Hey,
On Fri, 29 Nov 2019, Takashi Iwai wrote:
On Fri, 29 Nov 2019 15:47:11 +0100, Kai Vehmanen wrote:
This second patch is however trickier. Nikhil your patch changed the default allocation a bit, so the routing might be difference also with snd-hda-intel (i.e. not SOF) for existing platforms and this may surprise users.
Well, but the allocation itself is dynamic for DP-MST, even on Intel, so user can't expect the completely persistent index assignment. That's the reason I took Nikhil's patch (even I suggested to simplify in that way).
[...]
We had a trick to assign the primary index. It still works, right? That should be the only concern.
[...]
This is a platform-specific part, and on Intel, the assumption has been that pipe is equivalent with dev_id. If this changed, of course, we must reconsider the whole picture.
hmm, the pipe equivalency should actually still hold. Looking at the code more, this could also be a lurking bug on graphics driver that had new side-effects with the recent ALSA side changes. E.g. I've received logs where dev_id=1 is for a single connected HDMI monitor. I need to investigate more whether this is an expected behaviour or a bug. :)
PS I did not have time to fully test the RFC patch, so this is just for discussion now...
Since the assignment should work with your patch somehow, I already applied it. Let's do fine tune-up during 5.5 rc cycles, if any.
Ack, ok. My commit message is a bit confusing (the wording about MST is not correct) but the actual code restores original behaviour so this should be good to apply. I'll continue testing and also dig a bit deeper into the bugreports w.r.t. what happens in the problematic non-MST cases. Thanks for the quick reviews!
Br, Kai