Re: AMD Athlon bogus performance value causing RCU stalls?

From: Rob Prowel
Date: Sun Sep 23 2018 - 17:55:32 EST

On 09/23/18 17:19, Thomas Gleixner wrote:

What is that value for the second Athlon core (seems extremely bogus), and
would/could that be the reason for the schedule_timeouts? This bogus value
also shows up in the bootup log when the second core is activated. Seems to
be AMD specific, as the values are correct on my Xeon machines.

That's a 32bit machine I assume.

No. Actually it is a 64 bit AMD Athlon. Might need to check the math being a
problem in the 64 bit domain as well.

Yes. This was decoded recently as an issue on 32bit due to a calculation
which is based on 'unsigned long' but requires to be 64bit wide.

It's in the 4.18.8 stable kernel, which should be available from your
fedora repo anytime soon.

Will check with the newer kernel when it's available but please do double check if this
problem might exist on 64 bit architectures as well...and of course I'm interested to
know whether it is the likely cuprit of my RCU problem.