Re: Processor cores not properly detected/activated?

From: Tim Bishop <tim-lists_at_bishnet.net>
Date: Sat, 24 May 2014 11:38:35 +0100
On Fri, May 23, 2014 at 09:03:12PM -0600, Alan Somers wrote:
> On Fri, May 23, 2014 at 8:42 PM, Tim Bishop <tim-lists_at_bishnet.net> wrote:
> > On Fri, May 23, 2014 at 08:07:03PM -0600, Alan Somers wrote:
> >> On Fri, May 23, 2014 at 7:47 PM, Tim Bishop <tim-lists_at_bishnet.net> wrote:
> >> > I have a new quad CPU system containing four of these processors:
> >> >
> >> > Intel(R) Xeon(R) CPU E7-4830 v2 _at_ 2.20GHz (2200.05-MHz K8-class CPU)
> >> >
> >> > I've tried FreeBSD 10.0, stable/10 and head, but all of them only detect
> >> > a maximum of 64 "CPUs". There should be 80. Here's the relevant dmesg
> >> > output (full output attached):
> >> >
> >> > FreeBSD/SMP: Multiprocessor System Detected: 64 CPUs
> >> > FreeBSD/SMP: 3 package(s) x 10 core(s) x 2 SMT threads
> >> > ...
> >>
> >> Try setting MAXCPU higher.  It's defined by default to 64 in,
> >> sys/amd64/include/param.h
> >
> > Ah! Thank you, yes, that fixed it:
> >
> > FreeBSD/SMP: Multiprocessor System Detected: 80 CPUs
> > FreeBSD/SMP: 4 package(s) x 10 core(s) x 2 SMT threads
> >
> > Given the number of "CPUs" in some top end processors (up to 30 per
> > socket), a limit of 64 is starting to seem low. Is it worth doubling it
> > to 128? Or even higher?
> 
> Yeah, I think so.  It seems like a GENERIC kernel ought to be able to
> handle the biggest commonly available quad socket systems.  Anything
> with more than 4 sockets, though, is probably too exotic to deserve
> such special treatment.

I submitted a PR to that effect:

http://www.freebsd.org/cgi/query-pr.cgi?pr=190169

Thanks again for your help.

Tim.

-- 
Tim Bishop
http://www.bishnet.net/tim/
PGP Key: 0x6C226B37FDF38D55
Received on Sat May 24 2014 - 08:38:37 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:49 UTC