[ViewVC] Diff of: cvs/OpenCL/OpenCL.pm

Comparing OpenCL/OpenCL.pm (file contents):
Revision 1.51 by root, Tue Apr 24 13:45:58 2012 UTC vs.
Revision 1.56 by root, Tue Apr 24 23:53:12 2012 UTC

…		…
376		376
377	For this to work, the OpenGL library must be loaded, a GLX context must	377	For this to work, the OpenGL library must be loaded, a GLX context must
378	have been created and be made current, and C<dlsym> must be available and	378	have been created and be made current, and C<dlsym> must be available and
379	capable of finding the function via C<RTLD_DEFAULT>.	379	capable of finding the function via C<RTLD_DEFAULT>.
380		380
		381	=head2 EVENT SYSTEM
		382
		383	OpenCL can generate a number of (potentially) asynchronous events, for
		384	example, after compiling a program, to signal a context-related error or,
		385	perhaps most important, to signal completion of queued jobs (by setting
		386	callbacks on OpenCL::Event objects).
		387
		388	To facilitate this, this module maintains an event queue - each
		389	time an asynchronous event happens, it is queued, and perl will be
		390	interrupted. This is implemented via the L<Async::Interrupt> module. In
		391	addition, this module has L<AnyEvent> support, so it can seamlessly
		392	integrate itself into many event loops.
		393
		394	Since this module is a bit hard to understand, here are some case examples:
		395
		396	=head3 Don't use callbacks.
		397
		398	When your program never uses any callbacks, then there will never be any
		399	notifications you need to take care of, and therefore no need to worry
		400	about all this.
		401
		402	You can achieve a great deal by explicitly waiting for events, or using
		403	barriers and flush calls. In many programs, there is no need at all to
		404	tinker with asynchronous events.
		405
		406	=head3 Use AnyEvent
		407
		408	This module automatically registers a watcher that invokes all outstanding
		409	event callbacks when AnyEvent is initialised (and block asynchronous
		410	interruptions). Using this mode of operations is the safest and most
		411	recommended one.
		412
		413	To use this, simply use AnyEvent and this module normally, make sure you
		414	have an event loop running:
		415
		416	use Gtk2 -init;
		417	use AnyEvent;
		418
		419	# initialise AnyEvent, by creating a watcher, or:
		420	AnyEvent::detect;
		421
		422	my $e = $queue->enqueue_marker;
		423	$e->cb (sub {
		424	warn "opencl is finished\n";
		425	})
		426
		427	main Gtk2;
		428
		429	Note that this module will not initialise AnyEvent for you. Before
		430	AnyEvent is initialised, the module will asynchronously interrupt perl
		431	instead. To avoid any surprises, it's best to explicitly initialise
		432	AnyEvent.
		433
		434	You can temporarily enable asynchronous interruptions (see next paragraph)
		435	by calling C<$OpenCL::INTERRUPT->unblock> and disable them again by
		436	calling C<$OpenCL::INTERRUPT->block>.
		437
		438	=head3 Let yourself be interrupted at any time
		439
		440	This mode is the default unless AnyEvent is loaded and initialised. In
		441	this mode, OpenCL asynchronously interrupts a running perl program. The
		442	emphasis is on both I<asynchronously> and I<running> here.
		443
		444	Asynchronously means that perl might execute your callbacks at any
		445	time. For example, in the following code (I<THAT YOU SHOULD NOT COPY>),
		446	the C<until> loop following the marker call will be interrupted by the
		447	callback:
		448
		449	my $e = $queue->enqueue_marker;
		450	my $flag;
		451	$e->cb (sub { $flag = 1 });
		452	1 until $flag;
		453	# $flag is now 1
		454
		455	The reason why you shouldn't blindly copy the above code is that
		456	busy waiting is a really really bad thing, and really really bad for
		457	performance.
		458
		459	While at first this asynchronous business might look exciting, it can be
		460	really hard, because you need to be prepared for the callback code to be
		461	executed at any time, which limits the amount of things the callback code
		462	can do safely.
		463
		464	This can be mitigated somewhat by using C<<
		465	$OpenCL::INTERRUPT->scope_block >> (see the L<Async::Interrupt>
		466	documentation for details).
		467
		468	The other problem is that your program must be actively I<running> to be
		469	interrupted. When you calculate stuff, your program is running. When you
		470	hang in some C functions or other block execution (by calling C<sleep>,
		471	C<select>, running an event loop and so on), your program is waiting, not
		472	running.
		473
		474	One way around that would be to attach a read watcher to your event loop,
		475	listening for events on C<< $OpenCL::INTERRUPT->pipe_fileno >>, using a
		476	dummy callback (C<sub { }>) to temporarily execute some perl code.
		477
		478	That is then awfully close to using the built-in AnyEvent support above,
		479	though, so consider that one instead.
		480
		481	=head3 Be creative
		482
		483	OpenCL exports the L<Async::Interrupt> object it uses in the global
		484	variable C<$OpenCL::INTERRUPT>. You can configure it in any way you like.
		485
		486	So if you want to feel like a real pro, err, wait, if you feel no risk
		487	menas no fun, you can experiment by implementing your own mode of
		488	operations.
		489
		490	=cut
		491
		492	package OpenCL;
		493
		494	use common::sense;
		495	use Async::Interrupt ();
		496
		497	our $POLL_FUNC; # set by XS
		498
		499	BEGIN {
		500	our $VERSION = '0.97';
		501
		502	require XSLoader;
		503	XSLoader::load (__PACKAGE__, $VERSION);
		504
		505	@OpenCL::Platform::ISA =
		506	@OpenCL::Device::ISA =
		507	@OpenCL::Context::ISA =
		508	@OpenCL::Queue::ISA =
		509	@OpenCL::Memory::ISA =
		510	@OpenCL::Sampler::ISA =
		511	@OpenCL::Program::ISA =
		512	@OpenCL::Kernel::ISA =
		513	@OpenCL::Event::ISA = OpenCL::Object::;
		514
		515	@OpenCL::Buffer::ISA =
		516	@OpenCL::Image::ISA = OpenCL::Memory::;
		517
		518	@OpenCL::BufferObj::ISA = OpenCL::Buffer::;
		519
		520	@OpenCL::Image2D::ISA =
		521	@OpenCL::Image3D::ISA =
		522	@OpenCL::Image2DArray::ISA =
		523	@OpenCL::Image1D::ISA =
		524	@OpenCL::Image1DArray::ISA =
		525	@OpenCL::Image1DBuffer::ISA = OpenCL::Image::;
		526
		527	@OpenCL::UserEvent::ISA = OpenCL::Event::;
		528	}
		529
381	=head2 THE OpenCL PACKAGE	530	=head2 THE OpenCL PACKAGE
382		531
383	=over 4	532	=over 4
384		533
385	=item $int = OpenCL::errno	534	=item $int = OpenCL::errno
…		…
403		552
404	Returns all available OpenCL::Platform objects.	553	Returns all available OpenCL::Platform objects.
405		554
406	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformIDs.html>	555	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetPlatformIDs.html>
407		556
408	=item $ctx = OpenCL::context_from_type $properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $notify = undef	557	=item $ctx = OpenCL::context_from_type $properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $callback->($err, $pvt) = $print_stderr
409		558
410	Tries to create a context from a default device and platform - never worked for me.	559	Tries to create a context from a default device and platform type - never worked for me.
411		560
412	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html>	561	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html>
413		562
		563	=item $ctx = OpenCL::context $properties, \@devices, $callback->($err, $pvt) = $print_stderr)
		564
		565	Create a new OpenCL::Context object using the given device object(s). This
		566	function isn't implemented yet, use C<< $platform->context >> instead.
		567
		568	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html>
		569
414	=item OpenCL::wait_for_events $wait_events...	570	=item OpenCL::wait_for_events $wait_events...
415		571
416	Waits for all events to complete.	572	Waits for all events to complete.
417		573
418	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html>	574	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html>
419		575
		576	=item OpenCL::poll
		577
		578	Checks if there are any outstanding events (see L<EVENT SYSTEM>) and
		579	invokes their callbacks.
		580
		581	=item $OpenCL::INTERRUPT
		582
		583	The L<Async::Interrupt> object used to signal asynchronous events (see
		584	L<EVENT SYSTEM>).
		585
		586	=cut
		587
		588	our $INTERRUPT = new Async::Interrupt c_cb => [$POLL_FUNC, 0];
		589
		590	&_eq_initialise ($INTERRUPT->signal_func);
		591
		592	=item $OpenCL::WATCHER
		593
		594	The L<AnyEvent> watcher object used to watch for asynchronous events (see
		595	L<EVENT SYSTEM>). This variable is C<undef> until L<AnyEvent> has been
		596	loaded I<and> initialised (e.g. by calling C<AnyEvent::detect>).
		597
		598	=cut
		599
		600	our $WATCHER;
		601
		602	sub _init_anyevent {
		603	$INTERRUPT->block;
		604	$WATCHER = AE::io ($INTERRUPT->pipe_fileno, 0, sub { $INTERRUPT->handle });
		605	}
		606
		607	if (defined $AnyEvent::MODEL) {
		608	_init_anyevent;
		609	} else {
		610	push @AnyEvent::post_detect, \&_init_anyevent;
		611	}
		612
420	=back	613	=back
421		614
		615	=head2 THE OpenCL::Object CLASS
		616
		617	This is the base class for all objects in the OpenCL module. The only
		618	method it implements is the C<id> method, which is only useful if you want
		619	to interface to OpenCL on the C level.
		620
		621	=over 4
		622
		623	=item $iv = $obj->id
		624
		625	OpenCL objects are represented by pointers or integers on the C level. If
		626	you want to interface to an OpenCL object directly on the C level, then
		627	you need this value, which is returned by this method. You should use an
		628	C<IV> type in your code and cast that to the correct type.
		629
		630	=cut
		631
		632	sub OpenCL::Object::id {
		633	ref $_[0] eq "SCALAR"
		634	? ${ $_[0] }
		635	: $_[0][0]
		636	}
		637
		638	=back
		639
422	=head2 THE OpenCL::Platform CLASS	640	=head2 THE OpenCL::Platform CLASS
423		641
424	=over 4	642	=over 4
425		643
426	=item @devices = $platform->devices ($type = OpenCL::DEVICE_TYPE_ALL)	644	=item @devices = $platform->devices ($type = OpenCL::DEVICE_TYPE_ALL)
427		645
428	Returns a list of matching OpenCL::Device objects.	646	Returns a list of matching OpenCL::Device objects.
429		647
430	=item $ctx = $platform->context_from_type ($properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $notify = undef)	648	=item $ctx = $platform->context_from_type ($properties, $type = OpenCL::DEVICE_TYPE_DEFAULT, $callback->($err, $pvt) = $print_stderr)
431		649
432	Tries to create a context. Never worked for me, and you need devices explicitly anyway.	650	Tries to create a context. Never worked for me, and you need devices explicitly anyway.
433		651
434	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html>	652	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContextFromType.html>
435		653
436	=item $ctx = $platform->context ($properties = undef, @$devices, $notify = undef)	654	=item $ctx = $platform->context ($properties, \@devices, $callback->($err, $pvt) = $print_stderr)
437		655
438	Create a new OpenCL::Context object using the given device object(s)- a	656	Create a new OpenCL::Context object using the given device object(s)- a
439	CL_CONTEXT_PLATFORM property is supplied automatically.	657	CL_CONTEXT_PLATFORM property is supplied automatically.
440		658
441	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html>	659	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clCreateContext.html>
…		…
781	=cut	999	=cut
782		1000
783	sub OpenCL::Context::build_program {	1001	sub OpenCL::Context::build_program {
784	my ($self, $prog, $options) = @_;	1002	my ($self, $prog, $options) = @_;
785		1003
		1004	require Carp;
		1005
786	$prog = $self->program_with_source ($prog)	1006	$prog = $self->program_with_source ($prog)
787	unless ref $prog;	1007	unless ref $prog;
788		1008
		1009	# we build separately per device so we instantly know which one failed
789	for my $dev ($self->devices) {	1010	for my $dev ($self->devices) {
790	eval { $prog->build ($dev, $options); 1 }	1011	eval { $prog->build ([$dev], $options); 1 }
791	or Carp::croak "Building OpenCL program for device '" . $dev->name . "' failed:\n"	1012	or Carp::croak ("Building OpenCL program for device '" . $dev->name . "' failed:\n"
792	. $prog->build_log ($dev);	1013	. $prog->build_log ($dev));
793	}	1014	}
794		1015
795	$prog	1016	$prog
796	}	1017	}
797		1018
…		…
1000		1221
1001	Yeah.	1222	Yeah.
1002		1223
1003	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferToImage.html>.	1224	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueCopyBufferToImage.html>.
1004		1225
		1226	=item $ev = $queue->enqueue_fill_buffer ($mem, $pattern, $offset, $size, ...)
		1227
		1228	Fills the given buffer object with repeated applications of C<$pattern>,
		1229	starting at C<$offset> for C<$size> octets.
		1230
		1231	L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueFillBuffer.html>
		1232
		1233	=item $ev = $queue->enqueue_fill_image ($img, $r, $g, $b, $a, $x, $y, $z, $width, $height, $depth, ...)
		1234
		1235	Fills the given image area with the given rgba colour components. The
		1236	components are normally floating point values between C<0> and C<1>,
		1237	except when the image channel data type is a signe dor unsigned
		1238	unnormalised format, in which case the range is determined by the format.
		1239
		1240	L<http://www.khronos.org/registry/cl/sdk/1.2/docs/man/xhtml/clEnqueueFillImage.html>
		1241
1005	=item $ev = $queue->enqueue_task ($kernel, $wait_events...)	1242	=item $ev = $queue->enqueue_task ($kernel, $wait_events...)
1006		1243
1007	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueTask.html>	1244	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clEnqueueTask.html>
1008		1245
1009	=item $ev = $queue->enqueue_nd_range_kernel ($kernel, @$global_work_offset, @$global_work_size, @$local_work_size, $wait_events...)	1246	=item $ev = $queue->enqueue_nd_range_kernel ($kernel, @$global_work_offset, @$global_work_size, @$local_work_size, $wait_events...)
…		…
1177	OpenCL::Image1DArray, OpenCL::Image1DBuffer, OpenCL::Image2D,	1414	OpenCL::Image1DArray, OpenCL::Image1DBuffer, OpenCL::Image2D,
1178	OpenCL::Image2DArray and OpenCL::Image3D.	1415	OpenCL::Image2DArray and OpenCL::Image3D.
1179		1416
1180	=over 4	1417	=over 4
1181		1418
1182	=item $packed_value = $ev->image_info ($name)	1419	=item $packed_value = $image->image_info ($name)
1183		1420
1184	See C<< $platform->info >> for details.	1421	See C<< $platform->info >> for details.
1185		1422
1186	The reason this method is not called C<info> is that there already is an	1423	The reason this method is not called C<info> is that there already is an
1187	C<< ->info >> method inherited from C<OpenCL::Memory>.	1424	C<< ->info >> method inherited from C<OpenCL::Memory>.
1188		1425
1189	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetImageInfo.html>	1426	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetImageInfo.html>
1190		1427
		1428	=item ($channel_order, $channel_data_type) = $image->format
		1429
		1430	Returns the channel order and type used to create the image by calling
		1431	C<clGetImageInfo> with C<CL_IMAGE_FORMAT>.
		1432
1191	=for gengetinfo begin image	1433	=for gengetinfo begin image
1192		1434
1193	=item $int = $image->element_size	1435	=item $int = $image->element_size
1194		1436
1195	Calls C<clGetImageInfo> with C<CL_IMAGE_ELEMENT_SIZE> and returns the result.	1437	Calls C<clGetImageInfo> with C<CL_IMAGE_ELEMENT_SIZE> and returns the result.
…		…
1268		1510
1269	=head2 THE OpenCL::Program CLASS	1511	=head2 THE OpenCL::Program CLASS
1270		1512
1271	=over 4	1513	=over 4
1272		1514
1273	=item $program->build ($device, $options = "")	1515	=item $program->build (\@devices = undef, $options = "", $cb->($program) = undef)
1274		1516
1275	Tries to build the program with the given options. See also the	1517	Tries to build the program with the given options. See also the
1276	C<$ctx->build> convenience function.	1518	C<$ctx->build> convenience function.
1277		1519
		1520	If a callback is specified, then it will be called when compilation is
		1521	finished. Note that many OpenCL implementations block your program while
		1522	compiling whether you use a callback or not. See C<build_async> if you
		1523	want to make sure the build is done in the background.
		1524
		1525	Note that some OpenCL implementations atc up badly, and don't call the
		1526	callback in some error cases (but call it in others). This implementation
		1527	assumes the callback will always be called, and leaks memory if this is
		1528	not so. So best make sure you don't pass in invalid values.
		1529
1278	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clBuildProgram.html>	1530	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clBuildProgram.html>
		1531
		1532	=item $program->build_async (\@devices = undef, $options = "", $cb->($program) = undef)
		1533
		1534	Similar to C<< ->build >>, except it starts a thread, and never fails (you
		1535	need to check the compilation status form the callback, or by polling).
1279		1536
1280	=item $packed_value = $program->build_info ($device, $name)	1537	=item $packed_value = $program->build_info ($device, $name)
1281		1538
1282	Similar to C<< $platform->info >>, but returns build info for a previous	1539	Similar to C<< $platform->info >>, but returns build info for a previous
1283	build attempt for the given device.	1540	build attempt for the given device.
…		…
1460		1717
1461	Waits for the event to complete.	1718	Waits for the event to complete.
1462		1719
1463	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html>	1720	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clWaitForEvents.html>
1464		1721
		1722	=item $ev->cb ($exec_callback_type, $callback->($event, $event_command_exec_status))
		1723
		1724	Adds a callback to the callback stack for the given event type. There is
		1725	no way to remove a callback again.
		1726
		1727	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetEventCallback.html>
		1728
1465	=item $packed_value = $ev->info ($name)	1729	=item $packed_value = $ev->info ($name)
1466		1730
1467	See C<< $platform->info >> for details.	1731	See C<< $platform->info >> for details.
1468		1732
1469	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetEventInfo.html>	1733	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clGetEventInfo.html>
…		…
1529		1793
1530	=over 4	1794	=over 4
1531		1795
1532	=item $ev->set_status ($execution_status)	1796	=item $ev->set_status ($execution_status)
1533		1797
		1798	Sets the execution status of the user event. Can only be called once,
		1799	either with OpenCL::COMPLETE or a negative number as status.
		1800
1534	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetUserEventStatus.html>	1801	L<http://www.khronos.org/registry/cl/sdk/1.1/docs/man/xhtml/clSetUserEventStatus.html>
1535		1802
1536	=back	1803	=back
1537		1804
1538	=cut	1805	=cut
1539
1540	package OpenCL;
1541
1542	use common::sense;
1543
1544	BEGIN {
1545	our $VERSION = '0.96';
1546
1547	require XSLoader;
1548	XSLoader::load (__PACKAGE__, $VERSION);
1549
1550	@OpenCL::Buffer::ISA =
1551	@OpenCL::Image::ISA = OpenCL::Memory::;
1552
1553	@OpenCL::BufferObj::ISA = OpenCL::Buffer::;
1554
1555	@OpenCL::Image2D::ISA =
1556	@OpenCL::Image3D::ISA =
1557	@OpenCL::Image2DArray::ISA =
1558	@OpenCL::Image1D::ISA =
1559	@OpenCL::Image1DArray::ISA =
1560	@OpenCL::Image1DBuffer::ISA = OpenCL::Image::;
1561
1562	@OpenCL::UserEvent::ISA = OpenCL::Event::;
1563	}
1564		1806
1565	1;	1807	1;
1566		1808
1567	=head1 AUTHOR	1809	=head1 AUTHOR
1568		1810

Diff Legend

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

Comparing OpenCL/OpenCL.pm (file contents): Revision 1.51 by root, Tue Apr 24 13:45:58 2012 UTC vs. Revision 1.56 by root, Tue Apr 24 23:53:12 2012 UTC

Diff Legend

Comparing OpenCL/OpenCL.pm (file contents):
Revision 1.51 by root, Tue Apr 24 13:45:58 2012 UTC vs.
Revision 1.56 by root, Tue Apr 24 23:53:12 2012 UTC