Remove the core permission check callback. We can now use the per
client permission configuration.
Rework the flatpak module to use the permissions. When a client
connects, do the portal call and iterate all globals, updating the
permissions. Also update the permissions of newly added globals.
The client is owner of itself.
Don't pass around the client object but keep track of the current
client in the core object. This way we don't need to add a client
argument to functions and can check security when needed.
Pass client to some core functions to make it possible to check
permissions when iterating globals.
Check permissions of factory before using it.
Check permissions in link factory, only allow linking of nodes we can
see.
Check permissions in the autolink module, only try to link to nodes
we can see.
Make client permissions update behave like an atomic update of an
unordered set of permissions, which is perhaps a bit more intuitive.
Load and register the module in one go because we need to have the
global set when calling the init function so that factories and nodes
can use this as the parent global.
Set the name in the global properties.
Make a steal_fd method on the remote.
Add a hello method that sends the initial core info and types to
the client.
With an explicit method we can do this multiple times when we steal
the fd from a remote and use it to make a new remote.
Add properties to global objects to make it easier to select what
nodes and objects to bind too. They can also be used to implement
permissions on the globals based on properties.
Add more error handling in _register.
Make more functions return a result code.
Make a separate pw_module_register, like all other objects.
Fix video in SDL examples by invoking rendering in the same thread
(pw_loop) that created the window, renderer and texture.
At least on some systems, no video is shown without this patch. In
those error cases, SDL_RenderPresent triggers an SDL error "The
specified window has not been made current". That, in turn, has to do
with how SDL uses the current thread for determing the current window.
In general, the page linked to below states that the SDL API is not
designed to be used from multiple threads.
Reference: https://wiki.libsdl.org/CategoryRender
Don't unlink the ports of an inactive link because then it might be
possible that the scheduler can't recycle a buffer, instead use
a port flag to mark the ports disabled and change to scheduler to
skip those ports.
Make a link factory and use create-object to make links. That way
we can have different kinds of links based on the factory and we
can also hide the factory when link creation should be blocked.
Add an interface that can manager a dbus connection.
Make a dbus interface in the core that can create connections running
in the core main loop. Keep this as support for spa plugins.
Make a new PropInfo parameter that allows us to iterate all
properties. Make some new fields to set labels and names for
properties. We will be able to add more things to describe the
properties this way.
Use the Props param to simply query or set properties.
We can now make int enum properties and describe all possible
values with labels, we don't need to register types anymore.
This then makes it possible to enumerate the v4l2 controls and make
them available as control params.
Add a permissions call that can be used to update the permissions
of global objects. This can be used to set specific permissions
on certain objects, hide objects and define the permissions on new
objects. The system is quite generic so that we can extend it later
with more complicated match rules or so.
Don't recurse into the directory of systemd when it was not found.
There are still some issues because libudev is actually provided
by systemd these days and we still need that for building the v4l2 and
alsa monitors.
Remove the per-port memory and only use per-node memory. This will
make it possible to share memory between ports in the future.
Keep refs to memory on the buffers and free (close) the memory
when no longer used.
Do not send the offset and size in the add_mem call, just send the
fd and the flags. The area that we need to map from this to find the
meta, chunk and data are sent in a separate call. This should make
it possible to truncate the memory to a larger size to dynamically
allocate more shared memory for a client.
Remove the Id data type, it's not needed.
Don't automatically map memory in remote.c
Pass the original memory type from server to client.
Handle DmaBuf mem in video-play now that the server passed it on.
Add control objects for all controllable properties on ports.
Try to link compatible control properties in autolink.
Allocate shared memory for the output property memory and configure
the io area on the ports when the controls are linked.
Send the shared memfd to clients when the io area is configured.
Add port_set_io support in remote.c, mmap the control io area and
set on the port.
Add some param helpers
Add volume control to export-source update the volume before
sending each buffer.
Make an int return from pw_node_for_each_port() so that we can return
errors or early exit.
Add pw_port_for_each_param() to iterate params. And use this to collect
the formats.
Reorganize the io parameter ids and objects. Make separate enumerations
for buffer, control, input and output properties.
Add a volume output property to export-source. This is still unused
but will eventually be routed to a PropsIn io area where it can
control the volume of a mixer, for example.
