Re: acpicpu slowing down computer [was Re: firefox slowness -- X server problem?]

[Paul Goyette <paul%whooppee.com@localhost>]

On Thu, 19 Aug 2010, Thomas Klausner wrote:

8<  Snip acpicpu stuff >8

> During boot, I always get:
> aibs0: warning over limit on 'MB Temperature'
> aibs0: warning under limit on 'CHASSIS1 FAN Speed'
> aibs0: warning under limit on 'CHASSIS2 FAN Speed'
> aibs0: warning under limit on 'CHASSIS3 FAN Speed'
> aibs0: warning under limit on 'POWER FAN Speed'
> but envstat output right (no acpicpu) now looks ok:
>
>                        Current  CritMax  WarnMax  WarnMin  CritMin Unit
> [aibs0]
>       Vcore Voltage:     0.936    1.600                      0.800    V
>        +3.3 Voltage:     3.216    3.630                      2.970    V
>          +5 Voltage:     4.878    5.500                      4.500    V
>         +12 Voltage:    12.137   13.800                     10.200    V
>     CPU Temperature:    49.000   75.000   60.000                   degC
>      MB Temperature:    46.000   75.000   45.000                   degC
>       CPU FAN Speed:      1318              7200      600           RPM
>  CHASSIS1 FAN Speed:         0              7200      600           RPM
>  CHASSIS2 FAN Speed:         0              7200      600           RPM
>  CHASSIS3 FAN Speed:         0              7200      600           RPM
>     POWER FAN Speed:         0                       7200           RPM

The problem here is that there is no way to tell if a fan is absent vs 
stalled (installed-but-not-spinning).  So when the BIOS specifies a 
WarnMin limit and the measured value is 0, we generate the "warning 
under limit" event.

Are you really running with only the CPU cooler fan, and no case fan? 
Or is your case fan an older 2-wire model which doesn't provide a Sense 
wire for measurement?

The CPU is
cpu0: "Intel(R) Xeon(R) CPU           W3520  @ 2.67GHz"
but it's often misregnized by NetBSD with a higher clockrate:
cpu0: Intel Core i7 (Nehalem) (686-class), 3207.40 MHz, id 0x106a5
(both lines from the same cpuctl output).

The first line is a text string set-up by your motherboard's BIOS, while 
the second line is NetBSD's own decode of the CPU signature based on the 
CPU_ID.  I don't have my decode-sheet handy to tell which is correct.
Based on the fact that the BIOS seems to have gotten the clock speed 
wrong, I'd suspect it also got the CPU model wrong.  (An updated BIOS 
might be able to decode the more-modern CPU_ID.)


-------------------------------------------------------------------------
| Paul Goyette     | PGP Key fingerprint:     | E-mail addresses:       |
| Customer Service | FA29 0E3B 35AF E8AE 6651 | paul at whooppee.com    |
| Network Engineer | 0786 F758 55DE 53BA 7731 | pgoyette at juniper.net |
| Kernel Developer |                          | pgoyette at netbsd.org  |
-------------------------------------------------------------------------