summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* sendmail: remove -bs and -bpmanuel2020-08-251-3/+3
|
* Revert "Rewrite 554 smtp code for web.de/gmx.net/1&1 servers"manuel2020-06-021-17/+1
| | | | This reverts commit f2ef25deb1aa356d41cdd3f6e46d9a68c48bfce0.
* Make sure to limit max alloc sizemanuel2020-05-191-0/+5
| | | | see https://www.qualys.com/2020/05/19/cve-2005-1513/remote-code-execution-qmail.txt
* Rewrite 554 smtp code for web.de/gmx.net/1&1 serversmanuel2019-12-021-1/+17
|
* increase recipients limit to 300manuel2019-06-121-1/+1
|
* SMTP: limit max errors to 20manuel2019-06-111-24/+40
|
* max recipients: make error message more consistentmanuel2019-06-111-1/+1
|
* Add max recipients limitmanuel2019-06-111-1/+10
|
* Fix compile warnings...Manuel Mausz2019-06-1194-124/+200
|
* Only do deferrals in case qmail-remote does not spawnmanuel2019-05-281-1/+2
|
* Fix CVE-2011-1431manuel2019-04-101-1/+4
|
* Remove client certificate left oversmanuel2019-02-012-24/+0
|
* Disable TLS 1.0 only for smtp auth/submissionmanuel2018-12-261-4/+7
|
* Disable TLS 1.0manuel2018-12-041-1/+1
|
* plus addressing improvementsmanuel2018-10-152-23/+34
| | | | | * first `+` splits local part from sub-address * prioritize plus addressing over domain wildcards
* Enable PRIORITIZE_CHACHA per defaultmanuel2018-09-173-3/+6
|
* More OpenSSL 1.1 compatibilitymanuel2018-06-281-0/+4
|
* OpenSSL 1.1 compatibilityManuel Mausz2018-06-278-137/+125
| | | | | | | | | | | | | | This adds compatibility for OpenSSL 1.1 Since renegotiation is removed from TLS 1.3 we also removed support for authentication via client certificates (control/tlsclients). In general this is still supported by TLS 1.3 however I'm just lazy and we don't need this feature anyway. This also adds optional support for OpenSSL configuration commands for qmail-smtpd and qmail-remote. Commands are loaded from control/opensslconf. For a list of supported commands see https://www.openssl.org/docs/man1.0.2/ssl/SSL_CONF_cmd.html#SUPPORTED-CONFIGURATION-FILE-COMMANDS
* work around a libval bugmanuel2018-04-161-0/+3
|
* Disable SSLv3 and enable prefer server ciphersmanuel2018-04-162-2/+3
|
* Initialize session early for loggingmanuel2018-04-161-0/+1
|
* ignore AUTH parameters in MAIL FROMmanuel2018-04-041-1/+2
|
* add X-UD-Smtp-Session and remove ip from authed clientsmanuel2018-04-043-2/+87
|
* realrcptto: drop default QMAILQUEUE for catch-allmanuel2018-02-261-8/+0
|
* realrcptto: extract username from mailbox forwardsmanuel2017-12-061-0/+5
|
* realrcptto: stop resolving after dovecot-rda has been foundmanuel2017-12-061-4/+12
|
* SMTPUTF8: substract null terminationmanuel2017-10-101-0/+1
| | | | | qmail-remote assumes host is not null terminated. IDN requires null termination so subtract the trailing NULL after appending.
* SMTPUTF8: convert UTF-8 domains to IDNAmanuel2017-08-182-2/+31
| | | | | | | | | | | | | This isn't entirely correct according to RFC 6531 as it's better to not convert at all. However this would require we additionally add the UTF-8 form to qmails local recipient users/domains. Plus our SMTPUTF8 implementation doesn't convert outgoing UTF-8 mails if receiving MTA doesn't support SMTPUTF8. So mail forwarding might generate a bounce the user very likely doesn't understand. As MTAs who support SMTPUTF8 most likely also support the IDNA form, always converting is probably a good trade-off.
* add support for SMTPUTF8manuel2017-08-176-13/+126
|
* Implement plus addressing for virtualdomainsmanuel2017-08-162-1/+43
|
* realrcptto: extract username from dovecot-rda callsmanuel2017-08-161-1/+20
|
* make qmail support EC keysmanuel2016-08-072-3/+3
|
* implement tls before auth as environment variablemanuel2016-03-211-8/+10
|
* don't announce SMTP AUTH commands if not availablemanuel2016-03-211-7/+10
|
* disable SSLv2manuel2016-03-012-0/+2
|
* [PATCH] qmail queue custom error patchmanuel2015-07-272-1/+26
| | | | see google qmail-queue-custom-error-v2.patch
* update DANE support to libval 2.1 (+fixes)manuel2015-07-241-20/+35
|
* fix missing check for invalid dane statusmanuel2015-07-061-13/+14
| | | | | libval returns VAL_DANE_NOERROR if domain is insecure and no TLSA RR exists. we should stop all further dane handling in this case
* add SMTP DANE supportmanuel2015-07-043-63/+121
|
* add gitignoremanuel2015-07-041-0/+405
|
* opportunistic TLSmanuel2015-06-303-8/+50
| | | | When the TLS handshake fails, retry delivery with TLS disabled
* add support for ECDH as well as custom ec+dh params in certificatemanuel2015-06-232-6/+71
|
* Use DH parameters from OpenSSL and remove support for ephemeral RSAmanuel2015-06-228-101/+66
| | | | This adds support for DH parameters from 1024 to 8192 bits.
* Treat ECONNREFUSED the same as ETIMEDOUTmanuel2014-05-203-1/+9
| | | | | | | | After an SMTP connection attempt times out, qmail-remote records the relevant IP address. If the same address fails again (after at least two minutes with no intervening successful connections), qmail-remote assumes that further attempts will fail for at least another hour.
* Remove qmail-remote workaround that Dan Bernstein added on 1996-10-03...manuel2014-05-201-1/+1
| | | | | | ...to work around a bug in BIND versions earlier than version 4.9.4 We don't use the code any more but it does no harm.
* Increase memory buffer from 256 to 8kmanuel2014-05-201-2/+2
| | | | | | | | This patch increases the size of the memory buffer that qmail uses when querying the system for a list of all local IP addresses. This mailing list message describes the problem and includes the patch file. see http://www.mail-archive.com/qmail@id.wustl.edu/msg20935.html
* Fix qmail-remote to handle multiple MX records correctly as in rfc2821manuel2014-05-201-2/+3
| | | | | | | | This patch changes qmail-remote to skip over MX servers that greet with codes 400 to 499 and to bounce mail when any MX server that qmail tries greets with a code 500 to 599. see http://www-dt.e-technik.uni-dortmund.de/~ma/qmail/patch-qmail-1.03-rfc2821.diff
* get rid of DNS ANY querymanuel2014-02-181-24/+4
| | | | | | | | UltraDNS started blocking UDP-ANY queries so this is a problem. This will disable domain canonicalization of the envelope address of outgoing messages. This is as specified by RFC 1123 section 5.2.2. However this requirement is obsolete and modern MTAs have disabled it too.
* export TLS-Session to environmentmanuel2014-02-141-0/+1
|
* replace \r\n with \n in sendmail wrappermanuel2014-01-302-8/+60
|
generated by cgit v1.2.3 (git 2.25.1) at 2026-01-13 04:11:16 +0000