On Saturday, Jun 19, 2004, at 01:26 Australia/Sydney, Richard Sims
wrote:

"Signal 11" is included in ADSM.QuickFacts, where it is explained to
be a
programming error: you may feel justified to report to IBM as a TSM
defect.
(The Linux TSM is still the junior member of the TSM family, going
through
growing pains which its brethern have experienced and overcome in the
past.)
Searching IBM for  +"segfault" +"producer thread"  turns up only an
old APAR, so
I would advise calling in this new one.

The manpages on the Linux box I have here don't seem to say, but my
understanding is that sig 11 (SIGSEGV) will be sent when the stack
ulimit is exceeded. The NetBSD manpages back this up.

Quick test program backs this up on a Linux 2.4 kernel and AIX 5.1.
Richard, you may want to add this to your excellent ADSM Quick Facts!
You mention UNIX resource limits, but no specific one. AFAIK, the
stack limit is the only one that causes a SEGV.

Of course, this is in addition to the more common cause of a SEGV,
which is accessing memory you don't have mapped, ie. a software bug.

Cheers,
--
Paul Ripke
Unix/OpenVMS/TSM/DBA
I love deadlines. I like the whooshing sound they make as they fly by.
-- Douglas Adams