This will make sure we spawn a new plugin scanner helper for each plugin
to be introspected, which helps with making sure we don't load too many
shared objects (libs, plugins) at the same time on systems where there
is a hard limit like on Android.
A better version might re-use the scanner for up to N times, though
it's not clear whether that would actually improve things dramatically.
https://bugzilla.gnome.org/show_bug.cgi?id=662091
Returning a newly allocated string makes no sense. It's unexpected for a
getter, and also this behaves differently in 0.10, so it would make future
merges harder.
Except for these two places here in core which were updated for the new
semantic, the return value is getting leaked all over the place.
The feature name is not supposed to change over time anyway. In order to enforce
this parentize features to the registry and make the feature->name pointing to
GstObject:name. In 0.11 we could consider of removing the feature->name variable
(FIXME comment added).
Fixes: #459466
When a plugin file no longer exists, e.g. because it's been removed or
renamed, don't remove all features in the registry based on the *name*
of the plugin they belong to, but only remove those who actually belong
to that particular plugin (object/pointer).
This fixes issues of plugin features disappearing when a plugin .so file
is renamed.
https://bugzilla.gnome.org/show_bug.cgi?id=604094
Only replace existing plugins by blacklisted ones if they correspond
to the exact same plugin. If they're not the same, keep the existing
valid one.
Fixes#638941
Ignore plugins which have been moved into coreelements, so it's
still possible to just upgrade GStreamer core without having to
upgrade the whole stack.
Adds that warning to configure.ac
Includes a tiny change of the GST_BOILERPLATE_FULL() macro:
The get_type() function is no longer declared before being defined.
https://bugzilla.gnome.org/show_bug.cgi?id=611692
This returns the internal feature list cookie, which changes every
time a feature is added or removed. This can be used by elements
to check if they should update their cached feature lists.
This avoids:
* triple-checking for the GType when type-checking is enabled (see #597260)
* Avoids going through an expensive no-argument checking which landed in
glib-2.22
* Avoids going through 2 extrac functions (g_object_new -> g_object_new_valist)
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.
This avoids having to do the sorting everytime we use typefind
The behaviour of gst_type_find_factory_get_list has subtlely changed
in the sense that the order was previously undefined, whereas now
it returns them sorted by rank and then by name.
Maintain a hashtable of the plugin basename. We can then use this
hashtable to speedup the search for an existing plugin and avoid
a whole lot of strcmp calls.
When the plugin-scanner load fails (because the helper can't be
spawned), make sure to load the plugin that failed in-process, so
that all plugins do get loaded.
Everyone running an uninstalled git setup is going to wonder about
this failure next time they update, so let's mention the solution
in the error message.
When updating the registry, we don't need to re-read the registry cache
and waste time replacing all our existing, hopefully identical, plugins
and features that we're about to re-scan anyway.
phase 2 - make the plugin loader receive the list of plugins to load and
send back the results asynchronously, so we don't context switch back
and forth so much.
Original commit message from CVS:
Patch by: Simon Holm Thøgersen <odie at cs dot aau dot dk>
* gst/gstregistry.c: (gst_registry_scan_path_level):
Reduce the number of stat() calls for every file from three times
to one time. Fixes bug #560360.
Original commit message from CVS:
* gst/gstplugin.c:
* gst/gstregistry.c:
GstRegistryPool doesn't exist; don't refer to it in docs.
Don't refer to functions that don't exist in docs, it's
unhelpful.
Original commit message from CVS:
Patch by: Rob Bradford <rob at robster dot org dot uk>
* gst/gstregistry.c: (gst_registry_scan_path_level):
Don't recurse into .debug directories as some distros install
the debugging symbols next to the plugins in .debug directories
and dlopen() crashes on them sometimes. Fixes bug #508070.
Add FIXME for 0.11 to not recurse into directories at all because
it's very inconsistent to the behaviour of other PATH environment
variables.
Original commit message from CVS:
* gst/gstregistry.c: (gst_registry_scan_path_level):
Run g_str_has_suffix() only on the file name, not the
entire file path.
Original commit message from CVS:
2008-04-11 Julien Moutte <julien@fluendo.com>
Mac OS X love...
* configure.ac: Merge platform specific defines, introduce a new
define on OS X to remember that forking when updating registry is
unsafe.
* docs/faq/gst-uninstalled: Updated to include gst-libs in the bad
module.
* gst/gst.c: Don't fork when updating registry if GST_HAVE_UNSAFE_FORK
is defined.
* gst/gstregistry.c: (gst_registry_scan_path_level): Fixed a bogus
condition that leads to absolutely no plugins being registered on
OS X.
Original commit message from CVS:
* configure.ac:
* gst/gstregistry.c:
MacOS has plugins under .so or under .dylib. Add detection for MacOS
and handle this case.
* gst/gst.c:
Add a comment here describing, why we stat each plugin and not try to
be smart.
Original commit message from CVS:
* gst/gst.c:
Fix type in log message and add one to ease seeing how long registry
cache verification takes.
* gst/gstregistry.c:
Only test plugin filenames against G_MODULE_SUFFIX.
Original commit message from CVS:
* gst/gstregistry.c: (gst_registry_add_feature):
When replacing an existing feature in the registry, make sure to
continue holding a reference until we've replaced the name string
within our feature hash table. Make sure to use g_hash_table_replace
instead of g_hash_table_insert to ensure the new name string is used
as a key instead of the old one that we're about to free.
Fixes: #462085
Original commit message from CVS:
* gst/gstregistry.c: (gst_registry_init), (gst_registry_finalize),
(gst_registry_remove_features_for_plugin_unlocked),
(gst_registry_add_feature), (gst_registry_remove_feature),
(gst_registry_lookup_feature_locked):
* gst/gstregistry.h:
Speed up gst_registry_lookup_feature_locked() by using a hashmap.
Fixes#459501.
Original commit message from CVS:
* gst/gstindexfactory.c: (gst_index_factory_get_type):
* gst/gstplugin.c: (gst_plugin_init):
* gst/gstpluginfeature.c: (gst_plugin_feature_init):
* gst/gstquery.c: (gst_query_get_type):
* gst/gstregistry.c: (gst_registry_init):
* gst/gsturi.c: (gst_uri_handler_base_init):
Remove empty instance_init() functions to save relocs and lessen the
noise. Remove some of the function prototypes that are doubled by
G_DEFINE_TYPE.
Original commit message from CVS:
* gst/gstplugin.c: (gst_plugin_load_file):
* gst/gstregistry.c: (gst_registry_scan_path_level):
Don't print a g_warning for any failure to load a shared object.
Instead, push this down into gstplugin.c, and warn _only_ if we
failed to open the module (i.e. failure to link).
Avoids warnings on normal, working, non-plugin .so files.
Original commit message from CVS:
* gst/gstplugin.c (gst_plugin_load_file):
* gst/gstregistry.c (GST_CAT_DEFAULT,
gst_registry_lookup_feature_locked, gst_registry_scan_path_level):
Print a g_warning if there was an error when loading a plugins during
registry scan. The shuld help beginners starting with gst-plugin
template.
Original commit message from CVS:
2007-02-09 Andy Wingo <wingo@pobox.com>
* gst/gstregistryxml.c (load_feature, load_plugin): Drop some
normal-case logs down to LOG, raise errors to WARNING.
(gst_registry_xml_read_cache): Don't log before calling a function
that logs.
* gst/gstregistry.c (gst_registry_finalize): Less debug on program
exit (registry finalize).
(gst_registry_add_plugin, gst_registry_add_feature): No need for a
DEBUG log when we emit signals that people don't even have the
chance to connect to.
(gst_registry_scan_path_level): Less logging in the normal case.
Original commit message from CVS:
* docs/gst/gstreamer.types.in:
* gst/gstregistry.c: (gst_registry_class_init):
Document registry signals and make gtk-doc pick them up (#401381).
Original commit message from CVS:
* gst/gstregistry.c: (gst_registry_scan_path_level):
Plugin extension on HP-UX is .sl, add that to the list of approved
plugin extensions (see #393796).
* tests/check/gst/gstpad.c: (GST_START_TEST):
ulong => gulong. Fixes compilation with HP-UX compiler.
* tests/check/pipelines/parse-launch.c: (GST_START_TEST):
Fix compilation if valgrind headers are not available.
Original commit message from CVS:
* gst/gst.c: (ensure_current_registry_forking):
Ignore EINTR when reading from the child registry pipe.
Explicitly ignore the return value from close, since it makes no
difference.
* gst/gstminiobject.c: (gst_mini_object_ref),
(gst_mini_object_unref):
When debugging refcounts, check GST_IS_MINI_OBJECT and warn.
* gst/gstregistry.c: (_priv_gst_registry_remove_cache_plugins):
When removing cached plugins, remove their features too, so they're
not visible after they've disappeared.
* gst/gstutils.c: (prepare_link_maybe_ghosting):
In the unlikely case that we are linking pads with no parents, don't
crash trying to get the non-existent parent bin.
* gst/parse/grammar.y:
Output debug in the PIPELINE category
Original commit message from CVS:
* gst/gst.c: (init_pre), (scan_and_update_registry),
(ensure_current_registry_nonforking),
(ensure_current_registry_forking), (ensure_current_registry),
(init_post), (gst_debug_help), (gst_deinit):
* gst/gst_private.h:
* gst/gstregistry.c: (gst_registry_finalize),
(gst_registry_remove_features_for_plugin_unlocked),
(gst_registry_remove_plugin), (gst_registry_scan_path_level),
(gst_registry_scan_path),
(_priv_gst_registry_remove_cache_plugins),
(_priv_gst_registry_cleanup):
* gst/gstregistry.h:
Re-commit the registry changes, along with an extra fix:
When a cached plugin is encountered at a different file path,
update the stored path in the registry cache so that the parent
process knows where it actually is now when it re-reads the registry
cache. Fixes the thing that broke distcheck with the previous commit.
* tests/check/Makefile.am:
Clean up files named 'core' too when running make clean.
* tests/examples/manual/Makefile.am:
Set up a registry path for running these tests, and clean it properly
for distcheck.
Original commit message from CVS:
* gst/gst.c: (init_pre), (scan_and_update_registry),
(ensure_current_registry_nonforking),
(ensure_current_registry_forking), (ensure_current_registry),
(init_post), (gst_debug_help), (gst_deinit):
Make init_pre and init_post take the full complement of GOptionFunc
args so they can return useful GErrors. Make the registry updating
functions do so.
Call _priv_gst_registry_remove_cache_plugins after scanning files to
ensure that the registry we're about to write out doesn't contain
stale information about old-deleted plugin files.
Make _priv_gst_registry_remove_cache_plugins return a boolean so
that deletion of plugin files is considered a registry change.
* gst/gst_private.h:
* gst/gstregistry.c: (gst_registry_finalize),
(gst_registry_remove_features_for_plugin_unlocked),
(gst_registry_remove_plugin), (gst_registry_scan_path_level),
(gst_registry_scan_path),
(_priv_gst_registry_remove_cache_plugins),
(_priv_gst_registry_cleanup):
* gst/gstregistry.h:
Rename _gst_registry_remove_cache_plugins and _gst_registry_cleanup
by adding _priv prefix, so that they won't appear in the global
symbol table. They still do atm though because of #318031. Move the
prototypes to gst_private.h
When removing a plugin, remove all features for that plugin too.
Fixes#340878.
Original commit message from CVS:
* gst/gst.c:
* gst/gstregistry.c: (gst_registry_scan_path_level),
(gst_registry_scan_path):
* gst/gstregistry.h:
only write out registry if it has changed, fixes#338339
Original commit message from CVS:
* gst/gstregistry.c: (gst_registry_get_default),
(_gst_registry_cleanup):
Protect default registry with lock and ref/sink it.
Fixes#324818, patch by Torsten Schoenfeld.
Original commit message from CVS:
2006-01-17 Andy Wingo <wingo@pobox.com>
* gst/gstregistry.c (_gst_registry_remove_cache_plugins): Lock the
registry while we remove plugins.
* tools/gst-inspect.c (print_element_info): Don't unref the
factory arg, that should be the responsibility of whatever code
received the ref. Fixes a double-free when called from
print_element_list via gst-inspect-0.10 -a. Fixes#327324.
(main): Unref the factory if we have one.
(print_element_list): No change -- relies on the
plugin_feature_list_free to free the list of features.
Original commit message from CVS:
2006-01-17 Andy Wingo <wingo@pobox.com>
* gst/gstregistry.c: Reflow design comment. Update so as to speak
in the present tense without reference to versions.
* gst/gstregistry.c (gst_registry_add_plugin)
(gst_registry_remove_plugin, gst_registry_remove_feature)
(gst_registry_find_feature, gst_registry_get_feature_list)
(gst_registry_get_plugin_list, gst_registry_lookup_feature)
(gst_registry_lookup, gst_registry_scan_path)
(_gst_registry_remove_cache_plugins)
(gst_registry_get_feature_list_by_plugin): Add argument
validation.
Original commit message from CVS:
* gst/gstregistry.c: (gst_registry_add_path):
* gst/gstregistry.h:
* gst/gstregistryxml.c:
Fix docs again.
Removed old code from gstregistryxml.c
Original commit message from CVS:
2005-11-21 Andy Wingo <wingo@pobox.com>
* *.h:
* *.c: Ran scripts/update-macros. Oh yes.
* gst/gstobject.h (GST_OBJECT_GET_LOCK, GST_OBJECT_LOCK)
(GST_OBJECT_TRYLOCK, GST_OBJECT_UNLOCK): Renamed from
GST_GET_LOCK, etc.
* scripts/update-macros: New script. Run it on your files to
change GST_LOCK to GST_OBJECT_LOCK, and the same for UNLOCK as
well.
Original commit message from CVS:
* docs/README:
* docs/gst/gstreamer-sections.txt:
* gst/gstbin.c:
doc updates
* gst/gstregistry.c: (gst_registry_scan_path_level):
fix for a nasty little missed situation where an installed plug-in
which was in the cache did not get overridden by an uninstalled one
which was earlier in the plugin path because the newly created plugin
for the uninstalled one (not in the registry) didn't get its
->registered set to TRUE
