MFC r311471:

Conditionalize wrap(3) use based on MK_TCP_WRAPPERS instead of
always building support into sendmail.

MFC r291348:

Use LIBEXECDIR for /usr/libexec.

MFC: Temporarily disable WARNS while addressing a non-issue with the upstream code

Approved by: re (gjb)

By default, sendmail 8.15 uses uncompressed IPv6 addresses. Keep
current FreeBSD 10 and earlier behavior of using compressed IPv6
addresses in configuration, maps, rulesets, etc. (FreeBSD 11 and
later will use the new default of uncompressed IPv6 addresses.)

Approved by: re (gjb)

Copy head (r256279) to stable/10 as part of the 10.0-RELEASE cycle.

Approved by: re (implicit)
Sponsored by: The FreeBSD Foundation

Since clang 3.2 now has an option to suppress warnings about implicitly
promoted K&R parameters, remove the workarounds added for sendmail
components in r228558.

MFC after: 1 week

Define several extra macros in bsd.sys.mk and sys/conf/kern.pre.mk, to
get rid of testing explicitly for clang (using ${CC:T:Mclang}) in
individual Makefiles.

Instead, use the following extra macros, for use with clang:
- NO_WERROR.clang (disables -Werror)
- NO_WCAST_ALIGN.clang (disables -Wcast-align)
- NO_WFORMAT.clang (disables -Wformat and friends)
- CLANG_NO_IAS (disables integrated assembler)
- CLANG_OPT_SMALL (adds flags for extra small size optimizations)

As a side effect, this enables setting CC/CXX/CPP in src.conf instead of
make.conf! For clang, use the following:

CC=clang
CXX=clang++
CPP=clang-cpp

MFC after: 2 weeks

Unfortunately, clang gives warnings about sendmail code that cannot be
turned off yet. Since this is contrib code, and we don't really care
about the warnings, just turn make them non-fatal for now.

MFC after: 1 week

Lower WARNS for sendmail.

Even though it builds with WARNS=2, some users link sendmail from the
base system against SASL. This doesn't build in this case.

Reported by: Andrzej Tobola <ato iem pw edu pl>

The last big commit: let usr.sbin/ use WARNS=6 by default.

This FFR is no longer needed in sendmail 8.14

Catch up with share/mk/bsd.sys.mk rev. 1.38.

Reimplementation of world/kernel build options. For details, see:

http://lists.freebsd.org/pipermail/freebsd-current/2006-March/061725.html

The src.conf(5) manpage is to follow in a few days.

Brought to you by: imp, jhb, kris, phk, ru (all bugs are mine)

Make links for hoststat(8) and purgestat(8) man pages.

PR: docs/85009
MFC after: 3 days

Remove kludges intended to support src trees with partial obj trees.

Discussed with: ru

NOINET6 -> NO_INET6

NOCRYPT -> NO_CRYPT

Enable _FFR_DEAL_WITH_ERROR_SSL to deal with OpenSSL error returns of 0.

PR: bin/73105
MFC after: 2 days

Add knob NO_NIS (fka NO_YP_LIBC) and make world compileable when set.
If turned on no NIS support and related programs will be built.

Lost parts rediscovered by: Danny Braniss <danny at cs.huji.ac.il>
PR: bin/68303
No objections: des, gshapiro, nectar
Reviewed by: ru
Approved by: rwatson (mentor)
MFC after: 2 weeks

Join the 21st century: Cryptography is no longer an optional component
of releases. The -DNOCRYPT build option still exists for anyone who
really wants to build non-cryptographic binaries, but the "crypto"
release distribution is now part of "base", and anyone installing from a
release will get cryptographic binaries.

Approved by: re (scottl), markm
Discussed on: freebsd-current, in late April 2004

Reflect changes in sendmail 8.13 source tree

Honor NOINET6 and disable IPv6 support in libmilter and sendmail if it
is set.

MFC after: 4 days

Don't check for the existance of src/crypto/ for building items that
may contain crypto. The days of ITAR paranoia are over, and the simple
macro tests that remain are sufficient.

Remove MAINTAINER= lines from individual Makefiles in favor of the
MAINTAINER file (which already had entries for sendmail).

CSTD=c89 due to mismatched declarations.

Retire the useless NOSECURE knob.

Approved by: re (scottl)

style.Makefile(5)

No need for the OPENSSL_NO_KRB5 switch anymore.

Fixed by: nectar

iAdjust for OpenSSL 0.9.7.

Update build infrastructure for sendmail 8.12.
Add new build knob, SENDMAIL_SET_USER_ID, which installs sendmail as a
set-user-ID root binary instead of the new method (set-group-ID smmsp).

Set BINOWN=root explicitly for setuid root binaries.

Convert to using the same method of detecting the secure directory as other
usr.sbin programs.

Noted by: markm

Perform a major cleanup of the usr.sbin Makefiles.
These are not perfectly in agreement with each other style-wise, but they
are orders of orders of magnitude more consistent style-wise than before.

- Backout botched attempt to introduce MANSECT feature.
- MAN[1-9] -> MAN.

Move installation of sendmail helpfile into etc/sendmail/Makefile. It is
a user configurable file and it would benefit from mergemaster.

Move creation of the sendmail statistics file from the usr.sbin/sendmail
Makefile to the etc/sendmail Makefile to be consistent with all of the
other /var file creations. In doing so, change the Makefile target from
etc-sendmail.cf to distribution as it installs more than just the sendmail.cf.

Add a MAINTAINER= line so people know who to blame

Fix up the build for the STARTTLS version of sendmail (again). This method
mimics that of tcpdump in that for normal builds, sendmail will only be
built once. For 'make release', it is built once for the bin dist and
once for the crypto dist. This method also removes the need for two separate
Makefiles (which could become out of sync).

Suggested by: bde
Assisted by: kris

../Makefile.inc was clobbering BINDIR so sendmail was being installed in
/usr/sbin/ instead of /usr/libexec/sendmail/

Submitted by: bde

Remove STARTTLS support as it breaks builds without crypto installed.
Waiting to hear back regarding the best way to do this.

With apoligies to Greg Shapiro, fix the world. The previous commit
lost -lutil and -lwrap by replacing $LDADD and $DPADD rather than
appending to them with +=.

Style fixes

NOCRYPT imples NO_OPENSSL.
Still need to solve the distribution problem.

Submitted by: kris

Build sendmail with STARTTLS support unless NO_OPENSSL is set.

Give users a way to alter the sendmail (and related utilities) build
environment so they can enable functionality such as SASL, LDAP, Hesiod.

Allow users to add libraries for sendmail (e.g. Cyrus SASL)

Obtained from: Sergei Vyshenski <svysh@pn.sinp.msu.ru>

Turn on support for IPv6

Get rid of the /etc/aliases -> /etc/mail/aliases hack. /etc/mail/aliases
now exists in the distribution.

The rest of the changes needed to support the new version of sendmail (8.11.0).
Beyond changes to the build system, this includes fixing up the sample
freebsd.mc configuration for changes in defaults and syntax, removing
outdated documentation, and updating the release notes.

Since /etc/sendmail.cf got moved to /etc/mail/sendmail.cf, a 'make world'
would leave you with a broken sendmail and local mail loss.
This evil hack moves sendmail.cf from the old location to the new one (if
required) at install time.

Install sendmail in it's new location.

$Id$ -> $FreeBSD$

Enable tcp_wrapper support by default.

MaxHeaderLines is now MaxHeadersLength (in bytes)

Support 'O MaxHeaderLines=' to override the default header count and line
length limits. The configuration keyword is: confMAX_HEADER_LINES

Update for 8.9.2 (new file, control.c)
Also, turn on support for the MaxMimeHeaderLength option in sendmail.cf.

Remove useless `BINOWN=root' now that it is the default.

Remove old sendmail (to the Attic)

Connect up sendmail-8.9.1

$@ is deprecated, use longer forms of single char macros

replaced check_rcpt ruleset with one
recommended by Gary Palmer <gpalmer@freebsd.org>
Obtained from: http://www.informatik.uni-kiel.de/%7Eca/email/check.html#check_rcpt

grr....forgot a bug fix in rewrite rules on hub
Reviewed by: jmb
Submitted by: Ruslan Ermilov <ru@ucb.crimea.ua>

update to match configuration currently in use
on hub.freebsd.org

setpwent now has a void return type, so remove the now bogus cast.

update hub's master configuraton file
limit sendmail to 500 messages per run
keeps the process from growing too large
sort the queue by host, not numeric priority
connecting to a host is too expensive,
once we got the connection, ram all the mail down his throat ;>

update hub.mc to match what we are currently using:

both /etc/aliases and /etc/majordomo.aliases
sendmail.cw is *not* optional
command and connection timeouts
common hoststat directory for all queues
add fakenames database
clean up spam filter rules
limit the number of message processed each queue run
dont need to check Vixie's RBL in both check_relay and check_mail

update hub.mc with the latest set of anti-spam rules

Submitted by: jmb

improve anti-spam rulesets: check_relay, check_mail, check_rcpt.
break each ruleset into identified sections. (called groups).
note which groups can be reordered.
each group accepts and returns the same strings,
as much as possible.
reactivate Paul Vixie's RBL (in check_mail)
add rules to limit mail relaying to a list of hosts and domains
in the R class (check_rcpt, not active on hub.freebsd.org)

Submitted by: jmb

remove #include <login_cap.h>, this is included as needed elsewhere now.

setusercontext() came into the 2.2 branch just before 2.2.2-release.

Merge sendmail 8.8.8 changes onto mainline. (For all of CVS's warts, when
things like this turn out this easy (ie: it took 30 seconds), it's nice :-)

This commit was generated by cvs2svn to compensate for changes in r31087,
which included commits to RCS files with non-trunk default branches.

Typo.

Hmm.. where did _that_ come from? :-)

record the envelope-from address in the received headers
for all mail received at hub.freebsd.org

Obtained from: Peter Wemm <peter@freebsd.org>

DOH! I cannot believe that none of us caught this. Do *not*
turn TCPWRAPPER support on by default, eh? :-)

Changes to support full make parallelism (-j<n>) in the world
target.
Reviewed by: <many different folks>
Submitted by: Nickolay N. Dudorov" <nnd@nnd.itfs.nsk.su>

Record the envelope from address in the received: line. This has turned
out to be very useful for spam tracking and other problem debugging.

I first noticed this in one of Paul Vixie's machine's headers a while ago.

Sort cross refereces in section SEE ALSO.

Add (commented out) macros to make compiling with tcpwrappers easier.

If we're going to blindly install freebsd.cf from the etc-sendmail.cf
rule then we'd better depend on it.

Set ForwardPath as presently inuse on hub

Snapshot the various mail configurations. hub.mc was reconstructed to
produce the same sendmail.cf since the original has been lost.

Merge sendmail-8.8.7 changes from vendor branch

Merge sendmail 8.8.7 changes from vendor branch

This commit was generated by cvs2svn to compensate for changes in r27876,
which included commits to RCS files with non-trunk default branches.

README is for Lite2. We don't want it.

This commit was generated by cvs2svn to compensate for changes in r27850,
which included commits to RCS files with non-trunk default branches.

Revive this file, it's come back from the dead in the 8.8.x dists.

Merge in sendmail-8.8.5 -> 8.8.6 changes to those files that have left the
vendor branch.

This commit was generated by cvs2svn to compensate for changes in r26986,
which included commits to RCS files with non-trunk default branches.

compare return value from getopt against -1 rather than EOF, per the final
posix standard on the topic.

Activate HASSETUSERCONTEXT to set resource limits while delivering
to pipes in .forward files, but with a slight bugfix to the code.

Revert $FreeBSD$ to $Id$

Replace my EOF fix with better one from sendmail-bugs discussion

Should go into 2.2

Back out clearerr fix after discussion in sendmail-bugs,
it gains nothing

clearerr() clears both EOF and error flags, so do it only for
error and not for EOF or loop test becomes while(1)

Should go into 2.2
Submitted by: Gregory Neil Shapiro <gshapiro@WPI.EDU>

Under some circumstanes sendmail can loop forever collecting input.
Fix uninitialized character to prevent it.

Should go into 2.2

Merge 8.8.4 -> 8.8.5 changes onto mainline

This commit was generated by cvs2svn to compensate for changes in r22028,
which included commits to RCS files with non-trunk default branches.

Make the long-awaited change from $Id$ to $FreeBSD$

This will make a number of things easier in the future, as well as (finally!)
avoiding the Id-smashing problem which has plagued developers for so long.

Boy, I'm glad we're not using sup anymore. This update would have been
insane otherwise.

Add two missing links (hoststat/purgestat)

Allow an optional setting to specify a .cf file to be built and installed
into /etc during the sendmail build. Not for the feint hearted.
(I have been using something like this for some time since I only ever
edit my system's .mc file)

Use the label that Bruce suggested, it seems the src/etc/Makefile "helpers"
are all prefixed with "etc-".

Add a hook for installing the /etc/sendmail.cf file

Submitted by: bde

Update this to what freefall's been running for a day or so.

Remove file that we don't have any way of using. We don't have the other
Makefiles for other OS's in our tree that this uses when not using bmake.

Remove file no longer shipped with sendmail

Remove defunct files that are no longer shipped (and don't work with this
version of sendmail any more)

Back out rev 1.6, it's in the official source now (a few lines up)

Merge 8.8.3->8.8.4 changes onto our mainline where we've edited the
file at some point in the past.

Obtained from: Eric Allman <eric@sendmail.org>

This commit was generated by cvs2svn to compensate for changes in r20103,
which included commits to RCS files with non-trunk default branches.

Use LC_TIME=C

Got the bugger. Eric's now passing in a hint structure to dbopen()
but forgot to zero the structure first, leaving hinfo.bsize undefined
and causing a loop in __log2().

Merge conflicts from 8.8.3 import onto mainline.

This commit was generated by cvs2svn to compensate for changes in r19841,
which included commits to RCS files with non-trunk default branches.

Disallow non-root users to start in daemon mode.

Obtained from: Eric Allman <eric@sendmail.org>

Back out freebsd local change which always forced SMTP to assume an 8-bit
clear channel. This change was originally put in for freefall, and is
completely irrelevant given that freefaill defines the generic SMTP
service to use the smtp8 mailer (which is standard mechanism.)

[The original patch violated RFC-821.]

Apply to: 2.2 (please)

make the /etc/sendmail.cw file optional by default.

Definately should go into 2.2..

Requested by: lots of people

The standard .mc files shipped with the new sendmail have an extra divert(0)
at the top. Make our freebsd-supplied files match the standard files.

Apply-to: 2.2

Fix base64->8bit conversion
Submitted by: Eric Allman <eric@sendmail.org>

Integrate our changes from bin/rmail. Mainly it causes immediate
delivery rather than queueing it and forgetting it.

(and some -Wall fixes where gcc is too dumb to see it's a "cant happen")

Convert the DONT_FSYNC compile time option to a runtime option (like
nobiff). The options to turn these on are specified in the
LOCAL_MAILER_ARGS define the the sendmail.mc that you build from.

Merge in our changes from src/libexec/mail.local. This version doesn't
violate disk quotas, has more robust locking, is still being developed,
etc. The main changes merged in were the ability to not fsync() the
mailbox, not do biff/comsat broadcasts, man page fixes.

Untangle NEWDB/NDBM options.
Enable praliases.c to not use NDBM.

missed this praliases.8 <-> praliases.0 change

Fold sendmail-8.8.2 changes into files that have been touched.

(^!&@$#&^! delete !!@^@^ trailing !@^&#$!& whitespace!!!)

This commit was generated by cvs2svn to compensate for changes in r19146,
which included commits to RCS files with non-trunk default branches.

Enable NIS support

add missing comma(s) in .Xr macros

Fix one byte buffer overflow
Submitted by: Eric Allman <eric@sendmail.org>

Fix & expansion (from pw_gecos field), sizeof buf used when
buf is pointer, not array, buflen must be used instead.
Charlie & from root name was expanded to junk as result.

Merge import conflicts

This commit was generated by cvs2svn to compensate for changes in r18366,
which included commits to RCS files with non-trunk default branches.

Add the -o option to the usage line.

Oops, forgot to actually add the new man page.

Add a man page for mailstats(8).

Obtained from: NetBSD

freefall's new sendmail.cf file.
Submitted by: gpalmer

typo shmrsh -> smrsh

Find official home of smrsh

Enable proper installation of sendmail restricted shell smrsh(8).

This program is a wrapper for the prog mailer in sendmail. It does shell
meta character masking and restricts the list of executables to those found
in /usr/libexec/sm.bin.

The default sendmail.cf file does not use this tool, however you can enable
it by either changing /bin/sh to /usr/libexec/smrsh or adding the line
FEATURE(smrsh) into your sendmail .mc file and rebuilding your .cf file.

For more info, RTFMP.

Remove the legacy -d option from the arguments to mail.local in 4.4bsd.

Add the '8' flag to the smtp and esmtp mailers. This causes sendmail-8.7.5
to behave like the older sendmails when talking to a peer that does not
have esmtp or does not advertise 8BITMIME. The old sendmail "just sent it
anyway", while the 8.7.x series mangle any extended character set mail
by conferting it to quoted-printable or base64. Freefall has been
running this for some time.

Fold in 8.7.5 change onto mainline (all one line).

This commit was generated by cvs2svn to compensate for changes in r14615,
which included commits to RCS files with non-trunk default branches.

Remove useless (for us) Makefiles. There were already other "Makefile.dist"
files missing, so these shouldn't hurt. If somebody wanted to use sendmail
8.7 on their machine, they should use a clean dist anyway, not this one.

Submitted by: wollman

Merge 8.7.3->8.7.4 changes onto mainline.

This commit was generated by cvs2svn to compensate for changes in r14182,
which included commits to RCS files with non-trunk default branches.

Correct a bunch of man page cross references and generally
try and silence "manck".

ncurses, rpc, and some of the gnu stuff are still a big mess, however.

Add some missing MLINKS, correct some cross references, correct some
file locations and some minor formatting/style problems.

Remove support for OSI networking in user-land (#ifdef OSI aor CCITT)
in preparation for its removal from the kernel source tree. NB: because
a function was deleted, libc is now at version 3.0 (was 2.2 previously).

Don't install sendmail setgid kmem since it doesn't really
require kmem access. It is currently configured to call getloadavg(3),
which uses sysctl(2) to determine the load average, which is the
only reason some sendmails require kmem access.

Whoops! Include paths.h instead of sys/cdefs.h - that results in
PATH_VARTMP having a somewhat saner value.
Submitted by: matthew green <mrg@eterna.com.au>
By way of: Eric Allman <eric@sendmail.org>

Get sendmail to detect when to use the setproctitle() in libutil

This uses osreldate.h and other stuff to determine whether it's on a
-current system. It should still compile cleanly on a -stable branch
system.

Revert from fiddling with the PS_STRINGS pointers back to the traditional
spamming of the argv, almost back to the original code (except for NULL
padding of the string).
This makes this version of the sendmail source compile on 2.1.X again.

These files are regeneratable and are making huge rcs deltas for no real
need. Note that "op" and "intro" are already being built under
share/doc/smm (08.sendmailop and 09.sendmail)

Change freefall's default mailer from smtp to smtp8 to try and avoid
the 8 to 7 bit conversion..

Back out "find itself" fix for security purist reasons from
sendmail mailing list. Our /etc/rc will be fixed instead.
It seems sendmail becomes more and more incompatible, f.e.
Return-Receipt-To not works anymore :-(

Fix SIGHUP restarting bug: sendmail can't find itself, it uses argv[0]
but don't uses PATH

Update sendmail's "knowledge" of the ps_strings structure.

This really is ugly, we need setproctitle() in a shared lib somewhere to
make these more transparent.

Specify Timeout.queuewarn.normal at 1day as well as Timeout.queuewarn.
For some reason, the documentation does not appear to be in sync with
reality.. :-)

Make sendmail's setproctitle() use PS_STRINGS rather than smash
the argv. It appears that this once didn't work, but now it seems
to work just fine.

Whitespace change - cosmetic alignment only.

Add in Gary's custom Received: line into the freefall.mc

Add "majordom" to trusted user class for freefall's sendmail.mc file.

Sigh, I really hate this "feature". I'm tempted to comment the test out
(like in sendmail 8.6.xx), especially since we log the info in the
Received: header line anyway.. Doing it like this only makes it painful
for things like uucp..

Fold in changes to files no longer on the vendor branch from 8.7.3 update.

This commit was generated by cvs2svn to compensate for changes in r12601,
which included commits to RCS files with non-trunk default branches.

*GULP* cvs remove the uncomfortably large list of files that are no longer
part of sendmail 8.7.2...

Update the freefall.mc file to (almost) match what is actually being
run on freefall..
Create freebsd.mc which is intended to be installed by src/etc/Makefile
onto generic freebsd systems.

Re-disable the cf/cf SUBDIR - we were not building it before anyway.
The Makefile down there does not handle the obj dir well..

Initial round of conflict resolutions.. I need to do another
pass over this and check it more closely, but my initial concern was
to get it to all compile again. Some of the obsoleted files 'cvs rm'ed.

This commit was generated by cvs2svn to compensate for changes in r12571,
which included commits to RCS files with non-trunk default branches.

If sendmail is collecting a message *not* via SMTP, then this patch will
turn off the read data block timeouts. This was inspired by the same fix
in sendmail-8.7, but the code in this area is somewhat different. (8.7
is much more sensible). This will stop the infamous
"timeout waiting for input during message body read" on long running jobs,
eg: /etc/daily | mail root.

change cross reference from binmail(1) to mail.local(8)

Obtained from: the NetBSD bugs mailinglist

Merge in the remaining differences from 8.6.11 to 8.6.12.
Note that conf.c, although there was an import conflict, it did not
require intervention, as it was the $Id$ tag. It would have become
rev 1.8 on checkout so there's no point changing it from 1.7 to
1.1.1.3 as the "-j" option wanted to do.. Trust me.. :-)

This commit was generated by cvs2svn to compensate for changes in r10088,
which included commits to RCS files with non-trunk default branches.

Change `install' to `${INSTALL}' so that default install flags can be
specified in the top level Makefiles.

Previously I missed dozens of Makefiles that skip the install after
using `cmp -s' to decide that the install isn't necessary.

Change `install' to `${INSTALL}' so that default install flags can be
specified in the top level Makefiles.

Remove trailing whitespace.

Fix nested includes for -Wcomment.

Remove Matt Day patch (handled elsewhere in 8.6.11)

Merge CSRG branch changes into main branch

This commit was generated by cvs2svn to compensate for changes in r6973,
which included commits to RCS files with non-trunk default branches.

Fix CurHostName ($m macro)
Submitted by: eric@cs.berkeley.edu

This commit was generated by cvs2svn to compensate for changes in r6653,
which included commits to RCS files with non-trunk default branches.

Cover 8.6.10 merge conflicts and warnings

This commit was generated by cvs2svn to compensate for changes in r6647,
which included commits to RCS files with non-trunk default branches.

Fix configurations in which mail route depends of the name
of sender host.
Submitted by: vak@cronyx.msk.su

from Matt Day <mday@artisoft.com>:

> So, the problem is hostnamebyanyaddr() assumes that _res is already
> initialized to proper values, but this is no longer true as of BIND
> 4.9.3. Here is a patch for sendmail 8.6.9 which changes
> hostnamebyanyaddr() to make sure res_init() has been called before using
> _res.retry.

Submitted by: Matt Day <mday@artisoft.com> via Terry Lambert

BSD 4.4 compatibility fix
Submitted by: Ollivier Robert

__386BSD__ -> __FreeBSD__

I know that many of these entries are bogus and need to be revisited,
but let's get the tree working again for now and then do a pass through
looking at all the __FreeBSD__ entries, shall we?

Deleted the private clean/cleandir targets in the Makefile. Delete
a hardcoded reference to the "obj" dir. So this makefile should now
work with and without an obj dir and deletes the obj dir with a
normal cleandir.

You need the freefall.mc file to build the release /etc/sendmail.cf file from.

Get rid of update. Make man page installation work with our scheme
(and rename a few in the process).

Converted 'vmunix' to 'kernel'.

Repeat after me, I will not edit files on Freefall, but bring patches
over instead.
Moved #ifdef to correct place.

Disable PS_STRINGS for now until I get the headers fixed.

This commit was generated by cvs2svn to compensate for changes in r1553,
which included commits to RCS files with non-trunk default branches.