11:19 <DavidHeidelberg> digetx: I've been looking into them. They share properties, but not always having same, putting them into one file would be a lot of ifs. I'm always concerned about readability of doc, when you have milions of `if`

11:19 <DavidHeidelberg> s/them./`tegra-audio-*`./

11:30 <digetx> all those properties are common and depend on a board configuration, so you should make those gpios optional and that's it

12:06 <DavidHeidelberg> yeah, did that, now testing how it'll work out with dt-schema checks

13:21 <DavidHeidelberg> dtbs_check seems to be happy, pushed :)

14:32 <DavidHeidelberg> any idea what is gpio ear_sel?

14:44 <DavidHeidelberg> found out, it's Sergey implementation of int/ext mic switching

14:47 <digetx> yes, don't add those new non-upstream bindings into that patch

14:48 <digetx> convert non-upstream bindings to the new common binding in a separate patches

15:17 <DavidHeidelberg> btw. with cc8f70f5603986a99e7775f3cc4a10d337b82a4d you kinda wiped one property which been parsed

15:17 <DavidHeidelberg> `nvidia,ac97-controller` no longer exist in unified version

15:18 <DavidHeidelberg> WM9712 seems to be upset about it ;-)

15:22 <DavidHeidelberg> I'll make GH issue for it

15:22 <DavidHeidelberg> and for now also add `ac97-controller` to the docs

15:26 <DavidHeidelberg> it's sad we have support for these devices in kernel, but no-one using them.. At least with postmarketOS, there is few users.

15:26 <digetx> add oneOf for nvidia,ac97 and nvidia,i2s

15:29 <DavidHeidelberg> ok

15:31 <DavidHeidelberg> * ok, done

15:32 <digetx> I actually broke nvidia,ac97-controller when unified sound drivers, need to fix it now

15:32 <DavidHeidelberg> digetx: should I make an issue or you're up to it right now?

15:35 <digetx> no need for issue

18:22 <DavidHeidelberg> it's kinda heartwarming that docs fixes helps discover regressions :)

18:31 <digetx> the ci is also handy, it helped to spot couple obscure problems that I would miss otherwise

19:28 <DavidHeidelberg> digetx: I added multiple routing info for each codec

19:29 <DavidHeidelberg> * each codec, it was incomplete before

20:13 <DavidHeidelberg> digetx: do u strongly object against adding definitions of not upstreamed drivers? I'm asking because as I noted, Rob generally says it's about HW not the drivers.

20:15 <digetx> you may add them, but in a separate patch; otherwise it looks like you trying to sneak in new props

20:18 <DavidHeidelberg> oh, right, I'll separate it

22:32 <DavidHeidelberg> digetx: do u insist on pattern for 3 sgtl5000 devices?

22:58 <digetx> no

23:03 <digetx> why both old and new drm uapis are broken in upstream

23:05 <digetx> pretty sure old worked a week ago

23:09 <DavidHeidelberg> digetx: then the PR is ready :)