I have been a proud owner of a DIY fw 13 AMD ryzen7840u (1TB SSD, 2x16GB ram, all fw branded/sold, batch3) for just over a week now, and am generally very happy with the laptop.
Yesterday and just now, however, I experienced a hard freeze, meaning that the pc was completely unresponsive (ctrl+alt+del didn’t work), the mouse cursor also got stuck. When it happened, I was basically just working on a document in Libreoffice, and yesterday I was typing an email in Thunderbird. The only thing I could do was a hard shutdown and then restart.
I am worried that it might be a hardware problem. After waiting half a year for this fw while struggling with a semi-functional thinkpad, that would be pretty frustrating…
What would you recommend I do to start diagnosing this problem?
Any advice/help would be greatly appreciated. I haven’t opened a support ticket yet, thought I’d try the community first.
I assume you’ve installed the driver package from Framework? Apparently, some people have also had success in solving some issues by installing additional drivers directly from AMD. Might be something to try, if you haven’t already.
For what it’s worth, the newer AMD drivers helped this for me (less frequent) but it still happens occasionally.
If you wait long enough on the hard freeze, in my experience after a couple of minutes Windows will BSOD with a DPC_WATCHDOG_VIOLATION. As a result, I think it’s still the drivers rather than hardware.
I had similar issues with the 7640U on Win11. Specifically, I had 4 freezes. A few points:
All freezes were on battery power.
Two were preceded by some stuttering.
Before updating to BIOS v3.03, the (two) freezes wouldn’t lead to a BSOD, at least before my patience ran out. After updating, I had two in one day, but they led to a BSOD within seconds (also DPC_WATCHDOG_VIOLATION), the PC rebooted, and I was able to get debug information. WinDbg traced it back to the GPU driver (amdkmdag).
Since updating to the AMD GPU drivers (just using the auto-detect tool), I haven’t had any issues (5 days now). However, the laptop was plugged in during most of this.
My dual boot Arch setup has been very stable (i.e., no hiccups at all so far). I’m using GPU acceleration in media playback as well as a terminal, and I’ve done a little gaming on Windows after the driver update. So, stable across a variety of GPU loads.
Given all that, I’m pretty sure it was (and possibly still is) a driver issue. My hunch is some kind of GPU driver power saving feature bugging out. I also sent a support ticket last week. Suppose we’ll know more soon.
Have you checked Event Viewer around the time the freeze happened? That would be a good way to figure out if the issue is a Hardware or Software problem.
Other than that, updating to BIOS 3.03, or reinstalling the driver bundle and OS are both good troubleshooting steps. Remember that there is a new driver bundle to go along with the new BIOS, so if you upgrade to 3.03 or have already and haven’t updated your drivers to go along with it, do that for sure! Also make sure to check Windows Update for any newer drivers there too (the driver bundle just gets you the necessary drivers to get Windows installed and connected to the internet, newer driver updates come through Windows Update).
If all of those steps don’t help you could be looking at a hardware issue, at which point you’ll need to contact support. Just remember that they are really busy right now and will take some time to respond, though they will get back to you eventually!
Hi Azure, thanks for the advice, I just checked it and there is indeed an entry for the time it happend yesterday. Here is the log, I myself can’t make any sense of it, maybe someone else can?
Critical 08/11/2023 10:39:10 Kernel-Power 41 (63)
Log Name: System
Source: Microsoft-Windows-Kernel-Power
Date: 08/11/2023 10:39:10
Event ID: 41
Task Category: (63)
Level: Critical
Keywords: (70368744177664),(2)
User: SYSTEM
Computer: framework
Description:
The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.
Event Xml:
I’m afraid this is only pointing at the forced shutdown (caused either by you holding the power button or the OS resetting), which is unsurprising.
In my case, details were easier to track down through Reliability History. There, I had “Windows has stopped working” errors that clearly pointed to the DPC_WATCHDOG_VIOLATION. What do you see there when you click on a critical Windows error and follow “see technical details”?
Log Name: System
Source: Microsoft-Windows-Kernel-Power
Date: 08/11/2023 10:39:10
Event ID: 41
Task Category: (63)
Level: Critical
Keywords: (70368744177664),(2)
User: SYSTEM
Computer: framework
Description:
The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.
Event Xml:
Just came to say I’m experiencing this with my fw13 amd7840u on Windows 11 too. I’m following the advice of @okko and updating AMD drivers. Hopefully that resolves the issue.
I’ve been reading these posts with interest because I have also experienced a couple freezes from which I could not recover without doing a hard reset. I am running Win 11 Pro & have NOT installed the updated BIOS yet.
Just got my first hang/BSOD after updating to the AMD drivers. Seems this is definitely a driver issue on the new AMD platform. I’m going to log a support request with Framework to ensure they’re getting the signal.
I have sort of narrowed it down just to the igpu/internal display. Whenever I hook up to my egpu dock I can run for hours as long as my internal display doesn’t turn on first. I have been able to reproduce this, pretty easy just turn the laptop on wait for the login screen then plug in the egpu, so I’m narrowing it down to just the iGPU. Almost all of my crashes/blackscreens are when I’m using it as a laptop so I’m going to assume this is driver related but I very, very, very much hope it isn’t hardware related. Best of luck to you all!
Bearing in mind the notes from my previous post, I’ve had a chance to use both Win11 and Arch on the machine, both plugged in and on battery, with no issues. Most I’ve had was a sub-second stutter or two during light gaming on Windows.
So, after another 5 days of stability, I’m even more confident that this was/is a GPU driver issue. Seems I’ve trapped myself in a win-win situation where the next BSOD will be another debug opportunity to cherish, and the absence of one will be a reward on its own!
