Still, would be good if you can share the various / multiple circumstances even if you, at the individual level, can’t determine a pattern. i.e. a larger data set might allow for a pattern to be draw…but we still need the data points to begin with.
After adding
i915.enable_psr=0to my kernel parameters, I’ve not had any more freezes.
This seems to solve it for me as well. Before that, I had 20sec freezes up to Wayland/KDE crashes every hour(?). Now, already one day without any freeze.
Linux fedora 5.19.16-301.fc37.x86_64 #1 SMP PREEMPT_DYNAMIC Fri Oct 21 15:55:37 UTC 2022 x86_64 x86_64 x86_64 GNU/Linux
2 Likes
At first on 6.0.6-300.fc37.x86_64 kernel my machine was fine. Ran great for over an hour, no freezes, but it seems like it’s still not fixed as the freezing has returned. 
Yeah…an hour is too short…reporting too soon.
Strangely every time I tried to remove the kernel param before (5.19.x kernels) I got freezing within minutes of login 
but agreed, too soon. A little too eager for this issue to be resolved.
I would be, too. But at least the psr setting is a temporary workaround.
1 Like
That’s unrelated to the enable_psr setting. That’s bluetooth autosuspend. See Ubuntu Bug#1801642 (Add btusb.enable_autosuspend=0 to your kernel command line to fix)
I haven’t had a Gnome shell freeze in the 18 days since my post #161 with the upgrade to kernel-5.19.15-301.fc37.x86_64 and blacklisting hid_sensor_hub.
I have not disabled PSR via the kernel command line.
(though I’ll admit I haven’t been exercising potential problematic code paths as I’ve had my Framework docked the whole time as a desktop replacement rather than using it as a laptop)
Having similar hard freezes with Asus Zenbook Fedora 36 kernel 5.19.8-200.fc36.x86_64 halt - Ask Fedora
Haven’t had any significant halt after blacklisting hid_sensor_hub.
just to add my experiences on i5-12th gen; hid_sensor_hub blacklisted (purely for brightness cntl) had no effect on stablity. i required i915.enable_psr=0.
only been 2 days since i added psr=0 but i’ve had multiple resume from sleep’s without issue where as before i’d be getting the issue every resume. i’m reasonably confident its stable for me now, hence this post.
2 Likes
@Aggraxis I tried the fix you suggested and unfortunately it didn’t work for me.
Up until yesterday, I was on Fedora 35. I saw random freezes occasionally, but not triggered by any particular action on my part. It did not happen frequently enough for me to care about trying to fix it (I saw <10 freezes in 2 months). I updated to Fedora 36 yesterday. I’m seeing freezes consistently when I am simultaneously listening to an audiobook on hoopla and playing the browser game tower swap. A freeze has not occurred with both websites simply open. I’m using google chrome with each website in a separate window on a separate desktop. My computer will run fine for a little while (length of time is inconsistent) and then freeze up. It’s happened 5 times today with that specific combination, including once after implementing the suggested fix by @Aggraxis. Neither hoopla nor the game has triggered a freeze when I’ve used them separately in the past.
Here’s my system information, copied from the page in settings:
Operating System: Fedora Linux 36
KDE Plasma Version: 5.25.5
KDE Frameworks Version: 5.99.0
Qt Version: 5.15.6
Kernel Version: 6.0.7-200.fc36.x86_64 (64-bit)
Graphics Platform: Wayland
Processors: 16 × 12th Gen Intel® Core™ i5-1240P
Memory: 15.3 GiB of RAM
Graphics Processor: Mesa Intel® Graphics
Manufacturer: Framework
Product Name: Laptop (12th Gen Intel Core)
System Version: A4
In hindsight, I wish I didn’t have the 12th gen intel chip. This is my first time using linux and it has caused a number of frustrating issues, many of which were difficult for me to even diagnose. It doesn’t make sense to me that Framework stopped offering the 11th gen chips when the 12th gen has so many issues.
@Tuesday I’ll try to beat my laptop up for a bit and see if I can get it to misbehave. Unrelated side note, I have a cousin once removed named Tuesday. 
@Iann_C Yeah, I should have mentioned that my instructions were specific for a Fedora system. as @PDXTabs noted, update-initramfs -u would have the same effect as the dracut business. See Ubuntu Manpage: update-initramfs - generate an initramfs image for more details if you are curious. I keep forgetting how fundamentally different various aspects of these two distributions can be.
2 Likes
in terminal/console/whatever you want to call it, does dmesg | grep -i "GPU HANG" show the issue?
if not then it most likely another issue. ‘dmesg’ might show something at the time you encounter the issues to help find the cause.
At the time i placed my 12th gen order (3-4weeks ago) the 11 gens were still available and even discounted. maybe that was to clear stock and now no longer available, but 11th’s have been available.
i dont beleive we’re dealing with a hardware fault in this thread as it’d affect windows too so once the devs at intel/linux kernel get to the bottom of it.
i’m sure i had a similar issue with gpu crash & psr on an 11th gen with my Dell XPS13, but that was eventually fixed.
As a general update, I had another hard freeze last night while listening to an audiobook on scribd and playing the tower swap game. I haven’t played the game without an audiobook going on Fedora 36, so possibly that website is causing the problem, however, I never saw freezes with the game on fedora 35.
I think that command isn’t working when I run it. Below is what I get back when I run it. If I switch the curly quotes to straight quotes, I get nothing back at all. Sorry if I’m misunderstanding your suggestion, but I can’t run anything in the terminal at the time I encounter the issue as my computer is completely frozen and unresponsive to anything other than me holding in the power button until it turns off.
grep: HANG”: No such file or directory
I ordered my 12th gen earlier than you did and at the time I placed my order, 11th gens were not available. The listings were still showing as available on the store, but they were no models actually available when I clicked into the listings (other than maybe just one with totally different specs than I wanted).
My understanding is that the 12th gen intel chips and the current linux kernel do not fully support each other, which leads to a variety of interesting issues.
wonder if the forum has messed with the formatting. my guess is the double quotes are wrong…
the following works for me:
dmesg | grep -i "GPU HANG"
If nothing gets outputted then your not experiencing this issue. if it returns something then it’s likely related to this thread.
I’m in the process of updating docs for Fedora 37. I about done with my 11th gen testing, with 12th gen testing coming up next. It’s good, really good. I’ll see if I can replicate anything there when I get my 12th gen board installed and configured later today/tomorrow.
4 Likes
Thank you! I just ran it and got the following output:
[72899.691144] i915 0000:00:02.0: [drm] GPU HANG: ecode 12:1:859ffffb, in chrome [65524]
[72899.793255] i915 0000:00:02.0: [drm] chrome[65524] context reset due to GPU hang
I should note that I got a soft freeze earlier today though. The computer froze temporarily, but unfroze itself while I was pressing the power button (so it never actually turned off). I think I ran that correctly this morning when I got nothing back. Again, I can’t actually run the command when the hard freeze issue happens as my computer ends up completely frozen and unresponsive, so I have no option other than to restart it.
Already on Fedora 37 myself. Had two hard freezes in one week. Both times it appeared to be related to 1) coming out of sleep following an extended screen lock, and 2) bluetooth mouse had just connected again and almost immediately a hard freeze. I have disabled suspend, sleep, hybrid-sleep, and hibernate by way of a custom sleep.conf for systemd. I don’t use them to begin with, so hopefully this takes care of my freezes. Other than the two hard freezes no glitches…except of course the usual weirdness with docks.
I have not had a chance to try any of the steps listed in this thread yet, but I saw there were some questions about whether anyone had tried Ubuntu yet. I have been having this happen intermittently for the last couple weeks.
I was running Chrome and Inkdrop.
[ 1751.221969] Asynchronous wait on fence 0000:00:02.0:gnome-shell[2840]:7602 timed out (hint:intel_atomic_commit_ready [i915])
[ 1755.070154] i915 0000:00:02.0: [drm] GPU HANG: ecode 12:0:00000000
[ 1755.070215] i915 0000:00:02.0: [drm] Resetting chip for stopped heartbeat on rcs0
[ 1755.073595] i915 0000:00:02.0: [drm] GuC firmware i915/adlp_guc_62.0.3.bin version 62.0 submission:enabled
[ 1755.073604] i915 0000:00:02.0: [drm] GuC SLPC: enabled
[ 1755.073608] i915 0000:00:02.0: [drm] HuC firmware i915/tgl_huc_7.9.3.bin version 7.9 authenticated:yes
- Distro: Ubuntu 22.04.1 LTS
- Kernel: 5.15.0-52-generic
- Adj. Params:
mem_sleep_default=deep - SSD: WDS500G1X0E; however it was NOT mounted at the time
- Repro: the only thing that has been consistent for me is that it always seems to happen when Chrome is the foreground app
- CPU: i5-1240P
Fedora 36 has seen a few kernel updates of late and other updates and I’m wondering if freezing is still going happening for people? At least on the latest kernel and updates.
My experience seems to have been improved. I am on kernel 6.08, recent Framework w/ i5-1240P
I did experience lockups , sometimes in clusters relatively close together, but have been trying some of the recommendations in this thread and they reduced in frequency.
As of now, I’ve been up for over 48 hours running like a charm.
I have been staying away from Settings and not leaving it running in the background. Not sure it helps but most my freezes were in Gnome Settings or Gnome Software, or where they where left running in the background. I’ve definitely seen “I was in settings” mentioned in may of the freeze experiences here.
Anyway, I will try using Settings more today after avoiding it to see if freezes return.