- DALnet-src - nomad.dal.net

vhost
by Sergey Shmyrov 19 Oct '10

19 Oct '10

Hello. Can I use virtual hosts (name.local, mysuper.host). If I use them, how? OS: Freebsd, bahamut-1.8.9, no services.

2 1

(no subject)
by Fuad Aliyev 11 Oct '10

11 Oct '10

hello.I need your help guys.im using bahamut with anope services.i got such problem,when im checking channel info it says -ChanServ- For more verbose information, type /msg ChanServ INFO #services ALL. But when i type /msg ChanServ INFO #services ALL it doesnt work.it says Error! "/msg ChanServ" is no longer supported. Use "/msg ChanServ(a)services.domain" or "/ChanServ" instead. when i use /msg chanserv(a)services.domain and /cs info #channel all it works.How to make the first one make working ? and 2nd question what means O *@* * Trusty 15728631 opers this numbers here.im root admin with this numbers.How to make this numbers more with 49 ?

2 2

initconf issue still persists
by Juan Baez 19 Aug '10

19 Aug '10

Back on June I had reported a problem with the function "initconf" from "src/confparse.c" This function basically calls fopen but never bothers to call fclose regardless whether the function was successful or not. I've attached the original message that I had sent that describes the problem in more detail, as well as a git formatted patch with the latest git repository to fix this. It may not seem as a problem, but this bug eventually caught up to our server; it made the IRCd incapable of accepting any new connections whatsoever. The fix is really simple, just call fclose() before every closing statement on the function. This bug still persists on bahamut 1.8.9. Sincerely, Juan Baez.

2 1

the bug of Bahamut 1.8.9
by ShuXin Zheng 17 Aug '10

17 Aug '10

Hi, I got a core dump from Bahamut 1.89 ircd. Shown below: Program terminated with signal 11, Segmentation fault. #0 0x0000003b206782d0 in strlen () from /lib64/libc.so.6 (gdb) bt #0 0x0000003b206782d0 in strlen () from /lib64/libc.so.6 #1 0x0000003b20646729 in vfprintf () from /lib64/libc.so.6 #2 0x0000003b20663689 in vsprintf () from /lib64/libc.so.6 #3 0x0000000000451479 in sendto_channel_butone (one=0x2ad5837e2fb8, from=0x2ad5837e2fb8, chptr=<value optimized out>, pattern=0x46996d ":%s %s %s :%s") at send.c:529 #4 0x0000000000448590 in m_message (cptr=0x2ad5837e2fb8, sptr=0x2ad5837e2fb8, parc=<value optimized out>, parv=0x7cf8e0, notice=0) at s_user.c:1612 #5 0x000000000042bd2e in client_dopacket (cptr=0x2ad5837e2fb8, buffer=<value optimized out>, length=<value optimized out>) at packet.c:234 #6 0x0000000000437831 in do_client_queue (cptr=0x2ad5837e2fb8) at s_bsd.c:1409 #7 0x0000000000437ffa in readwrite_client (cptr=0x2ad5837e2fb8, isread=1, iswrite=667256208) at s_bsd.c:1691 #8 0x0000000000457e55 in engine_read_message (delay=4) at socketengine_epoll.c:167 #9 0x0000000000419db8 in main (argc=0, argv=<value optimized out>) at ircd.c:1150 -- zsxxsz

1 0

Re: ADMIN: Bahamut 1.8.7
by Aaron Wiebe 18 Jul '10

18 Jul '10

patch -p1 from the root of the tree, -p2 in src/ It could be that the cut/paste isnt working - in which case, just take a look at src/zlink.c and move the "#include <zlib.h>" down under memcount.h -Aaron On Sun, Jul 18, 2010 at 2:33 PM, Chris Malayter <mustangthz(a)gmail.com> wrote: > What's the syntax to apply that patch? > > If I apply it in the root directory of 1.8.9 it can't find zlib.h > > If I apply it in /src: > > %patch < patch > Hmm... Looks like a unified diff to me... > The text leading up to this was: > -------------------------- > |diff --git a/src/zlink.c b/src/zlink.c > |index f724125..7d46ff7 100644 > |--- a/src/zlink.c > |+++ b/src/zlink.c > -------------------------- > Patching file zlink.c using Plan A... > patch: **** malformed patch at line 6: #include <stdio.h> > > > -Chris > > > > > On Jul 18, 2010, at 1:24 PM, Aaron Wiebe wrote: > >> diff --git a/src/zlink.c b/src/zlink.c >> index f724125..7d46ff7 100644 >> --- a/src/zlink.c >> +++ b/src/zlink.c >> @@ -26,8 +26,8 @@ >> #include <stdio.h> >> #include <stdlib.h> >> #include <string.h> >> -#include <zlib.h> >> #include "memcount.h" >> +#include <zlib.h> >> >> #define COMPRESSION_LEVEL 3 /* 0 to 9, 0 = none */ >> #define ZIP_MIN_BLOCK 1024 /* smallest block to compress */ > >

1 0

Re: ADMIN: Bahamut 1.8.7
by Aaron Wiebe 18 Jul '10

18 Jul '10

Hey Chris, This is a known issue - I had forgotten to get the patch into the mainline before the release. It's specifically related to old versions of gcc. Apply the following patch - that should resolve it. -Aaron diff --git a/src/zlink.c b/src/zlink.c index f724125..7d46ff7 100644 --- a/src/zlink.c +++ b/src/zlink.c @@ -26,8 +26,8 @@ #include <stdio.h> #include <stdlib.h> #include <string.h> -#include <zlib.h> #include "memcount.h" +#include <zlib.h> #define COMPRESSION_LEVEL 3 /* 0 to 9, 0 = none */ #define ZIP_MIN_BLOCK 1024 /* smallest block to compress */

1 0

Bahamut 1.8.9 Released
by Aaron Wiebe 22 Jun '10

22 Jun '10

Greetings all, The Bahamut team is proud to announce the release of Bahamut 1.8.9. This release is a small patch release, consisting of several small bug fixes and updates. Most notably, this resolves an issue with SSL connections being dropped incorrectly under certain load situations. The updated software can be downloaded at: http://code.dal.net/release/bahamut-1.8.9.tar.gz md5: e1a2b6ffab3c7a88bdd0a3561da4e5ee sha1: 64b783ed04983bb1f5ea2b75b3a479888413fb3e Note, due to recent problems with Unreal IRCd's code security[1][2], we're now distributing all releases with appropriate sums. We also recommend grabbing the code tree via git, at http://code.dal.net/bahamut.git - this release is tagged v1.8.9-forreal. This release consists of the following changes: Kobi Shmueli (2): Made make install copy make-cert.cnf to INSTALL_DIR. Fixed a memleak bug on SVSHOST if used twice on the same user (reported by DiAvOl). Ned T. Crigler (1): Fix safe_ssl_read/safe_ssl_write to return -1 for EWOULDBLOCK/EAGAIN. epiphani (4): Fix regression in make install Fix badly declared bircmodule_free() in hooks header Update build_version for more common version string Significant update of autoconf/configure scripts Thank you all! -epiphani [1] - http://slashdot.org/story/10/06/13/0046256 [2] - Who uses Unreal anyway? Dirty, dirty codebase.

2 1

Modules and 64-bit Warnings/Errors
by Meekys 16 Jun '10

16 Jun '10

Hey, I've been working on a module for Bahamut, and I've noticed some discrepancies in the latest release (1.8.8) for the declarations of a few functions. Specifically the functions "bircmodule_malloc" and "bircmodule_free" in "include/hooks.h" and "src/modules.c" hooks.h extern int bircmodule_malloc(int); extern int bircmodule_free(void *); modules.c void *bircmodule_malloc(int); void bircmodule_free(void *); These signatures still work fine for 32-bit compilation, but can produce warnings and/or errors when compiling for a 64-bit environment. I believe the hooks.h just got missed and need to be updated to have the same signatures as in modules.c Hope this makes sense. Cheers, Sean (Meekys)

1 0

SSL - /links Suggestion
by PapaSmurf 31 May '10

31 May '10

Would it be possible to update the static /links output with servers that support SSL connections? It would save we helpers from having to run /trace commands to find what servers are using 1.8.7/1.8.8 and which are not. Maybe after all servers have the latest bahamut (with SSL support), you can leave it on the /links output. aqua.se.eu.dal.net (0) [Blubb Blubb] Maybe instead of the 0 hop/static output, (SSL) or (NONE)? Thanks, PapaSmurf

1 0

Some Comments about +S Channel Mode
by Coca Cola 30 May '10

30 May '10

I beleive that we need some improvements for this particular channel mode that does not exist in the latest version of bahamut: 1) +S should be available only for channel opers connected through SSL (In a case of a new channel with one single chanoper). 2) +S should be available for use ONLY if ALL users in the channel are connected through SSL. 3) Invites should NOT over-ride +S and if for any reason someone non-ssl joins, S mode should be removed by the server. 4) No exceptions/overrides for /samode. (Perhaps only ulines should override it but at as far as i know ircservices at least "enforce" the mode and kicks out non-ssl users). In my opinion, these checks should be done before the mode set to ensure that a +S channel is a channel that only SSL users are actually there. We have to be sure that when i'm on a +S channel, the channel is actually secure. Because even if only one user in the channel is not connected with SSL, it breaks the whole porpuse of this mode -> the conversation in the channel is transmitted somewhere (to a non-ssl user) unencrypted. Thanks :)

9 16