Original commit message from CVS:
2005-06-08 Andy Wingo <wingo@pobox.com>
* gst/gstutils.c: RPAD fixes all around.
(gst_element_link_pads): Refcounting fixes.
* tools/gst-inspect.c:
* tools/gst-xmlinspect.c:
* parse/grammar.y:
* gst/base/gsttypefindhelper.c:
* gst/base/gstbasesink.c:
* gst/gstqueue.c: RPAD fixes.
* gst/gstghostpad.h:
* gst/gstghostpad.c: New ghost pad implementation as full proxy
pads. The tricky thing is they provide both source and sink
interfaces, since they proxy the internal pad for the external
pad, and vice versa. Implement with lower-level ProxyPad objects,
with the interior proxy pad as a child of the exterior ghost pad.
Should write a doc on this.
* gst/gstpad.h: s/RPAD/PAD/, s/RealPad/Pad/.
(gst_pad_set_name, gst_pad_set_parent): Macros removed, use
gst_object API.
* gst/gstpad.c: Big changes. No more stub base GstPad, now all
pads are real pads. No ghost pads in this file. Not documenting
the myriad s/RPAD/PAD/ and REALIZE fixes.
(gst_pad_class_init): Add properties for "direction" and
"template". Both are construct-only, so they can't change during
the life of the pad. Fixes properly deriving from GstPad.
(gst_pad_custom_new, gst_pad_custom_new_from_template): Gone. For
derived objects, just set properties when creating the objects via
g_object_new.
(gst_pad_get_parent): Implement as a function, return NULL if the
parent is not an element.
(gst_pad_get_real_parent, gst_pad_add_ghost_pad)
(gst_pad_remove_ghost_pad, gst_pad_realize): Removed.
* gst/gstobject.c (gst_object_class_init): Make name a construct
property. Don't set it in the object init.
* gst/gstelement.c (gst_element_add_pad): Don't allow adding pads
with UNKNOWN direction.
(gst_element_add_ghost_pad): Remove non-orthogonal API. Replace
with gst_element_add_pad (e, gst_ghost_pad_new (name, pad)).
(gst_element_remove_pad): Remove ghost-pad special cases.
(gst_element_pads_activate): Remove rpad cruft.
* gst/gstbin.c (gst_bin_change_state): Use gst_pad_get_parent to
catch the pad's-parent-not-an-element case.
* gst/gst.h: Include gstghostpad.h.
* gst/gst.c (init_post): No more real, ghost pads.
* gst/Makefile.am: Add gstghostpad.[ch].
* check/Makefile.am:
* check/gst/gstbin.c:
* check/gst/gstghostpad.c (test_ghost_pads): Check that linking
into a bin creates ghost pads, and that the refcounts are right.
Partly moved from gstbin.c.
Original commit message from CVS:
* check/gst-libs/.cvsignore:
* check/gst/.cvsignore:
* check/pipelines/.cvsignore:
ignore more
* check/pipelines/cleanup.c: (setup_pipeline), (run_pipeline),
(START_TEST), (cleanup_suite), (main):
add some tests related to cleanup after running pipelines
Original commit message from CVS:
* gst/elements/gstcapsfilter.c: (gst_capsfilter_set_property):
Free existing caps if the capsfilter changes. Add a FIXME about
setting those caps on the pads.
* gst/gstutils.c: (gst_element_get_compatible_pad), (ghost_up):
Before adding a ghost pad to a parent bin, check that there isn't
already one for the element on the bin. Prevents infinite recursion
when using decodebin in parse pipelines. Andy says he'll rewrite the
way this works anyway, so ignore the hack.
Original commit message from CVS:
2005-06-02 Andy Wingo <wingo@pobox.com>
* gst/elements/gsttypefindelement.c (do_pull_typefind): Query the
file size, pass it on to the type find helper.
* gst/base/gstbasesrc.c (gst_basesrc_do_seek): Set the
segment_start and segment_end properly according to the seek
method. Segment_end is still a bit flaky because offset can be
negative for CUR and END cases, but it takes -1 as an "unset"
value.
Original commit message from CVS:
* gst/elements/gstidentity.c: (gst_identity_event),
(gst_identity_transform), (gst_identity_get_property):
Protect last_message property as it is accessed from
multiple threads.
Original commit message from CVS:
* gst/gstelement.c: (gst_element_init),
(gst_element_pads_activate), (gst_element_change_state):
Slicker pad activation code.
Original commit message from CVS:
* gst/Makefile.am:
* gst/gstelement.h:
* gst/gstelementfactory.h:
* gst/gsttypes.h:
Move elementfactory methods to separate .h file.
Original commit message from CVS:
* gst/gstbuffer.c: (gst_subbuffer_finalize),
(gst_buffer_create_sub), (gst_buffer_is_span_fast):
* gst/gstbuffer.h:
Avoid typechecking in places where not needed.
Added accessor for malloc_data.
Original commit message from CVS:
* gst/gstpad.c: (gst_real_pad_init), (gst_pad_set_active),
(gst_pad_link_prepare), (gst_pad_link), (gst_pad_accept_caps),
(gst_pad_configure_sink), (gst_pad_configure_src),
(gst_pad_alloc_buffer), (gst_pad_query), (gst_pad_send_event),
(gst_pad_start_task):
Propagate errors from _set_caps() in configure_src/sink
functions instead of returning TRUE.
FLUSH events can travel up and downstream
Original commit message from CVS:
* gst/gstqueue.c: (gst_queue_handle_sink_event), (gst_queue_chain),
(gst_queue_loop), (gst_queue_handle_src_event):
Remove old pieces of code
Flushing the queue in an upstream event is a very bad idea.
Original commit message from CVS:
2005-05-26 Andy Wingo <wingo@pobox.com>
* gst/gstminiobject.c (gst_value_mini_object_collect): Use
gst_value_set_mini_object so as to add a ref on the object (which
will be removed when the value is unset).
* gst/elements/gstfakesink.c (gst_fakesink_class_init): Fix signal
arg type in ::handoff.
* gst/gstelement.c (gst_element_change_state): Also deactivate
pads in READY->NULL, just in case the element didn't make it to
PAUSED. Wingo tested, Wim approved.
Original commit message from CVS:
* gst/gstpad.c: (gst_pad_set_active), (gst_pad_link_prepare),
(gst_pad_link), (gst_pad_accept_caps), (gst_pad_alloc_buffer),
(gst_pad_query), (gst_pad_send_event), (gst_pad_start_task):
A flushing pad cannot be used to alloc_buffer from.
Original commit message from CVS:
* gst/gstbus.c: (gst_bus_init), (gst_bus_dispose), (gst_bus_post),
(gst_bus_pop), (gst_bus_source_prepare), (gst_bus_source_check),
(gst_bus_source_dispatch), (gst_bus_source_finalize),
(gst_bus_create_watch), (gst_bus_add_watch_full):
* gst/gstbus.h:
Implement a real GSource and use g_main_context_wakeup() to
signal new messages instead of the socketpair.
Original commit message from CVS:
* gst/gstbin.c: (bin_element_is_sink), (has_ancestor),
(bin_element_is_semi_sink), (append_child), (gst_bin_change_state):
* gst/gstpad.c: (gst_pad_set_active), (gst_pad_link_prepare),
(gst_pad_link), (gst_pad_accept_caps), (gst_pad_query),
(gst_pad_send_event), (gst_pad_start_task):
* gst/gstqueue.c: (gst_queue_init), (gst_queue_locked_flush),
(gst_queue_handle_sink_event), (gst_queue_chain), (gst_queue_loop),
(gst_queue_sink_activate), (gst_queue_src_activate),
(gst_queue_change_state):
* gst/gstqueue.h:
Fix state changes for non sinks. We now change sinks, then elements
with unconnected srcpads, then the rest.
More efficient queue unlocking in flush and state changes.
Set the pad activate mode even if it does not have an activate
function.
Original commit message from CVS:
* gst/base/gstbasesrc.c: (gst_basesrc_activate):
Don't go in pull mode for non-seekable sources.
* gst/elements/gsttypefindelement.c: (gst_type_find_element_init),
(gst_type_find_element_dispose), (gst_type_find_handle_src_query),
(free_entry), (stop_typefinding),
(gst_type_find_element_handle_event), (find_peek),
(gst_type_find_element_chain), (do_pull_typefind),
(gst_type_find_element_change_state):
Allow typefinding (w/o seeking) in push-mode, simplified version
of what was in 0.8.
* gst/gstutils.c: (gst_buffer_join):
* gst/gstutils.h:
gst_buffer_join() from 0.8.
Original commit message from CVS:
* gst/gstpad.c: (gst_pad_set_active), (gst_pad_link_prepare),
(gst_pad_link), (gst_pad_accept_caps), (gst_pad_query),
(gst_pad_send_event), (gst_pad_start_task):
Disable attempt at mode switching until it is figured out.
Original commit message from CVS:
* configure.ac: Use -no-common everywhere. In a sane world, it
would be the default in libtool, because without it, you can't
build DLLs on Windows.
* docs/gst/gstreamer-docs.sgml: Remove GstCpu, GstData, GstThread
* docs/gst/gstreamer-sections.txt:
* docs/gst/tmpl/gstcpu.sgml:
* docs/gst/tmpl/gstdata.sgml:
* docs/gst/tmpl/gstthread.sgml:
Original commit message from CVS:
* gst/gst.c:
* gst/gstmessage.c:
Make sure the class is reffed/unreffed once before threads can be
used. Fixes#304551.
Original commit message from CVS:
* gst/base/gstbasesink.c: (gst_basesink_finish_preroll),
(gst_basesink_chain_unlocked), (gst_basesink_activate):
* gst/gstminiobject.c: (gst_mini_object_get_type),
(gst_mini_object_free):
* gst/gstpad.c: (gst_pad_accept_caps), (gst_pad_query),
(gst_pad_push), (gst_pad_push_event):
* gst/gstqueue.c: (gst_queue_change_state):
Don't queue buffers in basesink when we are flushing.
Unref buffer when flushing in basesink.
Flush queue when going to READY
Unref buffer when _push() returns an error.
Don't free MiniObject instance when refcount is incremented
in _finalize() so that we can recover objects.
Original commit message from CVS:
* gst/gstpad.c: (gst_pad_accept_caps), (gst_pad_set_caps),
(gst_pad_alloc_buffer), (gst_pad_query), (gst_pad_push):
* gst/gstqueue.c: (gst_queue_change_state):
Clear queue when going to READY.
Remove IN_SETCAPS flag too.
Original commit message from CVS:
* gst/base/gstbasesrc.c: (gst_basesrc_change_state):
Remove implicit cast from gboolean to GstElementStateReturn;
make sure we still return failure in paused => ready case if
the parent class fails to change state and our own stop
vfunc succeeds.
Original commit message from CVS:
2005-05-16 Andy Wingo <wingo@pobox.com>
* gst/gstbin.c (sink_iterator_filter): Err... um...
* check/gst/gstbin.c (test_ghost_pads): New test for the
ghosting-if-elements-not-in-same-bin behavior.
Original commit message from CVS:
* gst/elements/Makefile.am: Remove some elements that don't
need to be in the core (or even exist at all).
* gst/elements/gstaggregator.c:
* gst/elements/gstaggregator.h:
* gst/elements/gstmd5sink.c:
* gst/elements/gstmd5sink.h:
* gst/elements/gstmultifilesrc.c:
* gst/elements/gstmultifilesrc.h:
* gst/elements/gstpipefilter.c:
* gst/elements/gstpipefilter.h:
* gst/elements/gstshaper.c:
* gst/elements/gstshaper.h:
* gst/elements/gststatistics.c:
* gst/elements/gststatistics.h:
* po/POTFILES.in: Remove above files.
Original commit message from CVS:
2005-05-14 Andy Wingo <wingo@pobox.com>
* gst/gstbin.c (gst_bin_iterate_sinks): Use sink_iterator_filter
so as to get the refs right.
(sink_iterator_filter): New function, wraps bin_element_is_sink,
unreffing objects that don't pass the filter.
Original commit message from CVS:
2005-05-14 Andy Wingo <wingo@pobox.com>
* gst/gstpipeline.c (gst_pipeline_init): Drop ref on bus after
gst_element_set_bus.
(gst_pipeline_dispose): Set the bus on the pipeline to NULL. In
normal cases, this will destroy the bus.
* gst/gstutils.c (prepare_link_maybe_ghosting): Drop ref on root
object.
Original commit message from CVS:
2005-05-14 Andy Wingo <wingo@pobox.com>
* gst/gstbin.c (gst_bin_change_state): Fix state changes if a bin
has no sinks.
Original commit message from CVS:
2005-05-13 Andy Wingo <wingo@pobox.com>
* gst/gstutils.c (gst_element_link_pads): Instead of calling
gst_pad_link, call pad_link_maybe_ghosting,
(pad_link_maybe_ghosting): Links pads, making sure that the
elements being linked are in the same bin.
(find_common_root, object_has_ancestor, ghost_up, remove_pad):
Helpers for pad_link_maybe_ghosting.
Original commit message from CVS:
* gst/base/gstbasesink.c: (gst_basesink_init),
(gst_basesink_activate):
* gst/base/gstbasesrc.c: (gst_basesrc_unlock),
(gst_basesrc_is_seekable):
* gst/gstbin.c: (gst_bin_add_func), (gst_bin_remove_func),
(bin_element_is_sink), (gst_bin_change_state):
* gst/gstelement.c: (gst_element_add_pad), (gst_element_query):
* gst/gstelement.h:
Identify sinks by their flag to avoid overly complicated
checks (fow now).
Do state changes even for elements not reachable from the
sinks.
BaseSink is a sink now :)
Some more debugging info in the basesrc.
Original commit message from CVS:
* gst/base/gstbasesrc.c: (gst_basesrc_do_seek):
Discont event offset format should be GST_FORMAT_BYTES,
not GST_FORMAT_TIME.
Original commit message from CVS:
* gst/gstbin.c: (gst_bin_add_func), (cb_parent_unset),
(gst_bin_get_state):
If a child is removed from a bin while we remove the child from
the bin and while we're retrieving its state, signal this to the
get_state function so we abort the wait (instead of waiting for
a timeout) and can immediately re-iterate over all other elements.
Original commit message from CVS:
* gst/gstelement.c: (gst_element_add_pad), (gst_element_query):
* gst/gstutils.c: (gst_element_get_compatible_pad_template),
(gst_element_link_pads), (gst_element_query_position),
(gst_element_query_convert), (intersect_caps_func),
(gst_pad_query_position), (gst_pad_query_convert):
Fix refcounting in utils function.
No point in trying to activate a pad when it's added, it could
be added from the state change function and then we deadlock, the
element has to decide what to do.
Original commit message from CVS:
2005-05-10 Andy Wingo <wingo@pobox.com>
* gst/elements/gstfakesink.c (gst_fakesink_render): Er, emit with
*all* the arguments.
* gst/base/gstbasetransform.c (gst_base_transform_event): Grab the
stream lock if it's a FLUSH_DONE; normal flushes don't get the
lock (according to the docs -- if this is wrong change the docs).
* gst/gstpipeline.c (gst_pipeline_change_state): Set the bus to
flush messages in the NULL state.
* gst/gstbus.c (gst_bus_post): If a bus is flushing, unref the
message immediately and return.
(gst_bus_set_flushing): New function. If a bus is flushing, it
flushes out any queued messages and immediately unrefs new
messages. This is so when an element goes to NULL, all of the
unhandled messages coming from it can be freed, and their
references to the element dropped. In other words: message source
ref considered harmful :P
* gst/gstbin.c (gst_bin_change_state): Unref peer element when
we're finished with it.
* gst/gstmessage.c (gst_message_new_state_changed):
Original commit message from CVS:
2005-05-09 Andy Wingo <wingo@pobox.com>
* gst/base/gstbasetransform.c (gst_base_transform_proxy_getcaps):
Intersect the peer's caps with our caps.
Original commit message from CVS:
Remove old query functions. Ported old code.
Added position/convert helper functions to gstutils.
Reordered gstpad.c code, grouping relevant things.
Remove gst_message_new(), always need to speficy a specific
message.
Original commit message from CVS:
2005-05-09 Andy Wingo <wingo@pobox.com>
* gst/gstiterator.h: Add some includes.
* gst/gstqueryutils.h: Include more headers.
* gst/gstpad.h:
* gst/gstpad.c (gst_pad_query_position): New routine, replaces
some uses of gst_pad_query.
* gst/gstqueryutils.c: Build fixes. Make parse functions ignore
NULL out parameters.
(gst_query_new_position): New proc, allocates a new position
query.
* gst/Makefile.am (libgstreamer_@GST_MAJORMINOR@_la_SOURCES): Add
gstqueryutils.c to the build.
* gst/gststructure.c (gst_structure_set_valist): Implement with
the generic G_VALUE_COLLECT.
Original commit message from CVS:
* gst/Makefile.am: (gst_headers):
Added gstqueryutils.h to the list of headers to install, that was
a 'nachty' move wingo :)
Original commit message from CVS:
2005-05-06 Andy Wingo <wingo@pobox.com>
* gst/gstquery.h
* gst/gstquery.c (_gst_query_initialize): Extend GstQuery from
GstData, init a memchunk.
(standard_definitions): Add a few query types, deprecate a few.
(gst_query_get_type): New proc.
(_gst_query_copy, _gst_query_free, gst_query_new): GstData
implementation.
(gst_query_new_application, gst_query_get_structure): New public
procs.
* docs/design/draft-query.txt: Removed LINKS from the query types,
because all the rest can be dispatched to other pads -- seemed
ugly to have a query that couldn't be dispatched. internal_links
is fine as a pad method.
* gst/gstpad.h: Add query2 as a pad method, add the new functions
in gstpad.c, but maintain binary compatibility for the moment.
Will fix before 0.9 is out.
* gst/gstqueryutils.c:
* gst/gstqueryutils.h: New files, implement 3 methods for each
query type: parse_query, parse_response, and set. Probably need an
allocator as well.
* gst/gst.h: Add gstquery.h and gstqueryutils.h to the list.
* gst/elements/gstfilesink.c (gst_filesink_query2):
* gst/base/gstbasesrc.c (gst_basesrc_query2): Replace old query,
query_types, and formats methods.
* gst/gstpad.c (gst_pad_query2, gst_pad_query2_default)
(gst_pad_set_query2_function): New functions.
(gst_real_pad_init): Set query2_default as the default query2
function. Basically just dispatches to internally linked pads.
Needs review!
* gst/gstdata_private.h (_GST_DATA_INIT): Set data->refcount to 1
without using the atomic operations. Only one thread can possibly
be accessing the data at this point. Changed so as to avoid
gst_atomic operations.
Original commit message from CVS:
* gst/base/gstbasetransform.c: (gst_base_transform_proxy_getcaps):
* gst/elements/gstcapsfilter.c: (gst_capsfilter_getcaps):
* gst/gstpad.c: (gst_pad_peer_get_caps):
* gst/gstqueue.c: (gst_queue_init), (gst_queue_getcaps),
(gst_queue_bufferalloc), (gst_queue_handle_sink_event),
(gst_queue_src_activate), (gst_queue_change_state):
* gst/gstqueue.h:
* gst/gstutils.c: (gst_element_get_compatible_pad_template),
(intersect_caps_func):
Fix gst_pad_peer_get_caps(), make it return NULL if no peer.
Always take QUEUE_LOCK after STREAM_LOCK or we might deadlock.
Some fixes for the peer_get_caps() change.
Original commit message from CVS:
* gst/base/gstbasesink.c: (gst_basesink_preroll_queue_empty),
(gst_basesink_handle_buffer), (gst_basesink_chain_unlocked),
(gst_basesink_activate):
Actually do something with error codes returned from the push
functions.
Original commit message from CVS:
2005-05-05 Andy Wingo <wingo@pobox.com>
* check/pipelines/simple_launch_lines.c (test_2_elements): "Fix"
the test suite.
Original commit message from CVS:
* gst/gstplugin.c: (gst_plugin_check_module),
(gst_plugin_check_file), (gst_plugin_load_file):
apply patch from #172526 to make register work on MacOSX
Original commit message from CVS:
* docs/design/draft-push-pull.txt:
* docs/design/draft-query.txt:
* gst/base/gstbasesrc.c: (gst_basesrc_get_range_unlocked),
(gst_basesrc_start):
Added draft for new query API.
Added draft for better selecting scheduling methods.
Make basesrc ignore length if the subclass does not support
it.
Original commit message from CVS:
* check/Makefile.am:
fix the broken rule for cleanup. Apparently this rule is
only needed on FC2, so maybe this warrants further autotool
inspection.
Original commit message from CVS:
* gst/gsttrashstack.h:
Ooohh. a nasty one! After having a failed pop() from the stack,
it's possible that the stack is empty. In that case, don't
follow the NULL pointer.
Original commit message from CVS:
* gst/gstpad.c: (gst_pad_set_active), (gst_pad_peer_set_active),
(gst_pad_set_checkgetrange_function),
(gst_pad_link_check_compatible_unlocked), (gst_pad_link_prepare),
(gst_pad_check_pull_range), (gst_pad_pull_range),
(gst_static_pad_template_get_caps), (gst_pad_start_task),
(gst_pad_pause_task), (gst_pad_stop_task):
* gst/gstplugin.c: (gst_plugin_load):
* gst/gstplugin.h:
Remove gst_library_load as it does more harm than good with
the new g_module flags.
Revert bogus caps template check in pad linking, pad caps
are important when linking not the template, which is more
general than the current caps.
