Hi,
> You are using me_cleaner.
Some of those issues remind me of ones I noticed when using me_cleaner (vendor firmware too only worked with the soft-disable-only flag). I'd be curious to know what ME_CLEANER_ARGS is set to.
I think it's somewhat common knowledge that removing the MFS/EFFS partition is a bad idea. Per this, I'm extending the above to assume that whitelisting the partition could work.
Hi Dalao,
On Fri, Mar 13, 2020 at 3:36 PM Dalao via coreboot
<coreboot@coreboot.org> wrote:
>
> I'm trying to build coreboot for T440p and still can't boot. I have tried the official repo's latest master branch, it can't boot. The power button led is on, keyboard led on for a second and then off, fan is on, but the screen is not. I don't have a debug device. Ordered a FT232H but it's on the way. I don't know what's the problem. So I looked around trying to find a working one. I also tried the official repo's 4.11_branch, it's the same problem.
I would suggest enabling SPI flash console, which writes logs to the
SPI flash chip. Build, flash, and try booting. Then, power off and
read the flash chip back. There should be a log somewhere inside the
CBFS.
> I also tried different configs use LIBGFXINIT or use VGAROM, and Tianocore or Seabios payload. Still the same problem. The most recent config is:
> https://pastebin.com/7vnji9i2
You are using me_cleaner. Try not using it, as it can break things.
Ideally, just selecting USE_BLOBS (needed for microcode), selecting
the mainboard and adding the mrc.bin should result in a bootable
build. It will print a large warning, though: since the IFD/ME/GbE
regions were left empty, flashing the result as-is will not work. On
the t440p with two flash chips, you only need to flash the last 4 MiB
of the 12 MiB coreboot.rom into the 4 MiB chip.
libgfxinit should just work. The VBIOS is less likely to work fine on
the first try, as extracting it is more error-prone.
> The sha1sum of the blobs I obtained are:
> $ sha1sum mrc.bin
> d18de1e3d52c0815b82ea406ca07897c56c65696 mrc.bin
Okay, that matches with the mrc.bin of peppy. The one I use (from
wolf) has a different hash, no idea as to why.
> $ sha1sum pci8086,0416.rom (obtained through linux kernel)
> 4074e1fa2f788f91d3612b6fe861c9c3faf5560a pci8086,0416.rom
>
> I also tried archfan's repo for T440p but still can't build. It has some issues with submodules https://github.com/archfan/coreboot/issues/12
>
> I flashed back backuped original bios, and it can boot. I assume it's still a software issue with my coreboot build. How to get a working snapshot of coreboot, submodules, and seabios/tianocore at the time when the T440p can work?
That's good to hear.
> _______________________________________________
> coreboot mailing list -- coreboot@coreboot.org
> To unsubscribe send an email to coreboot-leave@coreboot.org
Best regards,
Angel Pons
_______________________________________________
coreboot mailing list -- coreboot@coreboot.org
To unsubscribe send an email to coreboot-leave@coreboot.org