[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-smp
Subject:    Re: Why so much cpu-time in system-mode?
From:       jakob () ostenfeld ! dk
Date:       1998-09-29 15:56:24
[Download RAW message or body]

On Tue, Sep 29, 1998 at 11:41:15AM +0200, Warner Bruns wrote:
> 
>  Dear smp-community,
> 
>  Since two weeks we have a smp with two Pentium-II 300 running.
>  /proc/version is:
> Linux version 2.0.32 (root@tetsmp1) (gcc version 2.7.2.3) #2 Tue Sep 15 15:17:32
>  CEST 1998

It's in the FAQ:   http://www.irisa.fr/prive/mentre/smp-faq/smp-faq-3.html

>  The performance of this beast is exactly as if only one processor would be
>  available. When I have two independent processes, "time" says that
>  both processes consume as much cpu-time in user mode as in system-mode.
>  When only one process is running, "time" says that almost all cpu-time
>  was consumed in user mode.
>  "top" says that both independent jobs get about 100% of cpu each.
>  The two independent processes do almost no I/O, but they both
>  need a memory bandwidth of about 120 MByte/s.
> 
>     Does this ring a bell to somebody?
>  What am i doing wrong?

The explanation is, that all time spent in processors other than the boot cpu
is accounted as system time.  If you time a program, add the user time and the
system time, then you timing will be almost right, except for also including the 
system time that is correctly accounted for.


David, perhaps this could be made a little clearer in the FAQ.


Regards,

................................................................
: jakob@ostenfeld.dtu.dk  : And I see the elder races,         :
:.........................: putrid forms of man                :
:   Jakob Østergaard      : See him rise and claim the earth,  :
:        OZ9ABN           : his downfall is at hand.           :
:.........................:............{Konkhra}...............:

[prev in list] [next in list] [prev in thread] [next in thread]