The other topic mentioned why it is not working someone sent the link here and bot deleted it
sorry, i was typed last reply on phone, i meant fingerprint driver it is from 2021 (3.12804.0.140) only not like 2024 or the version mentioned in the bundle. doesn’t it mean bundle didn’t update driver?
As i said previously (refer screenshot) in the gives an error message. The post @Second_Coming linked he mentioned that framework is using some option which was removed in latest insider build.
I didn’t quite understand your post initially, so I held off replying as I was waiting to check the driver version from the BIOS. Since WMIC is enabled on my system, that’s not the issue. I ran the bundle.exe again a few minutes ago and encountered the same “folder missing” error. Meanwhile, three unrelated Intel drivers appeared in Windows Update, which I’ve just installed. One of those was the Intel Display Driver, version 32.xxx.xx.xx. Regarding the Intel Management Engine, I’m currently on version 2406.5.5.0 from August. I posted to ask for suggestions on how to confirm whether the driver package bundle ran successfully, I had no ideas. but I’m unsure if the issue lies with just that driver or the entire package.
I also didn’t fully understand the part where you said, “That also refers to the driver package not running. You didn’t mention that, even though I asked.” Could you clarify that?
Worrying to see that even more people have had their devices soft-bricked due to the S3 / “mem_sleep_default=deep” and 3.20 BIOS fiasco…
The LogoFail vulnerability is still not fixed on 11th gen. It is 10 months after it got public.
Is there an ETA to fix this vulnerability (and others)?
In early 2022, it was found that s2idle and deep resulted in the same battery drain (dependent on which expansion cards we had plugged in) if we just added “nvme.noacpi=1” to the kernel parameters [1]: [TRACKING] Linux battery life tuning - #225 by tim300
I’ve had mine set to s2idle ever since. What benefit have you found when using deep?
Anyway, due to what seemed like ongoing concerns about the the 3.19 (with EFI installer) and 3.20 BIOS updates, I delayed doing them myself until this past week. The 3.19 update early in the week was uneventful. After detecting no issues for a few days, I did the 3.20 yesterday. I realized after one failed attempt that I needed to disable secure boot, and then it went smoothly. GNOME Settings → System → About → System Details reports that my firmware version is now 3.20. I enabled secure boot again, turned down the power button LED brightness again, and everything seems fine so far. Thanks again to the Framework team for continuing to support this three year-old hardware!
[1] Edited to add: I see now that the kernel parameter is no longer recommended (nor is TLP), so I’ve made changes to my system accordingly. We’ll see if I notice any difference: Ubuntu 24.04 on the Framework Laptop 13 - #32 by Phillip_Elvey
For me with s2idle (and nvme.noacpi=1) laptop lasts only ~16-18 hours in sleep, whereas with deep it lasts 2-3days; I’ve also noticed that in s2idle the laptop keeps warm, whereas with deep it cools down to room temperature (i.e. that there seems to be a difference in what systems get suspended too); In any case… It’s extremely unlikely that I will ever buy or recommend framework to anyone again, even if you completely ignore all the documented sleep issues they have, purely on the basis of poor customer support and non-existent communication to their customers of significant changes. Also, even with 3.20 their firmware still reports “deep” as a supported state, however if you enable that, it soft-bricks your laptop requiring a forced reset by taking the battery out. This is not acceptable.
1 Like
Hi,
I think I understand now why framework has been silent on the s2idle issue with the newest BIOS. They were working on a different solution already.
Look at these 2 posts on mastodon:
TLDR: They are working on a tool to automatically enable hibernation and be independent of suspend altogether. The key information is this part
Now, that said, hibernate would address battery drain while in suspend for Fedora (and later Ubuntu) as we’re not relying on suspend.
We’re putting the lift on hibernate instead of only relying on suspend.
While I think this new tool is a fantastic thing, it kind of throws the people not using Fedora/Ubuntu running Gnome under the bus.
I assume that it will all be open source, so there would be nothing preventing anyone from porting it to a different distro.
True, but if I understand it correctly it is essentially a GNOME Extension right now and it requires BTRFS. So it only makes sense for new installs. So it is in its current form not a generic solution for all Linux systems/distributions. Others will have to fiddle and do it by hand.
And suspend is still broken for my system…
Well, they only officially support Fedora and Ubuntu, so that’s nothing new. And if it requires a particular partition setup then the logical time to do it is on initial install.
Setting up hibernate is not that difficult. It takes a bit of effort, but it’s worth it. I have been very happy with suspend-then-hibernate on my systems.
Maybe we should take it to a different thread, but what is broken with suspend on your system, and what distro are you running?
I am in the same club as these fine people over here:
As this thread is focused on BIOS 3.20, if you have issues with suspend and would like help with that, please move to another thread or open a new one.
1 Like
Well, my issue with suspend was introduced after updating my from BIOS 3.17 to 3.20. Which soft-bricked my computer until I switched suspend mode from “deep” to “s2idle”
Deep suspend was working perfectly fine before and the battery drain was tolerable. Now it is not anymore as most of the time the laptop doesn’t really go to sleep - it just pretends it does by flashing the powerbutton. But if you listen closely you can still hear the fan spinning. Which tells me it is not sleeping at all. In that state it lasts maybe a few of hours instead of days.
Using hibernation is fine by me - but it isn’t fixing the underlying problem. It is just covering things up.
But fine, I will watch how this goes and if framework is providing a proper solution for this with the next BIOS version and then I decide if it is worth spending more money.
@lvdd deep sleep won’t be fixed anymore.
I was in contact with support and since 3.20 deep sleep is not supported anymore:
“Spoke with engineering for clarification on deep sleep not working. Yes, you will be using s2idle as we have found that the later BIOS are not supporting deep.”
That’s a pity for everyone with this issue, especially since there was no mention anywhere. At the least they added it now to the “Known Issues” for bios 3.20.
1 Like
Well, that’s a pity for them as they won’t see any more money from me, if that turns out to be true.
They remove an essential functionality and then afterwards tell me “sorry, bad luck”? That for me is a trust problem which they are not going to recover from in my book.
My laptop is still working and I can use it with the hibernate workaround to circumvent the battery drain issues. But the fiddling required to just get the most fundamental thing to work on this hardware is a waste of lifetime for me at this point.
Maybe I am just going to print a case and use the board as a homelab server (that doesn’t require any sleep functionality) or hand this thing to my wife replacing her x230 as a “couch surfing” machine that’s connected to power most of the time anyway and go back to my MacBook.
As much as I love the framework idea, I hate being kicked in the nuts after 3 years being left with less than before. Good luck to everyone - updating BIOS can leave you with a handicapped laptop with no way back - good work framework… 
1 Like
@lvdd Did you try to use nvme.noacpi=1 as boot parameter? It drastically reduced the battery drain in s2idle for me to a similar level as deep (~ 0.5%/h).
Which of the Windows states s0, (. . . . s3 s4 etc.) do you consider ‘deep’
I don’t know the Windows states, but it should be s3.
1 Like
I’m stubborn and still trying to figure out why both of my 11th gen systems suspend and resume with no issues. Poor @lvdd ran through about eleventybillion different scenarios and didn’t get anywhere with determining why his system does not come back. It feels like a driver not reloading properly, but I am way out over my skis on this. I am just a dumb end user, tinkering and plinking away on the machine. I would love it if we could get an engineer (not Support - nothing against Support, but this is out of their realm as well) that would work with us to test working and non-working systems to try to narrow things down.
3 Likes
Damn, so much friction just to even ‘attempt’ to get it to sleep (and wake) right. (successfully or not).