Eggdev: Re: Bugs: Bug report

Will Buckner wcc at techmonkeys.org
Fri Sep 10 15:56:37 CST 2004


on 09/10/04 02:28 Juha Ekholm said the following:
<snip>
> =====================================================================
> 1) INFORMATION ABOUT YOUR EGGDROP
> 
> 1.1) Eggdrop version:
>      1.6.17
<snip>
> 
>      Debug (eggdrop v1.6.17) written Fri Sep 10 09:04:35 2004
> Full Patch List:
> Tcl library: /usr/lib/tcl8.3
> Tcl version: 8.3.3 (header version 8.3.3)
> Compile flags: gcc -pipe -g -O2 -Wall -I.. -I.. -DHAVE_CONFIG_H
> Link flags: gcc -pipe
> Strip flags: touch
> Context: tclhash.c/688, []
>          tclhash.c/680, []
>          tclhash.c/684, [Tcl proc: <null>, param:  $_raw1 $_raw2 $_raw3]
>          tclhash.c/688, []
>          tclhash.c/680, []
>          tclhash.c/684, [Tcl proc: *raw:318, param:  $_raw1 $_raw2 $_raw3]
>          tclhash.c/688, []
>          tclhash.c/680, []
>          tclhash.c/684, [Tcl proc: <null>, param:  $_raw1 $_raw2 $_raw3]
>          tclhash.c/688, []
>          tclhash.c/680, []
>          tclhash.c/684, [Tcl proc: *raw:PING, param:  $_raw1 $_raw2 $_raw3]
>          tclhash.c/688, []
>          tclhash.c/680, []
>          tclhash.c/684, [Tcl proc: <null>, param:  $_raw1 $_raw2 $_raw3]
>          tclhash.c/688 []
> 
Eggdev: What's with proc being NULL when those Context's hit? It seems to me 
that it shouldn't happen. I noticed this with some 1.6.16 bug reports with 
the add_delay crash.

> SOCK ADDR     PORT  NICK      HOST              TYPE
> ---- -------- ----- --------- ----------------- ----
> 3    3EDCEB1A  4433 (users)   *                 lstn  4433
> 4    3EDCEB1A  4432 (bots)    *                 lstn  4432
> 5    00000000     0 (dns)                       dns   (ready)
> 8    51664709 13371 Buttercup idd.cable.ntl.com bot   flags: pUcosgwlia
> 7    82E930F2  6667 (server)  efnet.cs.hut.fi   serv  (lag: 0)
> 10   3EDCEB1A 42991 quaz      quaz at sivari.fi    pass  waited 0s
> 
> Compiled without extensive memory debugging (sorry).
> Open sockets: 3 (listen), 4 (listen), 5 (passed on), 7, 8, 10, 11 (file), done.
> 
> 4.3) Your comments and a description of the bug:
> 
> Opening a telnet connection from localhost (or foreign hosts) crashes the bot.
> 
> 4.4) Can you cause the bug condition to repeat? If so, please outline
>      step by step what causes the error:
> 
> The issue is possible to repeat.
> 
> ./eggdrop durruti.conf
> [ starts ok ]
> telnet localhost 4433
> [ prompts for username and crashes before asking for a password ]     
>      


What username did you enter?


<snip>

> 
> #0  0x400ea899 in nis_clone_result () from /lib/libc.so.6
> #1  0x0000003c in ?? ()
> #2  0x400f4a4a in __gconv_get_alias_db () from /lib/libc.so.6
> #3  0x400f5006 in __gconv_get_alias_db () from /lib/libc.so.6
> #4  0x4016d88e in wcsncasecmp_l () from /lib/libc.so.6
> #5  0x4016da12 in wcsncasecmp_l () from /lib/libc.so.6
> #6  0x4015998c in mbrtowc () from /lib/libc.so.6
> #7  0x401597b9 in mbrlen () from /lib/libc.so.6
> #8  0x4013b9b2 in funlockfile () from /lib/libc.so.6
> #9  0x40125f75 in vfprintf () from /lib/libc.so.6
> #10 0x401442b0 in vsnprintf () from /lib/libc.so.6
> #11 0x08064616 in dprintf_eggdrop (arg1=5) at dccutil.c:113
> #12 0x08062b66 in dcc_telnet_pass (idx=5, atr=4) at dcc.c:1482
> #13 0x0806286f in dcc_telnet_id (idx=5, buf=0xbffff6b0 "quaz", atr=4)
>     at dcc.c:1394
> #14 0x0806afbc in main (argc=2, argv=0xbffff914) at ./main.c:991
>            

Eggdev: Maybe something in add_note()? There was an invalid read that 
valgrind reported that was causing similar crashing on MacOS, but it's 
already been fixed. I'll take a look later.

Wcc



More information about the Eggdev mailing list