System halts [solved]

[JohnJanzen]

After exactly 7260 seconds the system comes to a complete halt ending the kern.log file with.

[drm:drm_atomic_helper_wait_for_flip_done] *ERROR* [CRTC:57:crtc-1] flip_done timed out
[drm:drm_atomic_helper_wait_for_flip_done] *ERROR* [CRTC:54:crtc-0] flip_done timed out
[drm:drm_atomic_helper_wait_for_dependencies] *ERROR* [CRTC:60:crtc-2] flip_done timed out

full kernel log of that run : https://pastebin.com/nygNQRMn

Im using amdgpu and an rx570 8GB [As this seems somehow GPU related]

[alamahant]

Hi
Welcome to Gentoo Forums!
Would it be ok with

[JohnJanzen]

[Banana]

you can search for the key CONFIG_DRM_KMS_HELPER in menuconfig and then use the result number to jump to the entry
_________________
My personal space
My delta-labs.org snippets do expire

PFL - Portage file list - find which package a file or command belongs to.

[JohnJanzen]

It seems the drm error wasn't the error that was crash/halt ing my system

However it halted again, I took a walk (that took longer than expected) and my System still works as I come back, however 3-4 minutes after using it, it halted again

However the time it failed wasn't logged at all


I will try to get a kernel log tomorrow as it's late here.

[Banana]

did you do some memory and disk check? Maybe something is corrupt and leads to this error.
_________________
My personal space
My delta-labs.org snippets do expire

PFL - Portage file list - find which package a file or command belongs to.

[JohnJanzen]

smartctl tells me my drive passed their check

Currently running a memtest

[JohnJanzen]

Also I've seen the new kernel log has some starts with root@livecd in it, despite me not booting from the livedisk today, (will post the log when my memtest finishes)

[JohnJanzen]

The Memtest was normal [no errors] with all my ram [16GB] detected and all 12 cores of my CPU [Ryzen 5 1600x] working
Disk also hassed passed the smartctl test
result here : https://pastebin.com/94wMzn0C

The kernel Logs of Today are here
https://pastebin.com/vEBC7q7U <- link edited as it didn't paste fully


It says root@livecd at the start as my System as I rebuild my kernel with a liveUSB [Recompiled right now, so on future logs it'll say root@gentuwu]
I used chroot and mounted partitions from my normal SSD + settings of course


The last time this happened, instead of the output just getting stuck, after pressing ALT+SHIFT+F2 [Something I always try] my screens went black [with backlight still on]
Although I don't know if this was caused by the keypresses, as me logging in blindly as root and touching a file didn't leave a file in the /root directory [I could have misspelled something tho]


I might try to run Audio nonstop next time to see if it continues [And maybe even accepts commands from my keyboard][Maybe remote control through spotify will work ?]

[Tony0945]

I've noted freezes with 5.10.27 Haven't checked kern.log
Try using 5.4.97 and see if the problem persists.

[JohnJanzen]

Idk if this system is trolling me, but since last rebuild I didn't encounter any freeze [Only 3h and 20minutes uptime yet, might run it through the night]

with constant mouse movement / drags over a browser window [Pyautogui script]

Maybe it has to do with going into Xorg power saving mode ?

Won't conduct many tests tomorrow as I am not going to be home after 1pm.

[JohnJanzen]

Okay it happened again, alltough after only very little time

I turned my PC on, connected my Phone and enabled tethering

with this log
https://pastebin.com/39Z5T4Gh


first my keyboard controls where not recognized but my mouse moved, but that stopped too shortly after

[JohnJanzen]

I have thought about this being an Xorg issue, as my Xorg logs are ended with "your system is too slow"
But after researching this further that doesn't seem to be the issue

Xorg log:
https://pastebin.com/U0YfFBqt


will try using prebuilt kernel / older kernel

[JohnJanzen]

It happened again, but now I can move the mouse with some "lag spikes" (freezes on spot but 1-2 seconds later I can move it again)

I can't press any keys / get no output from that, it's just a moving mouse on a background. with everything being redrawn (aka if I move my mouse, the background reapears without the mouse making a "trail")

[Tony0945]

Can you click on anything and see an effect?

[JohnJanzen]

No, neither dwm nor any Graphical programs, seem to register anything

[Tony0945]

Same here, but not as regular as yours. The first time I used the power button but thereafter, I've gone to another computer, ssh'd in and rebooted. The fact that we can ssh means the processor isn't frozen. Yes, it must be Xorg. There were a lot of changes in the last year. I wonder if I (we?) revert to an older version, the problems will dsappear.

[JohnJanzen]

I actually found out what caused this a while ago.

It was the C states of the ryzen 1x00 series not being correct / non existant at level 6
Which is used for powersaving, wich explains why the freeze didn't occur under full load but only when the system goes idle.

And it is fixed by disabling the C6 state. (Using tools like zenstates.py)

[Tony0945]

Ah! Unfortunately, it doesn't happen on my 2700X but on my Phenom II X6