Possible conflicts and error output in dmesg

[LIsLinuxIsSogood]

Hey so,
I am trying to make sense of what appears to be some conflict issues with the hardware and kernel modules, but I can't say I really understand what's going on with these messages, and other than occasional connectivity concerns with the wifi (usually only after services are being started and stopped and system becomes unstable, with a tainted kernel message from the BROADCOM-STA driver that is wl.ko.

My past habits with buying hardware was usually to read about it at the time and sort of hope it will work on the linux builds I have, like on Gentoo and ubuntu or other debian based systems.

But based on the information in the kernel ring buffer: https://paste.pound-python.org/show/qg9dZ2O3eWyfI6lxq5m1/ that is why I am posting the entire thing to see if someone could check out based on a list of modules here: https://paste.pound-python.org/show/xVVZY7Dgojwv2PtdwMop/
Here's a list of the hardware with kernel modules used, https://paste.pound-python.org/show/YT2OGQTB0L6S1D4EVBIu/


As for the graphics card that is one area where I have had really no complaints (none really).

But if the messages might suggest possible conflicts with that between kvm and nvidia, then I would like to know if these have issues in combination with one another... Also what does the line that says NVIDIA UNIX have to do with is it because of the proprietary drivers?


Then if broadcom's proprietary wl driver is conflicting then what is it in conflict with? Could that also be in conflict with kvm? I looked up this information about the wifi card from the linux wireless wiki it looks like I might be stuck with the proprietary module (not built-into kernel)

PCI-ID Supported? Chip ID Modes PHY version Alternative
14e4:43b1 no BCM4352 a/b/g/n/ac AC (r3) wl
https://wireless.wiki.kernel.org/en/users/Drivers/b43#Supported_devices

Would it be wise for me to purchase a new wireless card so that it won't taint the kernel?

In general the wireless connection had been really pretty stable, but then again I need to make sure that it isn't somehow compromising more critical applications like the ones I now want to run over the network, locally. One thing I really don't like that I am seeing in the kernel messages are the last two lines of the kernel log!
[ 7.970040] ERROR @wl_notify_scan_status :
[ 7.970044] wlp4s0 Scan_results error (-22)

[bunder]

[Hu]

I wouldn't buy a replacement card without more definitive evidence that the physical card, or the proprietary driver you are stuck using, are clearly responsible for an intolerable problem. However, if you had free access, even if only for a few days, to a supported wireless card, temporarily switching to it could be useful from a diagnostic perspective.

"Literally everything" is an overstatement, particularly with the "literally" qualifier. It's very possible to run an untainted kernel, although that does sometimes involve compromising on choice of hardware. Some kernel taints can be very important. Nothing shown so far leads me to believe this particular taint is critically important.

The kernel can be tainted by many things. Generally, taint due to loading proprietary drivers will make it difficult to get support from upstream, because some proprietary drivers have historically been such a mess that some developers treat "Loaded proprietary driver" as equivalent to "Started flipping random bits in important data" and then refuse to debug anything in such an undefined state.

If the machine is not too badly crippled by removal of the proprietary components, I would suggest trying to reproduce any significant problems in an untainted kernel. For the wireless problems, this is obviously impossible, since the untainted kernel will not be able to operate its wireless card at all. However, for other types of problems, it can be useful to try to determine whether the problem is related to a proprietary driver.