Top command crashing with floating point exception!
dirkw at rentec.com
Tue Mar 5 11:45:01 CST 2002
did you check the library checksums? where does 'strace -o file top'
stop ? another possibility is that top maybe reads something out of
the procfs which it doesn't understand. this could be the case e.g.
if you use a pretty recent kernel, but an ancient top version.
Dirk Wetter @ Renaissance Techn.
mailto:<dirkw at rentec dot com>
btw: a "very good" intruder modifies everything on your system, so that
he is hiding his traces. this includes rpm checksums for libraries
and binaries as well as e.g. hiding sniffers in the process list or
on the filesystem. but in most cases intruders leave traces there or
somewhere else ... fortunately :)
On Tue, 5 Mar 2002, Basil Hussain wrote:
> > Well, that is unless "checksums for the binary" means that you did this
> > already--sorry.
> Yes, like I said, I've already done this...
> > It would be a good idea to verify that binary first.
> > rpm -vVf $(which top) && echo OK || echo FAILED
> I downloaded a fresh copy of top from the procps RPM for RH7.0 from my local
> RedHat mirror (www.mirror.ac.uk - they have a great facility to 'browse'
> inside RPMs, zips, etc.) onto my Windows desktop, got the MD5 sum of it and
> compared it to the one on the server. It matches, so I conclude the binary
> is not trojanised.
> So, anybody have any ideas why top is suddenly throwing tantrums?
> Basil Hussain
> Internet Developer, Kodak Weddings
> E-Mail: basil.hussain at kodakweddings.com
> > Basil Hussain wrote:
> > >Hi,
> > >
> > >Just today, whilst winging my way around a command-line session
> > on one of my
> > >Linux servers (Dell-installed RH7.0, 2.2.16-enterprise) I ran the 'top'
> > >command and receieved, much to my puzzlement, the following error:
> > >
> > > 12:09pm up 252 days, 20:29, 1 user, load average: 0.00, 0.00, 0.00
> > >51 processes: 50 sleeping, 1 running, 0 zombie, 0 stopped
> > >Floating point exception (core dumped)
> > >
> > >Why on earth would top crash with such an error? Any specific
> > reasons this
> > >would happen? I've tried using 'ps', which seems to work fine.
> > >
> > >By the way, before anyone talks about intrusions, I've already had that
> > >thought (with top being trojanised, albeit malfunctioning). I've
> > checked all
> > >the usuals - checksums for the binary, syslogs (or lack of), unusual
> > >listening ports, etc. and nothing is out of the ordinary. So,
> > this doesn't
> > >appear to be the case. Must be just something ordinary like a bug...
> > >
> > >Regards,
> > >
> > >Basil Hussain
> > >---------------------------------------
> > >Internet Developer, Kodak Weddings
> > >E-Mail: basil.hussain at kodakweddings.com
> Linux-PowerEdge mailing list
> Linux-PowerEdge at dell.com
> Please read the FAQ at http://lists.us.dell.com/faq or search the list archives at http://lists.us.dell.com/htdig/
More information about the Linux-PowerEdge