modules/coreboot: set Dasharo coreboot fork rev to the main dasharo branch

[mkopec]

Here are the changes (mostly irrelevant to Heads) that were merged after #1846:

• Support for capsule updates (inactive for Heads build)
• Added NovaCustom MTL dGPU variants
• EC firmware update in coreboot (inactive for Heads build)

Changes that affect V560TU:

• Added support for reading power adapter wattage via ACPI
• Fixes for USB not working
• Add AC/DC loadline programming, for better stability
• Fixes for ME disable
• Fixes for Ethernet detection

I've tested booting into Qubes, ethernet detection, and USB type-A.

tlaurion edit: took #1889 (comment) content

[mkopec]

@tlaurion Ideally Dasharo/coreboot#609 shouldve been merged before #1846 , but it went the other way around, so I'm bumping the coreboot rev to what was merged in the end: https://github.com/Dasharo/coreboot/commits/048ca832325d716fcab596822b10f5d493fc2312/

[macpijan]

Can we explain briefly here in heads upstream MR what has changed since that revision, and which models are affected?
Is there any potential regression, anything to be tested specifically - or what was already tested on your side?

[mkopec]

Mostly changes irrelevant to Heads that were merged in the meantime:

• Support for capsule updates (inactive for Heads build)
• Added NovaCustom MTL dGPU variants
• EC firmware update in coreboot (inactive for Heads build)

Changes that affect V560TU:

• Added support for reading power adapter wattage via ACPI
• Fixes for USB not working
• Add AC/DC loadline programming, for better stability
• Fixes for ME disable
• Fixes for Ethernet detection

I've tested booting into Qubes, ethernet detection, and USB type-A.

[tlaurion]

This is the changeset between pointed commit change
Dasharo/coreboot@94e5f5d...048ca83

[tlaurion]

@mkopec

• just received v560t replacement. Flashed at 3.3v Found XMC flash chip "XM25QU256C/XM25QU256D" (32768 kB, SPI) on ft2232_spi.
• tested on nv41 confirmed pr0 works
• tested on v560tu confirming previous changes changes (functionnaly) and that PRO works (couldn't flash config changes after manually calling lock_chip from recovery shell)
  • boot delays less then 30s after initial flashing to Heads
    • Note that memory training happens twice, once prior of setting system clock, and once after.
    • Note that initial booting time is scary from stock: maybe 3 minutes with an automatic reboot leading to initial ram training
    • Booting time once that is done is ~3 seconds to bootsplash, and from bootsplash to Heads ~7s
  • installing QubesOS works with latest kernel option as per doc for meteor lake
    • Head factory reset works, can boot QubesOS second stage install
      • Second stage install installation of debian-12-xfce seems to take forever but I might be impatient, will see if IO slow later on)
    • Heads can boot final QubesOS install with TPM Disk Unlock Key configured for Default boot option
      • basic functions there including suspend/reume works, network connect, browsing, audio
  • comment on Busybox's fdisk calls warns on >2tb disks #1883 replication on issue with 4tb m2 drive
• merge in master
• merge under TESTING NEEDED: STAGING PR (quiet mode + diceware + nk3 fixes) #1875

[tlaurion]

Something is wrong with second stage QubesOS install, i'm stuck at installing debian-12-xfce template for more than an hour and none of the available terminals provide any interesting output.

Since I can boot into the second stage installer and none of this is related to proposed changes, I'll merge.

[tlaurion]

Merging

[tlaurion]

Something is wrong with second stage QubesOS install, i'm stuck at installing debian-12-xfce template for more than an hour and none of the available terminals provide any interesting output.

Since I can boot into the second stage installer and none of this is related to proposed changes, I'll merge.

Unfortunately, my nv41 did an "instant reboot" today. I have to revert the changes on master and #1875 @mkopec