summaryrefslogtreecommitdiff
path: root/src/sockets.cpp
AgeCommit message (Collapse)Author
2013-09-19Make os_session akill ip ranges and not individual user's ipsAdam
2013-07-01Change extensible keys to require explicitly having a type defined for it. ↵Adam
Completely modularize more features like bs_kick, entrymsg, log, mode, etc. Move fantasy to its own module. Move greet to its own module.
2013-06-15Fix cidr::match with odd cidr rangesAdam
2013-05-08Make sockaddrs/cidr not throw on invalid ips to give us an easier/cheaper ↵Adam
way to test for a valid IP
2013-03-13Refactor mask/entry code, allow full matching (against users real host/ip) ↵Adam
if their displayed host is their real real host. Also match against cloaked host even if full matching is not being done
2013-02-25Don't enforce session limit on clients with no IP on Unreal, fix typo in /cs ↵Adam
down syntax, fix os_session messages to reference ip
2013-01-21Merge usefulness of Flags and Extensible classes into Extensible, made most ↵Adam
flags we have juse strings instead of defines/enums
2013-01-09Update copyright to 2013. This was done with:Adam
find include/ src/ lang/ docs/ modules/ *.* Config -exec sed -i 's/-2012 Anope Team/-2013 Anope Team/i' {} \;
2012-12-13Optimize much of the database code and serialize code.Adam
2012-11-22Pretty large coding style cleanup, in source docAdam
cleanup, and allow protocol mods to depend on each other
2012-11-06Use std::tr1::unordered_map for a few of the larger mapsAdam
2012-11-01Fixed quite a bit of dumbness with m_ssl. Had toAdam
modify socketengines to allow polling for write & no read, but is it cleaner now. Made m_httpd able to listen using SSL.
2012-10-24BIND's forward ability did not work as I expectedAdam
because it will not forward non recursive queries. So, added support for SOA, NS, and AXFR requests.
2012-09-07Made session tracking ip based, not host based, and allow using CIDR to ↵Adam
group multiple ips from one subnet to one session
2012-09-02Fix WindowsAdam
2012-02-14Clean up and reorganize our header filesAdam
2012-01-10Removed this "valid" ip check in cidr::cidr, is wrong for IPv6 and ::pton ↵Adam
checks this anyway using inet_pton. Also fixed a comment Robby broke in chanserv.conf
2011-11-08WindowsAdam
2011-10-24Fixed bug #1349 (m_sqlite compiles without error under FreeBSD), as well as ↵Naram Qashat
use C99's stdint.h (or cstdint if available) to get (u)intX_t types instead of our stupid typedefs. pstdint.h included in case there is no cstdint or stdint.h available.
2011-10-09Give more verbose messages on startupAdam
2011-09-10Cleaned up some of the socket code, cleaned up the pipe engines, added ↵Adam
support for binary sockets, and cleaned up the asynch connect/accept code
2011-08-14Revert "Fixed resolving hosts on connect"Adam
This reverts commit 11619be8b97d7d8c28bc1ebbf6d412dff298b23d.
2011-08-14Fixed resolving hosts on connectDukePyrolator
2011-08-10Fixed windows buildAdam
2011-07-31Fixed error message from being unable to connectAdam
2011-07-31Fixed a few small problems, including m_ssl's connect feature sometimes ↵Adam
failing for no good reason
2011-06-20Cleaned up some of the logger code which fixes not logging debug logs to ↵Adam
files etc when debug is enabled, and some other small things
2011-05-23Rewrote the signal handling to use sigactionAdam
2011-05-21Made Anope able to process normally when disconnected from the uplink and ↵Adam
not sleep(), enable usage of non-blocking connect() and default all sockets to non blocking mode. Some cleanup to m_ssl and some cleanup to main.cpp.
2011-05-16Expand more on m_alias and changed some std::string usage in sockets.cpp to ↵Adam
use Anope::string
2011-05-16Unmodularized the socket engine because its causing problems and really is ↵Adam
unnecessary
2011-04-29Temporary fix for /os restart with m_async_commands loaded, is fixed ↵Adam
properly in 1.9.5
2011-04-25Don't attempt to connect to the uplink if given invalid hostnamesAdam
2011-01-08Made the Flag class able to convert flags to strings and backAdam
2010-12-23Do not clear want write on a socket if we are unable to completely write its ↵Adam
buffer
2010-12-12Handle not being able to completely flush our write buffer correctlyAdam
2010-12-12Added cs_mode, rewrote the old list mode code, and added CIDR supportAdam
2010-10-30Made gettext work on most OSs. Tested on Debian, FreeBSD, Gentoo, and Windows.Adam
Added a search path option to the Config script for cmake to use when finding libraries for modules or for gettext. Fixed m_mysql and m_ssl to work under Windows, made the Windows Config program remember the last used options, and fixed Windows release builds.
2010-10-01Rewrote some of the socket code to allow m_ssl to be a service.Adam
This allows modules (xmlrpc) to create and accept SSL connections. Also fixed unloading m_mysql at certain times and made the threading engine always work correctly on Windows.
2010-09-17Rejig of some of the socket stuff. Fixed marking sockets as nonblocking on ↵Adam
Windows. Added in a LastError function to keep having to use strerror/GetLastError everywhere.
2010-09-17Connect using nonblocking socketsAdam
2010-09-10Added Anope::CurTime to keep us from calling time() everywhereAdam
2010-09-09Added an asynchronous DNS system and m_dnsbl, which checks clients against ↵Adam
DNS blacklists. Rewrote internal handling of IPs, we now properly support users using IPv6. Fixed a few problems with the UnrealIRCd protocol module.
2010-08-27Added a new logging systemAdam
2010-08-15Rewrote the MySQL API to use threads. This acts similar to before, but is ↵Adam
faster. Removed db_mysql_execute for now.
2010-07-30Fixed some Windows issues with hashing and sockets. This currently limits ↵Adam
building to VS 2010, will look at 2008 later
2010-07-26Trying to make things a little more const-safe, a work in progress but this ↵Naram Qashat
is a bit better.
2010-07-25Epic commit to replace most of the strings in Anope with a single ↵Naram Qashat
Anope::string class, plus some other little fixes here and there. If you follow 1.9.x development and are testing things, THIS is one of those things that NEEDS testing.
2010-07-24Always use non-blocking socketsAdam
2010-07-24Properly handle single messages received that have no newlineAdam