Windows programs report 1 cpu, Debian and Ubuntu report 2.
Prior art:
I've scanned the subjects of debian-user emails from
Apr to Nov, but didn't see this problem discussed.
Hardware:
2 hard disks, Win2K on 1 and the following on the other...
Install history:
Thu Nov 1: Debian 4.0
Fri Nov 2: Ubuntu 7.10
Sat Nov 3: Ubuntu 7.04
and then Debian 4.0
and then Ubuntu 7.10
Sun Nov 4: FreeBSD 6.2
and then Debian 4.0
Reports:
Hardware as seen by Belarc Security Advisor under Windows:
http://savage.net.au/debian.ubuntu/win2k.belarc.report.html (inc mobo)
And as seen by CPUINFO under Windows:
http://savage.net.au/debian.ubuntu/win2k.cpuinfo.png
And as seen by lspci under Debian 4.0:
http://savage.net.au/debian.ubuntu/debian.4.0.lspci.log
And as seen by lspci under Ubuntu 7.04:
http://savage.net.au/debian.ubuntu/ubuntu.7.04.lspci.log
And as seen by lspci under Ubuntu 7.10:
http://savage.net.au/debian.ubuntu/ubuntu.7.10.lspci.log
And as seen by /proc/cpuinfo under Debian 4.0:
http://savage.net.au/debian.ubuntu/debian.4.0.proc.cpuinfo.log
And as seen by /proc/cpuinfo under Ubuntu 7.04:
http://savage.net.au/debian.ubuntu/ubuntu.7.04.proc.cpuinfo.log
And as seen by /proc/cpuinfo under Ubuntu 7.10:
http://savage.net.au/debian.ubuntu/ubuntu.7.10.proc.cpuinfo.log
And as seen by system monitor under Debian 4.0:
http://savage.net.au/debian.ubuntu/debian.4.0.system.monitor.png
And as seen by system monitor under Ubuntu 7.04:
http://savage.net.au/debian.ubuntu/ubuntu.7.04.system.monitor.png
And as seen by system monitor under Ubuntu 7.10:
http://savage.net.au/debian.ubuntu/ubuntu.7.10.system.monitor.png
Any ideas why the OSes don't agree on the CPU count?
Please don't bother posting cynicism about Windows, or wild guesses.
And, yes, I've heard some dual-core chips with 1 faulty core are
sold as mono-core, but even that doesn't explain the discrepancy.
TIA.
--
Ron Savage
http://savage.net.au/index.html
--
Bookmark/Search this post with:
Windows programs report 1 cpu, Debian and Ubuntu report 2
Hyperthreading.
--
John Hasler
--
Windows programs report 1 cpu, Debian and Ubuntu report 2
John Hasler wrote:
> Hyperthreading.
Hmmm. I see bit 28 - the HyperThreading bit - is set in the feature
flags register of the cpu, so that looks like it.
I guess I'll let Debian utilize the feature without any input from me :-).
As we say in Perl: $many x $thanx;
--
Ron Savage
http://savage.net.au/index.html
--