1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
17 PP/01 Solaris build fix for Oracle's LDAP libraries.
18 Bugzilla 1109, patch from Stephen Usher.
20 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
22 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
25 TF/02 Fix a couple more cases where we did not log the error message
26 when unlink() failed. See also change 4.74-TF/03.
28 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
29 lock up or crash if it happened to be inside a call to libc when it
30 got a SIGUSR1 from exiwhat.
32 The SIGUSR1 handler appends the current process status to the process
33 log which is later printed by exiwhat. It used to use the general
34 purpose logging code to do this, but several functions it calls are
37 The new output code in the SIGUSR1 handler is specific to the process
38 log, and simple enough that it's easy to inspect for signal safety.
39 Removing some special cases also simplifies the general logging code.
40 Removing the spurious timestamps from the process log simplifies
43 TF/04 Improved ratelimit ACL condition.
45 The /noupdate option has been deprecated in favour of /readonly which
46 has clearer semantics. The /leaky, /strict, and /readonly update modes
47 are mutually exclusive. The update mode is no longer included in the
48 database key; it just determines when the database is updated. (This
49 means that when you upgrde Exim will forget old rate measurements.)
51 Exim now checks that the per_* options are used with an update mode that
52 makes sense for the current ACL. For example, when Exim is processing a
53 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
54 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
55 must specify per_mail/readonly. If you omit the update mode it defaults to
56 /leaky where that makes sense (as before) or /readonly where required.
58 The /noupdate option is now undocumented but still supported for
59 backwards compatibility. It is equivalent to /readonly except that in
60 ACLs where /readonly is required you may specify /leaky/noupdate or
61 /strict/noupdate which are treated the same as /readonly.
63 A useful new feature is the /count= option. This is a generalization
64 of the per_byte option, so that you can measure the throughput of other
65 aggregate values. For example, the per_byte option is now equivalent
66 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
68 The per_rcpt option has been generalized using the /count= mechanism
69 (though it's more complicated than the per_byte equivalence). When it is
70 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
71 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
72 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
73 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
74 Note that using per_rcpt with a non-readonly update mode in more than
75 one ACL will cause the recipients to be double-counted. (The per_mail
76 and per_byte options don't have this problem.)
78 The handling of very low rates has changed slightly. If the computed rate
79 is less than the event's count (usually one) then this event is the first
80 after a long gap. In this case the rate is set to the same as this event's
81 count, so that the first message of a spam run is counted properly.
83 The major new feature is a mechanism for counting the rate of unique
84 events. The new per_addr option counts the number of different
85 recipients that someone has sent messages to in the last time period. It
86 behaves like per_rcpt if all the recipient addresses are different, but
87 duplicate recipient addresses do not increase the measured rate. Like
88 the /count= option this is a general mechanism, so the per_addr option
89 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
90 example, measure the rate that a client uses different sender addresses
91 with the options per_mail/unique=$sender_address. There are further
92 details in the main documentation.
94 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
96 TF/06 Removed a few PCRE remnants.
98 TF/07 Automatically extract Exim's version number from tags in the git
99 repository when doing development or release builds.
101 PP/02 Raise smtp_cmd_buffer_size to 16kB.
102 Bugzilla 879. Patch from Paul Fisher.
104 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
105 Heavily based on revision 40f9a89a from Simon Arlott's tree.
108 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
110 PP/05 Variable $av_failed, true if the AV scanner deferred.
111 Bugzilla 1078. Patch from John Horne.
113 PP/06 Stop make process more reliably on build failure.
114 Bugzilla 1087. Patch from Heiko Schlittermann.
116 PP/07 Make maildir_use_size_file an _expandable_ boolean.
117 Bugzilla 1089. Patch from Heiko Schlittermann.
119 PP/08 Handle ${run} returning more data than OS pipe buffer size.
120 Bugzilla 1131. Patch from Holger Weiß.
122 PP/09 Handle IPv6 addresses with SPF.
123 Bugzilla 860. Patch from Wolfgang Breyha.
125 PP/10 GnuTLS: support TLS 1.2 & 1.1.
127 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
130 PP/11 match_* no longer expand right-hand-side by default.
131 New compile-time build option, EXPAND_LISTMATCH_RHS.
132 New expansion conditions, "inlist", "inlisti".
134 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
136 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
138 PP/14 fix log_write() format string regression from TF/03.
139 Bugzilla 1152. Patch from Dmitry Isaikin.
145 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
147 PP/02 Harmonised TLS library version reporting; only show if debugging.
148 Layout now matches that introduced for other libraries in 4.74 PP/03.
150 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
152 PP/04 New "dns_use_edns0" global option.
154 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
157 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
158 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
160 TK/01 Updated PolarSSL code to 0.14.2.
161 Bugzilla 1097. Patch from Andreas Metzler.
163 PP/07 Catch divide-by-zero in ${eval:...}.
166 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
169 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
170 format-string attack -- SECURITY: remote arbitrary code execution.
172 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
173 time unintentionally subject to list matching rules, letting the header
174 cause arbitrary Exim lookups (of items which can occur in lists, *not*
175 arbitrary string expansion). This allowed for information disclosure.
177 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
178 INT_MIN/-1 -- value coerced to INT_MAX.
184 NM/01 Workround for PCRE version dependency in version reporting
187 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
188 This fixes portability to compilers other than gcc, notably
189 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
191 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
192 makefiles for portability to HP-UX and POSIX correctness.
194 PP/01 Permit LOOKUP_foo enabling on the make command-line.
195 Also via indented variable definition in the Makefile.
196 (Debugging by Oliver Heesakkers).
198 PP/02 Restore caching of spamd results with expanded spamd_address.
199 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
201 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
202 Improves build reliability. Fix from: Frank Elsner
204 NM/02 Fix wide character breakage in the rfc2047 coding
205 Fixes bug 1064. Patch from Andrey N. Oktyabrski
207 NM/03 Allow underscore in dnslist lookups
208 Fixes bug 1026. Patch from Graeme Fowler
210 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
211 Code patches from Adam Ciarcinski of NetBSD.
213 NM/04 Fixed exiqgrep to cope with mailq missing size issue
216 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
217 is logged, to avoid truncation. Patch from John Horne.
219 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
220 Patch from Jakob Hirsch.
222 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
223 SQL string expansion failure details.
224 Patch from Andrey Oktyabrski.
226 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
227 Patch from Simon Arlott.
229 PP/09 New lookups functionality failed to compile on old gcc which rejects
230 extern declarations in function scope.
231 Patch from Oliver Fleischmann
233 PP/10 Use sig_atomic_t for flags set from signal handlers.
234 Check getgroups() return and improve debugging.
235 Fixed developed for diagnosis in bug 927 (which turned out to be
238 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
239 Patch from Mark Zealey.
241 PP/12 Bugzilla 1056: Improved spamd server selection.
242 Patch from Mark Zealey.
244 PP/13 Bugzilla 1086: Deal with maildir quota file races.
245 Based on patch from Heiko Schlittermann.
247 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
248 Patch from Uwe Doering, sign-off by Michael Haardt.
250 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
251 variable declaration deep within a block. Bug and patch from
254 PP/15 lookups-Makefile IRIX compatibilty coercion.
256 PP/16 Make DISABLE_DKIM build knob functional.
258 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
259 Patch by Simon Arlott
261 TF/03 Fix valgrind.h portability to C89 compilers that do not support
262 variable argument macros. Our copy now differs from upstream.
268 TF/01 Failure to get a lock on a hints database can have serious
269 consequences so log it to the panic log.
271 TF/02 Log LMTP confirmation messages in the same way as SMTP,
272 controlled using the smtp_confirmation log selector.
274 TF/03 Include the error message when we fail to unlink a spool file.
276 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
277 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
278 for maintaining out-of-tree patches for some time.
280 PP/01 Bugzilla 139: Documentation and portability issues.
281 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
282 Handle per-OS dynamic-module compilation flags.
284 PP/02 Let /dev/null have normal permissions.
285 The 4.73 fixes were a little too stringent and complained about the
286 permissions on /dev/null. Exempt it from some checks.
287 Reported by Andreas M. Kirchwitz.
289 PP/03 Report version information for many libraries, including
290 Exim version information for dynamically loaded libraries. Created
291 version.h, now support a version extension string for distributors
292 who patch heavily. Dynamic module ABI change.
294 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
295 privilege escalation vulnerability whereby the Exim run-time user
296 can cause root to append content of the attacker's choosing to
299 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
302 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
303 If dropping privileges for untrusted macros, we disabled normal logging
304 on the basis that it would fail; for the Exim run-time user, this is not
305 the case, and it resulted in successful deliveries going unlogged.
306 Fixed. Reported by Andreas Metzler.
312 PP/01 Date: & Message-Id: revert to normally being appended to a message,
313 only prepend for the Resent-* case. Fixes regression introduced in
314 Exim 4.70 by NM/22 for Bugzilla 607.
316 PP/02 Include check_rfc2047_length in configure.default because we're seeing
317 increasing numbers of administrators be bitten by this.
319 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
321 PP/03 Bugzilla 994: added openssl_options main configuration option.
323 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
325 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
327 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
329 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
330 without a peer certificate, leading to a segfault because of an
331 assumption that peers always have certificates. Be a little more
332 paranoid. Problem reported by Martin Tscholak.
334 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
335 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
336 NB: ClamAV planning to remove STREAM in "middle of 2010".
337 CL also introduces -bmalware, various -d+acl logging additions and
338 more caution in buffer sizes.
340 PP/09 Implemented reverse_ip expansion operator.
342 PP/10 Bugzilla 937: provide a "debug" ACL control.
344 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
346 PP/12 Bugzilla 973: Implement --version.
348 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
350 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
352 PP/15 Bugzilla 816: support multiple condition rules on Routers.
354 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
355 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
356 ignore trailing whitespace.
358 JJ/02 prevent non-panic DKIM error from being sent to paniclog
360 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
363 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
364 Notification from Dr Andrew Aitchison.
366 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
367 ExtendedDetectionInfo response format.
368 Notification from John Horne.
370 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
373 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
374 XSL and documented dependency on system catalogs, with examples of how
377 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
380 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
381 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
382 configuration file which is writeable by the Exim user or group.
384 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
385 of configuration files to cover files specified with the -C option if
386 they are going to be used with root privileges, not just the default
389 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
390 option (effectively making it always true).
392 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
393 files to be used while preserving root privileges.
395 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
396 that rogue child processes cannot use them.
398 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
399 run-time user, instead of root.
401 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
402 Exim run-time user without dropping privileges.
404 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
405 result string, instead of calling string_vformat() twice with the same
408 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
409 for other users. Others should always drop root privileges if they use
410 -C on the command line, even for a whitelisted configure file.
412 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
414 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
420 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
421 $data_path, and $header_path variables; fixed documentation bugs and
424 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
425 exipick to access non-standard spools, including the "frozen" queue
428 NM/01 Bugzilla 965: Support mysql stored procedures.
429 Patch from Alain Williams
431 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
433 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
434 Patch from Andreas Metzler
436 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
437 Patch from Kirill Miazine
439 NM/05 Bugzilla 671: Added umask to procmail example.
441 JJ/03 installed exipick 20100323.0, fixing doc bug
443 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
444 directory. Notification and patch from Dan Rosenberg.
446 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
448 TK/02 Improve log output when DKIM signing operation fails.
450 MH/01 Treat the transport option dkim_domain as a colon separated
451 list, not as a single string, and sign the message with each element,
452 omitting multiple occurences of the same signer.
454 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
455 Bugzilla 985, 986. Patch by Simon Arlott
457 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
458 Patch by Simon Arlott
460 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
461 MBX locking. Notification from Dan Rosenberg.
467 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
469 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
471 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
473 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
475 NM/04 Bugzilla 915: Flag broken perl installation during build.
481 TK/01 Added patch by Johannes Berg that expands the main option
482 "spamd_address" if it starts with a dollar sign.
484 TK/02 Write list of recipients to X-Envelope-Sender header when building
485 the mbox-format spool file for content scanning (suggested by Jakob
488 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
489 (http://www.dcc-servers.net/) support via dccifd. Activated by
490 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
492 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
493 by Mark Daniel Reidel <mr@df.eu>.
495 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
496 When building exim an external PCRE library is now needed -
497 PCRE is a system library on the majority of modern systems.
498 See entry on PCRE_LIBS in EDITME file.
500 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
501 conversation. Added nologin parameter to request.
502 Patch contributed by Kirill Miazine.
504 TF/01 Do not log submission mode rewrites if they do not change the address.
506 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
508 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
509 log files in place. Contributed by Roberto Lima.
511 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
513 TF/03 Bugzilla 615: When checking the local_parts router precondition
514 after a local_part_suffix or local_part_prefix option, Exim now
515 does not use the address's named list lookup cache, since this
516 contains cached lookups for the whole local part.
518 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
519 Robert Millan. Documentation is in experimental-spec.txt.
521 TF/04 Bugzilla 668: Fix parallel build (make -j).
523 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
525 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
526 Patch provided by Jan Srzednicki.
528 TF/05 Leading white space used to be stripped from $spam_report which
529 wrecked the formatting. Now it is preserved.
531 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
532 that they are available at delivery time.
534 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
536 TF/08 TLS error reporting now respects the incoming_interface and
537 incoming_port log selectors.
539 TF/09 Produce a more useful error message if an SMTP transport's hosts
540 setting expands to an empty string.
542 NM/06 Bugzilla 744: EXPN did not work under TLS.
543 Patch provided by Phil Pennock.
545 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
546 Patch provided by Richard Godbee.
548 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
549 acl_smtp_notquit, added index entry.
551 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
552 Patch provided by Eugene Bujak.
554 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
555 accept(). Patch provided by Maxim Dounin.
557 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
558 Patch provided by Phil Pennock.
560 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
562 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
563 Patch provided by Brad "anomie" Jorsch.
565 NM/14 Bugzilla 622: Added timeout setting to transport filter.
566 Patch provided by Dean Brooks.
568 TK/05 Add native DKIM support (does not depend on external libraries).
570 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
571 Patch provided by Graeme Fowler.
573 NM/16 Bugzilla 851: Documentation example syntax fix.
575 NM/17 Changed NOTICE file to remove references to embedded PCRE.
577 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
580 NM/19 Bugzilla 745: TLS version reporting.
581 Patch provided by Phil Pennock.
583 NM/20 Bugzilla 167: bool: condition support.
584 Patch provided by Phil Pennock.
586 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
587 clients. Patch provided by Phil Pennock.
589 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
590 Patch provided by Brad "anomie" Jorsch.
592 NM/23 Bugzilla 687: Fix misparses in eximstats.
593 Patch provided by Heiko Schlittermann.
595 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
596 Patch provided by Heiko Schlittermann.
598 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
599 plus update to original patch.
601 NM/26 Bugzilla 799: Documentation correction for ratelimit.
603 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
604 Patch provided by David Brownlee.
606 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
608 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
610 NM/30 Bugzilla 888: TLS documentation bugfixes.
612 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
614 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
615 Unlike the original bugzilla I have changed all shell scripts in src tree.
617 NM/33 Bugzilla 898: Transport filter timeout fix.
618 Patch by Todd Rinaldo.
620 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
621 Patch by Serge Demonchaux.
623 NM/35 Bugzilla 39: Base64 decode bug fixes.
624 Patch by Jakob Hirsch.
626 NM/36 Bugzilla 909: Correct connect() call in dcc code.
628 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
630 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
632 NM/39 Bugzilla 911: Fixed MakeLinks build script.
638 TK/01 Add preliminary DKIM support. Currently requires a forked version of
639 ALT-N's libdkim that I have put here:
640 http://duncanthrax.net/exim-experimental/
642 Note to Michael Haardt: I had to rename some vars in sieve.c. They
643 were called 'true' and it seems that C99 defines that as a reserved
644 keyword to be used with 'bool' variable types. That means you could
645 not include C99-style headers which use bools without triggering
646 build errors in sieve.c.
648 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
649 as mailq or other aliases. Changed the --help handling significantly
650 to do whats expected. exim_usage() emits usage/help information.
652 SC/01 Added the -bylocaldomain option to eximstats.
654 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
656 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
658 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
664 PH/01 Another patch from the Sieve maintainer.
666 PH/02 When an IPv6 address is converted to a string for single-key lookup
667 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
668 dots are used instead of colons so that keys in lsearch files need not
669 contain colons. This was done some time before quoting was made available
670 in lsearch files. However, iplsearch files do require colons in IPv6 keys
671 (notated using the quote facility) so as to distinguish them from IPv4
672 keys. This meant that lookups for IP addresses in host lists did not work
673 for iplsearch lookups.
675 This has been fixed by arranging for IPv6 addresses to be expressed with
676 colons if the lookup type is iplsearch. This is not incompatible, because
677 previously such lookups could never work.
679 The situation is now rather anomolous, since one *can* have colons in
680 ordinary lsearch keys. However, making the change in all cases is
681 incompatible and would probably break a number of configurations.
683 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
686 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
687 conversion specification without a maximum field width, thereby enabling
688 a rogue spamd server to cause a buffer overflow. While nobody in their
689 right mind would setup Exim to query an untrusted spamd server, an
690 attacker that gains access to a server running spamd could potentially
691 exploit this vulnerability to run arbitrary code as the Exim user.
693 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
694 $primary_hostname instead of what libspf2 thinks the hosts name is.
696 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
697 a directory entry by the name of the lookup key. Previously, if a
698 symlink pointed to a non-existing file or a file in a directory that
699 Exim lacked permissions to read, a lookup for a key matching that
700 symlink would fail. Now it is enough that a matching directory entry
701 exists, symlink or not. (Bugzilla 503.)
703 PH/03 The body_linecount and body_zerocount variables are now exported in the
706 PH/04 Added the $dnslist_matched variable.
708 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
709 This means they are set thereafter only if the connection becomes
712 PH/06 Added the client_condition to authenticators so that some can be skipped
713 by clients under certain conditions.
715 PH/07 The error message for a badly-placed control=no_multiline_responses left
716 "_responses" off the end of the name.
718 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
720 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
721 (without spaces) instead of just copying the configuration text.
723 PH/10 Added the /noupdate option to the ratelimit ACL condition.
725 PH/11 Added $max_received_linelength.
727 PH/12 Added +ignore_defer and +include_defer to host lists.
729 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
730 way in which PCRE > 7.0 is built.
732 PH/14 Implemented queue_only_load_latch.
734 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
735 MAIL command. The effect was to mangle the value on 64-bit systems.
737 PH/16 Another patch from the Sieve maintainer.
739 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
741 PH/18 If a system quota error occurred while trying to create the file for
742 a maildir delivery, the message "Mailbox is full" was not appended to the
743 bounce if the delivery eventually timed out. Change 4.67/27 below applied
744 only to a quota excession during the actual writing of the file.
746 PH/19 It seems that peer DN values may contain newlines (and other non-printing
747 characters?) which causes problems in log lines. The DN values are now
748 passed through string_printing() before being added to log lines.
750 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
751 and InterBase are left for another time.)
753 PH/21 Added message_body_newlines option.
755 PH/22 Guard against possible overflow in moan_check_errorcopy().
757 PH/23 POSIX allows open() to be a macro; guard against that.
759 PH/24 If the recipient of an error message contained an @ in the local part
760 (suitably quoted, of course), incorrect values were put in $domain and
761 $local_part during the evaluation of errors_copy.
767 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
768 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
771 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
772 issue a MAIL command.
774 PH/02 In an ACL statement such as
776 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
778 if a client was not listed at all, or was listed with a value other than
779 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
780 the condition was not true (as it should be), so access was not denied.
781 The bug was that the ! inversion was incorrectly passed on to the second
782 item. This has been fixed.
784 PH/03 Added additional dnslists conditions == and =& which are different from
785 = and & when the dns lookup returns more than one IP address.
787 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
788 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
790 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
791 FSYNC, which compiles an option called disable_fsync that allows for
792 bypassing fsync(). The documentation is heavily laced with warnings.
794 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
796 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
797 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
798 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
799 including adding "make clean"; (3) Added -fPIC when compiling the test
800 dynamically loaded module, to get rid of a warning.
802 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
803 message fails, move_frozen_messages = true and ignore_bounce_errors_after
804 = 0s. The bug is otherwise harmless.
806 PH/07 There was a bug in the dovecot authenticator such that the value of
807 $auth1 could be overwritten, and so not correctly preserved, after a
808 successful authentication. This usually meant that the value preserved by
809 the server_setid option was incorrect.
811 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
813 PH/09 Installed PCRE release 7.0.
815 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
816 run for batched SMTP input. It is now run at the start of every message
817 in the batch. While fixing this I discovered that the process information
818 (output by running exiwhat) was not always getting set for -bs and -bS
819 input. This is fixed, and it now also says "batched" for BSMTP.
821 PH/11 Added control=no_pipelining.
823 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
824 patch, slightly modified), and move the expansion of helo_data till after
825 the connection is made in the smtp transport (so it can use these
828 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
830 PH/14 Added log_selector = +pid.
832 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
834 PH/16 Add ${if forany and ${if forall.
836 PH/17 Added dsn_from option to vary the From: line in DSNs.
838 PH/18 Flush SMTP output before performing a callout, unless control =
839 no_callout_flush is set.
841 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
842 was true (the default) a successful delivery failed to delete the retry
843 item, thus causing premature timeout of the address. The bug is now
846 PH/20 Added hosts_avoid_pipelining to the smtp transport.
848 PH/21 Long custom messages for fakedefer and fakereject are now split up
849 into multiline reponses in the same way that messages for "deny" and
850 other ACL rejections are.
852 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
853 with slight modification.
855 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
856 draft, changing the syntax and factoring some duplicate code".
858 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
859 for deliveries of the second and subsequent messages over the same SMTP
862 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
863 ${reduce, with only minor "tidies".
865 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
867 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
868 expansion side effects.
870 PH/27 When a message times out after an over-quota error from an Exim-imposed
871 quota, the bounce message says "mailbox is full". This message was not
872 being given when it was a system quota that was exceeded. It now should
875 MH/03 Made $recipients available in local_scan(). local_scan() already has
876 better access to the recipient list through recipients_list[], but
877 $recipients can be useful in postmaster-provided expansion strings.
879 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
880 in the case of a MAIL command with additional options following the
881 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
882 were accidentally chopped off.
884 PH/29 SMTP synchronization checks are implemented when a command is read -
885 there is a check that no more input is waiting when there shouldn't be
886 any. However, for some commands, a delay in an ACL can mean that it is
887 some time before the response is written. In this time, more input might
888 arrive, invalidly. So now there are extra checks after an ACL has run for
889 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
890 pipelining has not been advertised.
892 PH/30 MH's patch to allow iscntrl() characters to be list separators.
894 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
895 returned in the SMTP response when smtp_return_error_details was false.
898 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
899 instead of the C I/O that was originally supplied, because problems were
902 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
903 Exim which did not show up earlier: it was assuming that a call to
904 SSL_CTX_set_info_callback() might give an error value. In fact, there is
905 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
906 was a macro that became an assignment, so it seemed to work. This has
907 changed to a proper function call with a void return, hence the compile
908 error. Exim's code has been fixed.
910 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
913 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
915 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
917 PH/37 If a message is not accepted after it has had an id assigned (e.g.
918 because it turns out to be too big or there is a timeout) there is no
919 "Completed" line in the log. When some messages of this type were
920 selected by exigrep, they were listed as "not completed". Others were
921 picked up by some special patterns. I have improved the selection
922 criteria to be more general.
924 PH/38 The host_find_failed option in the manualroute router can now be set
925 to "ignore", to completely ignore a host whose IP address cannot be
926 found. If all hosts are ignored, the behaviour is controlled by the new
927 host_all_ignored option.
929 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
930 homing or because of multiple MX records with /mx) generated more than
931 one IP address, and the following item turned out to be the local host,
932 all the secondary addresses of the first item were incorrectly removed
933 from the list, along with the local host and any following hosts (which
934 is what is supposed to happen).
936 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
937 whoever called Exim into the -H file. In the case of the daemon it was
938 behaving confusingly. When first started, it used values for whoever
939 started the daemon, but after a SIGHUP it used the Exim user (because it
940 calls itself on a restart). I have changed the code so that it now always
943 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
944 message are rejected with the same error (e.g. no authentication or bad
945 sender address), and a DATA command is nevertheless sent (as can happen
946 with PIPELINING or a stupid MUA), the error message that was given to the
947 RCPT commands is included in the rejection of the DATA command. This is
948 intended to be helpful for MUAs that show only the final error to their
951 PH/42 Another patch from the Sieve maintainer.
953 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
954 Eximstats - Fixed some broken HTML links and added missing column headers
956 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
957 columns for Rejects, Temp Rejects, Ham, and Spam rows.
959 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
961 PH/43 Yet another patch from the Sieve maintainer.
963 PH/44 I found a way to check for a TCP/IP connection going away before sending
964 the response to the final '.' that terminates a message, but only in the
965 case where the client has not sent further data following the '.'
966 (unfortunately, this is allowed). However, in many cases there won't be
967 any further data because there won't be any more messages to send. A call
968 to select() can be used: if it shows that the input is "ready", there is
969 either input waiting, or the socket has been closed. An attempt to read
970 the next input character can distinguish the two cases. Previously, Exim
971 would have sent an OK response which the client would never have see.
972 This could lead to message repetition. This fix should cure that, at
973 least in a lot of common cases.
975 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
976 advertised in response to EHLO.
982 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
983 fixed by 4.65/MH/01 (is this a record?) are fixed:
985 (i) An empty string was always treated as zero by the numeric comparison
986 operators. This behaviour has been restored.
988 (ii) It is documented that the numeric comparison operators always treat
989 their arguments as decimal numbers. This was broken in that numbers
990 starting with 0 were being interpreted as octal.
992 While fixing these problems I realized that there was another issue that
993 hadn't been noticed. Values of message_size_limit (both the global option
994 and the transport option) were treated as octal if they started with 0.
995 The documentation was vague. These values are now always treated as
996 decimal, and I will make that clear in the documentation.
1002 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1003 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1006 MH/01 Don't check that the operands of numeric comparison operators are
1007 integers when their expansion is in "skipping" mode (fixes bug
1008 introduced by 4.64-PH/07).
1010 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1011 child addresses, Exim now panics and dies. Previously, because the count
1012 is held in a short int, deliveries were likely to be lost. As such a
1013 large number of recipients for a single message is ridiculous
1014 (performance will be very, very poor), I have chosen to impose a limit
1015 rather than extend the field.
1021 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1022 leftover -K file (the existence of which was triggered by #402).
1023 While we were at it, introduced process PID as part of the -K
1024 filename. This should rule out race conditions when creating
1027 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1028 processing considerably. Previous code took too long for large mails,
1029 triggering a timeout which in turn triggers #401.
1031 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1032 in the DK code in transports.c. sendfile() is not really portable,
1033 hence the _LINUX specificness.
1035 TF/01 In the add_headers option to the mail command in an Exim filter,
1036 there was a bug that Exim would claim a syntax error in any
1037 header after the first one which had an odd number of characters
1040 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1041 callout verification, Exim cached a "reject" for the entire domain. This
1042 is correct for most verifications, but it is not correct for a recipient
1043 verification with use_sender or use_postmaster set, because in that case
1044 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1045 case of MAIL FROM:<> rejection from other early rejections (e.g.
1046 rejection of HELO). When verifying a recipient using a non-null MAIL
1047 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1048 Whatever the result of the callout, the value of the domain cache is
1049 left unchanged (for any other kind of callout, getting as far as trying
1050 RCPT means that the domain itself is ok).
1052 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1055 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1056 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1057 session, the daemon ignores ECONNECT errors and logs others; it now
1058 ignores EPIPE as well.
1060 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1061 (quoted-printable decoding).
1063 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1064 later the small subsequent patch to fix an introduced bug.
1066 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1068 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1070 PH/08 An error is now given if message_size_limit is specified negative.
1072 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1073 to be given (somewhat) arbitrary names.
1075 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1078 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1079 miscellaneous code fixes
1081 PH/10 Added the log_reject_target ACL modifier to specify where to log
1084 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1085 hostname. This is wrong, because it relates to the incoming message (and
1086 probably the interface on which it is arriving) and not to the outgoing
1087 callout (which could be using a different interface). This has been
1088 changed to use the value of the helo_data option from the smtp transport
1089 instead - this is what is used when a message is actually being sent. If
1090 there is no remote transport (possible with a router that sets up host
1091 addresses), $smtp_active_hostname is used.
1093 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1094 tweaks were necessary in order to get it to work (see also 21 below):
1095 (a) The code assumed that strncpy() returns a negative number on buffer
1096 overflow, which isn't the case. Replaced with Exim's string_format()
1098 (b) There were several signed/unsigned issues. I just did the minimum
1099 hacking in of casts. There is scope for a larger refactoring.
1100 (c) The code used strcasecmp() which is not a standard C function.
1101 Replaced with Exim's strcmpic() function.
1102 (d) The code set only $1; it now sets $auth1 as well.
1103 (e) A simple test gave the error "authentication client didn't specify
1104 service in request". It would seem that Dovecot has changed its
1105 interface. Fortunately there's a specification; I followed it and
1106 changed what the client sends and it appears to be working now.
1108 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1111 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1112 address A is aliased to B and C, where B exists and C does not. Without
1113 -v the output is "A verified" because verification stops after a
1114 successful redirection if more than one address is generated. However,
1115 with -v the child addresses are also verified. Exim was outputting "A
1116 failed to verify" and then showing the successful verification for C,
1117 with its parentage. It now outputs "B failed to verify", showing B's
1118 parentage before showing the successful verification of C.
1120 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1121 look up a TXT record in a specific list after matching in a combined
1124 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1125 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1126 they consult the DNS. I had assumed they would set it the way they
1127 wanted; and indeed my experiments on Linux seem to show that in some
1128 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1129 To be on the safe side, however, I have now made the interface to
1130 host_find_byname() similar to host_find_bydns(), with an argument
1131 containing the DNS resolver options. The host_find_byname() function now
1132 sets these options at its start, just as host_find_bydns() does. The smtp
1133 transport options dns_qualify_single and dns_search_parents are passed to
1134 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1135 of host_find_byname() use the default settings of RES_DEFNAMES
1136 (qualify_single) but not RES_DNSRCH (search_parents).
1138 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1139 spool_read_header() do less string testing, by means of a preliminary
1140 switch on the second character of optional "-foo" lines. (This is
1141 overdue, caused by the large number of possibilities that now exist.
1142 Originally there were few.) While I was there, I also converted the
1143 str(n)cmp tests so they don't re-test the leading "-" and the first
1144 character, in the hope this might squeeze out yet more improvement.
1146 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1147 flag allowing group syntax was set by the header_syntax check but not
1148 turned off, possible causing trouble later; (2) The flag was not being
1149 set at all for the header_verify test, causing "group"-style headers to
1150 be rejected. I have now set it in this case, and also caused header_
1151 verify to ignore an empty address taken from a group. While doing this, I
1152 came across some other cases where the code for allowing group syntax
1153 while scanning a header line wasn't quite right (mostly, not resetting
1154 the flag correctly in the right place). These bugs could have caused
1155 trouble for malformed header lines. I hope it is now all correct.
1157 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1158 with the "reply" argument non-NULL. The code, however (which originally
1159 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1160 but it didn't always do it. This confused somebody who was copying the
1161 code for some other use. I have removed all the tests.
1163 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1164 feature that was used to support insecure browsers during the U.S. crypto
1165 embargo. It requires special client support, and Exim is probably the
1166 only MTA that supported it -- and would never use it because real RSA is
1167 always available. This code has been removed, because it had the bad
1168 effect of slowing Exim down by computing (never used) parameters for the
1169 RSA_EXPORT functionality.
1171 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1172 authenticator to fail if there's a tab character in the incoming data
1173 (there should never be unless someone is messing about, as it's supposed
1174 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1175 if the connection is using TLS or if the remote IP is the same as the
1176 local IP, and the "valid-client-cert option" if a client certificate has
1179 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1180 authenticators. This can be used for authorization after authentication
1181 succeeds. (In the case of plaintext, it servers for both authentication
1184 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1185 if any retry times were supplied.
1187 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1188 connection, where there is no client IP address to check. In this
1189 situation, the verify now always succeeds.
1191 PH/25 Applied John Jetmore's -Mset patch.
1193 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1195 PH/27 In a string expansion for a processed (not raw) header when multiple
1196 headers of the same name were present, leading whitespace was being
1197 removed from all of them, but trailing whitespace was being removed only
1198 from the last one. Now trailing whitespace is removed from each header
1199 before concatenation. Completely empty headers in a concatenation (as
1200 before) are ignored.
1202 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1203 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1205 PH/29 [Removed. This was a change that I later backed out, and forgot to
1206 correct the ChangeLog entry (that I had efficiently created) before
1207 committing the later change.]
1209 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1210 address errors (4xx response to RCPT) over the same connection as other
1211 messages routed to the same hosts. Such deliveries are always "forced",
1212 so retry times are not inspected. This resulted in far too many retries
1213 for the affected addresses. The effect occurred only when there were more
1214 hosts than the hosts_max_try setting in the smtp transport when it had
1215 the 4xx errors. Those hosts that it had tried were not added to the list
1216 of hosts for which the message was waiting, so if all were tried, there
1217 was no problem. Two fixes have been applied:
1219 (i) If there are any address or message errors in an SMTP delivery, none
1220 of the hosts (tried or untried) are now added to the list of hosts
1221 for which the message is waiting, so the message should not be a
1222 candidate for sending over the same connection that was used for a
1223 successful delivery of some other message. This seems entirely
1224 reasonable: after all the message is NOT "waiting for some host".
1225 This is so "obvious" that I'm not sure why it wasn't done
1226 previously. Hope I haven't missed anything, but it can't do any
1227 harm, as the worst effect is to miss an optimization.
1229 (ii) If, despite (i), such a delivery is accidentally attempted, the
1230 routing retry time is respected, so at least it doesn't keep
1231 hammering the server.
1233 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1234 in ${readsocket because some servers need this prod.
1236 PH/32 Added some extra debug output when updating a wait-xxx database.
1238 PH/33 The hint "could be header name not terminated by colon", which has been
1239 given for certain expansion errors for a long time, was not being given
1240 for the ${if def:h_colon_omitted{... case.
1242 PH/34 The spec says: "With one important exception, whenever a domain list is
1243 being scanned, $domain contains the subject domain." There was at least
1244 one case where this was not true.
1246 PH/35 The error "getsockname() failed: connection reset by peer" was being
1247 written to the panic log as well as the main log, but it isn't really
1248 panic-worthy as it just means the connection died rather early on. I have
1249 removed the panic log writing for the ECONNRESET error when getsockname()
1252 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1253 runs only) independently of the message's sender address. This meant
1254 that, if the 4xx error was in fact related to the sender, a different
1255 message to the same recipient with a different sender could confuse
1256 things. In particualar, this can happen when sending to a greylisting
1257 server, but other circumstances could also provoke similar problems.
1258 I have changed the default so that the retry time for these errors is now
1259 based a combination of the sender and recipient addresses. This change
1260 can be overridden by setting address_retry_include_sender=false in the
1263 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1264 remote server are returned as part of bounce messages. This was not
1265 happening for LMTP over a pipe (the lmtp transport), but now it is the
1266 same for both kinds of LMTP.
1268 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1269 in header lines that were in fact CNAMEs. This is no longer the case.
1271 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1272 and queue runs started by the daemon processed all messages. This has
1273 been fixed so that -R and -S can now usefully be given with -q<time>.
1275 PH/40 Import PCRE release 6.7 (fixes some bugs).
1277 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1279 PH/42 Give an error if -q is specified more than once.
1281 PH/43 Renamed the variables $interface_address and $interface_port as
1282 $received_ip_address and $received_port, to make it clear that these
1283 values apply to message reception, and not to the outgoing interface when
1284 a message is delivered. (The old names remain recognized, of course.)
1286 PH/44 There was no timeout on the connect() call when using a Unix domain
1287 socket in the ${readsocket expansion. There now is.
1289 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1290 be meaningful with "accept".
1292 SC/01 Eximstats V1.43
1293 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1295 SC/02 Eximstats V1.44
1296 Use a glob alias rather than an array ref in the generated
1297 parser. This improves both readability and performance.
1299 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1300 Collect SpamAssassin and rejection statistics.
1301 Don't display local sender or destination tables unless
1302 there is data to show.
1303 Added average volumes into the top table text output.
1305 SC/04 Eximstats V1.46
1306 Collect data on the number of addresses (recipients)
1307 as well as the number of messages.
1309 SC/05 Eximstats V1.47
1310 Added 'Message too big' to the list of mail rejection
1311 reasons (thanks to Marco Gaiarin).
1313 SC/06 Eximstats V1.48
1314 Mainlog lines which have GMT offsets and are too short to
1315 have a flag are now skipped.
1317 SC/07 Eximstats V1.49 (Alain Williams)
1318 Added the -emptyok flag.
1320 SC/08 Eximstats V1.50
1321 Fixes for obtaining the IP address from reject messages.
1323 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1324 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1325 whitesspace changes from 4.64-PH/27
1327 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1330 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1331 are found, allow negative numbers in numeric criteria)
1333 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1335 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1336 to match changes made in 4.64-PH/43
1338 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1340 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1341 "domain" is actually an IP address, and force a failure. This locks out
1342 those revolvers/nameservers that support "A-for-A" lookups, in
1343 contravention of the specifications.
1345 PH/48 When a host name was looked up from an IP address, and the subsequent
1346 forward lookup of the name timed out, the host name was left in
1347 $sender_host_name, contrary to the specification.
1349 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1350 restricted to single-key lookups, Exim was not diagnosing an error if
1351 * or *@ was used with a query-style lookup.
1353 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1355 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1356 long ago, but noone interested enough thought of it. Let's just say that
1357 the "1.1" means that there are some new functions that weren't there at
1358 some point in the past.
1360 PH/51 Error processing for expansion failure of helo_data from an smtp
1361 transport during callout processing was broken.
1363 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1364 tested/used via the -bh/-bhc/-bs options.
1366 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1367 bug, fixed in subsequent PCRE releases).
1369 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1370 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1372 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1378 SC/01 Use a glob alias rather than an array ref in eximstats generated
1379 parser. This improves both readability and performance.
1381 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1382 Don't display local sender or destination tables in eximstats unless
1383 there is data to show.
1384 Added average volumes into the eximstats top table text output.
1386 SC/03 Collect data on the number of addresses (recipients) as well
1387 as the number of messages in eximstats.
1389 TF/01 Correct an error in the documentation for the redirect router. Exim
1390 does (usually) call initgroups() when daemonizing.
1392 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1393 with consistent privilege compared to when running as a daemon.
1395 TF/03 Note in the spec that $authenticated_id is not set for local
1396 submissions from trusted users.
1398 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1399 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1401 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1402 by adding some example configuration directives to the default
1403 configuration file. A little bit of work is required to uncomment the
1404 directives and define how usernames and passwords are checked, but
1405 there is now a framework to start from.
1407 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1408 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1409 without this. I don't know how relevant this is to other LDAP libraries.
1411 PH/02 Add the verb name to the "unknown ACL verb" error.
1413 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1415 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1417 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1418 directory not expanded when it should be if an expanded home directory
1419 was set for the address (which is overridden by the transport).
1421 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1424 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1425 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1426 because it is too late at that time, and has no effect.
1428 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1429 security issue with \' (bugzilla #107). I could not use the
1430 PQescapeStringConn() function, because it needs a PGconn value as one of
1433 PH/08 When testing addresses using -bt, indicate those final addresses that
1434 are duplicates that would not cause an additional delivery. At least one
1435 person was confused, thinking that -bt output corresponded to deliveries.
1436 (Suppressing duplicates isn't a good idea as you lose the information
1437 about possibly different redirections that led to the duplicates.)
1439 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1440 systems where poll() doesn't work, in particular OS X.
1442 PH/10 Added more information to debugging output for retry time not reached.
1444 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1445 operations in malware.c.
1447 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1450 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1451 syslog_duplication set false, log lines that would normally be written
1452 both the the main log and to the reject log were not written to syslog at
1455 PH/14 In the default configuration, change the use of "message" in ACL warn
1456 statements to "add_header".
1458 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1459 not followed by a command (e.g. "seen endif").
1461 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1462 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1465 PH/17 Added extra conditions to the default value of delay_warning_condition
1469 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1470 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1471 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1474 The Auto-Submitted: and various List- headers are standardised, whereas I
1475 don't think Precedence: ever was.
1477 PH/18 Refactored debugging code in route_finduser() to show more information,
1478 in particular, the error code if getpwnam() issues one.
1480 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1481 This is apparently needed in addition to the PH/07 change above to avoid
1482 any possible encoding problems.
1484 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1485 but not after initializing Perl.
1487 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1488 output them only if debugging. By default they are written stderr,
1489 apparently, which is not desirable.
1491 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1494 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1497 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1499 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1500 authenticated or an ident call has been made. Suppress the default
1501 values for $authenticated_id and $authenticated_sender (but permit -oMai
1502 and -oMas) when testing with -bh.
1504 PH/24 Re-jigged the order of the tests in the default configuration so that the
1505 tests for valid domains and recipients precede the DNS black list and CSA
1506 tests, on the grounds that those ones are more expensive.
1508 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1509 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1510 This bug exists in every version of Exim that I still have, right back to
1513 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1514 However, an attempt to turn on case-sensitivity in a regex key by
1515 including (?-i) didn't work because the subject string was already
1516 lowercased, and the effects were non-intuitive. It turns out that a
1517 one-line patch can be used to allow (?-i) to work as expected.
1523 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1524 other effects) broke the use of negated acl sub-conditions.
1526 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1529 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1530 "Deny" causes Exim to reject the incoming connection with a 554 error.
1531 Unfortunately, if there is a major crisis, such as a disk failure,
1532 tcp-wrappers gives "deny", whereas what one would like would be some
1533 kind of temporary error. A kludge has been added to help with this.
1534 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1535 554 error is used if errno is still zero or contains ENOENT (which occurs
1536 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1539 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1541 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1542 errors. Otherwise a message that provokes a temporary error (when other
1543 messages do not) can cause a whole host to time out.
1545 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1546 addresses were routed directly to files or pipes from a redirect router.
1547 File deliveries just didn't batch; pipe deliveries might have suffered
1550 PH/06 A failure to get a lock for a hints database would erroneously always say
1551 "Failed to get write lock", even when it was really a read lock.
1553 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1554 of 0600. This has been changed to use the value of the lockfile_mode
1555 option (which defaults to 0600).
1557 PH/08 Applied small patch from the Sieve maintainer.
1559 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1560 folder from quota calculations, a direct delivery into this folder messed
1561 up the contents of the maildirsize file. This was because the regex was
1562 used only to exclude .Trash (or whatever) when the size of the mailbox
1563 was calculated. There was no check that a delivery was happening into an
1564 excluded directory. This bug has been fixed by ignoring all quota
1565 processing for deliveries into excluded directories.
1567 PH/10 Added the maildirfolder_create_regex option to appendfile.
1573 PH/01 The code for finding all the local interface addresses on a FreeBSD
1574 system running IPv6 was broken. This may well have applied to all BSD
1575 systems, as well as to others that have similar system calls. The broken
1576 code found IPv4 interfaces correctly, but gave incorrect values for the
1577 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1578 that it would not match correctly against @[] and not recognize the IPv6
1581 PH/02 The ipliteral router was not recognizing addresses of the form user@
1582 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1584 PH/03 Added disable_ipv6.
1586 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1587 decoded form, because it is most often used to construct To: headers
1588 lines in autoreplies, and the decoded form may well be syntactically
1589 invalid. However, $reply_address has leading white space removed, and all
1590 newlines turned into spaces so that the autoreply transport does not
1593 PH/05 If group was specified without a user on a router, and no group or user
1594 was specified on a transport, the group from the router was ignored.
1596 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1597 for visible compile-time settings that can be used to change these
1598 numbers, for those that want even more. Backwards compatibility with old
1599 spool files has been maintained. However, going back to a previous Exim
1600 release will lost any variables that are in spool files.
1602 PH/07 Two small changes when running in the test harness: increase delay when
1603 passing a TCP/IP connection to a new process, in case the original
1604 process has to generate a bounce, and remove special handling of
1605 127.0.0.2 (sic), which is no longer necessary.
1607 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1608 be the same on different OS.
1610 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1613 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1614 whether --show-vars was specified or not
1616 JJ/02 exipick: Added support for new ACL variable spool format introduced
1619 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1620 syntactically invalid From: or Reply-to: line, and a filter used this to
1621 generate an autoreply, and therefore failed to obtain an address for the
1622 autoreply, Exim could try to deliver to a non-existent relative file
1623 name, causing unrelated and misleading errors. What now happens is that
1624 it logs this as a hard delivery error, but does not attempt to create a
1627 PH/11 The exinext utility has a -C option for testing purposes, but although
1628 the given file was scanned by exinext itself; it wasn't being passed on
1629 when Exim was called.
1631 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1632 an end-of-file indication when reading a command response.
1634 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1635 compiled. In many other places in Exim, IPv6 addresses are always
1636 recognized, so I have changed this. It also means that IPv4 domain
1637 literals of the form [IPV4:n.n.n.n] are now always recognized.
1639 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1640 used if the router is not running as root, for example, when verifying at
1641 ACL time, or when using -bh. The debugging output from this situation was
1642 non-existent - all you got was a failure to exec. I have made two
1645 (a) Failures to set uid/gid, the current directory, or a process leader
1646 in a subprocess such as that created by queryprogram now generate
1647 suitable debugging ouput when -d is set.
1649 (b) The queryprogram router detects when it is not running as root,
1650 outputs suitable debugging information if -d is set, and then runs
1651 the subprocess without attempting to change uid/gid.
1653 PH/15 Minor change to Makefile for building test_host (undocumented testing
1656 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1657 additional section of a DNS packet that returns MX or SRV records.
1658 Instead, it always explicitly searches for A/AAAA records. This avoids
1659 major problems that occur when a DNS server includes only records of one
1660 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1661 fixed another bug: if SRV records were looked up and the corresponding
1662 address records were *not* found in the additional section, the port
1663 values from the SRV records were lost.
1665 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1666 using the correct key (the original address) when searching the retry
1667 rules in order to find which one to use for generating the retry hint.
1669 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1670 adding the default one. Similarly, if it contains a Reply-To: header, the
1671 errors_reply_to option, if set, is not used.
1673 PH/19 When calculating a retry time, Exim used to measure the "time since
1674 failure" by looking at the "first failed" field in the retry record. Now
1675 it does not use this if it is later than than the arrival time of the
1676 message. Instead it uses the arrival time. This makes for better
1677 behaviour in cases where some deliveries succeed, thus re-setting the
1678 "first failed" field. An example is a quota failure for a huge message
1679 when small messages continue to be delivered. Without this change, the
1680 "time since failure" will always be short, possible causing more frequent
1681 delivery attempts for the huge message than are intended.
1682 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1684 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1685 $1, $2, $3) because the numerical variables can be reset during some
1686 expansion items (e.g. "match"), thereby losing the authentication data.
1688 PH/21 Make -bV show the size of off_t variables so that the test suite can
1689 decide whether to run tests for quotas > 2G.
1691 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1692 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1693 filecount value is greater than 2G or if a quota value is greater than 2G
1694 on a system where the size of off_t is not greater than 4, a panic error
1697 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1698 never match. The debug and -bh output now contains an explicit error
1699 message indicating a malformed IPv4 address or mask.
1701 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1702 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1703 PH/23 above applies.
1705 PH/25 Do not write to syslog when running in the test harness. The only
1706 occasion when this arises is a failure to open the main or panic logs
1707 (for which there is an explicit test).
1709 PH/26 Added the /no_tell option to "control=freeze".
1711 PH/27 If a host name lookup failed very early in a connection, for example, if
1712 the IP address matched host_lookup and the reverse lookup yielded a name
1713 that did not have a forward lookup, an error message of the form "no IP
1714 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1715 could be logged. Now it outputs the IP address instead of "NULL".
1717 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1718 allows the sender and the authenticated sender to be set when
1719 submitting a message from within Exim. Since child_open_exim() is
1720 documented for local_scan(), the new function should be too.
1722 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1723 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1724 results in an empty string is now treated as unset.
1726 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1728 PH/31 Added sender_verify_fail logging option.
1730 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1731 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1732 tidied the source and removed it altogether.
1734 PH/33 When a queue run was abandoned because the load average was too high, a
1735 log line was always written; now it is written only if the queue_run log
1736 selector is set. In addition, the log line for abandonment now contains
1737 information about the queue run such as the pid. This is always present
1738 in "start" and "stop" lines but was omitted from the "abandon" line.
1740 PH/34 Omit spaces between a header name and the colon in the error message that
1741 is given when verify = headers_syntax fails (if there are lots of them,
1742 the message gets confusing).
1744 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1745 names, as there are now some PTR records that contain slashes. This check
1746 is only to protect against broken name servers that fall over on strange
1747 characters, so the fact that it applies to all lookups doesn't matter.
1749 PH/36 Now that the new test suite is complete, we can remove some of the
1750 special code in Exim that was needed for the old test suite. For example,
1751 sorting DNS records because real resolvers return them in an arbitrary
1752 order. The new test suite's fake resolver always returns records in the
1755 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1756 bounces) except when queue_only is set, to avoid logging races between
1757 the different processes.
1759 PH/38 Panic-die if .include specifies a non-absolute path.
1761 PH/39 A tweak to the "H" retry rule from its user.
1763 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1764 a label. They prevented compilation on older perls.
1766 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1767 a warning to be raised on newish perls.
1769 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1770 on queue. Changes to match documented behaviour of showing count of
1771 messages matching specified criteria.
1773 PH/40 Changed the default ident timeout from 30s to 5s.
1775 PH/41 Added support for the use of login_cap features, on those BSD systems
1776 that have them, for controlling the resources used by pipe deliveries.
1778 PH/42 The content-scanning code uses fopen() to create files in which to put
1779 message data. Previously it was not paying any attention to the mode of
1780 the files. Exim runs with umask(0) because the rest of the code creates
1781 files with open(), and sets the required mode explicitly. Thus, these
1782 files were ending up world-writeable. This was not a big issue, because,
1783 being within the spool directory, they were not world-accessible. I have
1784 created a function called modefopen, which takes an additional mode
1785 argument. It sets umask(777), creates the file, chmods it to the required
1786 mode, then resets the umask. All the relevant calls to fopen() in the
1787 content scanning code have been changed to use this function.
1789 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1790 to 24 hours. This avoids potential overflow problems when processing G
1791 and H retry rules. I suspect nobody ever tinkers with this value.
1793 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1795 PH/45 When the plaintext authenticator is running as a client, the server's
1796 challenges are checked to ensure they are valid base64 strings. By
1797 default, the authentication attempt is cancelled if an invalid string is
1798 received. Setting client_ignore_invalid_base64 true ignores these errors.
1799 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1800 they are received. Thus, the responses can be made to depend on the
1801 challenges. If an invalid string is ignored, an empty string is placed in
1804 PH/46 Messages that are created by the autoreply transport now contains a
1805 References: header, in accordance with RFCs 2822 and 3834.
1807 PH/47 Added authenticated_sender_force to the smtp transport.
1809 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1811 PH/49 Installed latest patch from the Sieve maintainer.
1813 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1814 was also set, the appendfile transport was unnecessarily scanning a
1815 directory of message files (e.g. for maildir delivery) to find the count
1816 of files (along with the size), even though it did not need this
1817 information. It now does the scan only if it needs to find either the
1818 size of the count of files.
1820 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1822 PH/52 Two bugs concerned with error handling when the smtp transport is
1825 (i) Exim was not creating retry information for temporary errors given
1826 for individual recipients after the DATA command when the smtp transport
1827 was used in LMTP mode. This meant that they could be retried too
1828 frequently, and not timed out correctly.
1830 (ii) Exim was setting the flag that allows error details to be returned
1831 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1832 recipients that were returned after the DATA command.
1834 PH/53 This is related to PH/52, but is more general: for any failing address,
1835 when detailed error information was permitted to be returned to the
1836 sender, but the error was temporary, then after the final timeout, only
1837 "retry timeout exceeded" was returned. Now it returns the full error as
1838 well as "retry timeout exceeded".
1840 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1841 do this, and (what is worse) MTAs that accept it.
1843 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1844 will now be deprecated.
1846 PH/56 New os.c-cygwin from the Cygwin maintainer.
1848 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1849 formats (previously only available in exim formats via -bpr, -bpru,
1850 and -bpra. Now also available in native and exiqgrep formats)
1852 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1853 with very large, slow to parse queues
1855 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1857 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1859 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1860 responses to authentication challenges, though it was showing the
1861 challenges; (ii) I've removed the CR characters from the debug output for
1864 PH/58 Allow for the insertion of a newline as well as a space when a string
1865 is turned into more than one encoded-word during RFC 2047 encoding. The
1866 Sieve code now uses this.
1868 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1869 data_4xx, lost_connection, tls_required.
1871 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1872 message was being sent as an SMTP response.
1874 PH/61 Add -l and -k options to exicyclog.
1876 PH/62 When verifying, if an address was redirected to one new address, so that
1877 verification continued, and the new address failed or deferred after
1878 having set something in $address_data, the value of $address_data was not
1879 passed back to the ACL. This was different to the case when no
1880 redirection occurred. The value is now passed back in both cases.
1882 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1883 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1884 use login_cap.h, so on its own it isn't the distinguishing feature. The
1885 new name refers directly to the setclassresources() function.
1887 PH/65 Added configuration files for NetBSD3.
1889 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1891 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1892 is preferred over IPv4.
1894 PH/68 The bounce_return_message and bounce_return_body options were not being
1895 honoured for bounces generated during the reception of non-SMTP messages.
1896 In particular, this applied to messages rejected by the ACL. This bug has
1897 been fixed. However, if bounce_return_message is true and bounce_return_
1898 body is false, the headers that are returned for a non-SMTP message
1899 include only those that have been read before the error was detected.
1900 (In the case of an ACL rejection, they have all been read.)
1902 PH/69 The HTML version of the specification is now built in a directory called
1903 spec_html instead of spec.html, because the latter looks like a path with
1904 a MIME-type, and this confuses some software.
1906 PH/70 Catch two compiler warnings in sieve.c.
1908 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1909 function verify_get_ident() calls ip_connect() to connect a socket, but
1910 if the "connect()" function timed out, ip_connect() used to close the
1911 socket. However, verify_get_ident() also closes the socket later, and in
1912 between Exim writes to the log, which may get opened at this point. When
1913 the socket was closed in ip_connect(), the log could get the same file
1914 descriptor number as the socket. This naturally causes chaos. The fix is
1915 not to close the socket in ip_connect(); the socket should be closed by
1916 the function that creates it. There was only one place in the code where
1917 this was missing, in the iplookup router, which I don't think anybody now
1918 uses, but I've fixed it anyway.
1920 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1921 well as to direct DNS lookups. Otherwise the handling of names in host
1922 lists is inconsistent and therefore confusing.
1928 PH/01 Two changes to the default runtime configuration:
1930 (1) Move the checks for relay_from_hosts and authenticated clients from
1931 after to before the (commented out) DNS black list checks.
1933 (2) Add control=submission to the relay_from_hosts and authenticated
1934 clients checks, on the grounds that messages accepted by these
1935 statements are most likely to be submissions.
1937 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1939 (1) Generate an error if the third argument for the ${prvs expansion is
1942 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1945 (3) Reset the variables that are obtained from the first argument of
1946 ${prvscheck and used in the second argument before leaving the code,
1947 because their memory is reclaimed, so using them afterwards may do
1950 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1951 one (it's much easier than Tom thought :-).
1953 (5) Because of (4), we can now allow for the use of $prvscheck_result
1954 inside the third argument.
1956 PH/03 For some reason, the default setting of PATH when running a command from
1957 a pipe transport was just "/usr/bin". I have changed it to
1960 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1961 anything to be listed in the output from -bV.
1963 PH/05 When a filter generated an autoreply, the entire To: header line was
1964 quoted in the delivery log line, like this:
1966 => >A.N.Other <ano@some.domain> <original@ddress> ...
1968 This has been changed so that it extracts the operative address. There
1969 may be more than one such address. If so, they are comma-separated, like
1972 => >ano@some.domain,ona@other.domain <original@ddress> ...
1974 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1975 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1976 not being looked up in the rDNS to get a host name, Exim was showing the
1977 IP address twice in Received: lines, even though the IP addresses were
1978 identical. For example:
1980 Received: from [1.2.3.4] (helo=[1.2.3.4])
1982 However, if the real host name was known, it was omitting the HELO data
1983 if it matched the actual IP address. This has been tidied up so that it
1984 doesn't show the same IP address twice.
1986 PH/07 When both +timestamp and +memory debugging was on, the value given by
1987 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1988 called by the debug printing, thereby overwriting the timestamp buffer.
1989 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1991 PH/08 When the original message was included in an autoreply transport, it
1992 always said "this is a copy of the message, including all the headers",
1993 even if body_only or headers_only was set. It now gives an appropriate
1996 PH/09 Applied a patch from the Sieve maintainer which:
1998 o fixes some comments
1999 o adds the (disabled) notify extension core
2000 o adds some debug output for the result of if/elsif tests
2001 o points to the current vacation draft in the documentation
2002 and documents the missing references header update
2006 o fixes a bug in processing the envelope test (when testing
2007 multiple envelope elements, the last element determinted the
2010 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2011 Electronic Mail") by including:
2013 Auto-submitted: auto-generated
2015 in the messages that it generates (bounce messages and others, such as
2016 warnings). In the case of bounce messages for non-SMTP mesages, there was
2017 also a typo: it was using "Auto_submitted" (underscore instead of
2018 hyphen). Since every message generated by Exim is necessarily in response
2019 to another message, thes have all been changed to:
2021 Auto-Submitted: auto-replied
2023 in accordance with these statements in the RFC:
2025 The auto-replied keyword:
2027 - SHOULD be used on messages sent in direct response to another
2028 message by an automatic process,
2030 - MUST NOT be used on manually-generated messages,
2032 - MAY be used on Delivery Status Notifications (DSNs) and Message
2033 Disposition Notifications (MDNs),
2035 - MUST NOT be used on messages generated by automatic or periodic
2036 processes, except for messages which are automatic responses to
2039 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2040 to the default Received: header definition.
2042 PH/12 Added log selector acl_warn_skipped (default on).
2044 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2045 variables because (a) they are in the wrong storage pool and (b) even if
2046 they were copied, it wouldn't work properly because of the caching.
2048 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2049 checking when decoding. Apparently there are clients that generate
2050 overlong encoded strings. Why am I not surprised?
2052 PH/15 If the first argument of "${if match_address" was not empty, but did not
2053 contain an "@" character, Exim crashed. Now it writes a panic log message
2054 and treats the condition as false.
2056 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2058 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2059 extension "envelope-auth". The code is finished and in agreement with
2060 other implementations, but there is no documentation so far and in fact,
2061 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2062 not changing the active code.
2064 Print executed "if" and "elsif" statements when debugging is used. This
2065 helps a great deal to understand what a filter does.
2067 Document more things not specified clearly in RFC3028. I had all this
2068 sorted out, when out of a sudden new issues came to my mind. Oops."
2070 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2073 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2074 canonical form (as documented). However, after a host name lookup from
2075 the IP address, check_host() was doing a simple string comparison with
2076 addresses acquired from the DNS when checking that the found name did
2077 have the original IP as one of its addresses. Since any found IPv6
2078 addresses are likely to be in abbreviated form, the comparison could
2079 fail. Luckily, there already exists a function for doing the comparison
2080 by converting both addresses to binary, so now that is used instead of
2081 the text comparison.
2083 PH/20 There was another similar case to PH/19, when a complete host name was
2084 given in a host list; looking up its IP address could give an abbreviated
2085 form, whereas the current host's name might or might not be abbreviated.
2086 The same fix has been applied.
2092 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2093 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2096 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2097 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2099 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2101 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2102 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2103 Also, for "Auto-Submitted", treat anything other than "no" as
2104 non-personal, in accordance with RFC 3834. (Previously it treated
2105 anything starting "auto-" as non-personal.)
2107 TF/01 The control=submission/name=... option had a problem with syntax
2108 errors if the name included a slash character. The /name= option
2109 now slurps the rest of the string, so it can include any characters
2110 but it must come last in the list of options (after /sender_retain
2113 PH/05 Some modifications to the interface to the fake nameserver for the new
2121 TK/01 Added the "success_on_redirect" address verification option. See
2122 NewStuff for rationale and an example.
2124 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2126 PH/02 Patch to exigrep to allow it to work on syslog lines.
2128 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2129 fread() to skip over the body file's header line, because in Cygwin the
2130 header line is locked and is inaccessible.
2132 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2133 co-exist for some time) to make it clear that it is the Exim ID that is
2134 referenced, not the Message-ID: header line.
2136 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2137 string_format() function, because snprintf() does not exist on all
2140 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2141 ${stat: expansion item.
2143 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2144 protocol synchronization error", to keep the pedants happy.
2146 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2147 well as for IRIX systems, when gcc is being used. See the host.c source
2150 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2152 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2155 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2156 transport and to the smtp transport in LMTP mode.
2158 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2160 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2161 run a filter in a subprocess. This could lead to confusion in subsequent
2162 lookups in the parent process. There should also be a search_tidyup() at
2163 the end of the subprocess.
2165 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2166 only if the host matched helo_try_verify_hosts, which caused the
2167 verification to occur when the EHLO/HELO command was issued. The ACL just
2168 tested the remembered result. Now, if a previous verification attempt has
2169 not happened, "verify = helo" does it there and then.
2171 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2173 TK/03 Fix log output including CR from clamd.
2175 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2176 exist provoked a memory error which could cause a segfault.
2178 PH/15 Installed PCRE 6.2
2180 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2182 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2183 of the problem. Specifically, suggested +O2 rather than +O1 for the
2186 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2188 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2189 record, the port was not correctly logged when the outgoing_port log
2190 selector was set (it logged the transort's default port).
2192 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2193 fallback_hosts, and "hosts" in the smtp transport.
2195 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2196 host errors such as "Connection refused".
2198 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2199 authentication with radiusclient 0.4.9:
2201 - Error returned from rc_read_config was caught wrongly
2202 - Username/password not passed on to radius server due to wrong length.
2204 The presumption is that some radiusclient API changes for 4.51/PH/17
2205 were not taken care of correctly. The code is still untested by me (my
2206 Linux distribution still has 0.3.2 of radiusclient), but it was
2207 contributed by a Radius user.
2209 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2210 expanding the "port" option of the smtp transport.
2212 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2213 while reading a MIME header. Thanks to Tom Hughes for a patch.
2215 PH/24 Include config.h inside local_scan.h so that configuration settings are
2218 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2219 that in an ACL for RCPT (for example), you can examine exactly what was
2222 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2223 commands, but it was not correctly comparing the address with the actual
2224 client host address. Thus, it would show the EHLO address in Received:
2225 header lines when this was not necessary.
2227 PH/27 Added the % operator to ${eval:}.
2229 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2230 it should be ignoring failures (because with -C, for example, it has lost
2231 privilege). It wasn't ignoring creation failures other than "already
2234 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2237 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2238 because an input line was too long, either on its own, or by virtue of
2239 too many continuations, the temporary file was not being removed, and the
2240 return code was incorrect.
2242 PH/31 Missing "BOOL" in function definition in filtertest.c.
2244 PH/32 Applied Sieve patches from the maintainer.
2246 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2248 PH/33 Added "verify = not_blind".
2250 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2251 Local/Makefile (with some defaults set). These are used in built scripts
2252 such as exicyclog, but they have never been used in the exim_install
2253 script (though there are many overriding facilities there). I have
2254 arranged that the exim_install script now takes note of these two
2257 PH/35 Installed configuration files for Dragonfly.
2259 PH/36 When a locally submitted message by a trusted user did not contain a
2260 From: header, and the sender address was obtained from -f or from an SMTP
2261 MAIL command, and the trusted user did not use -F to supply a sender
2262 name, $originator_name was incorrectly used when constructing a From:
2263 header. Furthermore, $originator_name was used for submission mode
2264 messages from external hosts without From: headers in a similar way,
2265 which is clearly wrong.
2267 PH/37 Added control=suppress_local_fixups.
2269 PH/38 When log_selector = +received_sender was set, and the addition of the
2270 sender made the log line's construction buffer exactly full, or one byte
2271 less than full, an overflow happened when the terminating "\n" was
2274 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2275 when the result of a list match is failure because a DNS lookup failed.
2277 PH/40 RM_COMMAND is now used in the building process.
2279 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2280 the "build-* directories that it finds.
2282 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2283 address in a domain literal was a prefix of an interface address.
2285 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2286 when verifying a sender address, unless rewrite_headers is false.
2288 PH/44 Wrote a long comment about why errors_to addresses are verified as
2289 recipients, not senders.
2291 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2292 the ratelimit ACL was added.
2294 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2296 PH/46 Added extra information about PostgreSQL errors to the error string.
2298 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2299 avoiding the need to install special zones in a real server. This is
2300 backwards compatible; if it can't find the fake resolver, it drops back.
2301 Thus, both old and new test suites can be run.
2303 TF/02 Added util/ratelimit.pl
2305 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2306 clock is set back in time.
2308 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2309 Candler <B.Candler@pobox.com>.
2311 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2312 OK for addresses that are the result of redirections.
2314 PH/48 A number of further additions for the benefit of the new test suite,
2315 including a fake gethostbyname() that interfaces to the fake DNS resolver
2318 TF/06 The fix for widen_domains has also been applied to qualify_single and
2319 search_parents which are the other dnslookup options that can cause
2322 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2325 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2327 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2328 service" if it encountered an MX record with an empty target hostname.
2329 The message is now "an MX or SRV record indicated no SMTP service".
2331 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2332 if the DNS of the sending site is misconfigured. This is quite a
2333 common situation. This change restores the behaviour of treating a
2334 helo verification defer as a failure.
2336 PH/51 If self=fail was set on a router, the bounce message did not include the
2337 actual error message.
2343 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2345 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2346 command itself ended in error, the underlying message about the transport
2347 filter timeout was being overwritten with the pipe command error. Now the
2348 underlying error message should be appended to the second error message.
2350 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2352 PH/02 Reduce the amount of output that "make" produces by default. Full output
2353 can still be requested.
2355 PH/03 The warning log line about a condition test deferring for a "warn" verb
2356 was being output only once per connection, rather than after each
2357 occurrence (because it was using the same function as for successful
2358 "warn" verbs). This seems wrong, so I have changed it.
2360 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2361 it should not have, which might have caused a crash in the right
2362 circumstances, but probably never did.
2364 PH/04 Installed a modified version of Tony Finch's patch to make submission
2365 mode fix the return path as well as the Sender: header line, and to
2366 add a /name= option so that you can make the user's friendly name appear
2369 TF/03 Added the control = fakedefer ACL modifier.
2371 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2372 Mark Lowes for thorough testing.
2374 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2376 TK/03 Merged latest SRS patch from Miles Wilton.
2378 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2379 with the definition in sysexits.h (which is #included earlier).
2380 Fortunately, Exim does not actually use EX_OK. The code used to try to
2381 preserve the sysexits.h value, by assumimg that macro definitions were
2382 scanned for macro replacements. I have been disabused of this notion,
2383 so now the code just undefines EX_OK before #including unistd.h.
2385 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2386 in the smtp transport. When a block could not be written in a single
2387 write() function, the timeout was being re-applied to each part-write.
2388 This seems wrong - if the receiver was accepting one byte at a time it
2389 would take for ever. The timeout is now adjusted when this happens. It
2390 doesn't have to be particularly precise.
2392 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2393 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2395 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2396 without a domain if the check to <postmaster@domain> fails.
2398 SC/01 Eximstats: added -xls and the ability to specify output files
2399 (patch written by Frank Heydlauf).
2401 SC/02 Eximstats: use FileHandles for outputing results.
2403 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2405 SC/04 Eximstats: fixed display of large numbers with -nvr option
2407 SC/05 Eximstats: fixed merging of reports with empty tables.
2409 SC/06 Eximstats: added the -include_original_destination flag
2411 SC/07 Eximstats: removed tabs and trailing whitespace.
2413 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2415 TK/06 MBOX spool code: Add real "From " MBOX separator line
2416 so the .eml file is really in mbox format (even though
2417 most programs do not really care). Patch from Alex Miller.
2419 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2420 The latter is generated from $received_to and is only set if the
2421 message has one envelope recipient. SA can use these headers,
2422 obviously out-of-the-box. Patch from Alex Miller.
2424 PH/08 The ${def test on a variable was returning false if the variable's
2425 value was "0", contrary to what the specification has always said!
2426 The result should be true unless the variable is empty.
2428 PH/09 The syntax error of a character other than { following "${if
2429 def:variable_name" (after optional whitespace) was not being diagnosed.
2430 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2431 accidental colon was present, for example, could give incorrect results.
2433 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2434 result is used (not including appendfile, where other changes are about
2437 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2438 This involved changing a lot of size variables from int to off_t. It
2439 should work with maildirs and everything.
2441 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2442 spamd dying while we are connected to it.
2444 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2447 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2448 function for BDB 4.3.
2450 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2452 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2453 the sources for PCRE have been re-arranged and more files are now
2456 PH/15 The code I had for printing potentially long long variables in PH/11
2457 above was not the best (it lost precision). The length of off_t variables
2458 is now inspected at build time, and an appropriate printing format (%ld
2459 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2460 to be "long long int" or "long int". This is needed for the internal
2461 formatting function string_vformat().
2463 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2464 the configuration file to be ":syslog", then the script "guesses" where
2465 the logs files are, rather than using the compiled in default. In our
2466 case the guess is not the same as the compiled default, so the script
2467 suddenly stopped working when I started to use syslog. The patch checks
2468 to see if log_file_path is "". If so, it attempts to read it from exim
2469 with no configuration file to get the compiled in version, before it
2470 falls back to the previous guessing code."
2472 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2473 implementing BATV in an Exim configuration. See NewStuff for the gory
2476 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2477 Makefile that are specific to HP-UX.
2479 PH/18 If the "use_postmaster" option was set for a recipient callout together
2480 with the "random" option, the postmaster address was used as the MAIL
2481 FROM address for the random test, but not for the subsequent recipient
2482 test. It is now used for both.
2484 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2485 patch removes a few documentation additions to RFC 3028, because the
2486 latest draft now contains them. It adds the new en;ascii-case comparator
2487 and a new error check for 8bit text in MIME parts. Comparator and
2488 require names are now matched exactly. I enabled the subaddress
2489 extension, but it is not well tested yet (read: it works for me)."
2491 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2492 rework some of the code of TK/09 above to avoid the hardwired use of
2493 "%lld" and "long long". Replaced the call to snprintf() with a call to
2496 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2497 records point to non-existent hosts", "retry timeout exceeded", and
2498 "retry time not reached for any host after a long failure period".
2500 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2501 experimental DomainKeys support:
2503 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2504 (2) On an error such as an illegally used "control", the wrong name for
2505 the control was given.
2507 These problems did NOT occur unless DomainKeys support was compiled.
2509 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2511 PH/24 Added ${if match_ip condition.
2513 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2514 databases so that it will be absolutely obvious if a crash occurs in the
2515 DB library. This is a regular occurrence (often caused by mis-matched
2518 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2519 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2520 picked up on a user's system that detects such things. There doesn't seem
2521 to be a gcc warning option for this - only an attribute that has to be
2522 put on the function's prototype. It seems that in Fedora Core 4 they have
2523 set this on a number of new functions. No doubt there will be more in due
2526 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2527 specify a transport. However, if an address that was verified by such a
2528 router was the subject of a callout, Exim crashed because it tried to
2529 read the rcpt_include_affixes from the non-existent transport. Now it
2530 just assumes that the setting of that option is false. This bug was
2531 introduced by 4.51/PH/31.
2533 PH/28 Changed -d+all to exclude +memory, because that information is very
2534 rarely of interest, but it makes the output a lot bigger. People tend to
2535 do -d+all out of habit.
2537 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2538 code in os-type was giving problems when libc.so lives in lib64, like on
2541 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2542 aren't the modern standard, and it seems that some systems' include files
2543 don't always have them. Exim was already checking for some of the newer
2544 ones like T_AAAA, and defining it itself. I've added checks for all the
2545 record types that Exim uses.
2547 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2548 not automatically generating a new one, as it is supposed to. This
2549 prevented TLS from working. If the file did exist, but contained invalid
2550 data, a new version was generated, as expected. It was only the case of a
2551 non-existent file that was broken.
2553 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2554 with a change in libdomainkeys > 0.64.
2556 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2557 from DNS. If the selector record carries the flag, it now has
2558 precedence over the domain-wide flag.
2560 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2562 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2563 the use of a port name that isn't defined in /etc/services) occurred, the
2564 message was deferred as in a normal delivery, and thus remained on the
2565 spool, instead of being failed because of the mua_wrapper setting. This
2566 is now fixed, and I tidied up some of the mua_wrapper messages at the
2569 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2570 the messages in a hash of arrays rather than using individual hashes.
2571 This is a bit cleaner and results in dramatic memory savings, albeit
2572 at a slight CPU cost.
2574 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2575 as requested by Marc Sherman.
2577 SC/10 Eximstats: added histograms for user specified patterns as requested
2580 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2582 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2583 fopen() in the content-scanning modules that did not already have it.
2589 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2590 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2592 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2594 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2596 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2597 to test Sieve filters that use "vacation".
2599 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2600 that changes the way the GnuTLS parameters are stored in the cache file.
2601 The new format can be generated externally. For backward compatibility,
2602 if the data in the cache doesn't make sense, Exim assumes it has read an
2603 old-format file, and it generates new data and writes a new file. This
2604 means that you can't go back to an older release without removing the
2607 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2608 work if there are any delivery delays because "one_time" forces the
2609 parent to be marked "delivered", so its unseen clone is never tried
2610 again. For this reason, Exim now forbids the simultaneous setting of
2613 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2614 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2615 entry if you want to know the details. The fix, however, neglected to
2616 consider the case when local delivery batching is involved. The test for
2617 "previously delivered" was not happening when checking to see if an
2618 address could be batched with a previous (undelivered) one; under
2619 certain circumstances this could lead to multiple deliveries to the same
2622 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2623 in its include files, and this causes problems building Exim.
2625 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2626 header_syntax) but Exim was just ignoring anything given after a slash.
2627 In particular, this caused confusion with an attempt to use "verify =
2628 reverse_host_lookup/defer_ok". An error is now given when options are
2629 supplied for verify items that do not have them. (Maybe reverse_host_
2630 lookup should have a defer_ok option, but that's a different point.)
2632 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2633 defined by RFC 821) to 2048, because there were problems with some AUTH
2634 commands, and RFC 1869 says the size should be increased for extended
2635 SMTP commands that take arguments.
2637 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2640 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2641 "unknown" error; now it says that the functionality isn't in the binary.
2643 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2644 an address' error message when a string expansion fails (syntax or
2645 whatever). Otherwise the password may appear in the log. Following change
2646 PH/42 below, there is no longer a chance of it appearing in a bounce
2649 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2651 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2652 of its addresses was ever tried. (Bugzilla bug #2.)
2654 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2655 the result incorrectly in the debug output. (It correctly added a newline
2656 to what was transported.)
2658 TF/01 Added $received_time.
2660 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2661 commented out examples of how to interface to a virus scanner and to
2662 SpamAssassin. Also added commented examples of av_scanner and
2663 spamd_address settings.
2665 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2666 and controls are allowed in which ACLs. There were a couple of minor
2667 errors. Some of the entries in the conditions table (which is a table of
2668 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2669 negation of where the condition IS allowed.
2671 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2673 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2674 header file does not have a version number, so I've had to invent a new
2675 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2676 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2677 radiusclient), but it was contributed by a Radius user.
2679 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2680 files or directories, for OpenSSL.
2682 PH/19 When an Exim process that is running as root has to create an Exim log
2683 file, it does so in a subprocess that runs as exim:exim so as to get the
2684 ownership right at creation (otherwise, other Exim processes might see
2685 the file with the wrong ownership). There was no test for failure of this
2686 fork() call, which would lead to the process getting stuck as it waited
2687 for a non-existent subprocess. Forks do occasionally fail when resources
2688 run out. I reviewed all the other calls to fork(); they all seem to check
2691 PH/20 When checking for unexpected SMTP input at connect time (before writing
2692 the banner), Exim was not dealing correctly with a non-positive return
2693 from the read() function. If the client had disconnected by this time,
2694 the result was a log entry for a synchronization error with an empty
2695 string after "input=" when read() returned zero. If read() returned -1
2696 (an event I could not check), uninitialized data bytes were printed.
2697 There were reports of junk text (parts of files, etc) appearing after
2700 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2702 PH/22 Added support for macro redefinition, and (re)definition in between
2703 driver and ACL definitions.
2705 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2706 forgetting to use the resulting value; it was using the unexpanded value.
2708 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2709 hadn't been configured. The fix is from Juergen Kreileder, who
2710 understands it better than I do:
2712 "Here's what I see happening with three configured cyrus_sasl
2713 authenticators configured (plain, login, cram-md5):
2715 On startup auth_cyrus_sasl_init() gets called for each of these.
2716 This means three calls to sasl_listmech() without a specified mech_list.
2717 => SASL tests which mechs of all available mechs actually work
2718 => three warnings about OTP not working
2719 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2721 With the patch, sasl_listmech() also gets called three times. But now
2722 SASL's mech_list option is set to the server_mech specified in the the
2723 authenticator. Or in other words, the answer from sasl_listmech()
2724 gets limited to just the mech you're testing for (which is different
2726 => the return list contains just 'plain' or 'login', 'cram-md5' or
2727 nothing depending on the value of ob->server_mech.
2729 I've just tested the patch: Authentication still works fine,
2730 unavailable mechs specified in the exim configuration are still
2731 caught, and the auth.log warnings about OTP are gone."
2733 PH/25 When debugging is enabled, the contents of the command line are added
2734 to the debugging output, even when log_selector=+arguments is not
2737 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2738 answer is "GNU", and only if the return is "GNU/something" is the answer
2741 PH/27 $acl_verify_message is now set immediately after the failure of a
2742 verification in an ACL, and so is available in subsequent modifiers. In
2743 particular, the message can be preserved by coding like this:
2745 warn !verify = sender
2746 set acl_m0 = $acl_verify_message
2748 Previously, $acl_verify_message was set only while expanding "message"
2749 and "log_message" when a very denied access.
2751 PH/28 Modified OS/os.c-Linux with
2753 -#ifndef OS_LOAD_AVERAGE
2754 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2756 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2757 nomenclature these days.)
2759 PH/29 Installed patch from the Sieve maintainer that adds the options
2760 sieve_useraddress and sieve_subaddress to the redirect router.
2762 PH/30 In these circumstances:
2763 . Two addresses routed to the same list of hosts;
2764 . First host does not offer TLS;
2765 . First host accepts first address;
2766 . First host gives temporary error to second address;
2767 . Second host offers TLS and a TLS session is established;
2768 . Second host accepts second address.
2769 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2770 and peerdn, if requested) that were in fact used only for the second
2773 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2774 attention to any local part prefix or suffix that was matched by the
2775 router that accepted the address. It now behaves in the same way as it
2776 does for delivery: the affixes are removed from the local part unless
2777 rcpt_include_affixes is set on the transport.
2779 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2780 timeout during the DATA phase of an incoming message.
2782 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2783 applied a patch sanctioned by the Sieve maintainer.
2785 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2786 the uid or gid is negative. A case of a negative gid caused this to be
2787 noticed. The fix allows for either to be negative.
2789 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2790 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2793 PH/36 The change PH/12 above was broken. Fixed it.
2795 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2796 the grounds that routing the same address twice would always produce the
2797 same answer. This might have been true once, but it is certainly no
2798 longer true now. Routing a child address may depend on the previous
2799 routing that produced that child. Some complicated redirection strategies
2800 went wrong when messages had multiple recipients, and made Exim's
2801 behaviour dependent on the order in which the addresses were given.
2803 I have moved the duplicate checking until after the routing is complete.
2804 Exim scans the addresses that are assigned to local and remote
2805 transports, and removes any duplicates. This means that more work will be
2806 done, as duplicates will always all be routed, but duplicates are
2807 presumably rare, so I don't expect this is of any significance.
2809 For deliveries to pipes, files, and autoreplies, the duplicate checking
2810 still happens during the routing process, since they are not going to be
2813 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2814 It corrects a timeout issue with spamd. This is Ian's comment: "The
2815 background is that sometimes spamd either never reads data from a
2816 connection it has accepted, or it never writes response data. The exiscan
2817 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2818 blindly assumes that writes won't block so it may never time out."
2820 PH/39 Allow G after quota size as well as K and M.
2822 PH/40 The value set for $authenticated_id in an authenticator may not contain
2823 binary zeroes or newlines because the value is written to log lines and
2824 to spool files. There was no check on this. Now the value is run through
2825 the string_printing() function so that such characters are converted to
2826 printable escape sequences.
2828 PH/41 $message_linecount is a new variable that contains the total number of
2829 lines in the message. Compare $body_linecount, which is the count for the
2832 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2833 bounce and delay warning messages, except in certain special cases, which
2836 (a) An SMTP error message from a remote host;
2837 (b) A message specified in a :fail: redirection;
2838 (c) A message specified in a "fail" command in a system filter;
2839 (d) A message specified in a FAIL return from the queryprogram router;
2840 (e) A message specified by the cannot_route_message router option.
2842 In these cases only, Exim does include the error details in bounce and
2843 warning messages. There are also a few cases where bland messages such
2844 as "unrouteable address" or "local delivery error" are given.
2846 PH/43 $value is now also set for the "else" part of a ${run expansion.
2848 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2849 being worked on, but at least Exim now implements the latest version to
2852 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2853 process to complete was treated as a delivery failure, a timeout while
2854 writing the message to the pipe was logged, but erroneously treated as a
2855 successful delivery. Such timeouts include transport filter timeouts. For
2856 consistency with the overall process timeout, these timeouts are now
2857 treated as errors, giving rise to delivery failures by default. However,
2858 there is now a new Boolean option for the pipe transport called
2859 timeout_defer, which, if set TRUE, converts the failures into defers for
2860 both kinds of timeout. A transport filter timeout is now identified in
2863 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2864 systems where "make" and "gmake" are different, calling "gmake" at top
2865 level broke things. I've arranged for the value of $(MAKE) to be passed
2866 from the Makefile to this script so that it can call the same version of
2870 A note about Exim versions 4.44 and 4.50
2871 ----------------------------------------
2873 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2874 changes of various kinds. As a consequence, a big documentation update was
2875 needed. This delayed the release for rather longer than seemed good, especially
2876 in the light of a couple of (minor) security issues. Therefore, the changes
2877 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2878 release. So 4.44 and 4.50 are in effect two different branches that both start
2881 I have left the 4.50 change log unchanged; it contains all the changes since
2882 4.43. The change log for 4.44 is below; many of its items are identical to
2883 those for 4.50. This seems to be the most sensible way to preserve the
2884 historical information.
2890 1. Minor wording change to the doc/README.SIEVE file.
2892 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2893 computation of the current number of files was incorrect.
2895 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2896 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2899 4. Give more explanation in the error message when the command for a transport
2900 filter fails to execute.
2902 5. There are several places where Exim runs a non-Exim command in a
2903 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2904 was being done only for the command run by the queryprogram router. It is
2905 now done for all such subprocesses. The other cases are: ${run, transport
2906 filters, and the commands run by the lmtp and pipe transports.
2908 6. Added CONFIGURE_GROUP build-time option.
2910 7. Some older OS have a limit of 256 on the maximum number of file
2911 descriptors. Exim was using setrlimit() to set 1000 as a large value
2912 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2913 systems. I've change it so that if it can't get 1000, it tries for 256.
2915 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2916 was an oversight, and furthermore, ever since the addition of extra
2917 controls (e.g. 4.43/32), the checks on when to allow different forms of
2918 "control" were broken. There should now be diagnostics for all cases when a
2919 control that does not make sense is encountered.
2921 9. Added the /retain_sender option to "control=submission".
2923 10. $recipients is now available in the predata ACL (oversight).
2925 11. Tidy the search cache before the fork to do a delivery from a message
2926 received from the command line. Otherwise the child will trigger a lookup
2927 failure and thereby defer the delivery if it tries to use (for example) a
2928 cached ldap connection that the parent has called unbind on.
2930 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2931 of $address_data from the recipient verification was clobbered by the
2932 sender verification.
2934 13. The value of address_data from a sender verification is now available in
2935 $sender_address_data in subsequent conditions in the ACL statement.
2937 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2939 15. Added a new option "connect=<time>" to callout options, to set a different
2942 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2943 was its contents. (It was OK if the option was not defined at all.)
2945 17. A "Completed" log line is now written for messages that are removed from
2946 the spool by the -Mrm option.
2948 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2949 information about exactly what failed.
2951 19. Added -dd to debug only the daemon process.
2953 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2954 handles timeouts, both on the server side and network timeouts. Renamed the
2955 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2957 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2958 It is now set to "smtps".
2960 22. $host_address is now set to the target address during the checking of
2961 ignore_target_hosts.
2963 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2964 being passed; this would have caused $sender_host_name to have been used if
2965 matching the list had actually called for a host name (not very likely,
2966 since this list is usually IP addresses). A host name is now passed as
2969 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2970 code that specifies a non-restarting handler (typically sigaction() in
2971 modern systems) in an attempt to fix a rare and obscure crash bug.
2973 25. Narrowed the window for a race in the daemon that could cause it to ignore
2974 SIGCHLD signals. This is not a major problem, because they are used only to
2975 wake it up if nothing else does.
2977 26. A malformed maildirsize file could cause Exim to calculate negative values
2978 for the mailbox size or file count. Odd effects could occur as a result.
2979 The maildirsize information is now recalculated if the size or filecount
2982 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2983 support for a long time. Removed HAVE_SYS_VFS_H.
2985 28. Installed the latest version of exipick from John Jetmore.
2987 29. In an address list, if the pattern was not a regular expression, an empty
2988 subject address (from a bounce message) matched only if the pattern was an
2989 empty string. Non-empty patterns were not even tested. This was the wrong
2990 because it is perfectly reasonable to use an empty address as part of a
2991 database query. An empty address is now tested by patterns that are
2992 lookups. However, all the other forms of pattern expect the subject to
2993 contain a local part and a domain, and therefore, for them, an empty
2994 address still always fails if the pattern is not itself empty.
2996 30. Exim went into a mad DNS loop when attempting to do a callout where the
2997 host was specified on an smtp transport, and looking it up yielded more
2998 than one IP address.
3000 31. Re-factored the code for checking spool and log partition space into a
3001 function that finds that data and another that does the check. The former
3002 is then used to implement four new variables: $spool_space, $log_space,
3003 $spool_inodes, and $log_inodes.
3005 32. The RFC2047 encoding function was originally intended for short strings
3006 such as real names; it was not keeping to the 75-character limit for
3007 encoded words that the RFC imposes. It now respects the limit, and
3008 generates multiple encoded words if necessary. To be on the safe side, I
3009 have increased the buffer size for the ${rfc2047: expansion operator from
3012 33. It is now permitted to omit both strings after an "if" condition; if the
3013 condition is true, the result is "true". As before, when the second string
3014 is omitted, a false condition yields an empty string. This makes it less
3015 cumbersome to write custom ACL and router conditions.
3017 34. Failure to deliver a bounce message always caused it to be frozen, even if
3018 there was an errors_to setting on the router. The errors_to setting is now
3021 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3022 canonical form (fully expanded) before being placed in
3023 $sender_host_address.
3025 36. The table in the code that translates DNS record types into text (T_A to
3026 "A" for instance) was missing entries for NS and CNAME. It is just possible
3027 that this could have caused confusion if both these types were looked up
3028 for the same domain, because the text type is used as part of Exim's
3029 per-process caching. But the chance of anyone hitting this buglet seems
3032 37. The dnsdb lookup has been extended in a number of ways.
3034 (1) There is a new type, "zns", which walks up the domain tree until it
3035 finds some nameserver records. It should be used with care.
3037 (2) There is a new type, "mxh", which is like "mx" except that it returns
3038 just the host names, not the priorities.
3040 (3) It is now possible to give a list of domains (or IP addresses) to be
3041 looked up. The behaviour when one of the lookups defers can be
3042 controlled by a keyword.
3044 (4) It is now possible to specify the separator character for use when
3045 multiple records are returned.
3047 38. The dnslists ACL condition has been extended: it is now possible to supply
3048 a list of IP addresses and/or domains to be looked up in a particular DNS
3051 39. Added log_selector=+queue_time_overall.
3053 40. When running the queue in the test harness, wait just a tad after forking a
3054 delivery process, to get repeatability of debugging output.
3056 41. Include certificate and key file names in error message when GnuTLS fails
3057 to set them up, because the GnuTLS error message doesn't include the name
3058 of the failing file when there is a problem reading it.
3060 42. Allow both -bf and -bF in the same test run.
3062 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3064 44. The "Exiscan patch" is now merged into the mainline Exim source.
3066 45. Sometimes the final signoff response after QUIT could fail to get
3067 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3068 < 0 before doing a fflush(). This bug looks as though it goes back to the
3069 introduction of TLS in release 3.20, but "sometimes" must have been rare
3070 because the tests only now provoked it.
3072 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3073 (this can affect the format of dates).
3075 47. exim_tidydb, when checking for the continued existence of a message for
3076 which it has found a message-specific retry record, was not finding
3077 messages that were in split spool directories. Consequently, it was
3078 deleting retry records that should have stayed in existence.
3080 48. Steve fixed some bugs in eximstats.
3082 49. The SPA authentication driver was not abandoning authentication and moving
3083 on to the next authenticator when an expansion was forced to fail,
3084 contradicting the general specification for all authenticators. Instead it
3085 was generating a temporary error. It now behaves as specified.
3087 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3088 (the order specifies the preference for clients). The order is now AES256,
3089 AES128, 3DES, ARCFOUR128.
3091 51. Small patch to Sieve code - explicitly set From: when generating an
3094 52. Exim crashed if a remote delivery caused a very long error message to be
3095 recorded - for instance if somebody sent an entire SpamAssassin report back
3096 as a large number of 550 error lines. This bug was coincidentally fixed by
3097 increasing the size of one of Exim's internal buffers (big_buffer) that
3098 happened as part of the Exiscan merge. However, to be on the safe side, I
3099 have made the code more robust (and fixed the comments that describe what
3102 53. Now that there can be additional text after "Completed" in log lines (if
3103 the queue_time_overall log selector is set), a one-byte patch to exigrep
3104 was needed to allow it to recognize "Completed" as not the last thing in
3107 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3108 patch that reportedly fixes this has been added. I am not expert enough to
3109 create a test for it. This is what the patch creator wrote:
3111 "I found a little strange behaviour of ldap code when working with
3112 Windows 2003 AD Domain, where users was placed in more than one
3113 Organization Units. When I tried to give exim partial DN, the exit code
3114 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3115 But simultaneously result of request was absolutely normal ldap result,
3116 so I produce this patch..."
3118 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3119 REFERENCE, so I have modified the code to exclude the patch when that macro
3122 55. Some experimental protocols are using DNS PTR records for new purposes. The
3123 keys for these records are domain names, not reversed IP addresses. The
3124 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3125 leaves it alone. Component reversal etc. now happens only for IP addresses.
3128 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3130 57. Double the size of the debug message buffer (to 2048) so that more of very
3131 long debug lines gets shown.
3133 58. The exicyclog utility now does better if the number of log files to keep
3134 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3136 59. Two changes related to the smtp_active_hostname option:
3138 (1) $smtp_active_hostname is now available as a variable.
3139 (2) The default for smtp_banner uses $smtp_active_hostname instead
3140 of $primary_hostname.
3142 60. The host_aton() function is supposed to be passed a string that is known
3143 to be a valid IP address. However, in the case of IPv6 addresses, it was
3144 not checking this. This is a hostage to fortune. Exim now panics and dies
3145 if the condition is not met. A case was found where this could be provoked
3146 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3147 components; fortuitously, this particular loophole had already been fixed
3148 by change 4.50/55 above.
3150 If there are any other similar loopholes, the new check in host_aton()
3151 itself should stop them being exploited. The report I received stated that
3152 data on the command line could provoke the exploit when Exim was running as
3153 exim, but did not say which command line option was involved. All I could
3154 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3155 running as the user.
3158 61. There was a buffer overflow vulnerability in the SPA authentication code
3159 (which came originally from the Samba project). I have added a test to the
3160 spa_base64_to_bits() function which I hope fixes it.
3163 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3164 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3166 63. The daemon start-up calls getloadavg() while still root for those OS that
3167 need the first call to be done as root, but it missed one case: when
3168 deliver_queue_load_max is set with deliver_drop_privilege. This is
3169 necessary for the benefit of the queue runner, because there is no re-exec
3170 when deliver_drop_privilege is set.
3172 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3173 This has been fixed.
3175 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3176 was in use, was not putting the data itself into the right store pool;
3177 consequently, it could be overwritten for a subsequent message in the same
3178 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3181 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3183 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3184 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3185 the function were treating the return as a boolean value, which happened to
3186 work because 0=false and not-0=true, but is not correct code.
3188 68. The host_aton() function was not handling scoped IPv6 addresses (those
3189 with, for example, "%eth0" on the end) correctly.
3191 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3192 negated items (that is, ~something) in unsigned ints. Some compilers
3193 apparently mutter when there is no cast.
3195 70. If an address verification called from an ACL failed, and did not produce a
3196 user-specific message (i.e. there was only a "system" message), nothing was
3197 put in $acl_verify_message. In this situation, it now puts the system
3200 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3201 session; change 4.31/43 added the unwanted input to the log line - except
3202 that it did not do this in the start of session case. It now does.
3204 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3205 This is wrong and can cause the other end to generate a synchronization
3206 error if it is another Exim or anything else that does the synchronization
3207 check. A QUIT command is no longer sent after a timeout.
3209 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3210 during host lookups.
3212 74. The defer_ok option of callout verification was not working if it was used
3213 when verifying addresses in header lines, that is, for this case:
3215 verify = header_sender/callout=defer_ok
3217 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3218 those file descriptors could be used for SMTP connections. If anything
3219 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3220 could be sent to the SMTP client, causing chaos. The daemon now opens
3221 stdin, stdout, and stderr to /dev/null when it puts itself into the
3224 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3225 log by default. The user can override this with suitable Perl magic.
3227 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3228 the log message when discard triggers, was not working for the DATA ACL or
3229 for the non-SMTP ACL.
3231 78. Error message wording change in sieve.c.
3233 79. If smtp_accept_max_per_host was set, the number of connections could be
3234 restricted to fewer than expected, because the daemon was trying to set up
3235 a new connection before checking whether the processes handling previous
3236 connections had finished. The check for completed processes is now done
3237 earlier. On busy systems, this bug wouldn't be noticed because something
3238 else would have woken the daemon, and it would have reaped the completed
3241 80. If a message was submitted locally by a user whose login name contained one
3242 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3243 It caused a spool format error. I have fixed the spool reading code. A
3244 related problem was that the "from" clause in the Received: line became
3245 illegal because of the space(s). It is now covered by ${quote_local_part.
3247 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3250 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3252 83. Patch from Sieve maintainer to fix unterminated string problem in
3253 "vacation" handling.
3255 84. Some minor changes to the Linux configuration files to help with other
3256 OS variants using glibc.
3258 85. One more patch for Sieve to update vacation handling to latest spec.
3261 ----------------------------------------------------
3262 See the note above about the 4.44 and 4.50 releases.
3263 ----------------------------------------------------
3269 1. Change 4.43/35 introduced a bug that caused file counts to be
3270 incorrectly computed when quota_filecount was set in an appendfile
3273 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3274 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3277 3. Give more explanation in the error message when the command for a transport
3278 filter fails to execute.
3280 4. There are several places where Exim runs a non-Exim command in a
3281 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3282 was being done only for the command run by the queryprogram router. It is
3283 now done for all such subprocesses. The other cases are: ${run, transport
3284 filters, and the commands run by the lmtp and pipe transports.
3286 5. Some older OS have a limit of 256 on the maximum number of file
3287 descriptors. Exim was using setrlimit() to set 1000 as a large value
3288 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3289 systems. I've change it so that if it can't get 1000, it tries for 256.
3291 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3292 was an oversight, and furthermore, ever since the addition of extra
3293 controls (e.g. 4.43/32), the checks on when to allow different forms of
3294 "control" were broken. There should now be diagnostics for all cases when a
3295 control that does not make sense is encountered.
3297 7. $recipients is now available in the predata ACL (oversight).
3299 8. Tidy the search cache before the fork to do a delivery from a message
3300 received from the command line. Otherwise the child will trigger a lookup
3301 failure and thereby defer the delivery if it tries to use (for example) a
3302 cached ldap connection that the parent has called unbind on.
3304 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3305 of $address_data from the recipient verification was clobbered by the
3306 sender verification.
3308 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3309 was its contents. (It was OK if the option was not defined at all.)
3311 11. A "Completed" log line is now written for messages that are removed from
3312 the spool by the -Mrm option.
3314 12. $host_address is now set to the target address during the checking of
3315 ignore_target_hosts.
3317 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3318 being passed; this would have caused $sender_host_name to have been used if
3319 matching the list had actually called for a host name (not very likely,
3320 since this list is usually IP addresses). A host name is now passed as
3323 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3324 code that specifies a non-restarting handler (typically sigaction() in
3325 modern systems) in an attempt to fix a rare and obscure crash bug.
3327 15. Narrowed the window for a race in the daemon that could cause it to ignore
3328 SIGCHLD signals. This is not a major problem, because they are used only to
3329 wake it up if nothing else does.
3331 16. A malformed maildirsize file could cause Exim to calculate negative values
3332 for the mailbox size or file count. Odd effects could occur as a result.
3333 The maildirsize information is now recalculated if the size or filecount
3336 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3337 support for a long time. Removed HAVE_SYS_VFS_H.
3339 18. Updated exipick to current release from John Jetmore.
3341 19. Allow an empty sender to be matched against a lookup in an address list.
3342 Previously the only cases considered were a regular expression, or an
3345 20. Exim went into a mad DNS lookup loop when doing a callout where the
3346 host was specified on the transport, if the DNS lookup yielded more than
3349 21. The RFC2047 encoding function was originally intended for short strings
3350 such as real names; it was not keeping to the 75-character limit for
3351 encoded words that the RFC imposes. It now respects the limit, and
3352 generates multiple encoded words if necessary. To be on the safe side, I
3353 have increased the buffer size for the ${rfc2047: expansion operator from
3356 22. Failure to deliver a bounce message always caused it to be frozen, even if
3357 there was an errors_to setting on the router. The errors_to setting is now
3360 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3361 canonical form (fully expanded) before being placed in
3362 $sender_host_address.
3364 24. Updated eximstats to version 1.33
3366 25. Include certificate and key file names in error message when GnuTLS fails
3367 to set them up, because the GnuTLS error message doesn't include the name
3368 of the failing file when there is a problem reading it.
3370 26. Expand error message when OpenSSL has problems setting up cert/key files.
3373 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3374 (this can affect the format of dates).
3376 28. exim_tidydb, when checking for the continued existence of a message for
3377 which it has found a message-specific retry record, was not finding
3378 messages that were in split spool directories. Consequently, it was
3379 deleting retry records that should have stayed in existence.
3381 29. eximstats updated to version 1.35
3382 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3383 1.35 - bugfix such that pie charts by volume are generated correctly
3385 30. The SPA authentication driver was not abandoning authentication and moving
3386 on to the next authenticator when an expansion was forced to fail,
3387 contradicting the general specification for all authenticators. Instead it
3388 was generating a temporary error. It now behaves as specified.
3390 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3391 (the order specifies the preference for clients). The order is now AES256,
3392 AES128, 3DES, ARCFOUR128.
3394 31. Small patch to Sieve code - explicitly set From: when generating an
3397 32. Exim crashed if a remote delivery caused a very long error message to be
3398 recorded - for instance if somebody sent an entire SpamAssassin report back
3399 as a large number of 550 error lines. This bug was coincidentally fixed by
3400 increasing the size of one of Exim's internal buffers (big_buffer) that
3401 happened as part of the Exiscan merge. However, to be on the safe side, I
3402 have made the code more robust (and fixed the comments that describe what
3405 33. Some experimental protocols are using DNS PTR records for new purposes. The
3406 keys for these records are domain names, not reversed IP addresses. The
3407 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3408 leaves it alone. Component reversal etc. now happens only for IP addresses.
3411 34. The host_aton() function is supposed to be passed a string that is known
3412 to be a valid IP address. However, in the case of IPv6 addresses, it was
3413 not checking this. This is a hostage to fortune. Exim now panics and dies
3414 if the condition is not met. A case was found where this could be provoked
3415 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3416 components; fortuitously, this particular loophole had already been fixed
3417 by change 4.50/55 or 4.44/33 above.
3419 If there are any other similar loopholes, the new check in host_aton()
3420 itself should stop them being exploited. The report I received stated that
3421 data on the command line could provoke the exploit when Exim was running as
3422 exim, but did not say which command line option was involved. All I could
3423 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3424 running as the user.
3427 35. There was a buffer overflow vulnerability in the SPA authentication code
3428 (which came originally from the Samba project). I have added a test to the
3429 spa_base64_to_bits() function which I hope fixes it.
3432 36. The daemon start-up calls getloadavg() while still root for those OS that
3433 need the first call to be done as root, but it missed one case: when
3434 deliver_queue_load_max is set with deliver_drop_privilege. This is
3435 necessary for the benefit of the queue runner, because there is no re-exec
3436 when deliver_drop_privilege is set.
3438 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3439 was in use, was not putting the data itself into the right store pool;
3440 consequently, it could be overwritten for a subsequent message in the same
3441 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3444 38. Sometimes the final signoff response after QUIT could fail to get
3445 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3446 < 0 before doing a fflush(). This bug looks as though it goes back to the
3447 introduction of TLS in release 3.20, but "sometimes" must have been rare
3448 because the tests only now provoked it.
3454 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3455 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3456 Now it returns an integer. A number of calls were still expecting a T/F
3457 return. Fortuitously, in all cases, the tests worked in OK situations,
3458 which is the norm. However, things would have gone wrong on any write
3459 failures on the smtp file descriptor. This function is used when sending
3460 messages over SMTP and also when doing verify callouts.
3462 2. When Exim is called to do synchronous delivery of a locally submitted
3463 message (the -odf or -odi options), it no longer closes stderr before doing
3466 3. Implemented the mua_wrapper option.
3468 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3470 5. Implemented the functions header_remove(), header_testname(),
3471 header_add_at_position(), and receive_remove_recipient(), and exported them
3474 6. If an ACL "warn" statement specified the addition of headers, Exim already
3475 inserted X-ACL-Warn: at the start if there was no header name. However, it
3476 was not making this test for the second and subsequent header lines if
3477 there were newlines in the string. This meant that an invalid header could
3478 be inserted if Exim was badly configured.
3480 7. Allow an ACL "warn" statement to add header lines at the start or after all
3481 the Received: headers, as well as at the end.
3483 8. Added the rcpt_4xx retry error code.
3485 9. Added postmaster_mailfrom=xxx to callout verification option.
3487 10. Added mailfrom=xxxx to the callout verification option, for verify=
3490 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3491 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3493 12. Inserted some casts to stop certain compilers warning when using pointer
3494 differences as field lengths or precisions in printf-type calls (mostly
3495 affecting debugging statements).
3497 13. Added optional readline() support for -be (dynamically loaded).
3499 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3500 same clock tick as a message's arrival, so that its received time was the
3501 same as the "first fail" time on the retry record, and that message
3502 remained on the queue past the ultimate address timeout, every queue runner
3503 would try a delivery (because it was past the ultimate address timeout) but
3504 after another failure, the ultimate address timeout, which should have then
3505 bounced the address, did not kick in. This was a "< instead of <=" error;
3506 in most cases the first failure would have been in the next clock tick
3507 after the received time, and all would be well.
3509 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3510 being recognized when the domain list was tested by the match_domain
3511 condition in an expansion string.
3513 16. Added the ${str2b64: operator.
3515 17. Exim was always calling setrlimit() to set a large limit for the number of
3516 processes, without checking whether the existing limit was already
3517 adequate. (It did check for the limit on file descriptors.) Furthermore,
3518 errors from getrlimit() and setrlimit() were being ignored. Now they are
3519 logged to the main and panic logs, but Exim does carry on, to try to do its
3520 job under whatever limits there are.
3522 18. Imported PCRE 5.0.
3524 19. Trivial typo in log message " temporarily refused connection" (the leading
3527 20. If the log selector return_path_on_delivery was set and an address was
3528 redirected to /dev/null, the delivery process crashed because it assumed
3529 that a return path would always be set for a "successful" delivery. In this
3530 case, the whole delivery is bypassed as an optimization, and therefore no
3533 21. Internal re-arrangement: the function for sending a challenge and reading
3534 a response while authentication was assuming a zero-terminated challenge
3535 string. It's now changed to take a pointer and a length, to allow for
3536 binary data in such strings.
3538 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3540 23. Exim was not respecting finduser_retries when seeking the login of the
3541 uid under which it was called; it was always trying 10 times. (The default
3542 setting of finduser_retries is zero.) Also, it was sleeping after the final
3543 failure, which is pointless.
3545 24. Implemented tls_on_connect_ports.
3547 25. Implemented acl_smtp_predata.
3549 26. If the domain in control=submission is set empty, Exim assumes that the
3550 authenticated id is a complete email address when it generates From: or
3551 Sender: header lines.
3553 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3554 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3555 chown and chgrp in /bin and hostname in /usr/bin.
3557 28. Exim was keeping the "process log" file open after each use, just as it
3558 does for the main log. This opens the possibility of it remaining open for
3559 long periods when the USR1 signal hits a daemon. Occasional processlog
3560 errors were reported, that could have been caused by this. Anyway, it seems
3561 much more sensible not to leave this file open at all, so that is what now
3564 29. The long-running daemon process does not normally write to the log once it
3565 has entered its main loop, and it closes the log before doing so. This is
3566 so that log files can straightforwardly be renamed and moved. However,
3567 there are a couple of unusual error situations where the daemon does write
3568 log entries, and I had neglected to close the log afterwards.
3570 30. The text of an SMTP error response that was received during a remote
3571 delivery was being truncated at 512 bytes. This is too short for some of
3572 the long messages that one sometimes sees. I've increased the limit to
3575 31. It is now possible to make retry rules that apply only when a message has a
3576 specific sender, in particular, an empty sender.
3578 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3579 it possible to be selective about when SMTP synchronization is enforced.
3581 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3583 32. Implemented hosts_connection_nolog.
3585 33. Added an ACL for QUIT.
3587 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3590 35. Added mailbox_size and mailbox_filecount to appendfile.
3592 36. Added control = no_multiline_responses to ACLs.
3594 37. There was a bug in the logic of the code that waits for the clock to tick
3595 in the case where the clock went backwards by a substantial amount such
3596 that the microsecond fraction of "now" was more than the microsecond
3597 fraction of "then" (but the whole seconds number was less).
3599 38. Added support for the libradius Radius client library this is found on
3600 FreeBSD (previously only the radiusclient library was supported).
3606 1. When certain lookups returned multiple values in the form name=value, the
3607 quoting of the values was not always being done properly. Specifically:
3608 (a) If the value started with a double quote, but contained no whitespace,
3610 (b) If the value contained whitespace other than a space character (i.e.
3611 tabs or newlines or carriage returns) it was not quoted.
3612 This fix has been applied to the mysql and pgsql lookups by writing a
3613 separate quoting function and calling it from the lookup code. The fix
3614 should probably also be applied to nisplus, ibase and oracle lookups, but
3615 since I cannot test any of those, I have not disturbed their existing code.
3617 2. A hit in the callout cache for a specific address caused a log line with no
3618 reason for rejecting RCPT. Now it says "Previous (cached) callout
3619 verification failure".
3621 3. There was an off-by-one bug in the queryprogram router. An over-long
3622 return line was truncated at 256 instead of 255 characters, thereby
3623 overflowing its buffer with the terminating zero. As well as fixing this, I
3624 have increased the buffer size to 1024 (and made a note to document this).
3626 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3627 when Exim is waiting for an SMTP response from a remote server, Exim
3628 restarts its select() call on the socket, thereby resetting its timeout.
3629 This is not a problem when such interrupts are rare. Somebody set up a cron
3630 job to run exiwhat every 2 minutes, which is less than the normal select()
3631 timeout (5 or 10 minutes). This meant that the select() timeout never
3632 kicked in because it was always reset. I have fixed this by comparing the
3633 time when an interrupt arrives with the time at the start of the first call
3634 to select(). If more time than the timeout has elapsed, the interrupt is
3635 treated as a timeout.
3637 5. Some internal re-factoring in preparation for the addition of Sieve
3638 extensions (by MH). In particular, the "personal" test is moved to a
3639 separate function, and given an option for scanning Cc: and Bcc: (which is
3640 not set for Exim filters).
3642 6. When Exim created an email address using the login of the caller as the
3643 local part (e.g. when creating a From: or Sender: header line), it was not
3644 quoting the local part when it contained special characters such as @.
3646 7. Installed new OpenBSD configuration files.
3648 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3649 try to make them clearer.
3651 9. Callout options, other than the timeout value, were being ignored when
3652 verifying sender addresses in header lines. For example, when using
3654 verify = header_sender/callout=no_cache
3656 the cache was (incorrectly) being used.
3658 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3659 only the Cygwin environment.
3661 11. When return_path_on_delivery was set as a log selector, if different remote
3662 addresses in the same message used different return paths and parallel
3663 remote delivery occurred, the wrong values would sometimes be logged.
3664 (Whenever a remote delivery process finished, the return path value from
3665 the most recently started remote delivery process was logged.)
3667 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3668 lines when AUTH and/or TLS are in use. This is the "received protocol"
3669 field. Exim used to use "asmtp" for authenticated SMTP, without any
3670 indication (in the protocol name) for TLS use. Now it follows the RFC and
3671 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3672 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3673 names appear in log lines as well as in Received: header lines.
3675 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3676 extensions, and comparison tests, and to fix some bugs.
3678 14. Changes to the "personal" filter test:
3680 (1) The test was buggy in that it was just doing the equivalent of
3681 "contains" tests on header lines. For example, if a user's address was
3682 anne@some.where, the "personal" test would incorrectly be true for
3684 To: susanne@some.where
3686 This test is now done by extracting each address from the header in turn,
3687 and checking the entire address. Other tests that are part of "personal"
3688 are now done using regular expressions (for example, to check local parts
3689 of addresses in From: header lines).
3691 (2) The list of non-personal local parts in From: addresses has been
3692 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3693 taken from the Sieve specification recommendations.
3695 (3) If the message contains any header line starting with "List-" it is
3696 treated as non-personal.
3698 (4) The test for "circular" in the Subject: header line has been removed
3699 because it now seems ill-conceived.
3701 15. Minor typos in src/EDITME comments corrected.
3703 16. Installed latest exipick from John Jetmore.
3705 17. If headers_add on a router specified a text string that was too long for
3706 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3707 of string_sprintf() is now avoided.
3709 18. $message_body_size was not set (it was always zero) when running the DATA
3710 ACL and the local_scan() function.
3712 19. For the "mail" command in an Exim filter, no default was being set for
3713 the once_repeat time, causing a random time value to be used if "once" was
3714 specified. (If the value happened to be <= 0, no repeat happened.) The
3715 default is now 0s, meaning "never repeat". The "vacation" command was OK
3716 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3717 (I found it when inspecting the code).
3719 20. There is now an overall timeout for performing a callout verification. It
3720 defaults to 4 times the callout timeout, which applies to individual SMTP
3721 commands during the callout. The overall timeout applies when there is more
3722 than one host that can be tried. The timeout is checked before trying the
3723 next host. This prevents very long delays if there are a large number of
3724 hosts and all are timing out (e.g. when the network connections are timing
3725 out). The value of the overall timeout can be changed by specifying an
3726 additional sub-option for "callout", called "maxwait". For example:
3728 verify = sender/callout=5s,maxwait=20s
3730 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3731 to the end before writing, but this should make it even safer).
3733 22. Exim was forgetting that it had advertised PIPELINING for the second and
3734 subsequent messages on an SMTP connection. It was also not resetting its
3735 memory on STARTTLS and an internal HELO.
3737 23. When Exim logs an SMTP synchronization error within a session, it now
3738 records whether PIPELINING has been advertised or not.
3740 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3741 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3742 rather than long int.
3744 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3746 26. Added the never_mail option to autoreply.
3752 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3753 crash if the getsockname() call failed; this can happen if a connection is
3754 closed very soon after it is established. The problem was simply in the
3755 order in which certain operations were done, causing Exim to try to write
3756 to the SMTP stream before it had set up the file descriptor. The bug has
3757 been fixed by making things happen in the correct order.
3763 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3764 before the connection was closed, thus losing the rejection response.
3766 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3767 some early Solaris releases, but causes trouble in current releases where
3768 socklen_t is defined.
3770 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3773 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3776 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3777 that releasing the top of it at the end releases what was used for sub-
3778 expansions (unless the block got too big). However, discard this block if
3779 the first thing is a variable or header, so that we can use its block when
3780 it is dynamic (useful for very large $message_headers, for example).
3782 6. Lookups now cache *every* query, not just the most recent. A new, separate
3783 store pool is used for this. It can be recovered when all lookup caches are
3784 flushed. Lookups now release memory at the end of their result strings.
3785 This has involved some general refactoring of the lookup sources.
3787 7. Some code has been added to the store_xxx() functions to reduce the amount
3788 of flapping under certain conditions.
3790 8. log_incoming_interface used to affect only the <= reception log lines. Now
3791 it causes the local interface and port to be added to several more SMTP log
3792 lines, for example "SMTP connection from", and rejection lines.
3794 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3796 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3798 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3799 could be overwritten at the end of the current message (or the start of a
3800 new message if it was set in a HELO ACL). The value is now preserved for
3801 the duration of the SMTP connection.
3803 12. If a transport had a headers_rewrite setting, and a matching header line
3804 contained an unqualified address, that address was qualified, even if it
3805 did not match any rewriting rules. The underlying bug was that the values
3806 of the flags that permit the existence of unqualified sender and recipient
3807 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3808 non-local messages, and by -bnq for local messages) were not being
3809 preserved with the message after it was received.
3811 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3812 "next input=" as part of the text comprising the host identity instead of
3813 the correct text. The code was using the same buffer for two different
3814 strings. However, depending on which order the printing function evaluated
3815 its arguments, the bug did not always show up. Under Linux, for example, my
3816 test suite worked just fine.
3818 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3819 doesn't work with some older versions of Perl. It has been changed to "my",
3820 which in any case is probably the better facility to use.
3822 15. A really picky compiler found some instances of statements for creating
3823 error messages that either had too many or two few arguments for the format
3826 16. The size of the buffer for calls to the DNS resolver has been increased
3827 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3828 for addresses that have a lot of PTR records. This alleviates a problem; it
3829 does not fully solve it.
3831 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3832 buffer now truncates the list and logs the incident, which is the same
3833 action as happens when Exim is looking up a host name and its aliases.
3834 Previously in this situation something unpredictable would happen;
3835 sometimes it was "internal error: store_reset failed".
3837 18. If a server dropped the connection unexpectedly when an Exim client was
3838 using GnuTLS and trying to read a response, the client delivery process
3839 crashed while trying to generate an error log message.
3841 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3842 string, for example:
3844 warn message = H1: something\nH2: something
3846 the text was added as a single header line from Exim's point of view
3847 though it ended up OK in the delivered message. However, searching for the
3848 second and subsequent header lines using $h_h2: did not work. This has been
3849 fixed. Similarly, if a system filter added multiple headers in this way,
3850 the routers could not see them.
3852 20. Expanded the error message when iplsearch is called with an invalid key to
3853 suggest using net-iplsearch in a host list.
3855 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3856 ACLs are no longer actually imposed (and a message to that effect is
3859 22. If a "gecos" field in a passwd entry contained escaped characters, in
3860 particular, if it contained a \" sequence, Exim got it wrong when building
3861 a From: or a Sender: header from that name. A second bug also caused
3862 incorrect handling when an unquoted " was present following a character
3863 that needed quoting.
3865 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3866 was not being matched caselessly.
3868 24. Arranged for all hyphens in the exim.8 source to be escaped with
3871 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3872 properly. Recipient callout cache records were still being keyed to include
3873 the sender, even when use_sender was set false. This led to far more
3874 callouts that were necessary. The sender is no longer included in the key
3875 when use_sender is false.
3877 26. Added "control = submission" modifier to ACLs.
3879 27. Added the ${base62d: operator to decode base 62 numbers.
3881 28. dnsdb lookups can now access SRV records.
3883 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3884 the configuration file.
3886 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3887 (-v) mode. This makes the output for a verbose queue run more intelligible.
3889 31. Added a use_postmaster feature to recipient callouts.
3891 32. Added the $body_zerocount variable, containing the number of binary zero
3892 bytes in the message body.
3894 33. The time of last modification of the "new" subdirectory is now used as the
3895 "mailbox time last read" when there is a quota error for a maildir
3898 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3900 35. Added +ignore_unknown as a special item in host lists.
3902 36. Code for decoding IPv6 addresses in host lists is now included, even if
3903 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3904 address was recognized as an IP address, but was then not correctly decoded
3905 into binary, causing unexpected and incorrect effects when compared with
3912 1. Very minor rewording of debugging text in manualroute to say "list of
3913 hosts" instead of "hostlist".
3915 2. If verify=header_syntax was set, and a header line with an unqualified
3916 address (no domain) and a large number of spaces between the end of the
3917 name and the colon was received, the reception process suffered a buffer
3918 overflow, and (when I tested it) crashed. This was caused by some obsolete
3919 code that should have been removed. The fix is to remove it!
3921 3. When running in the test harness, delay a bit after writing a bounce
3922 message to get a bit more predictability in the log output.
3924 4. Added a call to search_tidyup() just before forking a reception process. In
3925 theory, someone could use a lookup in the expansion of smtp_accept_max_
3926 per_host which, without the tidyup, could leave open a database connection.
3928 5. Added the variables $recipient_data and $sender_data which get set from a
3929 lookup success in an ACL "recipients" or "senders" condition, or a router
3930 "senders" option, similar to $domain_data and $local_part_data.
3932 6. Moved the writing of debug_print from before to after the "senders" test
3935 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3936 problems for message scanning, either using a data ACL, or using
3937 local_scan() because the Received: header was not generated till after they
3938 were called (in order to set the time as the time of reception completion).
3939 I have revised the way this works. The header is now generated after the
3940 body is received, but before the ACL or local_scan() are called. After they
3941 are run, the timestamp in the header is updated.
3947 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3948 before starting a queue runner without re-exec. This happened only when
3949 deliver_drop_privilege was set or when the Exim user was set to root. The
3950 effect of the bug was that timeouts during subsequent deliveries caused
3951 crashes instead of being properly handled. The handler is now left at its
3952 default (and expected) setting.
3954 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3955 message, again when deliver_drop_privilege is set or Exim is run as root.
3956 The bug described in (1) was not present in this case, but the tidying up
3957 of the other signals was missing. I have made the two cases consistent.
3959 3. The ignore_target_hosts setting on a manualroute router was being ignored
3960 for hosts that were looked up using the /MX notation.
3962 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3965 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3966 operated on the sender address. After changing the $sender_address to <>
3967 for the sender address verify, Exim was re-instated it as the original
3968 (before rewriting) address, but remembering that it had rewritten it, so it
3969 wasn't rewriting it again. This bug also had the effect of breaking the
3970 sender address verification caching when the sender address was rewritten.
3972 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3973 This has been changed so that if the ip literal address matches
3974 ignore_target_hosts, the router declines.
3976 7. Added expansion conditions match_domain, match_address, and match_local_
3977 part (NOT match_host).
3979 8. The placeholder for the Received: header didn't have a length field set.
3981 9. Added code to Exim itself and to exim_lock to test for a specific race
3982 condition that could lead to file corruption when using MBX delivery. The
3983 issue is with the lockfile that is created in /tmp. If this file is removed
3984 after a process has opened it but before that process has acquired a lock,
3985 there is the potential for a second process to recreate the file and also
3986 acquire a lock. This could lead to two Exim processes writing to the file
3987 at the same time. The added code performs the same test as UW imapd; it
3988 checks after acquiring the lock that its file descriptor still refers to
3989 the same named file.
3991 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3992 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3995 11. Added the smtp_active_hostname option. If used, this will typically be made
3996 to depend on the incoming interface address. Because $interface_address is
3997 not set up until the daemon has forked a reception process, error responses
3998 that can happen earlier (such as "too many connections") no longer contain
4001 12. If an expansion in a condition on a "warn" statement fails because a lookup
4002 defers, the "warn" statement is abandoned, and the next ACL statement is
4003 processed. Previously this caused the whole ACL to be aborted.
4005 13. Added the iplsearch lookup type.
4007 14. Added ident_timeout as a log selector.
4009 15. Added tls_certificate_verified as a log selector.
4011 16. Added a global option tls_require_ciphers (compare the smtp transport
4012 option of the same name). This controls incoming TLS connections.
4014 17. I finally figured out how to make tls_require_ciphers do a similar thing
4015 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4016 before starting the TLS session.
4018 18. Tabs are now shown as \t in -bP output.
4020 19. If the log selector return_path_on_delivery was set, Exim crashed when
4021 bouncing a message because it had too many Received: header lines.
4023 20. If two routers both had headers_remove settings, and the first one included
4024 a superfluous trailing colon, the final name in the first list and the
4025 first name in the second list were incorrectly joined into one item (with a
4026 colon in the middle).
4032 1. Added -C and -D options to the exinext utility, mainly to make it easier
4033 to include in the automated testing, but these could be helpful when
4034 multiple configurations are in use.
4036 2. The exinext utility was not formatting the output nicely when there was
4037 an alternate port involved in the retry record key, nor when there was a
4038 message id as well (for retries that were specific to a specific message
4039 and a specific host). It was also confused by IPv6 addresses, because of
4040 the additional colons they contain. I have fixed the IPv4 problem, and
4041 patched it up to do a reasonable job for IPv6.
4043 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4044 delivery, the log line now contains "pipelined" if PIPELINING was used.
4046 4. An SMTP transport process used to panic and die if the bind() call to set
4047 an explicit outgoing interface failed. This has been changed; it is now
4048 treated in the same way as a connect() failure.
4050 5. A reference to $sender_host_name in the part of a conditional expansion
4051 that was being skipped was still causing a DNS lookup. This no longer
4054 6. The def: expansion condition was not recognizing references to header lines
4055 that used bh_ and bheader_.
4057 7. Added the _cache feature to named lists.
4059 8. The code for checking quota_filecount in the appendfile transport was
4060 allowing one more file than it should have been.
4062 9. For compatibility with Sendmail, the command line option
4070 and sets the incoming protocol and host name (for trusted callers). The
4071 host name and its colon can be omitted when only the protocol is to be set.
4072 Note the Exim already has two private options, -pd and -ps, that refer to
4073 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4074 "s", but I don't think that's a major issue.
4076 10. A number of refactoring changes to the code, none of which should affect
4079 (a) The number of logging options was getting close to filling up the
4080 32-bit word that was used as a bit map. I have split them into two classes:
4081 those that are passed in the argument to log_write(), and those that are
4082 only ever tested independently outside of that function. These are now in
4083 separate 32-bit words, so there is plenty of room for expansion again.
4084 There is no change in the user interface or the logging behaviour.
4086 (b) When building, for example, log lines, the code previously used a
4087 macro that called string_cat() twice, in order to add two strings. This is
4088 not really sufficiently general. Furthermore, there was one instance where
4089 it was actually wrong because one of the argument was used twice, and in
4090 one call a function was used. (As it happened, calling the function twice
4091 did not affect the overall behaviour.) The macro has been replaced by a
4092 function that can join an arbitrary number of extra strings onto a growing
4095 (c) The code for expansion conditions now uses a table and a binary chop
4096 instead of a serial search (which was left over from when there were very
4097 few conditions). Also, it now recognizes conditions like "pam" even when
4098 the relevant support is not compiled in: a suitably worded error message is
4099 given if an attempt is made to use such a condition.
4101 11. Added ${time_interval:xxxxx}.
4103 12. A bug was causing one of the ddress fields not to be passed back correctly
4104 from remote delivery subprocesses. The field in question was not being
4105 subsequently used, so this caused to problems in practice.
4107 13. Added new log selectors queue_time and deliver_time.
4109 14. Might have fixed a bug in maildirsizefile handling that threw up
4110 "unexpected character" debug warnings, and recalculated the data
4111 unnecessarily. In any case, I expanded the warning message to give more
4114 15. Added the message "Restricted characters in address" to the statements in
4115 the default ACL that block characters like @ and % in local parts.
4117 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4118 Three changes have been made:
4120 (a) There was a serious bug; a negative response to MAIL caused the whole
4121 recipient domain to be cached as invalid, thereby blocking all messages
4122 to all local parts at the same domain, from all senders. This bug has
4123 been fixed. The domain is no longer cached after a negative response to
4124 MAIL if the sender used is not empty.
4126 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4129 (c) A new callout option, "use_sender" has been added for people who want
4130 the modified behaviour.
4136 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4139 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4140 indeed breaks things for older releases.
4142 3. Added additional logging to the case where there is a problem reading data
4143 from a filter that is running in a subprocess using a pipe, in order to
4144 try to track down a specific problem.
4146 4. Testing facility fudge: when running in the test harness and attempting
4147 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4148 getting "No route to host". Convert this to a timeout.
4150 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4153 6. Some OS don't have socklen_t but use size_t instead. This affects the
4154 fifth argument of getsockopt() amongst other things. This is now
4155 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4156 can be set for individual OS. I have set it for SunOS5, OSF1, and
4157 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4158 some earlier ones do not.
4160 7. Change 4.30/15 was not doing the test caselessly.
4162 8. The standard form for an IPv6 address literal was being rejected by address
4163 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4164 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4165 this, as well as the form without the "IPv6" on the front (but only when
4166 address literals are enabled, of course).
4168 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4170 10. Exim crashed if a message with an empty sender address specified by -f
4171 encountered a router with an errors_to setting. This could be provoked only
4172 by a command such as
4176 where an empty string was supplied; "<>" did not hit this bug.
4178 11. Installed PCRE release 4.5.
4180 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4181 remained set. It is now erased.
4183 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4184 times from message ids (which are base 36 rather than the normal 62).
4186 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4187 were being counted as actual protocol errors, and logged if the log
4188 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4189 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4190 and DATA following a set of rejected RCPTs do not count as protocol errors.
4191 In other words, Exim assumes they were pipelined, though this may not
4192 actually be the case. Of course, in all cases the client gets an
4193 appropriate error code.
4195 15. If a lookup fails in an ACL condition, a message about the failure may
4196 be available; it is used if testing the ACL cannot continue, because most
4197 such messages specify what the cause of the deferral is. However, some
4198 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4199 that caused an old message to be retained and used if a later statement
4200 caused a defer, replacing the real cause of the deferral.
4202 16. If an IP address had so many PTR records that the DNS lookup buffer
4203 was not large enough to hold them, Exim could crash while trying to process
4204 the truncated data. It now detects and logs this case.
4206 17. Further to 4.21/58, another change has been made: if (and only if) the
4207 first line of a message (the first header line) ends with CRLF, a bare LF
4208 in a subsequent header line has a space inserted after it, so as not to
4209 terminate the header.
4211 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4212 unnecessarily, used atoi() instead of strtol(), and didn't check the
4213 termination when getting file sizes from file names by regex.
4215 19. Completely re-implemented the support for maildirsize files, in the light
4216 of a number of problems with the previous contributed implementation
4217 (4.30/29). In particular:
4219 . If the quota is zero, the maildirsize file is maintained, but no quota is
4222 . If the maildir directory does not exist, it is created before any attempt
4223 to write a maildirsize file.
4225 . The quota value in the file is just a cache; if the quota is changed in
4226 the transport, the new value overrides.
4228 . A regular expression is available for excluding directories from the
4231 20. The autoreply transport checks the characters in options that define the
4232 message's headers; it allows continued headers, but it was checking with
4233 isspace() after an embedded newline instead of explicitly looking for a
4236 21. If all the "regular" hosts to which an address was routed had passed their
4237 expiry times, and had not reached their retry times, the address was
4238 bounced, even if fallback hosts were defined. Now Exim should go on to try
4241 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4242 equivalent code in the SMTP transport. Some hosts send humungous responses
4243 to HELO/EHLO, more than 1024 it seems.
4245 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4246 gives compiler warnings in some environments. I've now done it "properly",
4249 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4250 (because of problems with the built-in one) was declared to return uschar *
4251 instead of char *, causing compiler failure.
4253 25. Fixed a file descriptor leak when processing alias/forward files.
4255 26. Fixed a minor format string issue in dbfn.c.
4257 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4259 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4260 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4263 29. When a sender address is verified, it is cached, to save repeating the test
4264 when there is more than one recipient in a message. However, when the
4265 verification involves a callout, it is possible for different callout
4266 options to be set for different recipients. It is too complicated to keep
4267 track of this in the cache, so now Exim always runs a verification when a
4268 callout is required, relying on the callout cache for the optimization.
4269 The overhead is duplication of the address routing, but this should not be
4272 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4273 to be verified with callout=postmaster, and the main callout worked but the
4274 postmaster check failed, the verification correctly failed. However, if a
4275 subsequent RCPT command asked for sender verification *without* the
4276 postmaster check, incorrect caching caused this verification also to fail,
4279 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4280 it was not caching the DNS options (qualify_single, search_parents) that
4281 were used when the lookup failed. A subsequent lookup with different
4282 options therefore always gave the same answer, though there were cases
4283 where it should not have. (Example: a "domains = !$mx_any" option on a
4284 dnslookup router: the "domains" option is always processed without any
4285 widening, but the router might have qualify_single set.) Now Exim uses the
4286 cached value only when the same options are set.
4288 32. Added John Jetmore's "exipick" utility to the distribution.
4290 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4291 than a timeout (e.g. a certificate is required, and is not provided), an
4292 Exim server now closes the connection immediately. Previously it waited for
4293 the client to close - but if the client is SSL, it seems that they each
4294 wait for each other, leading to a delay before one of them times out.
4296 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4297 maintained 0.8.x compatibility because I don't think many are using it, and
4298 it is clearly obsolete.
4300 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4303 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4304 client certificate was expired. A simple patch fixes this, though I don't
4305 understand the full logic of why the verify callback is called multiple
4308 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4309 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4310 which causes problems with some clients (such as the Certicom SSL Plus
4311 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4312 disables the coutermeasure allowing Eudora to connect."
4314 38. Exim was not checking that a write() to a log file succeeded. This could
4315 lead to Bad Things if a log got too big, in particular if it hit a file
4316 size limit. Exim now panics and dies if it cannot write to a log file, just
4317 as it does if it cannot open a log file.
4319 39. Modified OS/Makefile-Linux so that it now contains
4321 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4323 The two -D definitions ensure that Exim is compiled with large file
4324 support, which makes it possible to handle log files that are bigger than
4327 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4328 instance) a domain was checked against a named list that involved a lookup,
4329 causing $domain_data to be set, then another domain was checked against the
4330 same list, then the first domain was re-checked, the value of $domain_data
4331 after the final check could be wrong. In particular, if the second check
4332 failed, it could be set empty. This bug probably also applied to
4335 41. The strip_trailing_dot option was not being applied to the address given
4336 with the -f command-line option.
4338 42. The code for reading a message's header from the spool was incrementing
4339 $received_count, but never initializing it. This meant that the value was
4340 incorrect (doubled) while delivering a message in the same process in which
4341 it was received. In the most common configuration of Exim, this never
4342 happens - a fresh exec is done - but it can happen when
4343 deliver_drop_privilege is set.
4345 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4346 it now includes up to 150 characters of the unexpected data in the log
4349 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4350 and building data strings. The size of both of these buffers was 10 000
4351 bytes - far larger than anybody would *ever* want, thought I. Needless to
4352 say, somebody hit the limit. I have increased the maximum line length to
4353 20 000 and the maximum data length of concatenated lines to 100 000. I have
4354 also fixed two bugs, because there was no checking on these buffers. Tsk,
4355 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4356 buffer is too small.
4358 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4359 lsearch lookups. Now it does.
4361 46. When parsing a route_list item in a manualroute router, a fixed-length
4362 buffer was used for the list of hosts. I made this 1024 bytes long,
4363 thinking that nobody would ever have a list of hosts that long. Wrong.
4364 Somebody had a whole pile of complicated expansion conditions, and the
4365 string was silently truncated, leading to an expansion error. It turns out
4366 that it is easier to change to an unlimited length (owing to other changes
4367 that have happened since this code was originally written) than to build
4368 structure for giving a limitation error. The length of the item that
4369 expands into the list of hosts is now unlimited.
4371 47. The lsearch lookup could not handle data where the length of text line was
4372 more than 4095 characters. Such lines were truncated, leading to shortened
4373 data being returned. It should now handle lines of any length.
4375 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4376 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4379 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4380 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4381 no longer happens in comment lines. A list of replacements is now placed
4382 at the head of all of the source files, except those whose only change is
4383 to replace PERL_COMMAND in the very first #! line.
4385 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4386 messages on the queue, with a bottom-up merge sort, using code contributed
4387 by Michael Haardt. This should make operations like -bp somewhat faster on
4388 large queues. It won't affect queue runners, except when queue_run_in_order
4391 51. Installed eximstats 1.31 in the distribution.
4393 52. Added support for SRV lookups to the dnslookup router.
4395 53. If an ACL referred to $message_body or $message_body_end, the value was not
4396 reset for any messages that followed in the same SMTP session.
4398 54. The store-handling optimization for building very long strings was not
4399 differentiating between the different store pools. I don't think this
4400 actually made any difference in practice, but I've tidied it.
4402 55. While running the routers to verify a sender address, $sender_address
4403 was still set to the sender address. This is wrong, because when routing to
4404 send a bounce to the sender, it would be empty. Therefore, I have changed
4405 it so that, while verifying a sender address, $sender_address is set to <>.
4406 (There is no change to what happens when verifying a recipient address.)
4408 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4409 target A or AAAA records (if not already returned) without resetting the
4410 qualify_single or search_parents options of the DNS resolver. These are
4411 inappropriate in this case because the targets of MX and SRV records must
4412 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4413 target that, when qualified, matched something in the local domain. These
4414 two options are now turned off when doing these lookups.
4416 57. It seems that at least some releases of Reiserfs (which does not have the
4417 concept of a fixed number of inodes) returns zero and not -1 for the
4418 number of available inodes. This interacted badly with check_spool_inodes,
4419 which assumed that -1 was the "no such thing" setting. What I have done is
4420 to check that the total number of inodes is greater than zero before doing
4421 the test of how many are available.
4423 58. When a "warn" ACL statement has a log_message modifier, the message is
4424 remembered, and not repeated. This is to avoid a lot of repetition when a
4425 message has many recipients that cause the same warning to be written.
4426 Howewer, Exim was preserving the list of already written lines for an
4427 entire SMTP session, which doesn't seem right. The memory is now reset if a
4428 new message is started.
4430 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4431 debugging output unless log_rewrite was also set.
4433 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4434 of (int)(handle) when we know that handle contains (void *)(-1).
4436 61. The Exim daemon panic-logs an error return when it closes the incoming
4437 connection. However "connection reset by peer" seems to be common, and
4438 isn't really an error worthy of noting specially, so that particular error
4441 62. When Exim is trying to find all the local interfaces, it used to panic and
4442 die if the ioctl to get the interface flags failed. However, it seems that
4443 on at least one OS (Solaris 9) it is possible to have an interface that is
4444 included in the list of interfaces, but for which you get a failure error
4445 for this call. This happens when the interface is not "plumbed" into a
4446 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4447 failure of the "get flags" call assumes that the interface is down.
4449 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4450 makes life easier for people who are doing arithmetic on fields extracted
4451 from dates, where you often get leading zeros that should not be
4452 interpreted as octal.
4454 64. Added qualify_domain to the redirect router, to override the global
4457 65. If a pathologically long header line contained very many addresses (the
4458 report of this problem mentioned 10 000) and each of them was rewritten,
4459 Exim could use up a very large amount of memory. (It kept on making new
4460 copies of the header line as it rewrote, and never released the old ones.)
4461 At the expense of a bit more processing, the header rewriting function has
4462 been changed so that it no longer eats memory in this way.
4464 66. The generation of the Received: header has been moved from the time that a
4465 message starts to be received, to the time that it finishes. The timestamp
4466 in the Received: header should now be very close to that of the <= log
4467 line. There are two side-effects of this change:
4469 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4470 logged header lines no longer include the local Received: line, because
4471 it has not yet been created. The same applies to a copy of the message
4472 that is returned to a non-SMTP sender when a message is rejected.
4474 (b) When a filter file is tested using -bf, no additional Received: header
4475 is added to the test message. After some thought, I decided that this
4478 This change does not affect the value of $received_for. It is still set
4479 after address rewriting, but before local_scan() is called.
4481 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4483 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4484 gave an unhelpful panic error message, and a defer error. I have managed to
4485 change this behaviour so that it now rejects any supplied certificate,
4486 which seems right, as the list of acceptable certificates is empty.
4488 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4489 gave an unhelpful defer error. I have not managed to make this reject any
4490 supplied certificates, but the error message it gives is "no certificate
4491 supplied", which is not helpful.
4493 70. exigrep's output now also includes lines that are not associated with any
4494 message, but which match the given pattern. Implemented by a patch from
4495 Martin Sluka, which also tidied up the Perl a bit.
4497 71. Recipient callout verification, like sender verification, was using <> in
4498 the MAIL FROM command. This isn't really the right thing, since the actual
4499 sender may affect whether the remote host accepts the recipient or not. I
4500 have changed it to use the actual sender in the callout; this means that
4501 the cache record is now keyed on a recipient/sender pair, not just the
4502 recipient address. There doesn't seem to be a real danger of callout loops,
4503 since a callout by the remote host to check the sender would use <>.
4504 [SEE ABOVE: changed after hitting problems.]
4506 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4507 temporary errors. However, in the case of such a code being given after
4508 the end of a data transmission (i.e. after ".") Exim was failing to write
4509 a retry record for the message. (Yes, there was some broken host that was
4510 actually sending 8xx at this point.)
4512 73. An unknown lookup type in a host list could cause Exim to panic-die when
4513 the list was checked. (An example that provoked this was putting <; in the
4514 middle of a list instead of at the start.) If this happened during a DATA
4515 ACL check, a -D file could be left lying around. This kind of configuration
4516 error no longer causes Exim to die; instead it causes a defer errror. The
4517 incident is still logged to the main and panic logs.
4519 74. Buglet left over from Exim 3 conversion. The message "too many messages
4520 in one connection" was written to the rejectlog but not the mainlog, except
4521 when address rewriting (yes!) was being logged.
4523 75. Added write_rejectlog option.
4525 76. When a system filter was run not as root (that is, when system_filter_user
4526 was set), the values of the $n variables were not being returned to the
4527 main process; thus, they were not subsequently available in the $sn
4530 77. Added +return_path_on_delivery log selector.
4532 78. A connection timeout was being treated differently from recipients deferred
4533 when testing hosts_max_try with a message that was older than the host's
4534 retry timeout. (The host should not be counted, thus allowing all hosts to
4535 be tried at least once before bouncing.) This may have been the cause of an
4536 occasionally reported bug whereby a message would remain on the queue
4537 longer than the retry timeout, but would be bounced if a delivery was
4538 forced. I say "may" because I never totally pinned down the problem;
4539 setting up timeout/retry tests is difficult. See also the next item.
4541 79. The ultimate address timeout was not being applied to errors that involved
4542 a combination of host plus message (for example, a timeout on a MAIL
4543 command). When an address resolved to a number of possible hosts, and they
4544 were not all tried for each delivery (e.g. because of hosts_max_try), a
4545 message could remain on the queue longer than the retry timeout.
4547 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4550 81. Fixed an obscure SMTP outgoing bug which required at least the following
4551 conditions: (a) there was another message waiting for the same server;
4552 (b) the server returned 5xx to all RCPT commands in the first message so
4553 that the message was not completed; (c) the server dropped the connection
4554 or gave a negative response to the RSET that Exim sends to abort the
4555 transaction. The observed case was a dropped connection after DATA that had
4556 been sent in pipelining mode. That is, the server had advertised PIPELINING
4557 but was not implementing it correctly. The effect of the bug was incorrect
4558 behaviour, such as trying another host, and this could lead to a crash.
4564 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4565 and daemon.c were passed as pointers to ints; they should have been
4566 pointers to socklen_t variables (which are typically unsigned ints).
4568 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4571 3. Fixed a really odd bug that affected only the testing scheme; patching a
4572 certain fixed string in the binary changed the value of another string that
4573 happened to be identical to the end of the original first string.
4575 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4576 name", it returns that address as the IP address. On some operating
4577 systems (e.g. Solaris), it also passes back the IP address string as the
4578 "host name". However, on others (e.g. Linux), it passes back an empty
4579 string. Exim wasn't checking for this, and was changing the host name to an
4580 empty string, assuming it had been canonicized.
4582 5. Although rare, it is permitted to have more than one PTR record for a given
4583 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4584 all the names associated with an address, because they do in Solaris.
4585 However, it seems that they do not in Linux for data that comes from the
4586 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4587 I found this out when I moved to a new Linux workstation and tried to run
4588 the Exim test suite.
4590 To get round this problem I have changed the code so that it now does its
4591 own call to the DNS to look up PTR records when searching for a host name.
4592 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4593 addresses that are only in /etc/hosts are still found.
4595 This behaviour is, however, controlled by an option called host_lookup_
4596 order, which defaults to "bydns:byaddr". If people want to use the other
4597 order, or indeed, just use one or the other means of lookup, they can
4598 specify it in this variable.
4600 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4601 some operating systems, this comes back from gethostbyaddr() as an empty
4602 string, and this is what Exim used to test for. However, it seems that in
4603 other systems, "." is yielded. Exim now tests for this case too.
4605 7. The values of check_spool_space and check_log_space are now held internally
4606 as a number of kilobytes instead of an absolute number of bytes. If a
4607 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4608 kilobyte. This means that much larger values can be stored.
4610 8. Exim monitor: an attempt to get the action menu when not actually pointing
4611 at a message produces an empty menu entitled "No message selected". This
4612 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4613 no entries in it ("Shell widget menu has zero width and/or height"). So I
4614 have added a single, blank menu entry in this case.
4616 9. Added ${quote_local_part.
4618 10. MIME decoding is now applied to the contents of Subject: header lines when
4621 11. Now that a reference to $sender_host_address automatically causes a reverse
4622 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4623 host lookup before query-style lookups in lists that might use this
4624 variable. This has therefore been abolished, and the "net-" prefix is no
4625 longer necessary for query-style lookups.
4627 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4628 have been a typo for LFLAGS, so it has been changed.
4630 13. The install script calls Exim with "-C /dev/null" in order to find the
4631 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4632 to be output. Howeve, since Exim outputs its version number before the
4633 error, it didn't break the script. It just looked ugly. I fixed this by
4634 always allowing "-C /dev/null" if the caller is root.
4636 14. Ignore overlarge ACL variable number when reading spool file - insurance
4637 against a later release with more variables having written the file.
4639 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4640 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4641 this, as well as the form without the "IPv6" on the front.
4643 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4644 OS/Makefile-Darwin file.
4646 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4647 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4649 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4650 variable. However, some OS use ints for inodes. I've added cast to long int
4651 to get rid of the compiler warning.
4653 19. I had forgotten to lock out "/../" in configuration file names when
4654 ALT_CONFIG_PREFIX was set.
4656 20. Routers used for verification do not need to specify transports. However,
4657 if such a router generated a host list, and callout was configured, Exim
4658 crashed, because it could not find a port number from the (non-existent)
4659 transport. It now assumes port 25 in this circumstance.
4661 21. Added the -t option to exigrep.
4663 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4664 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4665 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4666 them is set, LOOKUP_LSEARCH is forced.
4668 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4670 24. Added sender and host information to the "rejected by local_scan()" log
4671 line; previously there was no indication of these.
4673 25. Added .include_if_exists.
4675 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4676 Linux. It turns out that not all file systems support this. Apparently some
4677 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4678 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4679 not supported on the file descriptor, is now ignored when Exim is trying to
4680 sync a directory. This applies only to Linux.
4682 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4684 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4685 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4688 29. Added support for maildirsize files from supplied patch (modified a bit).
4690 30. The use of :fail: followed by an empty string could lead Exim to respond to
4691 sender verification failures with (e.g.):
4693 550 Verification failed for <xxx>
4694 550 Sender verify failed
4696 where the first response line was missing the '-' that indicates it is not
4697 the final line of the response.
4699 31. The loop for finding the name of the user that called Exim had a hardwired
4700 limit of 10; it now uses the value of finduser_retries, which is used for
4701 all other user lookups.
4703 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4706 33. Exim was neglecting to zero errno before one call of strtol() when
4707 expanding a string and expecting an integer value. On some systems this
4708 resulted in spurious "integer overflow" errors. Also, it was casting the
4709 result into an int without checking.
4711 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4712 did not work. The code looks as if it has *never* worked, though it appears
4713 to have been documented since at least releast 1.62. I have made it work.
4715 35. The "timeout_DNS" error in retry rules, also documented since at least
4716 1.62, also never worked. As it isn't clear exactly what this means, and
4717 clearly it isn't a major issue, I have abolished the feature by treating it
4718 as "timeout", and writing a warning to the main and panic logs.
4720 36. The display of retry rules for -brt wasn't always showing the error code
4723 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4724 timeout_connect_A, timeout_connect_MX.
4726 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4727 to the empty sender.
4729 39. The daemon was not analysing the content of -oX till after it had closed
4730 stderr and disconnected from the controlling terminal. This meant that any
4731 syntax errors were only noted on the panic log, and the return code from
4732 the command was 0. By re-arranging the code a little, I've made the
4733 decoding happen first, so such errors now appear on stderr, and the return
4734 code is 1. However, the actual setting up of the sockets still happens in
4735 the disconnected process, so errors there are still only recorded on the
4738 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4739 connections (as happens on some IP stacks) was logged at start up time as
4740 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4741 it from "IPv6 and IPv4", which means that two separate sockets are being
4744 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4745 says whether AF_INET or AF_INET6 was passed as an argument.
4747 42. Exiwhat output was messed up when time zones were included in log
4750 43. Exiwhat now gives more information about the daemon's listening ports,
4751 and whether -tls-on-connect was used.
4753 44. The "port" option of the smtp transport is now expanded.
4755 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4756 silently ignored. Now an error message is written to the main and panic
4759 46. There's a new ACL modifier called "logwrite" which writes to a log file
4760 as soon as it is encountered.
4762 47. Added $local_user_uid and $local_user_gid at routing time.
4764 48. Exim crashed when trying to verify a sender address that was being
4767 49. Exim was recognizing only a space character after ".include". It now also
4768 recognizes a tab character.
4770 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4771 extracting the relevant information from the specification. The man page no
4772 longer contains scrambled data for the -d option, and I've added a section
4773 at the front about calling Exim under different names.
4775 51. Added "extra_headers" argument to the "mail" command in filter files.
4777 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4780 53. Installed eximstats 1.29.
4782 54. Added transport_filter_timeout as a generic transport option.
4784 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4785 Cc: header lines. This was required by RFC 822, but it not required by RFC
4788 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4789 incoming messages that did not have them. Now it does so only if the
4790 message originates locally, that is, if there is no associated remote host
4791 address. When Resent- header lines are present, this applies to the Resent-
4792 lines rather than the non-Resent- lines.
4794 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4795 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4797 58. Messages for configuration errors now include the name of the main
4798 configuration file - useful now that there may be more than one file in a
4799 list (.included file names were always shown).
4801 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4802 for those rare installations that do not start the daemon as root or run it
4803 setuid root. I've cut out the call to initgroups() if the daemon is not
4806 60. The Exim user and group can now be bound into the binary as text strings
4807 that are looked up at the start of Exim's processing.
4809 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4811 62. Added $mailstore_basename variable.
4813 63. Installed patch to sieve.c from Michael Haardt.
4815 64. When Exim failed to open the panic log after failing to open the main log,
4816 the original message it was trying to log was written to stderr and debug
4817 output, but if they were not available (the usual case in production), it
4818 was lost. Now it is written to syslog before the two lines that record the
4819 failures to open the logs.
4821 65. Users' Exim filters run in subprocesses under the user's uid. It is
4822 possible for a "deliver" command or an alias in a "personal" command to
4823 provoke an address rewrite. If logging of address rewriting is configured,
4824 this fails because the process is not running as root or exim. There may be
4825 a better way of dealing with this, but for the moment (because 4.30 needs
4826 to be released), I have disabled address rewrite logging when running a
4827 filter in a non-root, non-exim process.
4833 1. The buildconfig auxiliary program wasn't quoting the value set for
4834 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4835 not defined. This bug was masked in 4.22 by the effect that was fixed in
4838 2. Some messages that were rejected after a message id was allocated were
4839 shown as "incomplete" by exigrep. It no longer does this for messages that
4840 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4842 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4843 have allow_domain_literals set, the ID did not get logged in the <= line.
4844 Domain literals are now always recognized in Message-ID: header lines.
4846 4. The first argument for a ${extract expansion item is the key name or field
4847 number. Leading and trailing spaces in this item were not being ignored,
4848 causing some misleading effects.
4850 5. When deliver_drop_privilege was set, single queue runner processes started
4851 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4852 same command in the process it spins off) were not dropping privilege.
4854 6. When the daemon running as "exim" started a queue runner, it always
4855 re-executed Exim in the spun-off process. This is a waste of effort when
4856 deliver_drop_privilege is set. The new process now just calls the
4857 queue-runner function directly.
4863 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4866 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4867 ignored. Though the use of -oP was forcing the writing of a pid file, it
4868 was always written to the default place.
4870 3. If the message "no IP address found for host xxxx" is generated during
4871 incoming verification, it is now followed by identification of the incoming
4872 connection (so you can more easily find what provoked it).
4874 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4876 5. Added some features to "harden" Exim a bit more against certain attacks:
4878 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4879 be put in Local/Makefile. This is like the never_users runtime option,
4880 but it cannot be overridden. The default setting is "root".
4882 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4883 prefix string with which any file named in a -C command line option
4886 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4887 is retained for -C and -D only if the caller of Exim is root. Without
4888 it, the exim user may also use -C and -D and retain privilege.
4890 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4891 command line option is disabled.
4893 6. Macro names set by the -D option must start with an upper case letter, just
4894 like macro names defined in the configuration file.
4896 7. Added "dereference=" facility to LDAP.
4898 8. Two instances of the typo "uknown" in the source files are fixed.
4900 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4901 the Configure-Makefile script screwed up while processing it.
4903 10. Incorporated PCRE 4.4.
4905 11. The SMTP synchronization check was not operating right at the start of an
4906 SMTP session. For example, it could not catch a HELO sent before the client
4907 waited for the greeting. There is now a check for outstanding input at the
4908 point when the greeting is written. Because of the duplex, asynchronous
4909 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4910 way, but not yet received, when the check is performed.
4912 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4913 on TCP/IP sockets, because this apparently causes some broken clients to
4916 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4917 unchanged) from the Cygwin maintainer.
4919 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4920 was supported instead of testing for mbx. Effectively a typo.
4922 15. The spa authenticator server code was not checking that the input it
4923 received was valid base64.
4925 16. The debug output line for the "set" modifier in ACLs was not showing the
4926 name of the variable that was being set.
4928 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4930 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4931 lookup on its own. Something else was needed to trigger the lookup. For
4932 example, a match in host_lookup or the need for a host name in a host list.
4933 Now, if $sender_host_name is referenced and the host name has not yet been
4934 looked up, a lookup is performed. If the lookup fails, the variable remains
4935 empty, and $host_lookup_failed is set to "1".
4937 19. Added "eqi" as a case-independent comparison operator.
4939 20. The saslauthd authentication condition could segfault if neither service
4940 nor realm was specified.
4942 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4943 error message that was logged was misleading, and incoming SMTP
4944 connections were dropped. The message is now more accurate, and temporary
4945 errors are given to SMTP connections.
4947 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4948 (except RSET) with a 421 error, until QUIT is received. However, it was
4949 failing to send a response to QUIT.
4951 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4952 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4953 ACL is now run after the helo_try_verify_hosts code.
4955 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4956 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4957 case-independent, so other case variants are also recognized). Apparently
4958 some systems use these upper case variants.
4960 25. If more than two messages were waiting for the same host, and a transport
4961 filter was specified for the transport, Exim sent two messages over the
4962 same TCP/IP connection, and then failed with "socket operation on non-
4963 socket" when it tried to send the third.
4965 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4967 27. The extern definition of crypt16() in expand.c was not being excluded when
4968 the OS had its own crypt16() function.
4970 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4971 as a preferred synonym for return_size_limit, both as an option and as an
4974 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4976 30. Changed the default configuration ACL to relax the local part checking rule
4977 for addresses that are not in any local domains. For these addresses,
4978 slashes and pipe symbols are allowed within local parts, but the sequence
4979 /../ is explicitly forbidden.
4981 31. SPA server authentication was not clearing the challenge buffer before
4984 32. log_message in a "warn" ACL statement was writing to the reject log as
4985 well as to the main log, which contradicts the documentation and doesn't
4986 seem right (because no rejection is happening). So I have stopped it.
4988 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4989 However, I am unable to do any testing of this.
4991 34. Fixed an infelicity in the appendfile transport. When checking directories
4992 for a mailbox, to see if any needed to be created, it was accidentally
4993 using path names with one or more superfluous leading slashes; tracing
4994 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4996 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4997 contents are added to the log line that is written for every discarded
4998 recipient. (Previously a log_message setting was ignored.)
5000 36. The ${quote: operator now quotes the string if it is empty.
5002 37. The install script runs exim in order to find its version number. If for
5003 some reason other than non-existence or emptiness, which it checks, it
5004 could not run './exim', it was installing it with an empty version number,
5005 i.e. as "exim-". This error state is now caught, and the installation is
5008 38. An argument was missing from the function that creates an error message
5009 when Exim fails to connect to the socket for saslauthd authentication.
5010 This could cause Exim to crash, or give a corrupted message.
5012 39. Added isip, isip4, and isip6 to ${if conditions.
5014 40. The ACL variables $acl_xx are now saved with the message, and can be
5015 accessed later in routers, transports, and filters.
5017 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5018 strings in the file are not string-expanded.
5020 42. If a MAIL command specified a SIZE value that was too large to fit into an
5021 int variable, the check against message_size_limit failed. Such values are
5022 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5023 day this will have to be increased, but I don't think I want to be around
5024 when emails are that large.
5031 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5032 iconv() is not standard in FreeBSD.
5034 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5035 IPv6 enabled. The observed symptom was a segmentation fault on return from
5036 the function os_common_find_running_interfaces() in src/os.c.
5038 3. In the check_special_case() function in daemon.c I had used "errno" as an
5039 argument name, which causes warnings on some systems. This was basically a
5040 typo, since it was named "eno" in the comments!
5042 4. The code that waits for the clock to tick (at a resolution of some fraction
5043 of a second) so as to ensure message-id uniqueness was always waiting for
5044 at least one whole tick, when it could have waited for less. [This is
5045 almost certainly not relevant at current processor speeds, where it is
5046 unlikely to ever wait at all. But we try to future-proof.]
5048 5. The function that sleeps for a time interval that includes fractions of a
5049 second contained a race. It did not block SIGALRM between setting the
5050 timer, and suspending (a couple of lines later). If the interval was short
5051 and the sigsuspend() was delayed until after it had expired, the suspension
5052 never ended. On busy systems this could lead to processes getting stuck for
5055 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5056 process, before it forks any delivery processes. The open lookup caching
5057 mechanism meant that the open file or database connection was passed into
5058 the delivery process. The problem was that delivery processes always tidy
5059 up cached lookup data. This could cause a problem for the next delivery
5060 process started by the queue runner, because the external queue runner
5061 process does not know about the closure. So the next delivery process
5062 still has data in the lookup cache. In the case of a file lookup, there was
5063 no problem because closing a file descriptor in a subprocess doesn't affect
5064 the parent. However, if the lookup was caching a connection to a database,
5065 the connection was closed, and the second delivery process was likely to
5066 see errors such as "PGSQL: query failed: server closed the connection
5067 unexpectedly". The problem has been fixed by closing all cached lookups
5068 in a queue runner before running a delivery process.
5070 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5071 seem to have the "const" qualifier which it has on other OS. I've
5074 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5075 *of the same type* (client or server) with the same public name that an
5076 error should be diagnosed.
5078 9. When Exim looked up a host name for an IP address, but failed to find the
5079 original IP address when looking up the host name (a safety check), it
5080 output the message "<ip address> does not match any IP for NULL", which was
5081 confusing, to say the least. The bug was that the host name should have
5082 appeared instead of "NULL".
5084 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5085 user that is built into the binary, and the -C or -D options is used, root
5086 privilege is dropped before the configuration file is read. In addition,
5087 logging is switched to stderr instead of the normal log files. If the
5088 configuration then re-defines the Exim user, the unprivileged environment
5089 is probably not what is expected, so Exim logs a panic warning message (but
5092 However, if deliver_drop_privilege is set, the unprivileged state may well
5093 be exactly what is intended, so the warning has been cut out in that case,
5094 and Exim is allowed to try to write to its normal log files.
5100 1. smtp_return_error_details was not giving details for temporary sender
5101 or receiver verification errors.
5103 2. Diagnose a configuration error if two authenticators have the same public
5106 3. Exim used not to create the message log file for a message until the first
5107 delivery attempt. This could be confusing when incoming messages were held
5108 for policy or load reasons. The message log file is now created at the time
5109 the message is received, and an initial "Received" line is written to it.
5111 4. The automatically generated man page for command line options had a minor
5112 bug that caused no ill effects; however, a more serious problem was that
5113 the procedure for building the man page automatically didn't always
5114 operate. Consequently, release 4.20 contains an out-of-date version. This
5115 shouldn't happen again.
5117 5. When building Exim with embedded Perl support, the script that builds the
5118 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5119 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5121 6. The freeze_tell option was not being used for messages that were frozen on
5122 arrival, either by an ACL or by local_scan().
5124 7. Added the smtp_incomplete_transaction log selector.
5126 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5127 was accepting AUTH without a new EHLO.
5129 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5130 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5133 10. Logging of TCP/IP connections (when configured) now happens in the main
5134 daemon process instead of the child process, so that the TCP/IP connection
5135 count is more accurate (but it can never be perfect).
5137 11. The use of "drop" in a nested ACL was not being handled correctly in the
5138 outer ACL. Now, if condition failure induced by the nested "drop" causes
5139 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5140 or "require"), the connection is dropped.
5142 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5143 that yield "discard" can now be used with an "accept" or a "discard" verb,
5144 but an error is generated for any others (because I can't see a useful way
5145 to define what should happen).
5147 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5148 are now processed in the same way as lines in the Exim configuration file.
5149 In particular, continuation lines are supported.
5151 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5153 15. Added -ti meaning -t -i.
5155 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5156 domains, and warn by logging if others are found.
5158 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5159 returned from GIFCONF when Exim is trying to find the list of interfaces on
5160 a host. The code in os.c has been modified to copy each ifreq to an aligned
5161 structure in all cases.
5163 Also, in some cases, the returned ifreq's were being copied to a 'struct
5164 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5165 means the last couple of bytes of an IPv6 address could be chopped if the
5166 ifreq contained only a normal sockaddr (14 bytes storage).
5168 18. Named domain lists were not supported in the hosts_treat_as_local option.
5169 An entry such as +xxxx was not recognized, and was treated as a literal
5172 19. Ensure that header lines added by a DATA ACL are included in the reject log
5173 if the ACL subsequently rejects the message.
5175 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5176 MD5 (which is deprecated).
5178 21. When testing a filter file using -bf, Exim was writing a message when it
5179 took the sender from a "From " line in the message, but it was not doing so
5180 when it took $return_path from a Return-Path: header line. It now does.
5182 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5183 with a valid header line field name (a series of printing characters
5184 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5186 23. Changed "disc" in the source to "disk" to conform to the documentation and
5187 the book and for uniformity.
5189 24. Ignore Sendmail's -Ooption=value command line item.
5191 25. When execve() failed while trying to run a command in a pipe transport,
5192 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5193 could be confused with a return value of 69 from the command itself. This
5194 has been changed to 127, the value the shell returns if it is asked to run
5195 a non-existent command. The wording for the related log line suggests a
5196 non-existent command as the problem.
5198 26. If received_header_text expands to an empty string, do not add a Received:
5199 header line to the message. (Well, it adds a token one on the spool, but
5200 marks it "old" so that it doesn't get used or transmitted.)
5202 27. Installed eximstats 1.28 (addition of -nt option).
5204 28. There was no check for failure on the call to getsockname() in the daemon
5205 code. This can fail if there is a shortage of resources on the system, with
5206 ENOMEM, for example. A temporary error is now given on failure.
5208 29. Contrary to the C standard, it seems that in some environments, the
5209 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5210 program. Exim now does this explicitly; it affects the formatting of
5211 timestamps using strftime().
5213 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5214 complaints. I've now initialized all the variables, to avoid this.
5216 32. Header lines added by a system filter were not being "seen" during
5217 transport-time rewrites.
5219 33. The info_callback() function passed to OpenSSL is set up with type void
5220 (*)(SSL *, int, int), as described somewhere. However, when calling the
5221 function (actually a macro) that sets it up, the type void(*)() is
5222 expected. I've put in a cast to prevent warnings from picky compilers.
5224 34. If a DNS black list lookup found a CNAME record, but there were no A
5225 records associated with the domain it pointed at, Exim crashed.
5227 35. If a DNS black list lookup returned more than one A record, Exim ignored
5228 all but the first. It now scans all returned addresses if a particular IP
5229 value is being sought. In this situation, the contents of the
5230 $dnslist_value variable are a list of all the addresses, separated by a
5233 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5234 to accept a host name and all its aliases if the forward lookup for any of
5235 them yielded the IP address of the incoming connection. Now it accepts only
5236 those names whose forward lookup yields the correct IP address. Any other
5237 names are discarded. This closes a loophole whereby a rogue DNS
5238 administrator could create reverse DNS records to break through a
5239 wildcarded host restriction in an ACL.
5241 37. If a user filter or a system filter that ran in a subprocess used any of
5242 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5243 the wrong values were passed to the pipe command ($thisaddress had the
5244 value of $0, $0 had the value of $1, etc). This bug was introduced by
5245 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5247 38. Improved the line breaking for long SMTP error messages from ACLs.
5248 Previously, if there was no break point between 40 and 75 characters, Exim
5249 left the rest of the message alone. Two changes have been made: (a) I've
5250 reduced the minimum length to 35 characters; (b) if it can't find a break
5251 point between 35 and 75 characters, it looks ahead and uses the first one
5252 that it finds. This may give the occasional overlong line, but at least the
5253 remaining text gets split now.
5255 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5256 file descriptors might be low, and that setting 1000 would always raise it.
5257 It turns out that in some environments, the limit is already over 1000 and
5258 that lowering it causes trouble. So now Exim takes care not to decrease it.
5260 40. When delivering a message, the value of $return_path is set to $sender_
5261 address at the start of routing (routers may change the value). By an
5262 oversight, this default was not being set up when an address was tested by
5263 -bt or -bv, which affected the outcome if any router or filter referred to
5266 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5267 log only when "message" or "log_message" are set. However, if one of the
5268 conditions was an address verification, or a call to a nested ACL, the
5269 messages generated by the underlying test were being passed through. This
5270 no longer happens. The underlying message is available in $acl_verify_
5271 message for both "message" and "log_message" expansions, so it can be
5272 passed through if needed.
5274 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5275 new expansion $bh_ to give the encoded byte string without charset
5276 translation. Translation happens only if iconv() is available; HAVE_ICONV
5277 indicates this at build time. HEADERS_CHARSET gives the charset to
5278 translate to; headers_charset can change it in the configuration, and
5279 "headers charset" can change it in an individual filter file.
5281 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5282 that creates RFC 2047 encoded "words" labels them as that charset instead
5283 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5284 expansion operator; (ii) when Exim creates a From: line for a local
5285 message; (iii) when a header line is rewritten to include a "phrase" part.
5287 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5288 buggy, causing it to skip the first lines of messages whose message ID
5289 ended in 'D'. This would not have bitten before Exim release 4.14, because
5290 message IDs were unlikely to end in 'D' before then. The effect was to have
5291 incorrect size information for certain domains.
5293 45. #include "config.h" was missing at the start of the crypt16.c module. This
5294 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5297 46. If there was a timeout during a "random" callout check, Exim treated it as
5298 a failure of the random address, and carried on sending RSET and the real
5299 address. If the delay was just some slowness somewhere, the response to the
5300 original RCPT would be taken as a response to RSET and so on, causing
5301 mayhem of various kinds.
5303 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5304 when I implemented it. It didn't allow for the fact that some option values
5305 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5306 the right test for positive values.
5308 48. Domain names in DNS records are case-independent. Exim always looks them up
5309 in lower case. Some resolvers return domain names in exactly the case they
5310 appear in the zone file, that is, they may contain uppercase letters. Not
5311 all resolvers do this - some return always lower case. Exim was treating a
5312 change of case by a resolver as a change of domain, similar to a widening
5313 of a domain abbreviation. This triggered its re-routing code and so it was
5314 trying to route what was effectively the same domain again. This normally
5315 caused routing to fail (because the router wouldn't handle the domain
5316 twice). Now Exim checks for this case specially, and just changes the
5317 casing of the domain that it ultimately uses when it transmits the message
5320 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5323 50. If a filter generated a file delivery with a non-absolute name (possible if
5324 no home directory exists for the router), the forbid_file option was not
5327 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5328 the existing equality matching.
5330 52. Exim was using ints instead of ino_t variables in some places where it was
5331 dealing with inode numbers.
5333 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5334 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5335 called TMPDIR, and if it finds it is different, it changes its value.
5337 54. The smtp_printf() function is now made available to local_scan() so
5338 additional output lines can be written before returning. There is also an
5339 smtp_fflush() function to enable the detection of a dropped connection.
5340 The variables smtp_input and smtp_batched_input are exported to
5343 55. Changed the default runtime configuration: the message "Unknown user"
5344 has been removed from the ACL, and instead placed on the localuser router,
5345 using the cannot_route_message feature. This means that any verification
5346 failures that generate their own messages won't get overridden. Similarly,
5347 the "Unrouteable address" message that was in the ACL for unverifiable
5348 relay addresses has also been removed.
5350 56. Added hosts_avoid_esmtp to the smtp transport.
5352 57. The exicyclog script was not checking for the esoteric option
5353 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5354 will work only if exicyclog is run under the appropriate euid.
5356 58. Following a discussion on the list, the rules by which Exim recognises line
5357 endings on incoming messages have been changed. The -dropcr and drop_cr
5358 options are now no-ops, retained only for backwards compatibility. The
5359 following line terminators are recognized: LF CRLF CR. However, special
5360 processing applies to CR:
5362 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5363 nor a local message in the state where . is a terminator.
5365 (ii) If a bare CR is encountered in a header line, an extra space is added
5366 after the line terminator so as not to end the header. The reasoning
5367 behind this is that bare CRs in header lines are most likely either
5368 to be mistakes, or people trying to play silly games.
5370 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5371 was being incorrectly given as 18 bytes larger than it should have been.
5372 This is a VOB (very old bug).
5374 60. This may never have affected anything current, but just in case it has:
5375 When the local host is found other than at the start of a list of hosts,
5376 the local host, those with the same MX, and any that follow, are discarded.
5377 When the list in question was part of a longer list of hosts, the following
5378 hosts (not currently being processed) were also being discarded. This no
5379 longer happens. I'm not sure if this situation could ever has previously
5382 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5385 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5386 header. This is something that is recommended in a new Internet Draft, and
5387 is something that is documented as being done by Sendmail. There are two
5388 possible values. For messages generated by the autoreply transport, Exim
5391 Auto-Submitted: auto-replied
5393 whereas for all other generated messages (e.g. bounces) it adds
5395 Auto-Submitted: auto-generated
5397 63. The "personal" condition in filters now includes a test for the
5398 Auto-Submitted: header. If it contains the string "auto-" the message it
5399 not considered personal.
5401 64. Added rcpt_include_affixes as a generic transport option.
5403 65. Added queue_only_override (default true).
5405 66. Added the syslog_duplication option.
5407 67. If what should have been the first header line of a message consisted of
5408 a space followed by a colon, Exim was mis-interpreting it as a header line.
5409 It isn't of course - it is syntactically invalid and should therefore be
5410 treated as the start of the message body. The misbehaviour could have
5411 caused a number of strange effects, including loss of data in subsequent
5412 header lines, and spool format errors.
5414 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5415 client host had authenticated. This control can now be exercised by an ACL
5416 for more flexibility.
5418 69. By default, callouts do not happen when testing with -bh. There is now a
5419 variant, -bhc, which does actually run the callout code, including
5420 consulting and updating the callout cache.
5422 70. Added support for saslauthd authentication, courtesy of Alexander
5425 71. If statvfs() failed on the spool or log directories while checking their
5426 size for availability, Exim confusingly gave the error "space shortage".
5427 Furthermore, in debugging mode it crashed with a floating point exception.
5428 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5429 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5430 serious problem, Exim now writes to the main and panic logs when this
5431 happens, with details of the failure. It then refuses to accept the
5432 incoming message, giving the message "spool directory problem" or "log
5433 directory problem" with a 421 code for SMTP messages.
5435 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5436 0, 1, and 2 exist, because some OS complain for execs without them (see
5437 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5438 descriptors. However, the code omitted to check that the open succeeded,
5439 causing mysterious errors if for some reason the permissions on /dev/null
5440 got screwed. Now Exim writes a message to the main and panic logs, and
5441 bombs out if it can't open /dev/null.
5443 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5444 interact so that it is all more flexible. It is supposed to remain
5445 backwards compatible. Also added extra_local_interfaces.
5447 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5448 to bomb out with an assertion failure - to the client this appears as a
5449 connection drop. This problem occurs in the part of the code that was taken
5450 from the Samba project. Fortunately, the assertion is in a very simple
5451 function, so I have fixed this by reproducing the function inline in the
5452 one place where it is called, and arranging for authentication to fail
5453 instead of killing the process with assert().
5455 75. The SPA client code was not working when the server requested OEM rather
5456 than Unicode encoding.
5458 76. Added code to make require_files with a specific uid setting more usable in
5459 the case where statting the file as root fails - usually a non-root-mounted
5460 NFS file system. When this happens and the failure is EACCES, Exim now
5461 forks a subprocess and does the per-uid checking as the relevant uid.
5463 77. Added process_log_path.
5465 78. If log_file_path was not explicitly set, a setting of check_log_space or
5466 check_log_inodes was ignored.
5468 79. If a space check for the spool or log partitions fails, the incident is now
5469 logged. Of course, in the latter case the data may get lost...
5471 80. Added the %p formatting code to string_format() so that it can be used to
5472 print addresses in debug_print(). Adjusted all the address printing in the
5473 debugging in store.c to use %p rather than %d.
5475 81. There was a concern that a line of code in smtp_in.c could overflow a
5476 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5477 initially expressed, the concern was not well-founded, because trailing
5478 spaces are removed early. However, if the trailing spaces were followed by
5479 a NULL, they did not get removed, so the overflow was possible. Two fixes
5482 (a) I re-wrote the offending code in a cleaner fashion.
5483 (b) If an incoming SMTP command contains a NULL character, it is rejected
5486 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5487 runs initgroups(), so that if the Exim user is in any additional groups,
5488 they will be used during message reception.
5494 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob