16:38 <chewitt> are there any known issues with ELD/EDID data on RK3588 ?

16:38 <chewitt> I'm working with a Rock 5B board, and it's spewing this into logs:

16:38 <chewitt> [ 32.054227] hdmi-audio-codec hdmi-audio-codec.0.auto: HDMI: Unknown ELD version 0

16:38 <chewitt> [ 32.054678] hdmi-audio-codec hdmi-audio-codec.0.auto: ASoC error (-19): at snd_soc_dai_prepare() on i2s-hifi

16:38 <chewitt> [ 32.054858] hdmi-audio-codec hdmi-audio-codec.0.auto: ASoC error (-19): at snd_soc_dai_prepare() on i2s-hifi

16:39 <CounterPillow> iirc the ELD version 0 warning is some overly verbose warning BS that was added recently

16:39 <chewitt> I've initially suspected the known-broken EDID data on the little 10" LCD that I use for general fiddling

16:40 <chewitt> but I've replaced that with an Raspberry Pi 15" screen (known not-broken) and it's still the same

16:40 <CounterPillow> HDMI audio also still does work if an HDMI connector is plugged in, I'm not sure why it logs the ENODEVs now if it's unplugged.

16:40 <chewitt> i'm unable to run "edid-decode" though .. it just hangs

16:41 <CounterPillow> that may be completely unrelated to those warnings

16:41 <chewitt> on other devices (Amlogic boards, RPi boards) with broadly the same distro image it outputs fine

16:42 <CounterPillow> anyway, I did look into the warnings myself: https://gitlab.collabora.com/hardware-enablement/rockchip-3588/linux/-/issues/40

16:44 <chewitt> "These log messages can be triggered very frequently in some situations, especially ELD parsing when there is nothing connected to a HDMI port"

16:44 <chewitt> there is definitely something connected to the HDMI port :)

16:45 <chewitt> I'm also seeing this error reported:

16:45 <chewitt> rockchip-pm-domain fd8d8000.power-management:power-controller: Failed to create device link (0x180) with supplier spi2.0 for /power-management@fd8d8000/power-controller/power-domain@12

16:46 <chewitt> I'm using 6.16-rc5

16:46 <mmind00> chewitt: for the edid thing, is the "recent" drive-strength change already in the kernel you built?

16:47 <mmind00> chewitt: https://git.kernel.org/pub/scm/linux/kernel/git/mmind/linux-rockchip.git/commit/?h=v6.16-armsoc/dtsfixes&id=af9feb0b85f92d2972061224839c5fea5ee39f6d

16:48 <chewitt> no .. I'll pick that too

16:57 <CounterPillow> chewitt: the power-controller device link messages are also inconsequential, they are produced as the kernel is actually breaking the dependency cycle so it's already handled properly.

16:58 <chewitt> ahh, good to know, thanks

16:59 <chewitt> I was seeing more of them on the collabora rockchip-devel branch so reverted to the default kernel

17:01 <chewitt> next up will be trying to figure out why Kodi cannot find a drm surface to render into

17:01 <chewitt> debug <general>: CDRMUtils::DrmFbGetFromBo - failed to add framebuffer: Invalid argument (22)

17:01 <chewitt> error <general>: CDRMAtomic::FlipPage - Failed to get a new FBO

17:02 <CounterPillow> you should make sure you build with panthor, and that you have the mali_csff firmware from linux-firmware

17:02 <CounterPillow> *build your kernel with panthor, that is

17:03 <chewitt> yup, that's present

17:03 <chewitt> I suspect something is needed in distro packaging on the mesa side

17:05 <chewitt> info <general>: GL_RENDERER = Mali-G610 (Panfrost) <= that's probably wrong

17:06 <CounterPillow> no that's correct

17:06 <CounterPillow> just the kernel part is called panthor

17:11 <chewitt> mesa driver is still 'panfrost' ?

17:23 <chewitt> still seeing that ELD/audio spam https://paste.libreelec.tv/fast-raccoon.log

17:27 <chewitt> hmm.. interesting

17:27 <chewitt> if I move the HDMI connection to HDMI-A-1 then the spam stops

17:28 <chewitt> if the HDMI connection is on HDMI-A-2 then it spams the same as if that patch isn't applied to the kernel

17:28 <chewitt> I still can't get output from "edid-decode" though

19:12 <helloRockiper>  Hello. Can I get some help getting U-Boot running on a Rockchip RK3188 device (Starway Andromeda S8)? U-Boot has SSD2828 display driver (same as my device) and I'm not sure how to make it all work.

19:13 <helloRockiper> I switched to U-Boot after failed attempt at booting mainline Linux. (I have no visible signs of life for my device such as USB or display output). Hence wanna try U-Boot

19:14 <diederik> I'm seeing a LOT of those ASoC errors on multiple devices. And the HDMI connector is plugged in

19:15 <helloRockiper> I said no visible signs of life for my device. I mean the device is working absolutely fine, it's just with mainline Linux it doesn't show a lot of life

19:19 <diederik> I've seen the ones that CounterPillow linked to, but also got other ones (with kernel 6.12.21): https://paste.sr.ht/~diederik/a0ca23a8439ade12d18a9b26997df32ac3b3aab8

19:21 <diederik> AFAICT Mark Brown's patch *reduces* the spam, but doesn't fix the (underlaying) issue