egghelp/eggheads community

Eggdrops later than 1.6.1 dies on /dcc chat

2002-06-28T15:05:31-04:00

Yes, well, of course I can reproduce it. It -never- fails to happen - /dcc chat is a 100% certain instant crash.

However, what with all the registering and crap I had to do to register a bug, this is it. Apparently it works for everyone else, so I'll just assume eggdrop is no longer for me (used it since pre-1 versions but since 1.6.1 works I may as well stay there).

I did report the simple_sprintf used with %lu once a ways back and that never disappeared (took like 10 versions) so I'm not all that confident in bug reports anyway =) Too bad noone else has this problem or it would probably have been fixed already.

/Micke

(gdb) bt
#0 0x4023f842 in channels_start () from /home/micke/Mii/modules/server.so
#1 0x4023e94c in channels_start () from /home/micke/Mii/modules/server.so
#2 0x80718ba in utf_converter (cdata=0x8188610, myinterp=0x8166100, objc=7, objv=0xbfffe9e0) at tcl.c:321
#3 0x400772e1 in EvalObjv () from /usr/lib/libtcl.so
#4 0x400779eb in Tcl_EvalEx () from /usr/lib/libtcl.so
#5 0x40077c65 in Tcl_Eval () from /usr/lib/libtcl.so
#6 0x4003e9d5 in Tcl_VarEvalVA () from /usr/lib/libtcl.so
#7 0x4003ea0f in Tcl_VarEval () from /usr/lib/libtcl.so
#8 0x80766d0 in trigger_bind (proc=0x81886d8 "*ctcp:server:DCC",
param=0x40241720 " $_ctcpr1 $_ctcpr2 $_ctcpr3 $_ctcpr4 $_ctcpr5 $_ctcpr6") at tclhash.c:680
#9 0x8076908 in check_tcl_bind (tl=0x8160708, match=0xbfffee80 "DCC", atr=0xbfffedf0,
param=0x40241720 " $_ctcpr1 $_ctcpr2 $_ctcpr3 $_ctcpr4 $_ctcpr5 $_ctcpr6", match_type=47) at tclhash.c:746
#10 0x40236a4a in channels_start () from /home/micke/Mii/modules/server.so
#11 0x40237b39 in channels_start () from /home/micke/Mii/modules/server.so
#12 0x4023ec23 in channels_start () from /home/micke/Mii/modules/server.so
#13 0x80718ba in utf_converter (cdata=0x8160740, myinterp=0x8166100, objc=4, objv=0xbffff250) at tcl.c:321
#14 0x400772e1 in EvalObjv () from /usr/lib/libtcl.so
#15 0x400779eb in Tcl_EvalEx () from /usr/lib/libtcl.so
#16 0x40077c65 in Tcl_Eval () from /usr/lib/libtcl.so
#17 0x4003e9d5 in Tcl_VarEvalVA () from /usr/lib/libtcl.so
#18 0x4003ea0f in Tcl_VarEval () from /usr/lib/libtcl.so
#19 0x80766d0 in trigger_bind (proc=0x8160808 "*raw:PRIVMSG", param=0x402416d3 " $_raw1 $_raw2 $_raw3")
at tclhash.c:680
#20 0x8076908 in check_tcl_bind (tl=0x8160678, match=0xbffff8c1 "PRIVMSG", atr=0x0,
param=0x402416d3 " $_raw1 $_raw2 $_raw3", match_type=41) at tclhash.c:746
#21 0x402368a3 in channels_start () from /home/micke/Mii/modules/server.so
#22 0x40239b82 in channels_start () from /home/micke/Mii/modules/server.so
#23 0x8069c48 in main (argc=2, argv=0xbffffb34) at ./main.c:975
#24 0x400fcb65 in __libc_start_main (main=0x8069124

, argc=2, ubp_av=0xbffffb34,
init=0x8049afc <_init>, fini=0x80815cc <_fini>, rtld_fini=0x4000df24 <_dl_fini>, stack_end=0xbffffb2c)
at ../sysdeps/generic/libc-start.c:111
(gdb)

Statistics: Posted by Guest — Fri Jun 28, 2002 3:05 pm

Eggdrops later than 1.6.1 dies on /dcc chat

2002-06-25T22:03:00-04:00

Try to reproduce this when you compile your bot (use the latest version) with make debug. If you can, run a gdb on the core file (gdb eggdrop -c core, then type bt at the prompt), fill out a bugreport, and send it to bugs@eggheads.org.

If you can't reproduce with make debug, I suppose sending in a bugreport might be helpful anyway.

Wcc

Statistics: Posted by Wcc — Tue Jun 25, 2002 10:03 pm

Eggdrops later than 1.6.1 dies on /dcc chat

2002-06-25T21:41:00-04:00

Somewhere, probably in the resolve code, is something that makes eggdrops later than 1.6.1 die horribly on /dcc chat eggie.

I'm using a Windows DNS server on a remote machine, I bet that's it, but eggie shouldn't die from it. The actual message from various versions point to a line in server.c but it makes no sense, probably due to macro expansion. For what it's worth, here's the error:

[00:32] Timeout/EOF ident connection
[00:32] * Last context: server:server.c/1395 []
[00:32] * Please REPORT this BUG!
[00:32] * Check doc/BUG-REPORT on how to do so.
[00:32] * Wrote DEBUG
[00:32] * SEGMENT VIOLATION -- CRASHING!

And just to be cool it also dies on +host DCC commands (I could dcc to it through a reverse DCC command within a script):

[01:09] #Micke# +host Konstiga_ *!konstiga_e@*.hs.bonet.se
[01:09] * Last context: tclhash.c/677 [Tcl proc: *dcc:+host, param: $_dcc1 $_dcc2 $_dcc3]
[01:09] * Please REPORT this BUG!
[01:09] * Check doc/BUG-REPORT on how to do so.
[01:09] * Wrote DEBUG
[01:09] * SEGMENT VIOLATION -- CRASHING!

Did I do the mistake of thinking older linux work work, that windows DNS servers would work or is there something else I could change? I have NO intention of going through 4 hours of registering pain JUST to let someone know about this bug, so this is it. Maybe the eggdrop peaked at 1.6.1 (I've tried 1.6.7, 1.6.10 - both dies on these things while 1.6.1 works perfectly).

Btw, linux version is 2.2.16-22, gotten from redhat something a while back.

/Micke

Statistics: Posted by Guest — Tue Jun 25, 2002 9:41 pm