Add refcounting to the GstAllocator object.
Remove const from functions because the allocator is refcounted now.
Rename the vmethods for consistency
Expose the constructor for GstAllocator and add a destroy notify for the
user_data. This should make it possible to create allocators that are not
registered and shared globally along with the possibility to destroy them
properly.
Update defs with new symbols.
Make the memory object simply manage the data pointer and the maxsize and move
the offset and size handling to common functionality.
Use the READONLY flag to set a readonly lock.
Remove the data and size fields from the unmap method. We need an explicit
resize operation instead of using the unmap function.
Make internal helper lock and unlock functions.
Update unit test and users of the old API.
Remove gst_mini_object_register() and add a GST_DEFINE_MINI_OBJECT macro to
define a _get_type() function for the boxed miniobject.
Remove a bunch of custom _get_type() functions and replace them with the
miniobject macro.
Rename some _init method to _priv_*_initialize() like the rest of them.
Inspired by patch from Johan Dahlin and see bug #657603
Avoid playing with the refcount to decide when a buffer has been recycled by the
dispose function. The problem is that we then temporarily can have a buffer with
a refcount > 1 being acquired from the pool, which is not writable. Instead use
a simple boolean return value from the dispose function to inform the called
that the object was recycled or not.
Don't remove memory blocks from the buffer when we clip and resize, instead set
the memory offset and size to 0. This allows us to make the buffer larger again
later.
Add more debug.
Alow resize to 0 bytes.
Do clipping correctly.
Add more unit tests. Also add a failing test: when we resize to 0 and then
try to resize back to the original size it fails because the memory was
removed.
Allow for negative offsets when doing memory copy and share.
Add fast path in the _get_sizes() function.
Fix resize for negative offset and expanding the buffer.
Add some unit tests.
Also return the offset in a GstMemory block with the get_sizes() method. This
allows us to figure out how much prefix there is unused.
Change the resize function so that a negative offset can be given. This would
make it possible to resize the buffer so that the prefix becomes available.
Add gst_buffer_get_sizes() to return the offset and maxsize as well as the size.
Also change the buffer resize method so that we can specify a negative offset
to remove prefix bytes.
Add an index to gst_buffer_take_memory() so that we can also insert memory at a
certain offset. This is mostly interesting to prepend a header memory block to
the buffer.
Fix the code to support allocating the buffer and memory in one memory block.
Add an extra variable to store the memory of the buffer.
This code is disabled still because of complications.
Make a new method to allocate a buffer + memory that takes the allocator and the
alignment as parameters. Provide a macro for the old method but prefer to use
the new method to encourage plugins to negotiate the allocator properly.
Rename the GstMemoryImpl to GstMemoryAllocator because that's really what it is.
Add an alloc vmethod to the allocator members.
Improve registration of allocators.
Add methods to get and set the default allocator
Always use an allocator to allocate memory, use the default allocator when NULL
is passed.
Add user_data to the allocator Info so that we can pass extra info to the
allocator new method.
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.
Allow for automatic merging of memory block in the _map function and automatic
freeing of the temporary memory.
Remove some unneeded functions.
Add possibility to force writable spanned memory.
Replace subbuffer and copy vmethods by a more generic transform function that
can then be parametrised by transform specific data. This should allow us to
implement make-writable and more future transform functions.
Keep a pointer to the bufferpool. Release the buffer to the pool when
finalizing. Make sure the pool sets itself as the pool member of buffers that it
sends out.
Make separate api for getting and adding metadata. This allows us to pass extra
parameters to the init functions when creating metadata, which is needed for
specific API implementations.
Add beginnings of memory metadata.
Add first implementation of arbitrary buffer metadata. We use a simple linked
linked of slice allocated metadata chunks. Future implementations could use
something more performant.
Add get, remove, iterate methods to handle the metadata.
Make sure we execute the same code path in git versions and in releases,
so just warn when metadata isn't writable when we want it to be instead
of bailing out.
The alignment guaranteed by malloc is not always sufficient. E.g. vector
instructions or hardware subsystems want specifically aligned buffers. The
attached patch will use posix_memalign if available to allocate buffers.
The desired alignment can be set when running configure using the new
--with-buffer-alignment option.
Move the parent buffer pointer into the GstBuffer struct so that we can
remove the subbuffer class and type. This is interesting because it allows us to
more naturally implement methods to get the real type and parent
of a subbuffer (See #545501).
It should also be slightly faster because there is no extra object hierarchy to
initialize and free.
Check for metadata writability when setting caps on buffer or when copying
metadata flags. Only enable these extra assertions in git versions.
This should help us find bad elements.
Original commit message from CVS:
* gst/gstbuffer.c: (gst_buffer_finalize):
Avoid costly typechecking for trivially correct pointers.
* gst/gstpoll.c: (gst_poll_wait):
Add some G_LIKELY here and there.
* libs/gst/base/gstadapter.c: (gst_adapter_push):
Add some debug info.
Original commit message from CVS:
2008-10-06 Andy Wingo <wingo@pobox.com>
* gst/gstbuffer.h (GST_BUFFER_FREE_FUNC): New API, a free function
that will be called on the malloc_data to free it. Basically a way
to avoid subclassing when all you need is a different free
function, i.e. free() instead of g_free().
* gst/gstbuffer.c (gst_buffer_finalize): Free malloc_data via
calling the free function.
(gst_buffer_init): Initialize the free function to g_free.
Original commit message from CVS:
* gst/gstbuffer.c:
Update some comments.
* tools/gst-inspect.c: (print_element_properties_info):
Improve printing of flags.
Original commit message from CVS:
* gst/gstbuffer.c: (gst_buffer_finalize):
Make it once again possible to free GstBuffers in the default
build.
The poisoning scribbles on parts of the miniobject we need in
order to free it.
Fixes#480341
Original commit message from CVS:
* gst/gstbuffer.c: (gst_buffer_finalize):
Implement poisoning for GstBuffer if --enable-poisoning is specified.
When finalizing a buffer the complete struct is filled with 0xff,
thus making a use of the buffer after the final unref impossible.
Original commit message from CVS:
* docs/gst/gstreamer-sections.txt:
* gst/gstbuffer.c: (gst_buffer_try_new_and_alloc):
* gst/gstbuffer.h:
* tests/check/gst/gstbuffer.c: (GST_START_TEST),
(gst_buffer_suite):
API: add gst_buffer_try_new_and_alloc() plus unit test (#431940).
Original commit message from CVS:
* gst/gstbuffer.c: (gst_buffer_make_metadata_writable),
(gst_buffer_create_sub):
Copy duration/offset_end/caps when creating a subbuffer of the
complete parent.
Make the subbuffer read-only when we make the metadata writable for
now. Fixes#351768.
* tests/check/gst/gstbuffer.c: (GST_START_TEST):
Added check for metadata copy when creating subbuffers.
Original commit message from CVS:
* gst/gstbuffer.c:
* gst/gsttask.c: (gst_task_join):
Small doc updates.
* gst/gstpad.c: (gst_pad_activate_pull), (gst_pad_activate_push),
(gst_pad_stop_task):
When pad (de)activation failed for some reason, restore the old
activation mode and set the pad to flushing instead of assuming the
pad is deactivated.
If the _task_join() failed, reinstall the task on the pad so that it can
be stopped later and return an error.
Original commit message from CVS:
* gst/gstbuffer.c: (gst_buffer_get_type), (gst_buffer_new),
(gst_subbuffer_get_type), (gst_buffer_create_sub):
Avoid function call for known types by keeping the buffer and
subbuffer GType global.
* plugins/elements/gstfilesrc.c: (gst_file_src_create_read):
Random silly optimisations in read() path.
Original commit message from CVS:
* gst/gstbuffer.c: (_gst_buffer_copy), (gst_buffer_create_sub):
* tests/check/gst/gstbuffer.c: (GST_START_TEST),
(gst_buffer_suite):
No point in checking if the size of the subbuffer > 0, the
code handles it correclty as demonstrated by unit test.
Also add a unit test for the zero sized _new_and_alloc and
_copy. Fixes#346663.
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/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.c: (gst_buffer_make_metadata_writable):
gst_buffer_make_metadata_writable() should maintain the
buffer flags (those that make sense at least) (see #340859).
Original commit message from CVS:
* gst/gstbuffer.c: (gst_buffer_is_metadata_writable),
(gst_buffer_make_metadata_writable):
* gst/gstbuffer.h:
* libs/gst/base/gstbasetransform.c:
(gst_base_transform_prepare_output_buf):
* plugins/elements/gstcapsfilter.c: (gst_capsfilter_prepare_buf):
* tests/check/gst/gstbuffer.c: (GST_START_TEST), (gst_test_suite):
Replace gst_buffer_(make|is)_metadata_writable patch now
that the release is out.
Original commit message from CVS:
* gst/gstbuffer.c:
* gst/gstbuffer.h:
* libs/gst/base/gstbasetransform.c:
(gst_base_transform_prepare_output_buf):
* plugins/elements/gstcapsfilter.c: (gst_capsfilter_prepare_buf):
* tests/check/gst/gstbuffer.c: (gst_test_suite):
Back out patch until after the release.
Original commit message from CVS:
Reviewed By: Andy Wingo
* gst/gstbuffer.c: (gst_buffer_is_metadata_writable),
(gst_buffer_make_metadata_writable):
* gst/gstbuffer.h:
Add gst_buffer_(is|make)_metadata_writable as analogues of
gst_buffer_(is|make)_writable.
* libs/gst/base/gstbasetransform.c:
(gst_base_transform_prepare_output_buf):
* plugins/elements/gstcapsfilter.c: (gst_capsfilter_prepare_buf):
Use name gst_buffer_(is|make)_metadata_writable functions.
* tests/check/gst/gstbuffer.c: (GST_START_TEST), (gst_test_suite):
Test gst_buffer_(is|make)_metadata_writable
(Closes: #324162)
Original commit message from CVS:
* gst/gstbuffer.c: (gst_buffer_span):
gst_buffer_span should copy the timestamp of the first buffer
if they were both originally overlapping subbuffers of the
same parent, using the same logic as the 'slow copy' case.
Original commit message from CVS:
* docs/random/ensonic/media-device-daemon.txt:
more ideas (dbus)
* gst/gstbuffer.c:
fix doc example, add clarification
* tools/gst-launch.1.in:
add initial info about GST_PLUGIN_PATH, needs more work
Original commit message from CVS:
* gst/gstbuffer.c:
Small docs update.
* gst/gstcaps.c: (gst_caps_is_equal):
Don't assert on NULL <--> X. Fixes#323260
* gst/gstminiobject.c: (gst_mini_object_replace):
If we're doing atomic operations, we might just as well use
the proper way to get an atomic pointer.
* libs/gst/base/gstbasesink.c: (gst_base_sink_get_position):
Clean up debugging.
Original commit message from CVS:
* gst/gstbuffer.c: (_gst_buffer_copy):
Copy more flags.
* gst/gstcaps.c: (gst_caps_is_equal):
Fix some docs.
Make _is_equal fast in the trivial cases.
* gst/gstminiobject.c:
* gst/gstminiobject.h:
More docs. Spifify .h file.
* gst/gstutils.c:
Small doc update.
Original commit message from CVS:
* check/gst/gstghostpad.c: (GST_START_TEST), (gst_ghost_pad_suite):
Added check for bug #317341
* gst/gstbuffer.c:
* gst/gstbuffer.h:
Some more spiffifying.
* gst/gstghostpad.c: (gst_ghost_pad_do_link):
Call peer linkfunction if we are a source pad. Totally fixes
#317341
* gst/gstpad.c:
Update docs, source pads should call the peer linkfunction
so they can atomically perform the pad link.
Original commit message from CVS:
* gst/gstbuffer.c: (gst_subbuffer_init):
Set READONLY flag on subbuffers, so that gst_buffer_make_writable()
works correctly in all circumstances (we could have just copied
the parent buffer's readonly flag, but conceptually it seems
cleaner to mark all subbuffers as read-only). (based on patch
by Alessandro Decina, #314710).
* check/gst/gstbuffer.c: (create_read_only_buffer),
(test_make_writable), (test_subbuffer_make_writable),
(gst_test_suite):
Add some tests for gst_buffer_make_writable().
