On Sat, 10 Nov 2001, Artovil wrote:
> while ((trigger_row = mysql_fetch_row(assigns))) {
> /*
> * Something goes wrong here when it assigns the trigger to the mob,
> * the mud SIGSEGV's on parse_object at db.c:1637
> * #0 parse_object (object_row=0x823d7b8, extradescs=0x8260c28,
>affects=0x81c7cb8, vnum=106) at db.c:1637
[...]
> dg_read_trigger(trigger_row, &mob_proto[i], MOB_TRIGGER);
Can you confirm 'i' is accurate?
Try adding '-lefence' to LIBS in Makefile as well. Then run CircleMUD
again. With luck, it'll crash where the problem is instead of the
symptom. If that doesn't work, try:
EF_PROTECT_BELOW=1 bin/circle -c
>I had a backtrace in my comments there... If I just comment out the line
>with dg_read_trigger, it works just fine. It passes the row instead of a
>file, and this works fine elsewhere, so I suppose it is something else,
>since I have it set up like a lot of other functions that work.
The other fun part about memory corruption is that you could have the
problem elsewhere that your trigger function just amplifies to cause a
crash in parse_object().
--
George Greer
greerga@circlemud.org
--
+---------------------------------------------------------------+
| FAQ: http://qsilver.queensu.ca/~fletchra/Circle/list-faq.html |
| Archives: http://post.queensu.ca/listserv/wwwarch/circle.html |
+---------------------------------------------------------------+
This archive was generated by hypermail 2b30 : 12/06/01 PST