dccm dumps core with dcc-dccd-1.1.35

Gary Mills mills@cc.UManitoba.CA
Thu Jul 10 14:54:49 UTC 2003

I had to back out of an upgrade from 1.1.35 to 1.1.42 because dccm
would dump core on startup.  This happens on my production mail server,
but not on my test server.  The command-line arguments are different
between the two servers.  On the production server, they are:

	DCCM_ARGS="-j 400 -S HELO -S mail_host -p inet:NNNN"

And, on the test server:

	DCCM_ARGS="-p inet:NNNN@localhost"

The only message in the log is:

  Jul  9 22:12:49 electra dccm[10968]: [ID 942344 mail.notice] 1.1.42 listening to inet:NNNN with /usr/local/dcc

The core is a result of a segmentation violation, probably by referencing
memory past the end of a segment.  It seems to be contexts that are
involved.  The traceback from the dump is:

  dcc_clnt_rdy(9edde,ff08795c,1,ff,0,1) + 504
  dcc_clnt_init(9edde,0,0,1,1,1) + bc
  dccm_conn(431b48,431c68,fe2ffb54,fe2ffaef,12,9eb28) + 2b8
  st_connectinfo(fe2ffc40,7aba0,36c00,f,17238,21) + 23c
  mi_engine(0,2,0,431c68,1a3b0,7) + 290
  mi_handle_session(431b48,0,0,0,0,0) + 40
  mi_thread_handle_wrapper(431b48,ff095d10,0,5,1,fe401000) + 4
  _thread_start(431b48,0,0,0,0,0) + 40
This is all on Solaris 8.

-Gary Mills-    -Unix Support-    -U of M Academic Computing and Networking-

More information about the DCC mailing list

Contact vjs@rhyolite.com by mail or use the form.