Original commit message from CVS:
* gst/gst.c: (scan_and_update_registry), (init_post):
If the fork()'ed child process can't write the updated registry cache
file to disk for some reason, make it exit with a failure exit code,
so that the parent can then re-scan the plugins itself and update the
registry structures in memory and work with that (rather than failing
when creating elements because seemingly no plugins are available).
Refactor registry scanning code into separate function for this and
also separate fork() and non-fork() code paths. Fixes#344748.
Original commit message from CVS:
* gst/gstregistryxml.c: (gst_registry_xml_write_cache):
Fix silly bug that prevented us from creating
~/.gstreamer-0.10 and writing the registry in one
go (the first call to g_mkstemp() would overwrite the
placeholder in the template string, so the second call
to g_mkstemp() after creating the missing directory
would then error out with 'invalid argument').
Original commit message from CVS:
* gst/gstsegment.c: (gst_segment_set_duration),
(gst_segment_set_last_stop), (gst_segment_set_seek),
(gst_segment_set_newsegment_full), (gst_segment_to_stream_time),
(gst_segment_to_running_time), (gst_segment_clip):
Use G_UNLIKELY to help the compiler a bit.
Original commit message from CVS:
* gst/gstbuffer.c: (gst_buffer_get_type), (gst_buffer_finalize),
(_gst_buffer_copy), (gst_buffer_is_metadata_writable),
(gst_subbuffer_finalize), (gst_buffer_create_sub),
(gst_buffer_is_span_fast), (gst_buffer_span):
Init _type for consistency.
Use _FLAGS macro to avoid type check.
Avoid unneeded type checks in subbufer code.
Original commit message from CVS:
* gst/gstobject.c: (gst_object_get_type), (gst_object_ref),
(gst_object_unref):
Add G_UNLIKELY in type registration.
Avoid type check in _ref/_unref since that is also
done in glib.
Original commit message from CVS:
* gst/gstinfo.c: (gst_debug_set_active),
(gst_debug_category_set_threshold), (_gst_debug_nameof_funcptr):
* gst/gstinfo.h:
Add some G_[UN]LIKELY.
Maintain __gst_debug_min to avoid formatting the arguments of
debug messages that will be dropped anyway to avoid a lot of
overhead from the debugging system.
Original commit message from CVS:
* gst/gst.c: (init_post):
use _exit() instead of exit() in our forked child; this ensures
that none of the registered exit handlers from whatever is using
GStreamer get executed. This fixes gnome-mixer-applet failing
to load, because ORBit would shut down.
Spotted by: Edward Hervey <edward@fluendo.com>
Fix suggested by: Tim-Philipp Müller <tim at centricular dot net>
Fixes#344474
Original commit message from CVS:
* gst/gstindex.c: (gst_index_gtype_resolver):
* tools/gst-xmlinspect.c: (print_plugin_info):
Fix leak spotted by coverity checker. Fixes#343827
Fix another other leak found by paolo borelli.
Original commit message from CVS:
* gst/gst.c:
plugin_paths is not used if we build without registry support.
* gst/gstsegment.c: (gst_segment_copy):
_copy() was always returning NULL...
Original commit message from CVS:
* gst/gst.c:
set #include <sys/wait.h> in a #ifdef #ifdef HAVE_FORK
* win32/common/libgstbase.def:
export gst_collect_pads_set_flushing
* win32/common/libgstreamer.def:
export gst_pad_set_acceptcaps_function, gst_structure_empty_new,
gst_value_fraction_multiply
* win32/vs6/gst_inspect.dsp:
add a link to intl.lib
Original commit message from CVS:
* gst/gstelement.c:
Clarify the use of _release_request_pad() and
_get_request_pad() a bit better.
* libs/gst/base/gstadapter.c: (gst_adapter_peek),
(gst_adapter_take_buffer):
Fix some doc and comment typos.
Original commit message from CVS:
* gst/gstsystemclock.c: (gst_system_clock_id_wait_unlocked):
Add debug that can be enabled using a #define at the top of the file,
for dumping stats about how late/early we were when waking up from
waiting on the clock.
Original commit message from CVS:
* gst/gst.c: (add_path_func), (init_post):
Don't scan registry paths passed via --gst-plugin--path immediately
(will crash, because absolutely nothing is set up and no types are
registered etc.); do this later in init_post(). Fixes#343057.
Original commit message from CVS:
* gst/gst.c: (init_post):
if we have fork, fork while reading/rebuilding the registry
so the parent doesn't take the hit of having all plugins loaded
in memory. Fixes#342777.
* configure.ac:
Check if we have fork()
* win32/common/config.h.in:
no fork() on win32
Original commit message from CVS:
* gst/gstvalue.c: (gst_value_deserialize_fraction):
Handle NULL input and output pointers silently as a failed conversion,
rather than g_warnings.
Original commit message from CVS:
* gst/gstpad.c: (gst_flow_get_name), (gst_flow_to_quark):
* gst/gstpad.h:
Added _CUSTOM error and success GstFlowReturn that can be
used be elements internally.
Added macro to check for SUCCESS flowreturns.
API: GST_FLOW_CUSTOM_SUCCESS
API: GST_FLOW_CUSTOM_ERROR
API: GST_FLOW_IS_SUCCESS
* tests/check/gst/gstpad.c: (GST_START_TEST), (gst_pad_suite):
Added check for GstFlowReturn sanity.
Original commit message from CVS:
* gst/gstpad.c: (gst_pad_set_caps):
Short-circuit gst_pad_set_caps if setting the existing
caps pointer again, and avoid printing debug and
reffing/unreffing the caps.
* plugins/elements/gstqueue.c: (gst_queue_push_one):
There's actually no need to set the caps before pushing -
the acceptcaps method will handle it anyway.
Original commit message from CVS:
* gst/gsttypefind.c: (gst_type_find_get_type):
* gst/gsttypefind.h:
Added GST_TYPE_TYPE_FIND and gst_type_find_get_type() so a GType gets
registered for GstTypeFind pointers. This allows wrapping the structure
in bindings (i.e. gst-python).
Original commit message from CVS:
* gst/gstregistryxml.c: (read_string), (load_pad_template),
(load_feature), (load_plugin):
Allow empty strings for some of the plugin fields so we don't
drop valid plugin entries that were written out correctly.
Original commit message from CVS:
* gst/gstregistryxml.c: (gst_registry_xml_write_cache):
Use g_remove and g_rename instead of remove and rename that don't
handle utf8 characters. rename was failing for users who had specific
characters in their name then the registry was built at each gstreamer init.
* win32/vs6/gst_inspect.dsp:
* win32/vs6/gst_launch.dsp:
* win32/vs6/libgstbase.dsp:
* win32/vs6/libgstcoreelements.dsp:
* win32/vs6/libgstreamer.dsp:
Use a debug version of libxml2 (libxml2D.lib,libxml2D.dll) for DEBUG build
of libgstreamer and clean unused libraries in project links settings.
Original commit message from CVS:
* gst/gstobject.h:
Add cast to make compiler happy - refcount variable was a gint
in GstObject but is a guint in GObject and g_atomic_int_get()
wants a gint *.
Original commit message from CVS:
* gst/gstbuffer.c: (_gst_buffer_initialize):
Register subbufer along with the buffer type so that
it does not accidentally gets registered from N
different streaming threads in a non threadsafe way.
Original commit message from CVS:
* gst/gstbuffer.h:
* gst/gstevent.h:
* gst/gstmessage.h:
Make gtk-doc generate docs for our inlined gst_buffer_ref(),
gst_event_ref() and gst_message_ref() functions again
(ugly hack, please do fix if there's a better way besides
overrides.txt, which doesn't seem to work).
Original commit message from CVS:
* gst/gstvalue.c: (gst_value_serialize_flags):
fix a leak if no flags are set
* tests/check/gst/gstvalue.c: (GST_START_TEST):
fix leak in tests
Original commit message from CVS:
* gst/gststructure.c: (gst_structure_gtype_from_abbr),
(gst_structure_parse_value):
add some comments
rename a method so that it actually says what it does better
