I noticed a bug in top where the idle% column sometimes contains bogus
numbers when running cpu intense apps like seti at home (as shown
below). I filed a bug report (bug#111779) in December and haven't seen
any action related to the bug. I would like to start investigating thi
bug myself. Does anyone have a pointer to info about the /proc
filesystem, specifically thet areas of /proc where top queries info? I
would like to write a script that polls /proc and mimicks top so I can
tell if this is a top reporting bug or a kernel bug. I can provide more
info about this bug on request or you can see bugs 109484/111779 at
bugzilla.redhat.com
sample top output...
17:55:47 up 1 day, 1:17, 1 user, load average: 3.92, 3.41, 3.63
49 processes: 44 sleeping, 5 running, 0 zombie, 0 stopped
CPU states: cpu user nice system irq softirq iowait idle
total 0.4% 396.0% 3.6% 0.0% 0.0% 0.0%
3681984845051806.0%
cpu00 0.0% 98.8% 1.3% 0.0% 0.0% 0.0%
3681984845051806.5%
cpu01 0.0% 99.0% 1.1% 0.0% 0.0% 0.0%
3681984845051806.5%
cpu02 0.5% 98.6% 0.7% 0.0% 0.0% 0.0% 0.0%
cpu03 0.0% 99.6% 0.3% 0.0% 0.0% 0.0% 0.0%
Mem: 1032080k av, 502588k used, 529492k free, 0k shrd,
76660k buff
115332k active, 259988k inactive
Swap: 1052632k av, 0k used, 1052632k free
230936k cached
PID USER PRI NI SIZE RSS SHARE STAT %CPU %MEM TIME CPU COMMAND
5081 root 26 1 15436 15M 676 R N 99.9 1.4 3:43 1 sah
5082 root 26 1 14404 14M 672 R N 99.9 1.3 3:43 3 sah
5079 root 26 1 15436 15M 676 R N 99.7 1.4 3:43 0 sah
5080 root 26 1 15432 15M 672 R N 99.3 1.4 3:42 2 sah
Thanks,
Dale
