Bug 119240 - top has WCPU over 100% on UP system
Summary: top has WCPU over 100% on UP system
Status: Closed FIXED
Alias: None
Product: Base System
Classification: Unclassified
Component: sparc64 (show other bugs)
Version: Unspecified
Hardware: Any Any
: Normal Affects Only Me
Assignee: freebsd-sparc64 (Nobody)
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2008-01-02 02:00 UTC by Stephen Hurd
Modified: 2011-02-23 01:14 UTC (History)
0 users

See Also:

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Stephen Hurd freebsd_committer freebsd_triage 2008-01-02 02:00:03 UTC 
On my blade 100, top can be seen showing WCPU of over 100% on a single process
59 processes:  2 running, 56 sleeping, 1 zombie
CPU states: 69.0% user,  0.0% nice, 29.5% system,  0.5% interrupt,  0.9% idle
Mem: 96M Active, 129M Inact, 97M Wired, 3528K Cache, 60M Buf, 162M Free
Swap: 1024M Total, 1024M Free

  PID USERNAME    THR PRI NICE   SIZE    RES STATE    TIME   WCPU COMMAND
81243 root          1 132    0 22776K 20296K RUN      0:01 118.00% cc1plus
81240 root          1  98    0  7952K  2280K RUN      0:00  1.54% top
81242 root          1   8    0  3944K  1184K wait     0:00  1.00% c++
 1046 haldaemon     1  96    0 21032K  4584K select   0:41  0.00% hald
  844 root          1  97    0  6072K  1160K select   0:15  0.00% moused
 1134 root          1  96    0 10880K  3520K select   0:09  0.00% sendmail
 1144 root          1   8    0  5944K  1416K nanslp   0:05  0.00% cron
 1011 root          1  96    0   261M  1408K select   0:05  0.00% rpc.statd
  959 root          1  96    0  4880K  1336K select   0:05  0.00% syslogd
79200 admin         1  -4    0 70816K 14392K nfs      0:05  0.00% xfce4-panel
 1200 root          1   8    0  4864K  1176K nanslp   0:03  0.00% getty
 1118 daemon        1   4    0  3768K  1184K sbwait   0:03  0.00% rwhod
 1017 root          1  96    0  5944K  1792K select   0:03  0.00% rpc.lockd
  977 root          1  96    0  5944K  1648K select   0:03  0.00% rpcbind
 1219 root          1  20    0  9856K  2928K pause    0:01  0.00% csh
 1032 messagebus    1  96    0  5280K  1872K select   0:01  0.00% dbus-daemon
63252 root          1   8    0  2592K  1688K wait     0:01  0.00% make
62944 admin         1  20    0  9856K  2872K pause    0:01  0.00% tcsh
63208 root          1   8    0  2592K  1488K wait     0:01  0.00% make
80343 root          1  20    0  9856K  3168K pause    0:00  0.00% csh
67357 root          1   8    0  6240K  1736K wait     0:00  0.00% gmake
81137 root          1   5    0  5504K  2376K ttyin    0:00  0.00% vi
62499 admin         1  -1    0  9856K  2856K nfsndl   0:00  0.00% tcsh
69551 root          1   8    0  6448K  1680K wait     0:00  0.00% sh
81012 root          1  20    0  9856K  2872K pause    0:00  0.00% csh
12591 root          1   8    0 19688K  1856K wait     0:00  0.00% login

How-To-Repeat: Watch top output during port builds.
Comment 1 craig001 2009-01-06 21:16:36 UTC 
Hello

I have also seen the same result on sparc64 and amd64 kit,  I have
looked into it a bit on google and the general feeling is that the WCPU
(or weighted CPU) count is not all that accurate.

A more accurate representation can be obtained by passing the -C flag
when starting top.  This is for raw CPU rather than a weighted average,
on some quick tests on a V120 whilst running buildworld I have not seen
a reading above 100%.

Kind Regards

Craig Butler
Comment 2 Eitan Adler freebsd_committer freebsd_triage 2011-02-23 01:13:31 UTC 
State Changed
From-To: open->closed

WCPU is a scheduler thing. This is not a bug