Commit graph

765 commits

Author SHA1 Message Date
Wim Taymans
62b1a5a7be pad: add 2 new caps methods
Add method to get the currently configured caps on the pad.
Add a method to check if caps are configured on a pad.
2011-05-06 19:03:33 +02:00
Wim Taymans
b35a700d3e pad: implement fixed caps with an object flag
Implement fixed caps with an object flag instead of a custom getcaps function.
2011-05-06 17:59:33 +02:00
Wim Taymans
92b4c8967b pad: don't use buffer caps for negotiation
Don't use the buffer caps for negotiation anymore but use the CAPS events.
Make the _set_caps method produce the CAPS event, add some backward
compatibility code to trigger the setcaps functions on src and sinkpads.
Remove all negotiation code from the chain functions.
Don't use the GST_PAD_CAPS variable anymore to store the caps but retrieve the
caps from the sticky event array.
2011-05-06 17:39:09 +02:00
Wim Taymans
4372869e80 pad: implement more sticky events
Remove the context again, adding an extra layer of refcounting and object
creation to manage an array is too complicated and inefficient. Use a simple
array again.
Also implement event updates when calling gst_pad_chain() and
gst_event_send_event() directly.
2011-05-06 11:14:02 +02:00
Wim Taymans
f663c9770b pad: simplify some more
If we get a context in the chain functions we always need to do a full update of
the context on the peer pad.
2011-05-05 19:24:46 +02:00
Wim Taymans
d3e0a37900 pad: improve context passing some more
Pass the context downstream when it got updated.
Have two ways of informing downstream of events, do a full context update when
the CONTEXT_PENDING flag is set and simply forward the event otherwise.
Set the CONTENT_PENDING flag when linking pads.
We don't need to old context anymore when updating the context of a pad.
2011-05-05 19:01:21 +02:00
Sebastian Dröge
292806f60e pad: Fix refcount leak of the parent in the default event dispatch function 2011-05-05 17:05:21 +02:00
Wim Taymans
dfef1d027d pad: improve passing around the context
Improve passing around the context, only send the context to the peer element
when the CONTEXT_PENDING flag is set.
2011-05-05 16:34:29 +02:00
Sebastian Dröge
98d1ee4a8f gst: Update everything for the new GstIterator API 2011-05-05 15:36:03 +02:00
Sebastian Dröge
5a7ec99214 pad: Keep track of reconfigure events and the pad-needs-reconfiguring status 2011-05-05 15:11:51 +02:00
Olivier Crête
40cff0eb6d pad: Send renegotiate event on link 2011-05-05 15:04:33 +02:00
Olivier Crête
581533cd00 pad: Drop renegotiate event if there is no getcaps function on a sink pad
If there is no custom getcaps function on a sink pad, then changes in
downstream caps will never be propagated, so there is no point in trying to
renegotiate the capabilities.
2011-05-05 15:04:25 +02:00
Wim Taymans
77569e0a9a pad: pass the context around
Pass the context from srcpad to sinkpad before dataflow when something
changed.
2011-05-05 13:17:08 +02:00
Wim Taymans
f16180a761 pad: update the context lazyly 2011-05-05 13:17:08 +02:00
Wim Taymans
c4751ec8c1 Revert "context: use context on buffers instead of caps"
This reverts commit 9ef1346b1f.

Way to much for one commit and I'm not sure we want to get rid of the pad caps
just like that. It's nice to have the buffer and its type in onw nice bundle
without having to drag the complete context with it.
2011-05-05 13:17:08 +02:00
Sebastian Dröge
c6e9ce64ae pad: Remove unnecessary FIXME
Resetting the result is not necessary when resyncing because
pads that previously got the event will be skipped and we
need to consider the results of the previous pushes.
2011-05-05 12:30:05 +02:00
Wim Taymans
9ef1346b1f context: use context on buffers instead of caps
Put the srcpad context on buffers instead of caps. This allows us to associate
all the relevant info contained in events with a buffer.
2011-05-04 18:59:47 +02:00
Wim Taymans
9136abf623 pad: use the context to store sticky events
Store the sticky events in the context of a source pad.
2011-05-04 15:53:21 +02:00
Wim Taymans
bc9cffda90 pad: more sticky events work
Copy the sticky events from the srcpad to the sinkpad when linking pads. Set the
STICKY_PENDING flag to make sure that the sticky events are dispatched before
pushing the next buffer to the element.
2011-05-03 18:58:18 +02:00
Wim Taymans
2243adffa1 event: add sticky flags to events
Add the sticky flag to events and a sticky index.
Keep sticky events in an array on each pad.
Remove GST_EVENT_SRC(), it is causing refcycles with sticky events, was not used
and is not very interesting anyway.
2011-05-02 18:34:18 +02:00
Wim Taymans
6ab7e6c15d Remove pad_alloc, this can now be done better
Remove pad_alloc and all references. This can now be done more efficiently and
more flexible with the ALLOCATION query and the bufferpool objects. There is no
reverse negotiation yet but that will be done with an event later.
2011-04-29 13:26:19 +02:00
Wim Taymans
8e76da20b6 Merge branch 'master' into 0.11 2011-04-19 19:30:28 +02:00
Havard Graff
547d451772 pad: unlock before freeing the cache to avoid deadlock
https://bugzilla.gnome.org/show_bug.cgi?id=648199
2011-04-19 15:15:34 +01:00
Sebastian Dröge
6e41a3cf2f pad: Make the size parameter of gst_pad_alloc_buffer() unsigned
Internally guints were used everywhere already.
2011-04-18 10:47:26 +02:00
Sebastian Dröge
9b2a8b8480 pad: Don't allow fixating ANY caps and remove FIXME 2011-04-18 10:41:18 +02:00
Sebastian Dröge
f51a23a83c Merge branch 'master' into 0.11 2011-04-16 08:59:58 +02:00
Robert Swain
fd5aa095da pad: Allow tracking of buffers in GST_SCHEDULING debug output
As GST_SCHEDULING reports when buffers pass through pads due to
gst_pad_push calls, they are a good way of tracking the progress of
buffers through pipelines. As such, adding output of the buffer pointers
to these messages allows tracking of specific buffers, easing debugging.
2011-04-11 11:05:24 +02:00
Wim Taymans
468ec5bc40 bufferlist: simplify bufferlists
We now have multiple memory blocks as part of the buffers and we can therefore
reduce the bufferlist to a simple array of buffers.
2011-03-31 17:51:02 +02:00
Wim Taymans
6da19ffeb2 buffer: more buffer updates 2011-03-29 11:07:36 +02:00
Wim Taymans
1eed0785af memory: more fixes
Fix span and is_span
Implement buffer memory
2011-03-28 20:08:45 +02:00
Sebastian Dröge
4e16347bfa Merge branch 'master' into 0.11
Conflicts:
	gst/gstbufferlist.c
2011-03-17 10:50:43 +01:00
Sebastian Dröge
b8d83f4e01 pad: Document that pad blocks only make sense for sink pads in pull mode and src pads in push mode
See bug #644907.
2011-03-16 12:01:21 +01:00
Wim Taymans
f505f778cc pad: more preroll lock to basesink
Move the preroll lock to basesink where it belongs.
2011-03-04 17:25:02 +01:00
Wim Taymans
509abb2162 pad: set boxed type correctly 2011-02-23 10:35:36 +01:00
Wim Taymans
6c18c9508d miniobject: more boxed type fixing
More miniobject fixing, leaks horribly somewhere..
2011-02-23 10:35:09 +01:00
Wim Taymans
317af67bc4 miniobject: make queries a boxed type
More minionject stuff.
2011-02-23 10:34:45 +01:00
Wim Taymans
238b9a57cc Merge branch 'master' into 0.11
Conflicts:
	configure.ac
	gst/gstelement.c
	gst/gstelement.h
	gst/gstpad.c
	gst/gstutils.c
	libs/gst/base/Makefile.am
	libs/gst/check/Makefile.am
	libs/gst/controller/Makefile.am
	libs/gst/dataprotocol/Makefile.am
	libs/gst/net/Makefile.am
	win32/common/libgstreamer.def
2011-02-22 14:11:59 +01:00
Wim Taymans
12d2d01663 pad: Check sinkpad for flushing
Check the sinkpad for the flushing state before calling the chainfunction on the
pad. We do this by checking the cache (which is also cleared on the srcpad when
the sink is set to flushing).

Fixes #641928
2011-02-14 17:31:25 +01:00
Tim-Philipp Müller
8a7fc1d8c9 Revert "micro-optim: if (x) is cheaper than if (x > 0) for unsigned integers"
This reverts commit 6aa8ca37ee.

See http://article.gmane.org/gmane.comp.video.gstreamer.devel/32282
2010-12-26 21:21:43 +00:00
Edward Hervey
3efb73c121 gst: documentation fixups and annotation
Reported by enabling the --warn-all option of g-ir-scanner
2010-12-17 19:14:41 +01:00
Edward Hervey
90fff577f0 Merge commit '0eaa25cbf5c0e4bf86545fb67c181a0ecd2f19c7' into 0.11 2010-12-08 17:51:10 +01:00
Tim-Philipp Müller
0eaa25cbf5 pad: register gst_pad_get_fixed_caps_func() with the debug log system 2010-12-07 18:37:04 +00:00
Tim-Philipp Müller
3256c708be docs: gst: more gobject introspection annotations
Many of these are superfluous, added for clarity.
2010-12-07 18:37:04 +00:00
Wim Taymans
b2ab72b916 pad: remove get_caps_reffed variants
Make the _get_caps functions behave like the _get_caps_reffed variants and
remove the _reffed variants. This means that _get_caps doesn't return a writable
caps anymore and an explicit _make_writable() is needed before modifying the
caps.
2010-12-07 18:14:38 +01:00
Wim Taymans
f558409fdc pad: Clean up .h file 2010-12-07 16:52:47 +01:00
Wim Taymans
993eda5004 remove deprecated symbols and methods 2010-12-06 19:18:31 +01:00
Mark Nauwelaerts
58868d4218 pad: add some debug to fast push path
... so we don't loose track at times it is needed the most.
2010-12-06 11:07:38 +01:00
Edward Hervey
6aa8ca37ee micro-optim: if (x) is cheaper than if (x > 0) for unsigned integers 2010-12-03 12:03:42 +01:00
Wim Taymans
cec2a42808 pad: add push cache to bufferlists
Add the push cahce for the bufferlist push code path as well.
2010-12-03 11:28:52 +01:00
Wim Taymans
59bc71c47a pad: don't cache the peer chainfunc
There is no need to cache the peer chainfunction as we can just as efficiently
get to it from the peer object. Also not caching the chain function works better
because then we automatically get the new chainfunctions when they change.
2010-12-03 11:28:52 +01:00
Wim Taymans
d3630379da pad: clear pad cache when installing probes
Move the method to clear the pad cache into _private.h
Clear the pad cache when installing pad probes.
2010-12-03 11:28:52 +01:00
Wim Taymans
2239038d76 pad: explicitly inline some functions 2010-12-03 11:28:52 +01:00
Wim Taymans
67d7c543b3 pad: remove unused variable 2010-12-03 11:28:52 +01:00
Wim Taymans
5e37ade932 pad: invalidate caches on flush and pad block 2010-12-03 11:28:52 +01:00
Wim Taymans
8abc14052a pad: don't unref NULL caps 2010-12-03 11:28:52 +01:00
Wim Taymans
14542a0d46 pad: add invalidate function
More small optimisations, remove the unneeded valid boolean.
Add function to invalide the cache.
Invalidate the cache on unlink.
2010-12-03 11:28:52 +01:00
Wim Taymans
1c79181afd pad: small cleanup 2010-12-03 11:28:52 +01:00
Wim Taymans
b83e66be46 pad: improve pad push caching
Build the cache while we push data. When we don't have a cache, we run the
slowpath and collect cacheable properties. When all conditions are met, keep the
cached data around so that we can more efficiently push data around.
2010-12-03 11:28:52 +01:00
Wim Taymans
d59b7f81b7 pad: prototype of pad push cache
Prototype of how we can cache the peer and caps for a pad link.
2010-12-03 11:28:52 +01:00
Stefan Kost
898583b2a0 docs: query doc improvements
More xrefs. Mentioned that some queries need a running pipeline.
2010-12-03 09:50:32 +02:00
Sebastian Dröge
f2f7842f6b pad: Set the event source object if none is set yet in gst_pad_push_event()
Otherwise the source will stay at NULL, the event is passed to the
peerpad via gst_pad_send_event() and then the peerpad is set as
source of the event instead of the originating pad.
2010-12-02 19:02:08 +01:00
Tim-Philipp Müller
fd6334cb7c pads: use new g_object_notify_by_pspec() for caps notifies if available
If we're building against GLib >= 2.26.0, we can use the more efficient
g_object_notify_by_caps(), which avoids the param spec lookup.
2010-10-07 19:03:42 +01:00
Stefan Kost
5550136fb2 Revert "pad: use a nested lock to avoid reffing the peer"
This reverts commit 9b424b1570.
2010-09-23 15:34:54 +03:00
Stefan Kost
9b424b1570 pad: use a nested lock to avoid reffing the peer
Fixes #503592
2010-09-22 10:22:40 +03:00
Thiago Santos
60fba4df8b gstpad: Fix flush-stop event handling
A flush-stop event would make a pad unflushing, causing it
to start acting as an activated pad. This, for example,
could lead to the chain function being called when stuff
isn't initialized.

This could happend when setting qtdemux to NULL while a seek
was being handled in the upstream filesrc (in push mode).

This patch makes it check if it is activated before setting
it to unflushing.
2010-09-13 20:52:03 -03:00
Stefan Kost
8ef7e46d24 docs: fix warnings pointed out by gtk-doc 2010-09-13 11:18:25 +03:00
Stefan Kost
770694f0f1 pad: add a unchecked variant for pull
Add internal _get_range_unchecked thats is called from _get_range and
_pull_range.
2010-09-08 10:25:55 +03:00
Stefan Kost
c94fbb5594 pad: log element:pad names with caps 2010-08-05 09:40:03 +03:00
Edward Hervey
21c8edca2a gstpad: Assume pads are compatible if we don't have templates
This is the same behaviour as if we had a pad template caps of
GST_CAPS_ANY on any of the pads (i.e. the actual check will be done
during caps negotiation).
2010-07-20 20:52:35 +02:00
Edward Hervey
7fa58dbebd GstPad: Do not call gst_pad_accept_caps() when caps change
Instead just check that the caps intersect with the pad template.
The elements should properly accept/refuse the caps in setcaps().

Shaves off calling the default implementation of acceptcaps which does
an expensive gst_pad_get_caps() (so if you have 50 of those elements in
a row, you'd be doing factorial(50) gst_pad_get_caps...).

Does not break any module unit test and most apps work fine.

https://bugzilla.gnome.org/show_bug.cgi?id=622740
2010-07-16 15:30:55 +02:00
Tim-Philipp Müller
71f3a6f0f7 Don't include <libxml/parser.h> from public headers if GST_DISABLE_DEPRECATED is defined
Since everything GstXML related has been deprecated, we can now skip the
libxml includes from the public headers when GST_DISABLE_DEPRECATED is
defined.

See #463435.
2010-06-26 10:35:38 +01:00
Edward Hervey
27b4868157 pad: more documentation regarding the new flags 2010-06-25 18:55:18 +02:00
Sebastian Dröge
5f4a965f67 gstxml: Deprecate GstXml and related functions
Pipeline serialisation to and from XML is horribly broken for all
but the most simple use cases, and will likely never be fixed.
Make sure everyone playing around with these tools is aware of
this, to avoid frustration. See countless bug reports in bugzilla.

Fixes bug #622685.
2010-06-25 18:25:40 +02:00
Edward Hervey
625c4e2d80 GstPad: Add new pad linking method with configurable checks.
To be used for cases where we don't need all checks to be validated.

API: gst_pad_link_full
API: GstPadLinkCheck

https://bugzilla.gnome.org/show_bug.cgi?id=622504
2010-06-23 18:42:35 +02:00
Edward Hervey
4a11063768 Revert "GstPad: Add GST_PAD_NEGOTIABLE GstPadFlag"
This reverts commit dc38e75d88.

boom
2010-06-15 11:54:19 +02:00
Edward Hervey
3df54c45bf Revert "gstpad: Return pad template in get_caps if pad is not negotiable"
This reverts commit 7460321a60.

crack
2010-06-15 11:54:19 +02:00
Edward Hervey
19334dbb85 Revert "pad: fix comment"
This reverts commit 8e92cb4a7d.

whatever...
2010-06-15 11:54:19 +02:00
Wim Taymans
8e92cb4a7d pad: fix comment 2010-06-14 17:33:16 +02:00
Edward Hervey
7460321a60 gstpad: Return pad template in get_caps if pad is not negotiable
https://bugzilla.gnome.org/show_bug.cgi?id=618644
2010-06-14 17:08:14 +02:00
Edward Hervey
dc38e75d88 GstPad: Add GST_PAD_NEGOTIABLE GstPadFlag
A pad is 'negotiable' when its container element is in a state greater
than GST_STATE_READY

API:gst_pad_is_negotiable
API:gst_pad_set_negotiable
API:GST_PAD_NEGOTIABLE

https://bugzilla.gnome.org/show_bug.cgi?id=618644
2010-06-14 17:08:14 +02:00
Sebastian Dröge
6ff314c75a pad: Fix iterator aggregation of all pads in the internal links fallback
g_list_prepend() returns the new head of the list and not
using this will create a memory leak and a single-element list.
2010-06-13 16:27:39 +02:00
Stefan Kost
21b4ef4d0f pads: Improve readability for gst_pad_fixate_caps()
Just truncate and then fixate. We check for empty caps in the begin and a
fixate-func that empties a caps would be broken. It also helps lazy caps impl.
in bug 618853 by avoiding the gst_caps_get_size().
2010-06-01 22:30:37 +03:00
Stefan Kost
edfbd90b29 docs: xref function name 2010-05-22 22:45:33 +03:00
Tim-Philipp Müller
4793930efb pad: don't print WARNING debug statements for normal things like EOS, part II 2010-05-17 13:09:15 +01:00
Wim Taymans
60499306a6 Revert "pad: don't check twice for changed caps per push"
We need to check the pad caps on the srcpad as well as on the sinkpad. Revert
this commit as it removes the check on the srcpad and can leave the srcpad
unnegotiated (or negotiated with wrong caps)

This reverts commit 07dc1e5b49.
2010-05-06 16:41:09 +02:00
Stefan Kost
07dc1e5b49 pad: don't check twice for changed caps per push
gst_pad_chain_data_unchecked() does the same check already.
2010-05-06 17:04:53 +03:00
Wim Taymans
6e4fde7195 docs: clarify the pull_range functions
Clarify the gst_pad_pull_range(), GstBaseSrc::create(), gst_pad_get_range()
and GstPadGetRange functions a little.

Fixes #617733
2010-05-05 12:01:50 +02:00
Benjamin Otte
b878069ef7 caps: Do not allow fixating empty caps
Passing empty caps to gst_pad_fixate_caps() is invalid, as empty caps
cannot be fixated.
2010-04-29 22:41:36 +02:00
Sebastian Dröge
fd68dbc08f gst: Use GSlice instead of normal g_malloc in more places 2010-03-28 19:48:45 +02:00
Wim Taymans
f089c3cceb pad: set a good name on the task of the pad
Use the element:pad names to configure a good name for the pad task.
2010-03-15 13:48:30 +01:00
Benjamin Otte
7e7f51f617 Fixes for -Wmissing-declarations -Wmissing-prototypes
Also adds those flags to the configure warning flags

https://bugzilla.gnome.org/show_bug.cgi?id=611692
2010-03-11 10:59:57 +01:00
Tim-Philipp Müller
c802301b69 pad: don't print WARN debug statements for normal things like EOS 2010-02-15 00:08:16 +00:00
Thijs Vermeir
053ada3308 gstpad: directly set the caps when pushing buffer with different caps.
This check is not necesarry as we are not negotiating anymore. And it can
be wrong if upstream can't produce this caps anymore, but downstream can
process them fine.
2010-01-19 14:13:25 +01:00
Wim Taymans
fc7dd46b20 avoid some more type checks 2009-12-24 14:40:54 +01:00
Wim Taymans
381d35fd1e pad: Fix problem with destroy callback not being called
When we unblock a pad with the same user_data, the destroy callback is not
called. This leads to refcounting leaks that cannot be avoided. Instead always
call the destroy notify whenever we install a new pad block.
In particular, this fixes a nasty pad leak in decodebin2.

Also update the unit test to have more accurate comments and test the required
behaviour.
2009-12-23 21:20:14 +01:00
Edward Hervey
4da0da822a gstpad: Only register debug funcptr once.
This makes pad initialization 2 times faster and without any contention
over the debug funcptr global lock.
2009-12-01 17:56:19 +01:00
Stefan Kost
23da3639f0 docs: fix xrefs in docs
Fix typos in xrefs, links to non existing functions and rework plural forms.
2009-11-25 16:59:50 +02:00
Stefan Kost
f63ecd04a8 pad: rename new api from _refed to _reffed.
Due to popular demand rename the new api as we still can.
API: gst_pad_get_caps_reffed(), gst_pad_peer_get_caps_reffed()
2009-11-05 12:54:32 +02:00
Rob Clark
746284f807 pad: make _fixate_caps() also truncate when needed
The default gst_pad_fixate_caps() previously would only fixate each individual
struct. In case there are multiple structs, the resulting caps would still not
be fixed. In the spirit of how individual structs are fixated, this patch
changes gst_pad_fixate_caps() to remove all but the first struct.

Fixes #595886
2009-10-28 16:36:13 +01:00
Tim-Philipp Müller
daecaf0e8a Remove GST_DEBUG_FUNCPTR where they're pointless
There's not much point in using GST_DEBUG_FUNCPTR with GObject
virtual functions such as get_property, set_propery, finalize and
dispose, since they'll never be used by anyone anyway. Saves a
few bytes and possibly a tenth of a polar bear.
2009-10-28 00:44:24 +00:00