What Logs Should I Collect To Troubleshoot an Issue?

[bob466]

I wouldn't. I would instead ask you to upload the .dmp file in your next reply, then run it through Blue Screen View. I'd have the cause of the BSOD in less than 20 seconds: far quicker than it would take you to even Google search the hex code.

The cause was always windoze...the reason it doesn't happen in Linux.

 

[Jack_Sparrow]

Thanks for those logs. After inspecting them, there did not appear to be anything critical in relation to freezing, except for one journal entry which was this:

nouveau 0000:01:00.0: drm: failed to create cd channel, -22

This error, or failure message as it is, comes from the nouveau gpu driver. The number 22 is an error code with the meaning that there's been an invalid argument in the code which has caused a failure. Error meaning check here:

[~]$ errno -l | grep 22
EINVAL 22 Invalid argument

The nouveau driver has tried to initialise part of the gpu processes, but the expected argument, or option the gpu expected, wasn't provided. The problem with this is simply that it leaves the gpu in a state that is less than optimal, which is going to make it vulnerable, and freezing is certainly going to be a possible behaviour as a result of the "Invalid argument".

This is the only vulnerability related to freezing that I could see in all the logs you provided, and so I guess that this is where I'd be applying some attention to try and resolve the issue.

The nouveau driver is used here on a number of machines, but they have more modern graphics cards than yours, and this issue hasn't arisen. The "NVIDIA GF119M [NVS 4200M]" graphics card was released in 2011 I found online, which makes it old.

The nvidia proprietary driver for that card, the 390.xx, is EOL (end of line). It was never used here anyway so I can't say anything about it.

The machine you have runs hybrid graphics with the intel and nvidia cards. They are the graphics devices 1 and 2 respectively in the inxi output in post #1.

It's probably worth trying to disable the nvidia card altogether and just run the intel card, since the intel card is supported with the i915 driver, and there are no errors associated with it anyway.

There are few ways of disabling nouveau. Nouveau can be blacklisted by creating a blacklist file named: blacklist-nouveau.conf, and placed in the directory: /etc/modprobe.d/. It'll look like:
/etc/modprobe.d/blacklist-nouveau.conf. This needs root or sudo permissions of course.
The contents of the file: blacklist-nouveau.conf should be:

blacklist nouveau
options nouveau modeset=0

Then one needs to update the initramfs as root or sudo running:

update-initramfs -u

Then reboot. On rebooting one can check to see if the nouveau driver has been loaded with:

lsmod | grep nouveau

The output should be null for nouveau, that is, no nouveau shown.

An alternative means of disabling nouveau is to add the following kernel parameters to the linux kernel line:

modprobe.blacklist=nouveau nouveau.modeset=0

If unsure about adding kernel parameters, see here: https://linuxconfig.org/how-to-set-kernel-boot-parameters-on-linux. Again, after booting, one can check to see if the nouveau driver is absent, with the above lsmod command.
The above is how I'd deal with the issue to see if it works. If not, it's all reversible, and one can start again.

Unfortunately, this is not a solution. HDMI breaks completely (no audio, no video). Like you said at the end, I was able to reverse everything, and using the nouveau driver, I get HDMI again. But the freezing still persists.