Oh man I feel for you. Also disturbing that on the newer BIOS version it is still happenening. I will go update my BIOS right away so that if I get another BSOD I can report that to Framework and show that the newest drivers and newest BIOS did not help.
Got the same. Seems it is running some tests or configurating something.
From the Release Notes of the just released version 24.4.1:
Fixed Issues
- Intermittent system or application crash when screen sharing using Microsoft Teams.
So there was a problem with the drivers on Teams …
I’m sure as hell not saying it’s fixed, but uptime of 6 days so far on/off battery, mostly light usage, (SSH, Firefox). (my previous best was 8 IIRC)
w11/7840/32gb crucial 5600/3.05/framework driver pack
Since several of you were asked by support to try without expansion cards (same with me), I am wondering: Has anyone been able to trace the problem down to a specific expansion card? I am having BSOD issues very rarely, and now tried 2 weeks without expansion cards, then 2 weeks with USB-C, then 2 weeks with additional USB-A, finally getting a BSOD. But one cannot confidently blame the USB-A card with this observations …
1 Like
I only had the hard freezes with BIOS 3.03.
BIOS 3.03b and 3.05 have fixed it for me.
But I haven’t experimented with different expansion “cards”.
I have almost always had USB-C in the top two and USB-A in the bottom two.
Note: I have also installed all the optional Windows Update driver updates.
1 Like
So far have not seen any hard freeze/resets under 3.05. In the logs there is an hardware error reporting in reliability report, but from a usability perspective it doesn’t seem to be a problem. At best I believe it is related to waking from sleep.
With BIOS 3.03, I had hard freezes (sometimes followed by blue screens) randomly. Sometimes multiple days in a row, but I also had a month without any freezes at all. BIOS 3.03b fixed that for me; I didn’t have a single freeze with it.
I installed 3.05 and the new Driver Package a few days ago, and the laptop froze twice again since then. Both times it froze just a few seconds after waking it up from hibernate.
Really disappointed in Framework at this point.
1 Like
What is everyone’s experiences of 3.05 like? I’m still running 3.03b because that’s been stable for me and I’m unwilling to go back to broke.
I’m new still to Adrenalin (since on desktop the last time I used an AMD GPU was an HD6950 and that used Catalyst Control Centre and even that was a bloody mare to use). I’ve only had Ryzen CPUs. I’m not sure how it updates any of the drivers at this point because I don’t see it provided in the software centre. I know my way around a manual update but if there’s software to do it, I’d rather it do it. I would also like the option at least of standalone drivers being available as well as a bundle. If you need to reinstall the sound driver or network driver, I want that to be individually available so I don’t need to reinstall everything
Framework does need to be clearer on updates, issues, where to find them.
1 Like
My experience so far has not been very good, I have noticed that the FW drains the battery excessively in standby (screenshot) which I did not experience with 3.03b. I think, it might be related to the Z sleep state enablement.
Also, I still get the LKD_0x141_IMAGE_amdkmdag.sys error: for me, it froze when the iGPU is somewhat overwhelmed with too many animations or at startup.
1 Like
Just to report back that this issue is totally solved for me. And no trackpad jumping either.
Just to add my experiences.
Since I bought the laptop in December 2023, I’ve had a laggy cursor. It lags when using the trackpad and when using a mouse. By lag I mean it takes a little while to get moving, then it ‘jumps’ to where it ought to be and is then fine, until I stop moving it. Then that pattern repeats. It’s very frustrating and makes precise cursor usage almost impossible.
On BIOS 3.03 I got all the same random freezes that are reported on here. Usually the system would lock up, but I could still move the cursor, then eventually the cursor would freeze and I’d either get a BSOD or I’d have to hard reset the laptop. I could get as many as ten freezes in a day, but sometimes none for a week or so. It didn’t matter if the laptop was running on AC power or battery, or what I was doing on the laptop.
I updated to BIOS 3.03b and that seemed to cure the freezes, but not the lag. Until earlier this week when, despite running the latest driver package, I had three freezes within about 30 minutes. I’ve now updated to 3.05 and I’m waiting to see whether it freezes again.
Support have had me run all sorts of memory diagnostics (all clear) and reset the mainboard. I just want a laptop that I can rely on not to freeze and where I can use the cursor. I’m beginning to doubt that Framework can supply me with that.
2 Likes
I think you should be polite yet assertive with the support and just demand a new motherboard to replace your existing one. Not to say that if you’ve been reporting issues since December 2023, in some countries, a warranty claim needs to be resolved within 30 days.
1 Like
I had the same problem after some time the full bios got corrupted the laptop would not boot anymore and I got a new motherboard
Is there any reason my laptop would do this on the battery? I’m still on 3.03b, the screen was off, the lid was closed and I would’ve expected the laptop to be asleep so I don’t expect it to discharge to 0% in the space of about 10 hours.
Past 7-day usage, it was pretty much only on the morning for a couple of minutes at most today.
Power options settings, closing the lid should put the laptop to sleep so I shouldn’t see the battery drain like that.
1 Like
Interesting… I installed OpenSuSE Leap 15.5 on a DIY FW13 (Intel Core i5 processor, single-boot O/S) and the system performs very well but for the same intermittent problem: it appears to completely freeze and has to be restarted by leaning on the ‘start’ button to power down.
-
I’ll begin keeping a log of such events, but it occurred last night (my user account, battery powered) and again this morning (my partner’s account, AC powered). Failures are usually less frequent though.
-
At this point, I strongly suspect lockups occur when the wired mouse is actively in use. (When configuring the system I expected to be using it with an external keyboard, mouse, & HP monitor, but I’m not sure I don’t just prefer the laptop & a proper mouse.)
-
Lockups are definitely not associated with heavy CPU activity. The initial O/S installation from a USB memory stick followed by a huge backlog of online updates ran perfectly.
-
The mouse, a printer and a scanner are connected to a Verbatim USB-A hub, which draws its power from the FW USB-A adapter. I’d like to replace this with a USB-C hub and will order a second USB-C laptop adapter.
-
The printer and scanner are normally off, however the two failures mentioned above both occurred when the scanner happened to be switched on. The WiFi & Bluetooth controllers, laptop camera & microphone, and touchpad are not used. The system also has an Ethernet and an (unused) HDMI adapter.
I’m beginning to think lockups are associated with some unusual interrupt state, maybe a lost or spurious interrupt from an external device, and I note most devices all hang off one Host Controller.
Hi here, first post on the Framework forum too.
The discussion here was about WIN11 at the beginning but I really have close symptoms so it will may be relevant to put this here.
I received few weeks ago a framework13 DIY laptop with this configuration (every components from the Framework shop):
- 32GB SSD
- 2TB NVME SSD
- BIOS 3.05
- 7040 AMD mainboard
I am posting here because of the same kind of problems concerning BSOD on a linux NixOS, where I can only press the power button until a hard shutdown. And to be honest it begins to be complicated to get where does this issue comes from.
Whats happens & observations:
- Random freezing with no real reasons, either with low or heavy workload (sometimes no problem at all to game for some hours but freezing on some trivial stuff)
- When I listen some music / video and the freeze happen, I cannot do anything and the screen turns black. But I can still hear the audio that continues for ~1 min before ending and have to hard shutdown
- Had a hard day yesterday because the laptop frozen every time I turned on a YouTube video on full-screen. On that case the screen was frozen and the audio doing infinite loops with the last second of sound until I shut it down.
- I noticed that when the charger is not plugged in, the laptop is really more stable, like for the last few days I was able to have no freeze when on battery while running video + Kali VM with ~12GB of RAM to make some HackTheBox stuff
The things I already tried:
- used fwupdmgr to update hardware
- Look for BIOS update (I think 3.05 is still the latest so far)
- Install and update the framework drivers for AMD 7040
- Update to Channel 24.05 of NixOS, rollback to 23.11
- Trying different GUI like plasma6, GNOME, with X11 or Wayland
- disable Hardware acceleration for every app I could
- Try to reset the mainboard state (disconnect the battery within the BIOS, unplug every adapters, shutdown the laptop, press the power button for few seconds and rebuild the whole)
- Not a professional at journalctl logging / debugging , but I did not found something relevant so far withing the log preceding the hard shutdowns
Next I will may try to setup power-profiles-daemon for power management and try to set PCI Express devices to performance mode. I really want to insist on the “when-plugged” issue because its really astonishing during the few past hours.
I will be grateful for any help but I am a little bit worried bc it seems that I got all last-updated drivers etc…
Thanks for your response… At this point it may be we need to better understand the circumstances in which the problem occurs and assemble some statistics. My Partner would like a FW laptop too, but not until this issue is fixed!
My system has some significant differences in configuration. It’s a FW13 with an Intel Core i5-1340P processor, and it runs OpenSuSE Leap 15.5 with the Plasma user interface. WiFi, Bluetooth, camera, and microphone devices are not used; the network connection is via Ethernet.
Yes… Failures have absolutely no correlation with processor load, and always seem to be associated with mouse movement on this system. The mouse, together with a printer and scanner which are normally powered off, and occasionally a memory stick for backups, are attached to a 4-port Verbatim USB-A hub. The hub is powered from its USB interface on the laptop. Having said that, failures also occur when the mouse is directly connected to the laptop.
That certainly sounds like a clue, but I can’t imagine to what. Have you tried repeating the YouTube full-screen video experiment using Wayland?
I suspected that too and tried running exclusively on battery, charging it while the laptop was not otherwise in use, no users were logged in, or it was shut down. I did have one lockup on battery, but they’re not frequent enough to get any meaningful statistics.
I wonder whether transients on the AC power supply are relevant? I could try running it on a UPS I suppose.
I configured this laptop expecting to use it with a large external screen, mouse and keyboard but so far prefer to use it just with an external mouse. Hence the system has an unused HDMI adapter which I’d like to replace with a USB-C adapter, and I wonder whether it’s generating spurious interrupts, etc.
(I’m surprised the touchpad still exists; the keyboard would be far more convenient if the battery & keyboard swapped places and the wretched touchpad were consigned to history.)
Hi, thanks for your answer @DavidL
Yep, at the very beginning I though that the last version of GNOME was broken within my config so I switched few times between GNOME / KDE & wayland / X11 to make some tries (even tried DWM and still had the issue).
So its a little bit too soon to say that everything is fine but it seems to gets better. With a little bit of ChatGPT carving on the AMD freeze topics, framework community forums and ArchWiki forums trying to deal with AMD it seems that I reached something way more stable. Also (see at the end of the post), a little manipulation on the battery management is undergoing.
I am not able to explain exactly why everything works better that way but here is the config that seems to be more reliable for my (I am running NixOS so the configuration is declarative in one file, but you could try google/chatGPT the list of options to get how to configure it with your OS):
imports =
[ # Include the results of the hardware scan.
./hardware-configuration.nix
<nixos-hardware/framework/13-inch/7040-amd> # Here adds the hardware support for FW13
];
# Enable power-profile-daemon
services.power-profiles-daemon.enable = true; # Also to avoid issues on framework13
# Enable powerManagement with powertop
powerManagement = {
enable = true;
powertop.enable = true;
}
# Using the last linux kernel
boot.kernelPackages = pkgs.linuxPackages_latest;
boot.kernelParams = [
"pcie_aspm=off" # Here are all the options I gathered on the Internet that
"amd_pstate=passive" # could help for the stability, related to AMD GPU
"amdgpu.dpm=1" # same
"amdgpu.aspm=1" # same
"mem_sleep_default=deep" # same
"kvm-amd" # Not related (virtualization stuff)
];
# I saw on some forums that the suspend mode of the audio card could cause stability issues:
hardware.pulseaudio.extraConfig = ''
load-module module-suspend-on-idle timeout=0
'';
### Enable the X11 windowing system.
services.xserver.enable = true; # remove if using wayland
### Enable the GNOME Desktop Environment.
services.xserver.displayManager.gdm.enable = true;
services.xserver.desktopManager.gnome.enable = true;
services.gnome.core-utilities.enable = false;
So this is a little patchwork of what seems to work for me.
I still have one freeze yesterday (so way more stable that the previous weekend) on charging but only when reaching 96% (and the external monitor plugged with a framework HDMI adapter in the right port has to be unplugged and plugged in after reboot to be detected). I tried to unplug the charger few times, discharge few percents and re-plug and freezes happened quite systematically.
So I also went into the BIOS to reduce the maximum battery charging point to 90% to see if it changes something. (Just press F2 and go in the Settings/Advanced tab to find the right option).
So it seems that the first part handled nicely the ‘classic’ cases, the battery thing the ‘fully-charged-and-plugged-in’ problem. I will update this post later if needed according to the results.
Maybe try turning off Memory integrity?
(I had a ProArt before and it would freeze every now and then whenever I turned on memory integrity.)