You are not logged in.
The previous time I had this issue, I had explained the issue in detail (here: https://bbs.archlinux.org/viewtopic.php?id=289309)
(and I got a solution)
For the next two months, there was no issue - it ran perfectly.
However, it is now happening again, and not just when playing fullscreen videos in a browser, but _randomly_. And the frequency has also increased.
Just like I mentioned in that post, there are no capturable logs or anything.
I have tried compositor-less openbox, I have gone to the other end of the spectrum and tried KDE plasma wayland -- it happens on all of them.
Is it a hardware issue?, or perhaps a mesa issue?
Should I try it with windows and check?
Thanks in advance
Last edited by porridgewithraisins (2023-11-10 10:38:48)
Offline
Trying some live-distro like grml or knoppix is probably easier.
And did you try the LTS kernel?
You didn't install xf86-video-intel, did you?
Offline
Hi again, seth.
You were the one who helped me in that previous post as well.
I will try in a live distro and tell you what happens.
Yes, I have tried mainline, zen and lts.
I have tried debian stable, fedora and arch.
On arch, I have tried all combinations, with and without xf86-video-intel and vulkan-intel.
It happens everywhere.
Offline
I have tried debian stable, fedora and arch.
…
It happens everywhere.
Feel free to also try windows, but
Is it a hardware issue?
If you're "lucky", it's RAM or simply a (D.U.S.T. driven) temperature issue.
Offline
All sensor output says its not a temperature issue. And I did a full memtest a while back, no issues with the RAM. My BIOS' diagnostic tests were all green (including fans)
So I will try a live distro, as well as windows, and let you know.
Last edited by porridgewithraisins (2023-11-12 04:43:56)
Offline
Is it possible that it is a mesa issue? Others have reported crashes with minor versions of mesa v22, as well as few with 23.0.x, this issue started happening for me around that time, now that I think of it.
Also, is it useful to try this? https://wiki.archlinux.org/title/Intel_ … Intel_CPUs
Offline
mesa versions would most likely clash w/ xf86-video-intel and otherwise we probably had seen more issues and not as sporadically.
You can try to keep the CPU busy, but those freezes would typically not lead to "flicker" (the output just freezes) and not be tied to the rather specific conditions of your first thread.
The aggrevating nature rather suggests a hardware issue now ![]()
If the pattern continues on other software stacks, "I did a full memtest a while back, no issues with the RAM" - useful memtest86 runs are measured in days. If you didn't run it overnight (16+h) you might want to look there again.
Offline
Hmm, alright. Yeah, this computer is from 2012 or so.
One of the RAM sticks is new (yes I made sure its the same frequency as the old one).
So there is an older RAM stick hanging around from 2012, should I remove that, and check if errors happen?
The CPU wasn't changed.
I upgraded the HDD to an SSD.
And the NVRAM coin cell battery is new as well.
Offline
should I remove that, and check if errors happen?
It's the next-easy test if limiting the c-states etc. (software fix) doesn't cut it (since memtest86 will last much longer to yield results)
Also/alternatively make sure to configure the most conservative RAM timings/frequency in yur BIOS/UEFI
Offline