Commit Graph

1581 Commits

Author SHA1 Message Date
Jilles Tjoelker 70fd7fc9fb BAN: Reject bans with insufficient non-wildcard characters.
Such bans are not applied locally, but are propagated normally.
They can only be removed on a server that applies them.

Note that normally KLINE will not accept such bans.
This is mainly for services, differing min_wildcard and
ircd changes.
2010-03-06 01:45:41 +01:00
Jilles Tjoelker 20eef93004 Add /stats g to show propagated (global) klines. 2010-03-06 00:31:51 +01:00
Jilles Tjoelker c177d078ea Among bans with the same creation time, prefer the one with longest lifetime. 2010-03-05 23:16:28 +01:00
Jilles Tjoelker 05114b16a3 BAN: Avoid fake direction. 2010-03-05 22:53:35 +01:00
Jilles Tjoelker e49d818579 Remove +/- from the BAN message, instead indicating unban with duration=0.
A kline must now last at least one second since its creation time.

Also add better logic for bans that have already expired
when they come in.
2010-03-05 22:51:47 +01:00
Jilles Tjoelker a4893a45be Use memmove instead of memcpy where there is overlap (modunload). 2010-03-05 22:05:15 +01:00
Jilles Tjoelker 65b8e0029e Add propagated klines.
A KLINE command without the ON clause now sets a propagated
("global") ban. KLINE commands with the ON clause work as
before.

Propagated klines can only be removed with an UNKLINE command
without the ON clause, and this removes them everywhere.
In fact, they remain in a deactivated state until the latest
expiry ever used for the mask has passed.

Propagated klines are part of the netburst using a new BAN
message and capab. If such a burst has an effect, both the
server name and the original oper are shown in the server
notice.

No checks whatsoever are done on bursted klines at this time.

The system should be extended to XLINE and RESV later.

There is currently no way to list propagated klines,
but TESTLINE works normally.
2010-03-05 18:36:44 +01:00
Jilles Tjoelker 90072e8be6 kline: Fix oper reasons. 2010-03-05 18:28:59 +01:00
Jilles Tjoelker dce31e2723 Fix --disable-balloc.
libratbox r26769
2010-03-05 17:28:47 +01:00
Jilles Tjoelker a491246f8f Skip propagated bans in rehash txlines/tresvs. 2010-03-05 00:37:56 +01:00
Jilles Tjoelker 12894c8888 Add code to expire "propagated" bans.
The data structure is very simple: a dlink list of all propagated bans.
2010-03-04 00:21:22 +01:00
Jilles Tjoelker ebae0e200e ConfItem.next is not used, remove it. 2010-03-03 23:36:28 +01:00
Jilles Tjoelker 8a1d7222f4 bandb: Group writes in transactions to reduce load on heavy kline activity.
When a ban is added or removed, open a transaction and
close it after 3 seconds; any bans in the 3 seconds
will not trigger another transaction (= 4 fsyncs with
sqlite).

The transaction is also committed if ircd closes the
connection, but not if bandb itself is terminated with
a signal.
2010-03-03 22:58:16 +01:00
Jilles Tjoelker cf06f92b78 Make struct operhash_entry private. 2010-03-01 01:24:00 +01:00
Jilles Tjoelker a0f4c418d3 Track who set a dline/kline/xline/resv as in ratbox3.
Like in ratbox3, there is no way to query this information
(other than bandb's tables, but they worked before this
commit).
2010-03-01 01:23:22 +01:00
Jilles Tjoelker 157c1f04bd Generate the "Temporary K-line %d min" part from aconf->hold - aconf->created. 2010-02-28 16:27:06 +01:00
Jilles Tjoelker ce60772db6 Store the creation time of klines and dlines as a time_t instead of as text.
The value 0 indicates the creation time is unknown (currently the case
for bandb).
Also store a creation time for xlines and resvs, but do not use it yet.
2010-02-28 00:46:56 +01:00
Jilles Tjoelker bebc68755d Add a missing comment for new /challenge numerics. 2010-02-27 23:56:56 +01:00
Alexander F?r?y 360742bdf4 Use RPL_QUIETLIST and RPL_ENDOFQUIETLIST instead of RPL_BANLIST and
RPL_ENDOFBANLIST for channel mode +q.
2010-02-27 01:08:41 +01:00
JD Horelick 88b89a4073 Update TODO a bit. 2010-03-26 00:52:42 -04:00
B.Greenham b791dea807 Some reference.conf cleanups. 2010-03-25 23:10:52 -04:00
JD Horelick 603c6d3c5a Add a small FAQ and modify the todo item about disabled_commands in
auth {} to disable_oper since that covers ~99% of the use-cases and
can be much cleaner.
2010-03-25 14:14:17 -04:00
JD Horelick 64dd2404ca Rewrite message when you don't have the privs to kick a user
(be it that the user is +a and you're +o/+h or you're not an op at all).
2010-03-25 12:31:12 -04:00
JD Horelick fb81a73307 Update NEWS. 2010-03-25 03:39:46 -04:00
JD Horelick 3232087a3c Make all snotes for mod* stuff netwide since unless you're opered
on more than one server, you won't know if the command worked or not.
2010-03-25 03:18:32 -04:00
JD Horelick 576bcf93e5 Add information to helpfiles about remote module loading. 2010-03-25 03:13:13 -04:00
JD Horelick e31b740dd1 Allow using all the mod* commands remotely. 2010-03-25 00:33:20 -04:00
JD Horelick 069f104a2e Add shared::modules option, remote module stuff coming later/soon. 2010-03-24 15:50:15 -04:00
JD Horelick d25c6eb1b2 Yay for documenting code.... 2010-03-23 21:20:07 -04:00
JD Horelick 4aa4fa57f2 I am made of fail. 2010-03-23 15:21:42 -04:00
JD Horelick 5ff039e923 Make remote restart and die actually...work... 2010-03-23 15:21:09 -04:00
JD Horelick 658a333d65 Pull in remote restart and die (from ircd-seven) allow opers
to remotely restart and die servers.
2010-03-23 14:21:31 -04:00
JD Horelick 74e3e4f1e2 Add DIE shared option for remote restart and die which are on their
way next.
2010-03-23 14:19:48 -04:00
JD Horelick 6f7a78f6aa Added check for flood settings in m_message. Why was this not here
before? What good are flood settings really if they're not applied
to messages?
2010-03-21 03:43:25 -04:00
JD Horelick 706f2ddf1a Add the note that no_oper_flood needs to be enabled for true_no_oper_flood
to work correctly.
2010-03-21 01:06:57 -04:00
JD Horelick 1cc810d3a7 Add general::true_no_oper_flood . I'm not going to explain it here.
Ask me what it does on IRC if you're curious.
2010-03-21 00:14:02 -04:00
B.Greenham aa9a8067b1 Use strip_unprintable for chmodes +K and +G, which makes evasion less trivial. 2010-03-20 12:49:57 -04:00
B.Greenham 114406d021 Add strip_unprintable, a function that strips all unprintable characters from a string. 2010-03-20 12:39:39 -04:00
JD Horelick be6960363c Bump the version in configure since we're now in development of 6.1.0.
We've been in 6.1.0 development for like a week now but i forgot to change this a week ago.
Better late than never...
2010-03-20 05:30:37 -04:00
JD Horelick bc9ccf4a0e Update NEWS. 2010-03-20 04:48:18 -04:00
JD Horelick 82b371736d Add showing modes to OLIST so you can more easily see if you're
about to join a +s channel and it can be useful as a oper to easily
see which channels have certain modes set (i'm thinking +PL and maybe more).
2010-03-20 04:34:44 -04:00
B.Greenham baca17aeb3 Some minor conf cleanups. 2010-03-18 13:50:47 -04:00
B.Greenham 2c0c690444 Fix some compile warnings. 2010-03-18 12:32:42 -04:00
JD Horelick 29b3f2e4e5 Fix a small typo that i noticed looking around the source. 2010-03-18 02:44:51 -04:00
Jilles Tjoelker cc7e9872be Fix a signedness comparison warning. 2010-03-16 19:13:09 +01:00
JD Horelick b705a46f43 Fix a signedness comparison warning. 2010-03-17 03:42:40 -04:00
B.Greenham 473d0842db Fix some compiler warnings that I generated by accidentally copying something twice. 2010-03-17 02:19:18 -04:00
B.Greenham 7469e4b0d2 Remove the ability of overriding opers to privmsg/notice @#channel while lacking ops.
Since this goes past +q/+m and so on, it could easily and silently be abused.
2010-03-16 13:47:03 -04:00
B.Greenham 7bd3701cd6 Remove override being able to automatically and silently go through +q/+m/etc.
This was previously doable without sending wallops, and had high abuse potential.
Overriding opers can set themselves +v/+o or remove the offending mode if they need to get past such things.
2010-03-16 13:44:44 -04:00
JD Horelick 853058531a Added m_cycle and added it to all the appropriate locations. 2010-03-15 15:35:32 -04:00