@Second_Coming both sides are green for a while, then turn off. If I plug in the power cord the light turns to orange/red.
Now the power light is turning off, then turning back on automatically with the fans turning on. After 30-45s it keeps repeating that…
Ubuntu 22.04, recent change would be BIOS update to 3.10 and the Ubuntu post install script/guide a few days ago.
Leave the laptop plugged in / charging for about 2 hours. Then try to power it on again.
Your RTC battery has depleted. The laptop is trying to go through memory training / testing (just like when you first got the laptop or when you’ve newly updated the BIOS). However, while in this memory training mode (green LED on)…the RTC battery doesn’t seem to have sufficient power to last till the end of training process.
Note: You can then continue to use the laptop once it has powered up successfully. But leave it plugged in and charging for a day.
Will do… I did just note that on turning it on the green light flashed a few times then the keyboard backlight turned on.
Will wait for a few hours. It’s just odd because I daily drive this machine and had it plugged in for a lot of hours yesterday and then powered it off… Will edit this post in a couple hours.
Take a look at this:
Laptop stuck in memory training, won’t boot - Community Support / Question - Framework Community
You may / may not need to do a mainboard reset if the board is in some weird purgatory state.
In your case, may not be drained RTC battery, now that you’ve mentioned it’s your daily driver. But you’re running into some kind of bug that requires mainboard reset.
Followed the main board reset guide and I’m back up and running again. Hope that doesn’t happen again…
Thanks for the tips @Second_Coming !
Edit: Booted into Ubuntu…closed the lid to tighten the back screws and now it’s doing the same thing again…did the mainboard reset again and going to keep it on for a while.
Edit 2: Still having the same issue. Submitted a ticket to the Framework team. In the meantime, I think I’m going to try to revert my BIOS back to 3.07 since I need this machine everyday for studying…
Regardless of the root cause, this inability to power on / unreliable behaviour can be really annoying. Imagine if you have meetings to attend, notes to take…etc.
I dont know if disabling secure boot is a hard requirement.
I successfully updated from 3.07 → 3.10 with secure boot enabled on Fedora 36.
Anyone else having issues booting Windows 10 after the update? I went from 3.07 → 3.10 using the fwupdmgr method in Fedora 36. If I set the EFI boot priority to use Windows 10 by default, Windows will boot. If I try to use Grub2 to chainload Windows 10, it just goes to a black screen with the Framework logo.
for those that had the problem of booting into the os after running fwupd to update the bios:
if you don’t have an option in grub/systemd-boot/refind/… about fwupd you have to do it manually byrebooting it into the uefi bootloader:
Failure to upgrade to BIOS 3.10! Running Fedora 36 (EDIT: resolved; see below)
Tried “fwupdmgr update”. After “Upgrade System Firmware from 0.0.3.7 to 0.0.3.10” it offers also an “Upgrade UEFI dbx from 33 to 77”, but this fails:
Blocked executable in the ESP, ensure grub and shim are up to date: GDBus.Error:org.freedesktop.DBus.Error.UnknownMethod: No such interface “org.freedesktop.UDisks2.Filesystem” on object at path /org/freedesktop/UDisks2/block_devices/nvme0n1p2
restarting the system does boot differently: some tiny-print version number appears in the top-left. But after that the system reboots and I’m back to my normal grub and my BIOS is still at 3.7.
EDIT: The “UEFI dbx” problems seems to have gone away by doing a “fwupdmgr update --force”. Now running “fwupdmgr update” or fwupdmgr install --allow-older" seems to correctly schedule the System Firmware Upgrade and offer to reboot (it didn’t before; apparently due to the dbx failure). Same problem occurs then, though: system reboots, there’s a tiny version number in the upper-left of the “Framework” boot screen, system reboots almost immediately, and I’m back at grub.
Also tried EFI Shell (which successfully got me to 3.7 in January). I can successfully boot into the EFI shell from USB. It gives me the screen:
Please do not remove the AC power!
Insyde H20FFT (Flash Firmware Tool) Version 2.12
Copyright (C) 2000-2020. Insyde Software Corp. All Rights Reserved
Current BIOS Model Name: GFW30
New BIOS Model Name: GFW30
Current BIOS Version: GFW30.03.07
New BIOS Version: GFW30.03.10
Error:
Update BIOS failed!
Shell> _
No indication about what failed, unfortunately.
EDIT: I finally did succeed via the fwupd route, but I’m not entirely sure what did the trick. Uninstalling an old kernel did not seem to do much for changing the space used on /boot/efi (/boot and /boot/efi are separate partitions), but there was a whole “Microsoft” subdirectory as well, and removing that did free up some space.
When I retried fwupdmgr after that, things did work. So perhaps filesystem space was the problem? It would mean that the firmware updater itself needs space, since I went from 70% full to 50% full on /boot/efi: by no means full! If the EFIShell updater failed for the same reason, then apparently it DOES end up using the disk as well, even though it starts from a USB drive.
My partition table looks like:
Device Start End Sectors Size Type
/dev/nvme0n1p1 2048 206847 204800 100M EFI System
/dev/nvme0n1p2 206848 2158591 1951744 953M EFI System
/dev/nvme0n1p3 2158592 4255743 2097152 1G Linux filesystem
/dev/nvme0n1p4 4255744 780214271 775958528 370G Linux filesystem
where the first partition is the one that is mounted as /boot/efi. The 100M is a bit small! I think it’s a remnant of the Windows install and that I made p2 when I partitioned the disk for linux (at that point it seemed the p1 partition was “protected”). Either it never used the p2 partition properly for EFI or the F35->F36 upgrade confused the system. It would explain why there was a Microsoft directory there in the first place … As a plus: by removing that Microsoft directory in /boot/efi I’m now finally rid of the defunct Windows boot entry.
This sounds similar to the problem I initially experienced. This is what solved it for me, maybe it’ll help you as well: 11th Gen Intel Core BIOS 3.10 Release - #39 by Jason_Hottelet
Thanks. I’ve tried scheduling the update from “Software Update” as well. That has the same result, unfortunately. Do you recall doing anything more?
Particularly the fact that the EFI-shell method doesn’t work for me either, which I assume should circumvent any installed operating systems, makes me suspect there’s some other system state in the way …
Unfortunately not, but you might want to check if your EFI partition has enough free space left (mine had, but others reported problems).
I reverted to BIOS version 3.07 and the issue seems to have gone away for anyone else having this issue. The only thing I can think that can differentiate me from others on Ubuntu 22.04 is if they didn’t use the post-install automated script for Ubuntu 20.04 and 22.04.
I’ll update this post pending further progress/discussion with Framework support.
Edit: Successfully reverted to 3.07 without any issues. I decided to EFI flash 3.10 instead of LFVS and it seems to be working now. Not sure of any root cause, just happy to have a working laptop. Hopefully it stays this way!
Running on Windows 11 I get “This platform does not support IHISI interface” - do this mean anything to anyone?
Under what circumstances are you getting this message?
Sorry, my mistake - I had not realised this was only 11th gen, and not 12th gen as well. Might be worth adding a check on that in case I’m not the only one making this silly mistake.
Thank you for the quick update. After going through this process several times now I have some suggestions. I use the EFI shell to preform my upgrades as I run an unsupported OS. This works well enough though having to remove the boot SSD seems like an unecessary step. Can we add that logic to the installer itself or just have an option in the bios to disable the m.2 slot so when someone goes in to disable secure boot it can all be taken care of in one step. Lastly can we add some type of binary verification method to these releases. Checksums or signatures that the community can use to verify the software being installed is as intended.
I updated to this BIOS version today, and my right USB ports stopped working for data transfer.
In Windows 11 I got an error that read “Power surge on the USB port – Unknown USB Device needs more power than the port can supply.”
Rolling back to 3.07 solved the issue for me.
Did anyone else get a boost when upgrading to 3.10?
My cinebench23 went from 4800 to 5003. My CPU is the 1135g7.
I can’t find anything in the release notes that indicates a performance increase.