Erlang Core dump

Valentin valentin@REDACTED
Sat Jun 14 02:17:12 CEST 2003


Any ideas what can cause segmentation violation (causing a core dump) in
ERTS (Erlang Release 8)?
We do have an Erlang production system that's running without a problem
(other than usual human factor) for the last 8 months, and I'm puzzled with
this problem that occurred on our test system -- we are busy testing an
application that provides a consolidation between multiple SCCP and TCAP
layers... anyway, the SCCP and TCAP drivers are passing binary PDUs and some
control information through ERLANG run-time system, and I was wondering if
there is a connection between garbage collection of the binary data and this
crash? Would it make any difference  if I change from binary data to list?

Would there be a performance penalty if I stop using:

 int ei_x_encode_binary(ei_x_buff* x, const void *p, long len)

and change to:

 int ei_x_encode_string_len(ei_x_buff* x, const char* s, int len)

Note that binary data is not interpreted by Erlang in any way but just
forwarded between layers.

Any takers?
Thanks in advance.


More information about the erlang-questions mailing list