Hi Dalao,
On 30.03.20 03:08, Dalao via coreboot wrote:
This might be easy to fix. What are the required IDs and what are the current ones with coreboot?
The normal id on vendor bios is: PCI\VEN_10EC&DEV_5227&SUBSYS_220E17AA but on coreboot it becomes PCI\VEN_10EC&DEV_5227&SUBSYS_522710EC . It seems wired because the coreboot's devicetree.cb file did have a line subsystemid 0x17aa 0x220e inherit.
this looks much like the ID of the card reader, and it didn't get the subsystem ID because it's not mentioned in the devicetree. Only devices that are explicitly mentioned will be considered "onboard" and inherit the IDs. You can add a `device pci 0.0 on end` below the `1c.0` bridge.
PCI Data Acquisition and Signal Processing Controller PCI\VEN_8086&DEV_0C03&SUBSYS_00000000&REV_06\3&B1BFB68...
No idea what this is.
I searched online on this VEN and DEV it seems it's the Intel DPTF stuff, but I don't know is it doing anything useful? I found some overclock fourm suggest to disable it. How to disable it in coreboot?
According to your lspci it's 00:04.0, you can just disable that in the devicetree as coreboot knows how to handle it. It probably doesn't show up with every processor SKU. And Intel didn't really document this device. So they probably were just experimenting for future platforms and it should always be disabled.
I also found another issue when i use a engineering sample cpu. When I use libgfxinit, the es cpu can boot in Windows but not in Manjaro. When I use extracted VGA BIOS, the es cpu can boot in Manjaro but not Windows... I know es cpu is known to have bugs, but it can boot in both using vendor bios. I have attached the journalctl logs if someone could help me.
It stops roughly at the point where it loaded the graphics driver `i915` in the `normal-cpu` log. You can try again with `drm.debug=0xe` in the kernel command line. It might give more output before it hangs.
Nico