02:57 <bamse> shawnguo: linux option installs RT_PROP and dtb in the configurationtable, alongside acpi table...so it will work just fine to boot both

03:00 <shawnguo> ah, nice! so the option can hopefully go away and just always be on in the future. thanks bamse, mani_s!

03:22 <shawnguo> hmm, then it means we will need to remove dtb from esp in order to boot Linux from acpi. that's should be fine though.

03:22 <bamse> we're not booting linux from acpi

03:22 <bamse> but for your experimentation...sure

03:22 <bamse> or you can disable the linux option in bios and boot with efi=noruntime and you should be good

03:32 <shawnguo> yeah, acpi can be enforced by setting `acpi=on` or `acpi=force` anyway

03:46 <bamse> ahh cool, didn't know

03:49 <steev> we still can't access that though right? not without qzed's driver?

03:52 <bamse> steev: efi variables?

03:52 <steev> yeah, i'm assuming we will want to install new dtbs into that?

03:53 <bamse> steev: that's unrelated, but there's no change on that front yet...so it's still required

10:57 <kettenis> so what path does it load the dtb from? and does it distinguish between the different models (display panels, touchscreen i2c address, etc.)?

14:37 <robclark> kettenis: in theory if the edid isn't busted the different panels should work with a single dtb.. you basically just need the dtb to have the worst-case delays configured for panel-edp, ie. hpd-reliable-delay-ms .. not sure about the other deltas, presumably the dtb can just have the superset (ie. include touchscreen even if device doesn't have it)??

15:49 <bamse> right, we just specify the panel as a generic "edp-panel" and edid does the rest

15:50 <bamse> but we have two touchpads, so that would be a good candidate for some switching logic...

20:12 <clover[m]> <shawnguo> "yeah, acpi can be enforced by..." <- tried. doesnt boot