Rg34xx-h HDMI initialization errors

Issues 2601 Jacaranda
,
1

Device: Anbernic RG34XX-H, /opt/muos/device/config/board/name reports rg34xx-h, 1GB RAM (973 MiB reported)
muOS: 2601.1 Funky Jacaranda (also reproduced on 2601.0), kernel 4.9.170, sun50iw9p1
Display tested: desktop monitor, input pinned to HDMI (not auto). Also tried an LG OLED TV.
Cable: mini-HDMI (Type C) to HDMI, single captive cable, no adapter. Only cable I have.

Symptom: muOS never enters HDMI mode. Internal screen stays on, no signal reaches the display, boot_hdmi=false every boot.

What I’ve verified:

  • /opt/muos/device/config/board/hdmi = 1
  • /opt/muos/device/config/screen/hdmi → /sys/class/extcon/hdmi/cable.0/state
  • With cable seated in a powered-on display, that node reads 1 and holds steady for 30+ seconds
  • HDMI Settings menu is present and populated; tried 720p60 and 480p60, RGB/Full and YUV/Limited
  • Cold shutdown from the device power menu, cable attached before power-on (not hotplug, not warm reboot)
  • Fresh 2601.1 flash, not an upgrade

dmesg with cable attached to a live display:

[    1.639078] boot_hdmi=false
[    1.639795] ERROR: pinctrl_get for HDMI2.0 DDC fail
[    1.641789] hpd status change!
[    1.662096] _read8-Error:ASSERT I2C DDC Read extended timeout - check system - exiting
[    1.663879] hpd status change!
[    1.682627] _read8-Error:ASSERT I2C DDC Read extended timeout - check system - exiting
... (repeats ~16x through 2.17s)
[   14.654304] hdmi_hpd_sys_config_release

Also Can not find the node of esm, and both HDMI regulators fall back to dummy (vcc-hdmi / vdd-hdmi not found).

/sys/class/disp/disp/attr/sys shows only screen 0 (internal LCD, 720x480). No screen 1 ever appears.

Questions:

  1. Does ERROR: pinctrl_get for HDMI2.0 DDC fail appear on your boot log if you have HDMI working on an RG34XX-H? Trying to tell whether it’s benign boilerplate or a real device-tree gap for this board string.
  2. Has anyone confirmed working HDMI output on rg34xx-h specifically (not 34XXSP or 35XX)?
  3. boot_hdmi is set pre-kernel by u-boot and logs before the first HPD event — is that expected ordering, or should u-boot be sampling HPD first?

Any insight is appreciated, thanks!

2

Are you sure about that? I feel like this issue is being hallucinated by something else and not actually tested properly since I have no issues here.