[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

getting a core dump

This seems like a dumb question, but after some time searching the HOWTOs
the kernel source docs, the net and even a couple of books(!) I've failed
to find the answer.

I'm trying to run/debug the arm version of the Citrix client. It doesn't
work and it's supplied as binary-only software (grumble). All I get is a
segfault. Citrix looked at the straces I sent them and now want core
dumps to work out what's going wrong.

But I can't for the life of me work out how to get a core dump out of
Debian-arm 3.0. My old red-hat box used to leave them lying all over the
place, but all I can find now is /proc/kcore. But a) this is stupendously
massive and b) I presumably need to get it copied somewhere at the moment
of the segfault otherwise it's going to represent the kernel state much
later than the point when the program crashes.

So a) how do I trigger a core dump (or whatever the more modern equivalent
is (oops text?)) on a segfault and b) what _is_ the accepted form for
generating a feasibly-sized file I can send to Citrix?

It seems that ksymoops might be a relavant tool, but I haven't got any oops
text to give it. Should there be some somewhere on getting a segfault? I
have modules and ksyms files being generated in /var/log/ksymoops at 06:25
(due to something anacron is running), but there is no correspondng oops
text in and of the logs I can find or in the kernel message buffer (dmseg)-
where does/should it go?

I'm hoping someone here can point me to the docs/commands I have failed to

Aleph One Ltd, Bottisham, CAMBRIDGE, CB5 9BA, UK  Tel +44 (0) 1223 811679
work: http://www.aleph1.co.uk/     play: http://www.chaos.org.uk/~wookey/

Reply to: