ndufresne: I really would like to see visl tested in test-media. Just a basic test running v4l2-compliance and checking that you can safely unbind visl, even if it is in use. Basically similar to the vicodec tests, but without the v4l2-ctl streaming tests.
test-media is also used for regression testing, making sure kAPI changes do not break drivers, and adding visl to test-media helps with that.
broonie: Re https://patchwork.linuxtv.org/project/linux-media/list/?series=11712 : I assume the idea is that the alsa bits are merged via the media subsystem once we're happy with this driver? The ASOC patches are all Acked by you, so that should be fine.
I hope to get this in for v6.9.
hverkuil: Yes.
hverkuil: we can cover unbind in that script, as for regression testing, this will be covered in Collabora gitlab CI (until Linux media cinis ready to have it) using fluster with modified md5, this is already in place in GStreamer gitlab ci
ndufresne: all you need is just run v4l2-compliance. All the codec specific bits are out of scope, but the basic compliance check should still take place in that script. It's easy enough to stick it in.
hverkuil: ack, I'll get someone to finish that
if i was gonna add bcm2835-codec to staging drivers in vc04_services, would i be trying to apply the relevant rpi patches in the media_tree git or the media_stage git? i'm assuming media_stage, because of staging driver?
alien_lappy: media_stage, but not because it's drivers/staging/. the media_stage tree is the first integration stage for all media drivers
this being said
we tried to submit bcm2835-isp to staging
and got told we have to clean up vc04_services and move it out of staging first
you're welcome to help with this work :-)
ic... :-(
pinchartl: as a newbie, i think it's difficult to move in on where others are already working; and i'm more interested in the codecs; i think i'll at least try to get that working for me; then i'll be more knowlegable about these things.
thanks for the help, i'll have to get more knowledgable about all of this, especially the vchiq and mmal stuff
uajain has submitted patches
some have been merged already
oic
i'll have to read up on these
you could maybe start by reviewing the ones that haven't been merged yet, and try to rebase the codec driver on them to make sure it still works
that would be useful
that seems like  a good plan
don't have much time for it atm, but we'll see
i guess the first part of this is to read up on the submitted and non-merged patches, which i assume is on the mailing list?
yes they are
uajain: do you have a recent branch ?