RELEASE TAG ircd-hybrid-7.2.2
bump version to 7.2.2
+ fixed a bug where auths/klines/glines with digit wildcards (#) were not handled correctly
- Updated messages/Makefile.in for ircd-brazilian_pt.lang
+ added DEAF isupport
- Rebuilt configure using autoconf 2.60
- Updated RELNOTES
- fixed core on MODRESTART as reported by xZ/CoolCold
- Also show the full path of .included configuration files when reporting syntax errors. This is just cosmetical.
- Updated RELNOTES
- Applied nenolod's parse_resvconf() fix
- Updated RELNOTES
- ACCEPT fixes as reported by ThaPrince
- Add Brazilian to 7.2 as well Marcelo Lipienski <und3rl1n3d@gmail.com>
- Fixed core in make_virthost() with host names that doesn't have a dot in it. Reported by CoolCold
- Fixed small memory leak in delete_auth()
- Backported r689 (the name of the source was appended to the RPL_TOPICTIME numeric instead of the target's name in mo_forcejoin)
- Updated RELNOTES
- Updated copyright years in version.c and INSTALL - Bumped version to 7.2.2 in configure.ac and doxygen.conf
- close_connection(): Fixed some harmless bug that could result in incorrect sent statistics in "STATS t" - Updated RELNOTES
+ patch against truncated NAMES output Submitted by: Murat Deligonul <deligonul@gmail.com>
- check_classes(): Only call destroy_cidr_class() on a class that is ready for deletion, otherwise the current cidr count would become unadjusted in case a disabled class gets re-enabled later on.
- Fixed possible core introduced by previous commit - Put some asserts into detach_conf()
- Fixed class limit bug as spotted by stu. ("REHASH" would always reset
current user count)
- Made "STATS y|Y" show inactive (deleted) classes that still have attached
clients
- balloc.(c|h): backported r544 (Killed Block::used_list) - Update RELNOTES
- Rebuilt y.tab.(c|h) using bison 2.3 - Bumped pcrelib version to 6.4. There were no changes since 6.3 that would have affected our hacked version of the pcre library, so this is mainly to satisfy paranoid I-always-need-the-latest-version people
Fix compile error. Be more careful with the so called stable branch :)
- Fixed compile error in s_auth.c
+ backported numeric changes
- Fixed bug that would exit unknown clients in check_unknowns_list() that are still processing auth/dns queries. Basically check_unknowns_list() should only exit clients that have finished auth/dns and didn't send a NICK/USER combination since they have been released from auth. Reported by ThaPrince.
- Fixed old bug that would forbid SPAMNUM to be disabled via "SET SPAMNUM 0"
- Updated example configuration files
+ added invisible character ^O (cancel all formatting)
- actually remove the chars that should not be marked as visible chan chars
- At least I noticed it myself, I got it completely backwards when folding in Michael's suggested optimisation.
- optimise Michael's fake channel catcher. Make it work first, then optimise.
- Added channel::disable_fake_channels which disallows creation of channels that have ascii 2, 3, 31 and 160 in their names. - Minor improvements and cleanups to channel name validation routines backported from 7.3
- Fixed compile warning in change_simple_umode as reported by Volk
- Fixed an off-by one bug in the ip limiting code as reported by Volk
- Unbroke KILLHOST
- Backported r376 (STATS y now shows the total number of clients in a class) - Rebuilt y.tab.c using bison 2.2
- attach_conf(): We even need to increment CurrUserCount if a class got attached to a server or operator. Spotted by stu. - Removed IRCD_SOCKET_ERROR define
- report_error() expects two %s conversion specifiers as reported by ThaPrince. This function really should get rewritten at given time. It just confuses people.
- Backported m_clearchan:kick_list fixes
- Fixed /dline nick for ipv6 servers
- ERR_NOSUCHNICK should actually be sent to local clients trying to PRIVMSG a non existant nick starting with a digit.
- Removed dalloca.h
- Added missing sigaction for SIGALRM
- Updated RELNOTES
- Cosmetical fix to "STATS k|K". temporary regexp based k-lines should be prefixed with 'kR' and permanent ones with 'KR'.
- Backported core fix for regexp k-lines
- Ignore signal SIGXFSZ which killed the ircd process on daxnet because of a file size limit set too low. That way, write() and friends will abort and errno will be set to EFBIG. We may think about a NOTICE which could be sent to operators on a failed write attempt to either the log files or k/x-line files.
- Backported RKLINE fix so the user and host portion of a banmask don't get cut off after 10 and 63 chars, respectively. A split_nuh() rewrite was required for this. - Removed now unused xstrldup() function
- Merged check_xline() and check_regexp_xline into a single function
- Clear the TS6 CAPAB bit possibly set on an unregistered client connection in mr_motd(), mr_admin() and register_local_user(). This is a kludge for now.
- m_restart(): don't send restart notices twice to locally connected clients
- Proper typecasting in newblock()
- Added new general::stats_e_disabled configuration option. Known from RB, all it does is to disable "STATS e", which is a good idea if you have any exempted server ips. - Updated RELNOTES
- Cleaned up example.conf - Copied example.conf over to example.efnet.conf and restored efnet specific values
- The 2nd and 3rd character of a SID may be a digit as well. Spotted by ThaPrince - Fixed compile warnings in log_user_exit()
- log_user_exit(): removed extranous newline character in log message. fixed invalid casts
- Update RELNOTES
- Fixed a bunch of compile warnings
- Updated INSTALL. --with-maxcon doesn't exist anymore
- ts6 fixes as explained on irc
- Backported changes made in HEAD to get rid of Channel::locmembers. This is mainly to save about 5megs of ram on networks like efnet where we have about 600k allocated Membership structures.
- Updated RELNOTES - Fixed mkpasswd help file
- Fixed core in m_mkpasswd.c as reported by Grigor Josifov
- Added missing bracket
- can_send: splitted up some bigass if() statement
- Allow servers to CHGHOST. This is mainly to fix an issue with ip_cloaking module
- Backported accept_connection() cleanup from HEAD
- Fixed negative "STATS ?" output as reported by ThaPrince - Updated RELNOTES
- Updated RELNOTED. Added missing entries etc.. BUGS, TODO, INSTALL: likewise - Fixed formatting in example.conf
- Don't send out multiple "requesting G-Line" notices in case the same oper is requesting the same gline over and over again.
- Forgot the half of the gline voting fix
- Fixed g-line voting as discussed on irc
- Backported add_connection() speedups
- Updated RELNOTES and example.efnet.conf
- added missing documentation for connect::vhost
- Fixed core on FORCEJOIN as reported by CoolCold.
** Fixed doesn't need to get forward-ported to HEAD
- Updated RELNOTES
bump version to 7.2.1 (release)
- Backported OJOIN fix from HEAD. Thanks for remembering me, CoolCold
- Can't use irccmp for qsort(), so let's use stricmp() when compiling on wins0x
+ fixed linking error on bcc/win32 about _strcasecmp
- Fixed compile error in m_operspy.c. Reported by Volk
- Added missing !u@h prefix when sending MODE message back to a client
- Updated RELNOTES
- Backported msg_tree_parse() speedup
- When checking for first " in a line, make sure it doesn't run off the end of the line before it finds one.
- Removed unused struct DNSQuery pointer from Listener structure
- Fixed static build for m_cap() (fix doesn't need to get forward-ported to HEAD) - Added BS alias for BOTSERV
- register_remote_user(): Fixed bug where we would increase Count.total and source_p->from->serv->dep_users for unregistered remote clients and don't decrease them if we exit the client due to non existand server or fake direction.
- Removed some outdated XXX comments
- channel_member_names(): properly count the number of prefixes that will be written to lbuf
- Backported CAP changes from HEAD since it doesn't affect any of the ircd's core components and should be supported as soon as possible.
- exit_client(): Don't bounce a QUIT back to origin's server Reported by adx
- Fixed bug where "OPERSPY NAMES #channel" would add a client twice onto a channel's member list if it is already on that channel.
- Break the loop if a client has reached its channel limit
- mo_force(): Don't allow servers to be added to a channel. (could possibly lead to a core)
- Rewrote m_join(). This was mainly done to fix some bug that can be easily abused. Details about this bug have been given on IRC.
- Fixed bug where "JOIN #keyedchan1,#keyedchan2 key1,key2" wouldn't work. Reported by ThaPrince
- Yet another mach_chan fix (broken #*chan vs. &chan match logic
as pointed out by adx)
- Don't give away IP information unless it's an admin doing /stats P
- Backported match_chan() fix to get /list *chan* to work
- Fixed RPL_WHOISACTUALLY as reported by CoolCold
- Cleaned up the resolver some more, added some comments
+ backported quit change
- Redo NXDOMAIN check so IPV6 still works
- Unbreak can_send
- Get rid of can_send_part. This also fixes a bug where voiced clients weren't able to send a PART message to the channel. Reported by ThaPrince.
- Properly close ssl clients
- Backported spy_trace_notice module unload fix
- reported by Jon, resolver was not checking for NXDOMAIN as well as SERVFAIL this caused odd DNS reports and took longer to connect. Tested by Jon (ThaPrince). (backported from head)
- rebuilt y.tab.c
- Fixed bug where "listen { flags = ssl; port 1,2,3; }" would result
in only port 1 being set ssl enabled.
- Fixed typo in example.conf and send.c
- Got rid of the last IVARIANT we had and replaced it with two asserts().
- Fixed bug in fd_close() that caused a core when calling restart_resolver()
- Fixed invalid output of the USERS command as reported by ThaPrince
- Don't try to keep pointers to already freed memory on temporary_resv list when manually removing tempoary N/C resvs.
- Fixed possible core in expire_temp_lines()
- Fixed same problem for SSL_write. Some interesting reference about the problem we experienced can be found at http://www.mail-archive.com/openssl-dev@openssl.org/msg13939.html
- Fixed bug where all ssl clients would get disconnected in certain situations. (reported by many people)
- Fixed compile errors
- Add BMASK and TMODE onto message tree when building with static modules. reported by hb505.
- Backported updated help file for "SET"
+ fixed unaligned memory reference in resolver Submitted by: Immo 'FaUl' Wehrenberg
- backported spelling fix
MFC r396 from ircd-hybrid 7.3 (s_auth.c reliability fixes)
- Fixed typo
- Fixed TBURST as reported by ThaPrince
- Backported max_chans_per_user fix from HEAD
- fixed compile error in sendnick_TS()
- Fixed typo in example configuration files
- Forward-port banfix from 7.2 (reported by ThaPrince)
+ this should fix sockhost relaying (ThaPrince #1)
- Soft caller ID fix as per ThaPrince (find_channel_link was called with wrong pointer)
- Fixed incorrect match handling in relay_xline as reported by ThaPrince. Damnit match() really shouldn't return 1 on match. This will be changed sooner or later.
- Imported recent channelban changes from HEAD which have been tested extensively. The match_cidr() fix didn't work in the first try, so I'm not going to bother with further debugging. - Removed match_cidr() and comp_with_mask()
- Forgot the half of the fix
- Fixed compile error on srv4 as reported by KEYofR
+ MFC support for halfop->op rewriting.
- Don't allow shared k-lines to be set without a valid reason
- Flag the client as being ip-spoofed if setting a new hostname via CHGHOST, that way non-opers can't figure out the IP with a simple WHOIS. Suggested by adx
- Fixed compile error reported by Volk and CoolCold
- As reported by CoolCold: - Updated cmode helpfile - Added support for old-style topicburst = yes|no
- Updated RELNOTES
- Someone fixed a core in ircd_parser.y but forgot to rebuild y.tab.c,
so ircd still kept on coring if using 'umodes' option in operator{}
- Don't allow other servers to send less than 4 parameters within TB
- Improved topic bursting. Read the comment in s_serv.c:send_tb for further info
- This should finally fix the TBURST issue as reported by ThaPrince. Untested as always
- Backported TBURST fix from HEAD
- Somehow we had a 7.2 release without mentioning the topicburst directive
in connect{} blocks -> updated example confs
- Updated RELNOTES
- Finish off TBURST (untested)
- propset svn:keywords "Id" modules/m_tburst.c - Added CAP_TBURST definition to get m_tburst.c compile
- Removed m_tb.c and updated Makefile.in
- Imported m_tburst.c from 7.3
- Updated RELNOTES
- Only locally connected clients should be checked for a valid channelmembership when processing the INVITE message. Remote clients that are outside a channel e.g. services, are allowed to send INVITEs. This also fixes a bug with anope services. Reported by r0d3nt
- Updated RELNOTES
- Fixed core in case aconf->cipher_preference and ConfigFileEntry.default_cipher_preference are both NULL pointers Reported by opteron-hani via Joan
- Updated RELNOTES
- Backported revision 315 (trimmed memory usage by moving invited dlink_list to LocalUser struct)
- Don't pace "WHO #somechan" and "WHO *" as this breaks several clients
+ clean up *.tds files after compile
+ fixed bugs related to win32 & static builds
- bump patchlevel to 7.2.1beta - stub 7.2.1beta1 in RELNOTES and add BETA verbage
- bump patchlevel to 7.2.0 - update RELNOTES - bump version in specfile
- fixed infinite loop spotted and fixed by *runedude* Jeremy Sayres and my email is runedude@runedude.net
- This is part of the nasty core fix; the y.tab.c and y.tab.h files were committed earlier, I have no idea why this file did not get committed.
- This BUG is gone so the ancillary files are no longer needed as well
+ fixed unportable usage of va_list that make ircd cry on amd64
+ MFC /dev/poll
- Back port whois fix from HEAD
+ more rate limiting fixes + update RELNOTES
+ add more load2hi protection + style fixes
MFC set +b on opers by default
- Fixed another core on dlclose()
- Fixed core on /modunload, /modrestart
- Fixed problem where configure wouldn't deal properly with openssl 0.9.8 - Rebuilt y.tab.c using bison 2.1
+ 7.1 -> 7.2
- Added missing 'return' after register_remote_user() call in nick_from_server()
- Use MODE_PRIVATE, MODE_PARANOID isn't defined until HEAD (7.3)
- make findmodule_byname case sensitive
+ reverted accidental invalid backport of win32build.bat
MFC:
+ removed invite_ops_only, controlled by +p now
+ sorted ircd.conf:channel{} a bit
- Fixed compile error in load_a_module() with glibc that don't have a Link_map typedef
- backported fix from HEAD done by Michael - Fixed bug in sendto_channel_remote() that would cause multiple copies of the same message to be sent to the uplinked server.
- Fixed gcc4 related compilewarnings in inet_ntop6(). Reported by ThaPrince
- ETRACE should spit out an error if used by a non-oper. Reported by nenolod
- Removed LocalUser.port - Removed two extra memsets in make_request()
- Fixed compilewarning about unused parc in a more or less clean way. Reported by Marines
Backported from HEAD - committed configure changes by nenolod (William Pitcock) to check for HAVE_DLINFO - fixed logic for picking up address for the mod_list
- MFC iorecv/iosend changes to allow charset recoding
- Fixed bug where we would increase Count.invisi even if invisible_on_connect is off
- MFC: fixed register_callback
(backported from HEAD) - irc_res.c would send out hundreds? of retry requests on clients that did not resolve, i.e. SERVFAIL. The code was essentailly ignoring SERVFAIL hence generating request after request after request. This appears to fix this major bug. It would be a good idea to add some logging here however.
- backported from HEAD flush_deleted_I_P is trying to treat MatchItems as if they are AccessItems I am almost sure this function is redundant as well.
- backport fix from HEAD to fix m_testline.c on 7.2 This bug is old. The code was missing a test for a kline conf and the sendto to print that kline out, in the case where it was a kline against and IP. i.e. foo@ip.ip.ip.ip it would work in the case where it was a kline against a dns.
- backport fix for memory leak and parv1 from HEAD
- Limit maximum WHOWAS replies for remote requests to 20, as per H6
- Fixed a bug where remoteservers would send us an ERROR on "WHOWAS nick count remoteserver.net". Propably one of the oldest bugs that got introduced in 7.0. (Backported from HEAD)
- fix compile warning
- backport fixes from 7.3 (HEAD) for testline and ~ flags
- Updated example.conf as suggested by pi
- Fixed compile error in m_ctrace.c
- MFC: fixed core on /links
- That one should be the propper fix for GDENY items. Previous fix broke the ability of stacking user/name/action combinations. That fix still needs to go into -HEAD, and VERY important, into 7.1.3. Someone please test this, e.g. write a small dump function that spits out GDENY items.
- CHANNELLEN vs. LOCAL_CHANNELLEN fix. Local JOINs are limited to 50chars whereas remote SJOINs/JOINs can be upto 200chars for backwards compatibility.
- And update the parser files
- At the very least, don't create a gdeny conf with a NULL name, this is bad manners.
- Updated 7.2 RELNOTES
- Fixed bug in m_whois() that would allow users to send a WHOIS to remote servers with an empty argument via "WHOIS someserver.org :"
- added /stats h to help
- Reverted oper_privs_as_string() back to its previous version which is easier to maintain when adding new operflags. Even the one who wrote the new version introduced a bug by not taking care to keep the oper_flags_table in sync with the OPER_FLAGS_* bits. (which also no one really wants to do ...) - Updated it to deal with new 'remoteban', 'operwall' and 'operspy' flags
- Re-added an equivalent of H6's sixth O-line field (Modes to get on operup). - Updated efnet's example.conf to reflect the new opers_bypass_callerid directive - Misc style corrections to s_user.c while on it
MFC: switch to static "SVN" string
- *hopefully* this hack fixes cryptlinks+ts6 instead of breaking them ;)
- documented cidr_bitlen
- MFC disallow ':' at the beginning and ' ' in server passwords
create 7.2 branch, we can move/rename it as needed.
- svn:keywords
- svn:keywords
- Fix svn:keywords
- imported sources - can be moved later according to the directory/branching scheme, but we need the svn up
x
test 12312431241
blah
xxx
zzz
test
another
test 123
this is a test
yet another test commit
yet another test commit
test commit. will it email?
Deleted ircd-hybrid/trunk
Deleted ircd-hybrid/tags
Deleted ircd-hybrid/branches
- test commit
Initial import