I don’t think it is related to the update, I’m still on an older version and recently experienced the same thing. In fact if you search the forum there are a number of posts that seem to be related to the same or similar problems with USB-A devices.
Is there anyone using NixOS that has found a way to do this?
Here is one way:
{
# bios/firmware update
# https://knowledgebase.frame.work/en_us/framework-laptop-bios-releases-S1dMQt6F
services.fwupd = {
enable = true;
};
environment.etc = {
# enable lvfs-testing remote to get the newest firmware from framework
"fwupd/remotes.d/lvfs-testing.conf" = lib.mkForce {
source = pkgs.runCommand "fwupd-lvfs-testing-enabled.conf" { } ''
sed "s,^Enabled=false,Enabled=true," \
"${pkgs.fwupd}/etc/fwupd/remotes.d/lvfs-testing.conf" > "$out"
'';
};
# recommended by framework for firmware 3.10
"fwupd/uefi_capsule.conf" = lib.mkForce {
source = pkgs.runCommand "fwupd-uefi-capsule-update-on-disk-disable.conf" { } ''
sed "s,^#DisableCapsuleUpdateOnDisk=true,DisableCapsuleUpdateOnDisk=true," \
"${pkgs.fwupd}/etc/fwupd/uefi_capsule.conf" > "$out"
'';
};
};
}
4 Likes
To make things slightly simpler, to enable lvfs-testing on NixOS all you should need to do is
services.fwupd = {
enable = true;
extraRemotes = [ "lvfs-testing" ];
};
You’d still need the second part to define DisableCapsuleUpdateOnDisk=true though.
4 Likes
Have the same issue after 3.10. Charge dropped to 15 percent and unable to charge. Fedora 37
EDIT: Shutting down and removing the little dime-sized CMOS battery worked like a charm.
2 Likes
I used fwdupmgr to move from 3.10 to 3.17 and now my power cord is no longer being recognized. I’ve used both the framework power supply and a unrelated wall wart to usb-c, and framework continues to believe it needs to burn through battery, and I’m sweating that in a few hours I’ll be unbootable.
fwupdmgr downgrade
Choose a device:
0. Cancel
- a45df35ac0e948ee180fe216a5f703f32dda163f (System Firmware)
- 362301da643102b9f38477387e2193e57abaa590 (UEFI dbx)
1
No downgrades for System Firmware: No releases found: Device System Firmware [a45df35ac0e948ee180fe216a5f703f32dda163f] does not currently allow updates: Cannot install update when not on AC power
I’d already rebooted 2x, and switched sides for my usb-c port when I posted this. however as of a few moments ago, with usb-c now in a 3rd slot, and then back to the framework power supply I finally got a light indicating power cord was working. I’m now back in business. hopefully unrelated to anything, and of no use to anyone. but just in case others begin to report similar.
@dmac8086
It’s seems to be a recurrent issue, and unrelated to this update. I have Ubuntu 22.04, with BIOS 3.10 and it regularly fails to detect my external HDD.
I’m also getting the 404 error when trying to refresh the lvfs-testing remote.
Same 404 error happened to me, it is a bug in the config file.
Had to add .gz to this line: MetadataURI=https://cdn.fwupd.org/downloads/firmware-testing.xml.gz, in file: /etc/fwupd/remotes.d/lvfs-testing.conf
1 Like
Is there a full changelog between 3.10 and 3.17 anywhere?
I had to revert to 3.10.
Issues I faced:
- After a while in hibernation and plugged there would be no charging, so the battery would drain
- Occasionally on resume from hibernation the touchpad would not work
- The right side expansion ports would not work.
After downgrading the expansion ports work, keeping an eye on the other issues to validate they do not also occur with 3.10.
I use Ubuntu 22.04.
Here is an example of ectool console for a case where the touchpad did not work on resume:
1042705.993548 HC 0x98]
[1042704.553326 PS2M renabling]
[1042704.686098 I2C4 port2 recov status 0x80, SDA:SCL=0x1]
[1042704.689015 I2C2 unwedge called with SDA held low]
[1042704.695302 I2C2 unwedge failed, SDA still low]
[1042704.752781 I2C4 port2 recov status 0x80, SDA:SCL=0x1]
[1042704.753931 I2C2 unwedge called with SDA held low]
[1042704.756193 I2C2 unwedge failed, SDA still low]
[1042704.786755 I2C4 port2 recov status 0x80, SDA:SCL=0x1]
[1042704.787821 I2C2 unwedge called with SDA held low]
[1042704.790262 I2C2 unwedge failed, SDA still low]
[1042704.822191 I2C4 port2 recov status 0x80, SDA:SCL=0x1]
[1042704.825216 I2C2 unwedge called with SDA held low]
[1042704.831488 I2C2 unwedge failed, SDA still low]
[1042704.865685 I2C4 port2 recov status 0x80, SDA:SCL=0x1]
[1042704.868810 I2C2 unwedge called with SDA held low]
[1042704.874881 I2C2 unwedge failed, SDA still low]
[1042704.911352 I2C4 port2 recov status 0x80, SDA:SCL=0x1]
[1042704.912481 I2C2 unwedge called with SDA held low]
[1042704.918610 I2C2 unwedge failed, SDA still low]
[1042704.954335 I2C4 port2 recov status 0x80, SDA:SCL=0x1]
[1042704.955377 I2C2 unwedge called with SDA held low]
[1042704.961621 I2C2 unwedge failed, SDA still low]
[1042704.994063 I2C4 port2 recov status 0x80, SDA:SCL=0x1]
[1042704.995197 I2C2 unwedge called with SDA held low]
[1042704.997585 I2C2 unwedge failed, SDA still low]
[1042705.028085 I2C4 port2 recov status 0x80, SDA:SCL=0x1]
[1042705.029291 I2C2 unwedge called with SDA held low]
[1042705.031552 I2C2 unwedge failed, SDA still low]
[1042705.062111 I2C4 port2 recov status 0x80, SDA:SCL=0x1]
[1042705.063233 I2C2 unwedge called with SDA held low]
[1042705.065625 I2C2 unwedge failed, SDA still low]
[1042705.096184 I2C4 port2 recov status 0x80, SDA:SCL=0x1]
[1042705.097366 I2C2 unwedge called with SDA held low]
[1042705.099627 I2C2 unwedge failed, SDA still low]
[1042705.110766 PS2M 1 Too many retries]
[1042705.864534 PS2M renabling]
[1042705.987347 HC 0x0b]
[1042705.990451 HC 0x97]
I also posted details of the charge on the ectool thread
1 Like
What is the rationale for this change? I haven’t tested yet, but IIUC the default power led state is also white, so essentially this removes the low battery indication? Does this also apply when the battery is so empty the machine can no longer boot, since that would be very confusing…
IIUC there are no versions between 3.10 and 3.17, so the changelog in this thread is all you need (see the note in the first post about skipping these versions).
3 Likes
I’ve just finished upgrading from 3.07 straight to 3.17 using the USB stick method, which seems to have worked fine. As expected, I had to reapply my BIOS settings (charge limit, power on AC, boot order).
My Linux/Windows dualboot still works, I only had to modify the order in the BIOS settings, since it had switched from booting Windows instead of Linux by default (I dualboot using F12 to switch to Windows if needed, I am not using the grub menu for dualboot).
During the update there was one weird thing: It booted into the updater app ok once, which very briefly showed some output (on a screen that looked like the screen that applies the actual update) and I think it said “Completed” somewhere, but then it rebooted before I could see. I thought the update was completed, but just to be sure I used F12 to boot from the USB stick again, and then it actually started updating (which took a minute or two). I’m not sure this is intended, but if so this should probably be documented. It might also have happened because before this, the first try was without AC plugged in, so I plugged in AC and rebooted, and after that got the short “update” I described above.
Also, after the update the first boot took a bit longer than normal (probably because of the RAM bank calibration or so - IIRC there was a green led on the side), which had me slightly worried - maybe also good to mention in the instructions.
I can confirm this is indeed fixed (I have observed this problem before). I also noticed that disabling quiet boot while quick boot is still enabled is fairly pointless, since every zaps by too fast. If I also disable quiet boot, the startup message can be read (and F12 still works).
What I did notice (which is likely not specific to this BIOS version, I’ve just never looked at the non-quiet startup messages before) is that it advertises F10 (for MEBx) and F9 for remote access, but pressing these does not do anything (the presses are registered because the messages are removed, but then it just boots Linux as normal). These are related to the Intel Management Engine, I think, which I’m not sure my i7-1165G7 even has, or if it has, it’s probably disabled (I’m on the default settings in this area).
The low battery power button LED change is because the chromebook version does not have a red LED on the power button. We unified the behavior across both products so it is consistent.
The power button will flash white now if you try to power on the system with the battery too low.
3 Likes
This only applies to the model with the i7-1185 CPU which has vPro support.
1 Like
Ah that makes sense, thanks for clarifying.
But I’m assuming the BIOS knows whether there is vPro support or not, so it would make sense (to me) that the BIOS would hide these confusing prompts if there is no vPro support?
2 Likes
I didn’t notice the issue until recently, as I only use USB-C and USB-A ports. After upgrading to BIOS 3.17, only the ports on the right side is able to mount USB storage. I tried using the USB-A → C cards on the left side, even using the expansion card A → C → C, but my system would still not be able to see it.
sudo dmesg shows that the USB is detected on the right side, but not on the left.
I haven’t tested with other expansion cards yet. Only USB A and C cards.
Is 3.17 still in Beta or has it moved to a full release?
1 Like
Still beta. 3.10 is the current stable per the driver and BIOS page here: Framework Laptop BIOS and Driver Releases (11th Gen Intel® Core™)