIGH LEVEL REQUEST API
145		193
146	Libeio has both a high-level API, which consists of calling a request	194	Libeio has both a high-level API, which consists of calling a request
…		…
153		201
154	You submit a request by calling the relevant C<eio_TYPE> function with the	202	You submit a request by calling the relevant C<eio_TYPE> function with the
155	required parameters, a callback of type C<int (eio_cb)(eio_req req)>	203	required parameters, a callback of type C<int (eio_cb)(eio_req req)>
156	(called C<eio_cb> below) and a freely usable C<void *data> argument.	204	(called C<eio_cb> below) and a freely usable C<void *data> argument.
157		205
158	The return value will either be 0	206	The return value will either be 0, in case something went really wrong
		207	(which can basically only happen on very fatal errors, such as C<malloc>
		208	returning 0, which is rather unlikely), or a pointer to the newly-created
		209	and submitted C<eio_req *>.
159		210
160	The callback will be called with an C<eio_req *> which contains the	211	The callback will be called with an C<eio_req *> which contains the
161	results of the request. The members you can access inside that structure	212	results of the request. The members you can access inside that structure
162	vary from request to request, except for:	213	vary from request to request, except for:
163		214
…		…
210	}	261	}
211		262
212	/* the first three arguments are passed to open(2) */	263	/* the first three arguments are passed to open(2) */
213	/* the remaining are priority, callback and data */	264	/* the remaining are priority, callback and data */
214	if (!eio_open ("/etc/passwd", O_RDONLY, 0, 0, file_open_done, 0))	265	if (!eio_open ("/etc/passwd", O_RDONLY, 0, 0, file_open_done, 0))
215	abort (); /* something ent wrong, we will all die!!! */	266	abort (); /* something went wrong, we will all die!!! */
216		267
217	Note that you additionally need to call C<eio_poll> when the C<want_cb>	268	Note that you additionally need to call C<eio_poll> when the C<want_cb>
218	indicates that requests are ready to be processed.	269	indicates that requests are ready to be processed.
		270
		271	=head2 CANCELLING REQUESTS
		272
		273	Sometimes the need for a request goes away before the request is
		274	finished. In that case, one can cancel the request by a call to
		275	C<eio_cancel>:
		276
		277	=over 4
		278
		279	=item eio_cancel (eio_req *req)
		280
		281	Cancel the request (and all its subrequests). If the request is currently
		282	executing it might still continue to execute, and in other cases it might
		283	still take a while till the request is cancelled.
		284
		285	Even if cancelled, the finish callback will still be invoked - the
		286	callbacks of all cancellable requests need to check whether the request
		287	has been cancelled by calling C<EIO_CANCELLED (req)>:
		288
		289	static int
		290	my_eio_cb (eio_req *req)
		291	{
		292	if (EIO_CANCELLED (req))
		293	return 0;
		294	}
		295
		296	In addition, cancelled requests will I<either> have C<< req->result >>
		297	set to C<-1> and C<errno> to C<ECANCELED>, or I<otherwise> they were
		298	successfully executed, despite being cancelled (e.g. when they have
		299	already been executed at the time they were cancelled).
		300
		301	C<EIO_CANCELLED> is still true for requests that have successfully
		302	executed, as long as C<eio_cancel> was called on them at some point.
		303
		304	=back
219		305
220	=head2 AVAILABLE REQUESTS	306	=head2 AVAILABLE REQUESTS
221		307
222	The following request functions are available. I<All> of them return the	308	The following request functions are available. I<All> of them return the
223	C<eio_req *> on success and C<0> on failure, and I<all> of them have the	309	C<eio_req *> on success and C<0> on failure, and I<all> of them have the
…		…
226	custom data value as C<data>.	312	custom data value as C<data>.
227		313
228	=head3 POSIX API WRAPPERS	314	=head3 POSIX API WRAPPERS
229		315
230	These requests simply wrap the POSIX call of the same name, with the same	316	These requests simply wrap the POSIX call of the same name, with the same
231	arguments. If a function is not implemented by the OS and cnanot be emulated	317	arguments. If a function is not implemented by the OS and cannot be emulated
232	in some way, then all of these return C<-1> and set C<errorno> to C<ENOSYS>.	318	in some way, then all of these return C<-1> and set C<errorno> to C<ENOSYS>.
233		319
234	=over 4	320	=over 4
235		321
236	=item eio_open (const char path, int flags, mode_t mode, int pri, eio_cb cb, void data)	322	=item eio_open (const char path, int flags, mode_t mode, int pri, eio_cb cb, void data)
…		…
317	char target = strndup ((char )req->ptr2, req->result);	403	char target = strndup ((char )req->ptr2, req->result);
318		404
319	free (target);	405	free (target);
320	}	406	}
321		407
		408	=item eio_realpath (const char path, int pri, eio_cb cb, void data)
		409
		410	Similar to the realpath libc function, but unlike that one, result is
		411	C<-1> on failure and the length of the returned path in C<ptr2> (which is
		412	not 0-terminated) - this is similar to readlink.
		413
322	=item eio_stat (const char path, int pri, eio_cb cb, void data)	414	=item eio_stat (const char path, int pri, eio_cb cb, void data)
323		415
324	=item eio_lstat (const char path, int pri, eio_cb cb, void data)	416	=item eio_lstat (const char path, int pri, eio_cb cb, void data)
325		417
326	=item eio_fstat (int fd, int pri, eio_cb cb, void *data)	418	=item eio_fstat (int fd, int pri, eio_cb cb, void *data)
327		419
328	Stats a file - if C<< req->result >> indicates success, then you can	420	Stats a file - if C<< req->result >> indicates success, then you can
329	access the C<struct stat>-like structure via C<< req->ptr2 >>:	421	access the C<struct stat>-like structure via C<< req->ptr2 >>:
330		422
331	EIO_STRUCT_STAT statdata = (EIO_STRUCT_STAT )req->ptr2;	423	EIO_STRUCT_STAT statdata = (EIO_STRUCT_STAT )req->ptr2;
332		424
333	=item eio_statvfs (const char path, int pri, eio_cb cb, void data)	425	=item eio_statvfs (const char path, int pri, eio_cb cb, void data)
334		426
335	=item eio_fstatvfs (int fd, int pri, eio_cb cb, void *data)	427	=item eio_fstatvfs (int fd, int pri, eio_cb cb, void *data)
336		428
337	Stats a filesystem - if C<< req->result >> indicates success, then you can	429	Stats a filesystem - if C<< req->result >> indicates success, then you can
338	access the C<struct statvfs>-like structure via C<< req->ptr2 >>:	430	access the C<struct statvfs>-like structure via C<< req->ptr2 >>:
339		431
340	EIO_STRUCT_STATVFS statdata = (EIO_STRUCT_STATVFS )req->ptr2;	432	EIO_STRUCT_STATVFS statdata = (EIO_STRUCT_STATVFS )req->ptr2;
341		433
342	=back	434	=back
343		435
344	=head3 READING DIRECTORIES	436	=head3 READING DIRECTORIES
345		437
346	Reading directories sounds simple, but can be rather demanding, especially	438	Reading directories sounds simple, but can be rather demanding, especially
347	if you want to do stuff such as traversing a diretcory hierarchy or	439	if you want to do stuff such as traversing a directory hierarchy or
348	processing all files in a directory. Libeio can assist thess complex tasks	440	processing all files in a directory. Libeio can assist these complex tasks
349	with it's C<eio_readdir> call.	441	with it's C<eio_readdir> call.
350		442
351	=over 4	443	=over 4
352		444
353	=item eio_readdir (const char path, int flags, int pri, eio_cb cb, void data)	445	=item eio_readdir (const char path, int flags, int pri, eio_cb cb, void data)
…		…
385		477
386	If this flag is specified, then, in addition to the names in C<ptr2>,	478	If this flag is specified, then, in addition to the names in C<ptr2>,
387	also an array of C<struct eio_dirent> is returned, in C<ptr1>. A C<struct	479	also an array of C<struct eio_dirent> is returned, in C<ptr1>. A C<struct
388	eio_dirent> looks like this:	480	eio_dirent> looks like this:
389		481
390	struct eio_dirent	482	struct eio_dirent
391	{	483	{
392	int nameofs; /* offset of null-terminated name string in (char )req->ptr2 /	484	int nameofs; /* offset of null-terminated name string in (char )req->ptr2 /
393	unsigned short namelen; /* size of filename without trailing 0 */	485	unsigned short namelen; /* size of filename without trailing 0 */
394	unsigned char type; /* one of EIO_DT_* */	486	unsigned char type; /* one of EIO_DT_* */
395	signed char score; /* internal use */	487	signed char score; /* internal use */
396	ino_t inode; /* the inode number, if available, otherwise unspecified */	488	ino_t inode; /* the inode number, if available, otherwise unspecified */
397	};	489	};
398		490
399	The only members you normally would access are C<nameofs>, which is the	491	The only members you normally would access are C<nameofs>, which is the
400	byte-offset from C<ptr2> to the start of the name, C<namelen> and C<type>.	492	byte-offset from C<ptr2> to the start of the name, C<namelen> and C<type>.
401		493
402	C<type> can be one of:	494	C<type> can be one of:
…		…
445	When this flag is specified, then the names will be returned in an order	537	When this flag is specified, then the names will be returned in an order
446	suitable for stat()'ing each one. That is, when you plan to stat()	538	suitable for stat()'ing each one. That is, when you plan to stat()
447	all files in the given directory, then the returned order will likely	539	all files in the given directory, then the returned order will likely
448	be fastest.	540	be fastest.
449		541
450	If both this flag and C<EIO_READDIR_DIRS_FIRST> are specified, then	542	If both this flag and C<EIO_READDIR_DIRS_FIRST> are specified, then the
451	the likely dirs come first, resulting in a less optimal stat order.	543	likely directories come first, resulting in a less optimal stat order.
452		544
453	=item EIO_READDIR_FOUND_UNKNOWN	545	=item EIO_READDIR_FOUND_UNKNOWN
454		546
455	This flag should not be specified when calling C<eio_readdir>. Instead,	547	This flag should not be specified when calling C<eio_readdir>. Instead,
456	it is being set by C<eio_readdir> (you can access the C<flags> via C<<	548	it is being set by C<eio_readdir> (you can access the C<flags> via C<<
457	req->int1 >>, when any of the C<type>'s found were C<EIO_DT_UNKNOWN>. The	549	req->int1 >>, when any of the C<type>'s found were C<EIO_DT_UNKNOWN>. The
458	absense of this flag therefore indicates that all C<type>'s are known,	550	absence of this flag therefore indicates that all C<type>'s are known,
459	which can be used to speed up some algorithms.	551	which can be used to speed up some algorithms.
460		552
461	A typical use case would be to identify all subdirectories within a	553	A typical use case would be to identify all subdirectories within a
462	directory - you would ask C<eio_readdir> for C<EIO_READDIR_DIRS_FIRST>. If	554	directory - you would ask C<eio_readdir> for C<EIO_READDIR_DIRS_FIRST>. If
463	then this flag is I<NOT> set, then all the entries at the beginning of the	555	then this flag is I<NOT> set, then all the entries at the beginning of the
…		…
501	as calling C<fdatasync>.	593	as calling C<fdatasync>.
502		594
503	Flags can be any combination of C<EIO_SYNC_FILE_RANGE_WAIT_BEFORE>,	595	Flags can be any combination of C<EIO_SYNC_FILE_RANGE_WAIT_BEFORE>,
504	C<EIO_SYNC_FILE_RANGE_WRITE> and C<EIO_SYNC_FILE_RANGE_WAIT_AFTER>.	596	C<EIO_SYNC_FILE_RANGE_WRITE> and C<EIO_SYNC_FILE_RANGE_WAIT_AFTER>.
505		597
		598	=item eio_fallocate (int fd, int mode, off_t offset, off_t len, int pri, eio_cb cb, void *data)
		599
		600	Calls C<fallocate> (note: I<NOT> C<posix_fallocate>!). If the syscall is
		601	missing, then it returns failure and sets C<errno> to C<ENOSYS>.
		602
		603	The C<mode> argument can be C<0> (for behaviour similar to
		604	C<posix_fallocate>), or C<EIO_FALLOC_FL_KEEP_SIZE>, which keeps the size
		605	of the file unchanged (but still preallocates space beyond end of file).
		606
506	=back	607	=back
507		608
508	=head3 LIBEIO-SPECIFIC REQUESTS	609	=head3 LIBEIO-SPECIFIC REQUESTS
509		610
510	These requests are specific to libeio and do not correspond to any OS call.	611	These requests are specific to libeio and do not correspond to any OS call.
…		…
551		652
552	eio_custom (my_open, 0, my_open_done, "/etc/passwd");	653	eio_custom (my_open, 0, my_open_done, "/etc/passwd");
553		654
554	=item eio_busy (eio_tstamp delay, int pri, eio_cb cb, void *data)	655	=item eio_busy (eio_tstamp delay, int pri, eio_cb cb, void *data)
555		656
556	This is a a request that takes C<delay> seconds to execute, but otherwise	657	This is a request that takes C<delay> seconds to execute, but otherwise
557	does nothing - it simply puts one of the worker threads to sleep for this	658	does nothing - it simply puts one of the worker threads to sleep for this
558	long.	659	long.
559		660
560	This request can be used to artificially increase load, e.g. for debugging	661	This request can be used to artificially increase load, e.g. for debugging
561	or benchmarking reasons.	662	or benchmarking reasons.
…		…
568		669
569	=back	670	=back
570		671
571	=head3 GROUPING AND LIMITING REQUESTS	672	=head3 GROUPING AND LIMITING REQUESTS
572		673
		674	There is one more rather special request, C<eio_grp>. It is a very special
		675	aio request: Instead of doing something, it is a container for other eio
		676	requests.
		677
		678	There are two primary use cases for this: a) bundle many requests into a
		679	single, composite, request with a definite callback and the ability to
		680	cancel the whole request with its subrequests and b) limiting the number
		681	of "active" requests.
		682
		683	Further below you will find more discussion of these topics - first
		684	follows the reference section detailing the request generator and other
		685	methods.
		686
		687	=over 4
		688
		689	=item eio_req grp = eio_grp (eio_cb cb, void data)
		690
		691	Creates, submits and returns a group request.
		692
		693	=item eio_grp_add (eio_req grp, eio_req req)
		694
		695	Adds a request to the request group.
		696
		697	=item eio_grp_cancel (eio_req *grp)
		698
		699	Cancels all requests I<in> the group, but I<not> the group request
		700	itself. You can cancel the group request via a normal C<eio_cancel> call.
		701
		702
		703
		704	=back
		705
		706
		707
573	#TODO	708	#TODO
574		709
575	/*****************************************************************************/	710	/*****************************************************************************/
576	/* groups */	711	/* groups */
577		712
578	eio_req eio_grp (eio_cb cb, void data);	713	eio_req eio_grp (eio_cb cb, void data);
579	void eio_grp_feed (eio_req grp, void (feed)(eio_req *req), int limit);	714	void eio_grp_feed (eio_req grp, void (feed)(eio_req *req), int limit);
580	void eio_grp_limit (eio_req *grp, int limit);	715	void eio_grp_limit (eio_req *grp, int limit);
581	void eio_grp_add (eio_req grp, eio_req req);
582	void eio_grp_cancel (eio_req grp); / cancels all sub requests but not the group */	716	void eio_grp_cancel (eio_req grp); / cancels all sub requests but not the group */
583		717
584		718
585	=back	719	=back
586		720
…		…
593	=head1 ANATOMY AND LIFETIME OF AN EIO REQUEST	727	=head1 ANATOMY AND LIFETIME OF AN EIO REQUEST
594		728
595	A request is represented by a structure of type C<eio_req>. To initialise	729	A request is represented by a structure of type C<eio_req>. To initialise
596	it, clear it to all zero bytes:	730	it, clear it to all zero bytes:
597		731
598	eio_req req;	732	eio_req req;
599		733
600	memset (&req, 0, sizeof (req));	734	memset (&req, 0, sizeof (req));
601		735
602	A more common way to initialise a new C<eio_req> is to use C<calloc>:	736	A more common way to initialise a new C<eio_req> is to use C<calloc>:
603		737
604	eio_req req = calloc (1, sizeof (req));	738	eio_req req = calloc (1, sizeof (req));
605		739
606	In either case, libeio neither allocates, initialises or frees the	740	In either case, libeio neither allocates, initialises or frees the
607	C<eio_req> structure for you - it merely uses it.	741	C<eio_req> structure for you - it merely uses it.
608		742
609	zero	743	zero
…		…
627	for example, in interactive programs, you might want to limit this time to	761	for example, in interactive programs, you might want to limit this time to
628	C<0.01> seconds or so.	762	C<0.01> seconds or so.
629		763
630	Note that:	764	Note that:
631		765
		766	=over 4
		767
632	a) libeio doesn't know how long your request callbacks take, so the time	768	=item a) libeio doesn't know how long your request callbacks take, so the
633	spent in C<eio_poll> is up to one callback invocation longer then this	769	time spent in C<eio_poll> is up to one callback invocation longer then
634	interval.	770	this interval.
635		771
636	b) this is implemented by calling C<gettimeofday> after each request,	772	=item b) this is implemented by calling C<gettimeofday> after each
637	which can be costly.	773	request, which can be costly.
638		774
639	c) at least one request will be handled.	775	=item c) at least one request will be handled.
		776
		777	=back
640		778
641	=item eio_set_max_poll_reqs (unsigned int nreqs)	779	=item eio_set_max_poll_reqs (unsigned int nreqs)
642		780
643	When C<nreqs> is non-zero, then C<eio_poll> will not handle more than	781	When C<nreqs> is non-zero, then C<eio_poll> will not handle more than
644	C<nreqs> requests per invocation. This is a less costly way to limit the	782	C<nreqs> requests per invocation. This is a less costly way to limit the

Diff Legend

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

Comparing libeio/eio.pod (file contents): Revision 1.10 by root, Sun Jun 5 23:22:04 2011 UTC vs. Revision 1.21 by root, Thu Jul 7 22:36:18 2011 UTC

Diff Legend

Comparing libeio/eio.pod (file contents):
Revision 1.10 by root, Sun Jun 5 23:22:04 2011 UTC vs.
Revision 1.21 by root, Thu Jul 7 22:36:18 2011 UTC