summaryrefslogtreecommitdiff
path: root/usr.sbin
AgeCommit message (Collapse)Author
2009-05-11This commit adds a new feature to sysmerge(8), usable in '-a' mode.Antoine Jacoutot
From now on, checksums of reference files (sets and/or src) will be stored under /var/db/sysmerge/{etcsum,xetcsum,srcsum} . This allows for "remembering last choice" several people have requested. This deprecates the need for -X and -S which have been removed (these switches may come back at some point for the auto-patch feature which halex@ proposed). It will only compare files which reference sources have changed since the last run and will attempt to automatically upgrade them to the newest version provided that they have no local changes. i.e. first run should dislay differences, second run should not display anything except if something changed between old and new sets/src. This way sysmerge can also warn you from files which may have been obsoleted. General idea taken from Net/FreeBSD but we're doing things differently. Intensively discussed with oga@, sthen@ and halex@ ok oga@ sthen@
2009-05-11remove an extraneous slash from the requested URL; some strict proxiesStuart Henderson
(and maybe http servers) have a problem with it. ok espie@
2009-05-11Make bgpctl show network print both AF_INET and AF_INET6 as mentioned in theClaudio Jeker
man page. Problem found by Rod Whitworth. OK sthen@
2009-05-10use both read and write timeout arguments in the bufferevent_settimeoutJacek Masiulaniec
call; ok gilles@
2009-05-10use rfc number 4954 in auth_plain handler name, rfc 4616 is impliedJacek Masiulaniec
because rfc 4954 references it.
2009-05-10assert copyright; ok gilles@Jacek Masiulaniec
2009-05-10RFC 4954 says the response to "AUTH PLAIN" must be "334 ", not "334";Jacek Masiulaniec
ok gilles@
2009-05-09correct AUTH PLAIN rfc number, no functional change.Jacek Masiulaniec
2009-05-09fix rare memleak; ok gilles@Jacek Masiulaniec
2009-05-09give few states more meaningful names; ok gilles@Jacek Masiulaniec
2009-05-09fix function names in log_debugJacek Masiulaniec
2009-05-09- New API to handle all DNS query types (A, MX, PTR) asynchronously.Jacek Masiulaniec
- Improve RFC compliance: CNAMEs are resolved, equal preference MXs are randomized, relaying via MX that has equal/lower preference than local server is prevented, decision on when to treat domain name as implicit MX is better. ok gilles@
2009-05-09zap trailing whitespace;Jason McIntyre
2009-05-09tell people about -F repair. ok phessler, espieStuart Henderson
2009-05-05fix a couple of comment typosStuart Henderson
2009-05-05check that an IMSG_CTL_SHOW_RIB_PREFIX request has a validStuart Henderson
address family before passing it to the RDE. ok henning@ claudio@
2009-05-01initialize secret to NULL otherwise we may try to free a junk pointer whenGilles Chehade
F_AUTH is not set. i was going to look into the issue, but i have received a diff from Simon Betrang <janus@errornet.de>, thanks !
2009-04-29de-static some functions. Helps handling the monster change I'm working on.Claudio Jeker
2009-04-29+.Xr ndp 8Jason McIntyre
from Bob Van Zant, freebsd docs/134053
2009-04-28smtp auth happened to work by luck because delivery to mbox action was theGilles Chehade
first action of an enum and we allocated a struct using calloc, but we did not properly initialize the action for authenticated users. while at it, change the name of path action flags so that we know at first sight that they are path related. this diff fixes the immediate issue but still needs work. ok jacekm@, "we'll work out a long term solution"
2009-04-28fix a bug where client with unresolvable revdns is never greeted; ok gilles@Jacek Masiulaniec
2009-04-28don't use strlcpy before calling a function that uses strlcpyJacek Masiulaniec
internally anyway; ok gilles@
2009-04-28in mta, change the batch status to S_BATCH_TEMPFAILURE by default so thatGilles Chehade
we need to explicitely set success and permanent failures, this way if we hit any condition that we don't explicitely flag as permanent failure the delivery will be rescheduled later. while at it, made various checks more strict to prevent hiding bugs and a bit of knr cleanups discussed with and "reads good" jacekm@
2009-04-28make S_DATAREQUEST state similar to S_{MAIL,RCPT}REQUEST; ok gilles@Jacek Masiulaniec
2009-04-28let mta, mda, and mfa become idle - no functional change; ok gilles@Jacek Masiulaniec
2009-04-28Add the possibility to go back from merging mode to the previous menu.Antoine Jacoutot
Move unset NO_INSTALLED where it belongs, at the start of the function. Several minor cleanups. from Alexander Hall ok sthen@
2009-04-27kill 2-line session_msg_submit, used just once; ok gilles@ long ago.Jacek Masiulaniec
2009-04-27fix unlikely fd leak; ok gilles@Jacek Masiulaniec
2009-04-27session_cleanup call in MAIL FROM handler may be removed as aJacek Masiulaniec
consequence of rev. 1.70. This makes session_destroy the only caller so merge the two; ok gilles@
2009-04-27check message_id received from smtp uses valid format.Jacek Masiulaniec
2009-04-27Disable EV_READ when setting F_QUIT flag up. Malicious client mayJacek Masiulaniec
send more commands after QUIT, these must not be processed; ok gilles@
2009-04-26sync with changes in relayd, diff by reyk@ and okayd by meGilles Chehade
2009-04-26Unused defines from a time long long ago.Claudio Jeker
2009-04-26typos fixedTobias Stoeckmann
ok jmc
2009-04-26comment typos; no binary changeStuart Henderson
2009-04-24oopsMarc Espie
2009-04-24in session_pickup, do not handle S_RCPT state, it must never beJacek Masiulaniec
called in this state; ok gilles@
2009-04-24Allow UDP and/or TCP redirections instead of just TCP.Reyk Floeter
Thanks to Marek Grzybowski for feedback and testing. ok jmc@ (manpage bits)
2009-04-24don't truncate http headers when no separator is given between key andPierre-Yves Ritschard
value. ok & ``makes sense'' reyk@
2009-04-24Make aliases case-insensitive, issue reported by Daniel Ouelett; ok gilles@Jacek Masiulaniec
2009-04-24Enclose common imsg handling code in a function, which additionallyJacek Masiulaniec
does some sanity checking. Fix a bug that could lead to fatal under rare circumstances, exposed by this newly added check; ok gilles@
2009-04-24use same timeout at each session state, and make use ofJacek Masiulaniec
bufferevent_settimeout instead of rolling our own; ok gilles@
2009-04-24kill unnecessary struct message_recipient; ok gilles@Jacek Masiulaniec
2009-04-23Rework the way we handle announced networks. Instead of two freak rde_peersClaudio Jeker
use one that is less freaky. Merge bgpctl and config networks into one tree. First step of a larger change in the RDE and this goes now in to allow to move forward.
2009-04-23allow bgpctl and bgpd.conf to contain 32-bit ASN written in ASPLAINStuart Henderson
format (RFC5396). ok claudio@ henning@
2009-04-23sort "show stats" outputJacek Masiulaniec
2009-04-22fix creating packages with long names in plist; from espie@Christian Weisgerber
2009-04-22ignore replies with timestamps after 2030 to prevent time_t / tv_sec wrapsHenning Brauer
input & ok theo
2009-04-21small corrections to log_warns and fatalsJacek Masiulaniec
2009-04-21Make /usr/sbin/sendmail not fail due to smtpd being down.Jacek Masiulaniec
The approach is to save cmdline + stdin in a file under a newly added directory /var/spool/smtpd/offline (uid 0 gid 0 mode 1777). Next time daemon starts, it uses information in that directory to replay sendmail on user's behalf. ok gilles@