[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: system/489: top(1) shown more than 100% CPU



The following reply was made to PR system/489; it has been noted by GNATS.

From: Felix Schroeter <uk1o_(_at_)_schlund_(_dot_)_de>
To: form_(_at_)_vs_(_dot_)_itam_(_dot_)_nsc_(_dot_)_ru, gnats_(_at_)_openbsd_(_dot_)_org
Cc:  Subject: Re: system/489: top(1) shown more than 100% CPU
Date: Tue, 7 Jul 1998 10:51:29 +0200

 Hello!
 
 On Thu, Jun 11, 1998 at 12:02:02PM +0700, Oleg Safiullin wrote:
 
 > [...]
 > >Synopsis:       top(1) shown more than 100% CPU
 > [...]
 
 > 	sometime top(1) shows more than 100% CPU:
 
 > 	$ top -Ibs1 -d1
 
 > 	load averages:  1.99,  1.80,  1.58    11:57:05
 > 	33 processes:  2 running, 30 idle, 1 zombie
 
 > 	Memory: Real: 4384K/26M act/tot  Free: 31M  Swap: 8K/128M used/tot
 
 
 > 	  PID USERNAME PRI NICE  SIZE   RES STATE WAIT     TIME    CPU COMMAND
 > 	25261 root     105   20  140K  428K run   -       27:59 104.25% cracker
 > [...]
 
 That's because of the way, the CPU percentage is calculated in the
 kernel. It's a decaying average where the decay factor depends on
 the load average. Look at the comments in /sys/kern/kern_synch.c and
 the locations messing with the p_pctcpu field of the proc structure
 for more details.
 
 Regards, Felix.