Make the alignment parameter optional when negotiating buffers.
Default to a 16 bytes alignment and adjust for the max cpu
alignment.
Remove the useless align buffer parameter in plugins, we always
set it to 16 anyway.
Gstreamer simply mmap()s the DMA buffers. This can be very very slow on
some hardware, so don't ask for it. For gstreamer to support DMA
buffers, it will need to support a proper way of transferring the
buffers from the GPU to the CPU, e.g. using EGL/OpenGL.
The current code has several issues if a pipewiresrc is stopped and deleted
while a buffer is still in use downstream.
- pw_stream_queue_buffer() is never called for the corresponding
pw_buffer. As a result, the busy counter is never decremented and
pw_stream_dequeue_buffer() of the corresponding pipewire output will
return NULL whenever it encounters this buffer.
- The pipewiresrc does not own the buffer reference unless the buffer is
queued in the source, so calling gst_buffer_unref() unconditionally
causes refcount issues for the buffer.
- buffer_recycle() can race with on_remove_buffer() and
gst_pipewire_src_close(). As a result, buffer_recycle() may access pwsrc
when it was already deleted.
The buffer has its own reference to the pool. So the pool object lock can
be used to ensure that the pwsrc, core and stream remain valid in.
buffer_recycle(). If the 'dispose' function pointer was already cleared,
then on_remove_buffer() has already finished, so abort early.
With the pool lock held, it is save to access the pipewire loop. Now the
loop lock can be used to synchronize with on_remove_buffer(). 'dispose'
must be checked again in case on_remove_buffer() was triggered by something
other than gst_pipewire_src_close().
In on_remove_buffer() unref the buffer if it is queued. Otherwise call
pw_stream_queue_buffer() to ensure that the busy counter is decremented
correctly.
PipeWire does not give DMABUF anymore to clients unless they
explicitly ask for it. This results in v4l2 sending an fd and
the mmapoffset + size fields in the buffer.
GStreamer does not handle the mmapoffset in the fd allocator and
thus runs into mmap problems when using the memory.
Enable DMABuf explicitly to work around the problem until
GStreamer is fixed.
Fixes#1115
Don't do any GStreamer data transport from the PipeWire callback
because it might block in preroll and block our communication with
PipeWire. Instead, take the caps and wake up the caller to continue
with negotiation.
Add a property to periodically send the last buffer to keep the
stream alive. Useful for sparse streams that need to keep the
encoder busy every once and a while.
Make all sources in the same process with the same fd share the
connection to the server. This makes it possible to set the same
fd on multiple sources/sinks and have them all use the same
connection, like when capturing multiple monitors from screencast
with the portal.
Fixes#241
Add an option to resend the last buffer on EOS with an updated
timestamp. This can be used to make sure encoders fill up the
gap between last buffer and EOS, like with sparse streams from
screen capture.
This makes sure we first nicely remove the stream from the server
and then close the socket.
If we don't do this, the disconnect might not have flushed out our
disconnect and the server is left with a non-responsive node,
especially if the disconnect on the core was done with a socket from
the portal that is still open.
Pipewire might update buffers that have not been recycled by GStreamer,
because they are still used downstream. There is nothing we can do about
it in the pipewiresrc.
If a buffer is sent downstream more than once, take an additional
reference to make sure that we don't queue a buffer that is still used
and print a warning.
By default, the pipewiresrc tries to negotiate 16 buffers. This value is
hard coded in the pipewiresrc. If the buffers are large, this could lead
to an undesirably high memory usage. Applications that know about the
buffer size and that fewer buffers are sufficient should be able to
configure the limits for the number of buffers that are negotiated.
Therefore, add the min-buffers and max-buffers properties to the
pipewiresrc to enable applications to configure limits for the number of
negotiated buffers.
Messages that are printed for every buffer should use the LOG debug
level, while messages that happen during setup and tear down should use
the DEBUG debug level.
Therefore, use LOG in on_process and when popping buffers and DEBUG when
Pipewire adds or removes buffers.
Make the thread_loop alloc its own loop by default to simplify
some core. Add extra new_full method to pass a custom pw_loop.
Make other loop implementations ready to support custom loops
if we want that later.
The proxy API is the one that we would like to expose for applications
and the other API is used internally when implementing modules or
factories.
The current pw_core object is really a context for all objects so
name it that way. It also makes it possible to rename pw_core_proxy
to pw_proxy later.
The pw_remote object is really a wrapper around the pw_core_proxy.
The events it emits are also available in the core proxy and are
generally awkward to use.
With some clever new pw_core_proxy_* methods and a pw_core_connect
to create the core_proxy, we can convert all code away from pw_remote.
This is a first step in this conversion, using the pw_remote behind
the scenes. It leaks into some places because it really needs to become
its own struct in a next step.
Remove some of the unused states in pw_stream. The app can know the
state by following the format and buffer events.
Make it possible to be notified of io are updates. This should make it
possible to follow the transport etc.
Make it possible to be notified of any param changes.
Rename finish_format to update_params because that is what it does.
Make this work in the same was as the filter: updating the params
removes all old params of the types and installs the new ones.
Don't get the Props and PropInfo from the node proxy, instead get them
directly from the adapter that we have locally. Update the controls
directly on the adapter instead of going to the server first.
Not closing the fd causes leaks in existing apps. It's probably better
to always close it and let apps deal with that by using dup or similar.
Make gst sink and source dup the fd before connect_fd().
Fixes#181
This is necessary for 'do-timestamp' to work if the source provides no
timestamps. Without this, the timestamp from the first use will remain,
because the basesrc only overwrites timestamps that are
GST_CLOCK_TIME_NONE.
Automatically parse and build key/value when in objects without having
to prefix the key with ":"
Automatically build control/value when in sequence without the "."
prefix.
Remove the builder with key/pod, taking a reference to the stack built
temporary pods is not allowed in c++. We can use the varargs version
with the same convenient syntax.
Remove the parser "*" option, it is unused.
Improve spa_pod_builder_add_* and spa_pod_parser_get_* and make them
look similar.
