https://bugs.kde.org/show_bug.cgi?id=500492

--- Comment #10 from [email protected] ---
I can add some more info on this issue now, which I think would count as a
cause but not an explanation (!?)

It seems I had a quite unusual motherboard fault on the i7-6700K based machine
that I initially installed KDE Neon on. What's remarkable is that when I pulled
the innards out of that machine and replaced them with a new AM5 motherboard,
Ryzen 7700 CPU, and supporting components like RAM, ANOTHER fault of the same
type occurred on the new system!

The nature of the fault(s) is-

1) A device on the motherboard (in the 2 examples I have, these were USB root
hub devices on the motherboard) fails to initialise properly, but fails in a
"noisy" way. This causes the USB driver stack to get stuck in a loop
continually re-enumerating the device tree. That's my non-expert
interpretation.

I had help from the kernel USB driver maintainers who looked at usbmon logs
from the newer machine and this is their expert analysis-

"The usbmon trace shows that port 4 on bus 2 generates a continual
stream of link-state-change events, constantly interrupting the system
and consuming computational resources.  That's why the performance
goes way down.

I can't tell what's causing those link-state changes.  It _looks_ like
what you would get if there was an intermittent electrical connection
causing random voltage fluctuations.
..."

So, it seems the underlying cause is a hardware fault in my case. However it is
noteworthy (I think) that the performance impacts seem to particularly affect
the desktop environment, including having the strange effect of tending to make
the refresh of panel widgets fail.

This issue clearly isn't something the desktop environment is responsible for,
so this bug needs to be closed. However I suggest it might be worth
considering-

1) Filing this info away in the memory banks (bug knowledge base?) in case
another issue with strange performance slowdowns comes up
2) Is there any way the DE could detect and highlight a situation like this? I
have to wonder if there are machines out there that have been running dog-slow
for months or years, or even driven the user to ditch Linux and/or KDE, with
the user unaware of the root cause because (like most people?) they don't make
a habit of looking at dmesg output.

-- 
You are receiving this mail because:
You are watching all bug changes.

Reply via email to