SUMMARY -unknown errors

From: David Robson (dave@lpdwhy.bhp.com.au)
Date: Thu Aug 04 1994 - 06:07:42 CDT


My original post:
> This message appears on a Sparc2 console (messing up xdm) and I have no
> idea what its about. SunOS 4.1.3, X11R4.
>
> sendsig: bad signal stack pid=29328, sig=11
> sigsp=0xef7ff6b8, action=0xef747ba0,upc=0xa8398
>
> If this is harmless, can this error be redirected in syslog.conf?
>

A couple of usful answers here, thanks to you both!

Looks like I'll have to review my swap.

This can be caused by running out of swap space, here's an old posting
that explains it very well:

----- Begin Included Message -----

In <1993Jun21.110921.16499@ericsson.se> lesley@aom.ericsson.se (Lesley Johansson) writes:

>We have a problem running some applications on a SS10/41 using
>ORACLE Database (SunOS 4.1.3).

>At least once a day the database server crashes and the following is written
>on the consol:

>May 26 13:24:31 teddy vmunix: sendsig: bad signal stack pid=5767, sig=11
>May 26 13:24:31 teddy vmunix: sigsp = 0xdfffff48, action = 0x24f668, upc = 0x23f320

>ORACLE say it's a SunOs problem, Sun that it's an application error.

You have almost surely run out of swap space. When a process runs out of swap
(usually by attempting to expand its stack), it gets a SIGSEGV. However, if
your application has specified a signal handler (with its own stack), the
kernel tries to use that stack. Since you are already out of swap, it can't
allocate space for that stack ("bad signal stack") and kills the process.

If this *is* the correct explanation, I am amazed that Sun support didn't
know about it... but then an identical problem (and Sun finger pointing)
has happened to another customer of mine, which is why I am posting the reply.
(This is actually logged as a Sun bug that states that running out of swap
doesn't give a descriptive error message).

 --------------------------------------------------------------------------
 | Adrian Booth | |
 | Adrian Booth Computing Consultants | Ph/Fax: (09) 354 4936 |
 | 7 Glenrowan Pl., Willetton WA 6155 | abcc@DIALix.oz.au |
 | AUSTRALIA | |
 --------------------------------------------------------------------------
for a in past present future; do
        for b in clients employers associates relatives neighbours pets; do
echo The opinions in this article in no way reflect the opinions of my $a $b.
done;done
>
>

----- End Included Message -----

Check your swap usage using pstat -T

regards,

--
Glenn Satchell                    glenn@uniq.com.au  | NO /\ PAIN
Uniq Professional Services Pty Ltd  ACN 056 279 335  |   /  \
PO Box 70, Paddington, NSW 2021, (Sydney) Australia  |   \  /
Phone 02 360 7434 Pager 016 287 000 Fax 02 331 2572  | NO \/ GAIN

---------------------------------------------------------------------------

I can't help with the actual error message, but as far as messing up xdm goes, are you aware of xconsole? This provides a small console window on the xdm login screen into which console messages are directed. xconsole is part of X11R5, but I have successfully implemented it under Solaris 2.[23]. I also use it to display our motd file (for general announcements), since xdm logins tend not to see motd.

Hope this helps,

Ray Brownrigg ray@isor.vuw.ac.nz

-- David Robson (ROBBO),BHP Information Technology, P.O.Box 21, Whyalla South Australia 5600, Ph: +61 86 404596 Fax:...404720, E-mail dave@lpdwhy.bhp.com.au "Life's battle's don't always go To the stronger or faster man, But soon or late, the man who wins, IS THE MAN WHO THINKS HE CAN." -- Walter W. Wintle



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:09:07 CDT