1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PP/01 Date: & Message-Id: revert to normally being appended to a message,
10 only prepend for the Resent-* case. Fixes regression introduced in
11 Exim 4.70 by NM/22 for Bugzilla 607.
13 PP/02 Include check_rfc2047_length in configure.default because we're seeing
14 increasing numbers of administrators be bitten by this.
16 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
18 PP/03 Bugzilla 994: added openssl_options main configuration option.
20 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
22 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
24 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
26 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
27 without a peer certificate, leading to a segfault because of an
28 assumption that peers always have certificates. Be a little more
29 paranoid. Problem reported by Martin Tscholak.
31 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
32 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
33 NB: ClamAV planning to remove STREAM in "middle of 2010".
34 CL also introduces -bmalware, various -d+acl logging additions and
35 more caution in buffer sizes.
37 PP/09 Implemented reverse_ip expansion operator.
39 PP/10 Bugzilla 937: provide a "debug" ACL control.
41 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
43 PP/12 Bugzilla 973: Implement --version.
45 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
47 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
49 PP/15 Bugzilla 816: support multiple condition rules on Routers.
51 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
52 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
53 ignore trailing whitespace.
55 JJ/02 prevent non-panic DKIM error from being sent to paniclog
57 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
60 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
61 Notification from Dr Andrew Aitchison.
63 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
64 ExtendedDetectionInfo response format.
65 Notification from John Horne.
67 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
70 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
71 XSL and documented dependency on system catalogs, with examples of how
74 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
77 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
78 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
79 configuration file which is writeable by the Exim user or group.
81 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
82 of configuration files to cover files specified with the -C option if
83 they are going to be used with root privileges, not just the default
90 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
91 $data_path, and $header_path variables; fixed documentation bugs and
94 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
95 exipick to access non-standard spools, including the "frozen" queue
98 NM/01 Bugzilla 965: Support mysql stored procedures.
99 Patch from Alain Williams
101 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
103 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
104 Patch from Andreas Metzler
106 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
107 Patch from Kirill Miazine
109 NM/05 Bugzilla 671: Added umask to procmail example.
111 JJ/03 installed exipick 20100323.0, fixing doc bug
113 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
114 directory. Notification and patch from Dan Rosenberg.
116 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
118 TK/02 Improve log output when DKIM signing operation fails.
120 MH/01 Treat the transport option dkim_domain as a colon separated
121 list, not as a single string, and sign the message with each element,
122 omitting multiple occurences of the same signer.
124 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
125 Bugzilla 985, 986. Patch by Simon Arlott
127 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
128 Patch by Simon Arlott
130 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
131 MBX locking. Notification from Dan Rosenberg.
137 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
139 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
141 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
143 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
145 NM/04 Bugzilla 915: Flag broken perl installation during build.
151 TK/01 Added patch by Johannes Berg that expands the main option
152 "spamd_address" if it starts with a dollar sign.
154 TK/02 Write list of recipients to X-Envelope-Sender header when building
155 the mbox-format spool file for content scanning (suggested by Jakob
158 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
159 (http://www.dcc-servers.net/) support via dccifd. Activated by
160 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
162 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
163 by Mark Daniel Reidel <mr@df.eu>.
165 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
166 When building exim an external PCRE library is now needed -
167 PCRE is a system library on the majority of modern systems.
168 See entry on PCRE_LIBS in EDITME file.
170 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
171 conversation. Added nologin parameter to request.
172 Patch contributed by Kirill Miazine.
174 TF/01 Do not log submission mode rewrites if they do not change the address.
176 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
178 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
179 log files in place. Contributed by Roberto Lima.
181 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
183 TF/03 Bugzilla 615: When checking the local_parts router precondition
184 after a local_part_suffix or local_part_prefix option, Exim now
185 does not use the address's named list lookup cache, since this
186 contains cached lookups for the whole local part.
188 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
189 Robert Millan. Documentation is in experimental-spec.txt.
191 TF/04 Bugzilla 668: Fix parallel build (make -j).
193 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
195 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
196 Patch provided by Jan Srzednicki.
198 TF/05 Leading white space used to be stripped from $spam_report which
199 wrecked the formatting. Now it is preserved.
201 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
202 that they are available at delivery time.
204 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
206 TF/08 TLS error reporting now respects the incoming_interface and
207 incoming_port log selectors.
209 TF/09 Produce a more useful error message if an SMTP transport's hosts
210 setting expands to an empty string.
212 NM/06 Bugzilla 744: EXPN did not work under TLS.
213 Patch provided by Phil Pennock.
215 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
216 Patch provided by Richard Godbee.
218 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
219 acl_smtp_notquit, added index entry.
221 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
222 Patch provided by Eugene Bujak.
224 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
225 accept(). Patch provided by Maxim Dounin.
227 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
228 Patch provided by Phil Pennock.
230 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
232 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
233 Patch provided by Brad "anomie" Jorsch.
235 NM/14 Bugzilla 622: Added timeout setting to transport filter.
236 Patch provided by Dean Brooks.
238 TK/05 Add native DKIM support (does not depend on external libraries).
240 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
241 Patch provided by Graeme Fowler.
243 NM/16 Bugzilla 851: Documentation example syntax fix.
245 NM/17 Changed NOTICE file to remove references to embedded PCRE.
247 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
250 NM/19 Bugzilla 745: TLS version reporting.
251 Patch provided by Phil Pennock.
253 NM/20 Bugzilla 167: bool: condition support.
254 Patch provided by Phil Pennock.
256 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
257 clients. Patch provided by Phil Pennock.
259 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
260 Patch provided by Brad "anomie" Jorsch.
262 NM/23 Bugzilla 687: Fix misparses in eximstats.
263 Patch provided by Heiko Schlittermann.
265 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
266 Patch provided by Heiko Schlittermann.
268 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
269 plus update to original patch.
271 NM/26 Bugzilla 799: Documentation correction for ratelimit.
273 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
274 Patch provided by David Brownlee.
276 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
278 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
280 NM/30 Bugzilla 888: TLS documentation bugfixes.
282 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
284 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
285 Unlike the original bugzilla I have changed all shell scripts in src tree.
287 NM/33 Bugzilla 898: Transport filter timeout fix.
288 Patch by Todd Rinaldo.
290 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
291 Patch by Serge Demonchaux.
293 NM/35 Bugzilla 39: Base64 decode bug fixes.
294 Patch by Jakob Hirsch.
296 NM/36 Bugzilla 909: Correct connect() call in dcc code.
298 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
300 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
302 NM/39 Bugzilla 911: Fixed MakeLinks build script.
308 TK/01 Add preliminary DKIM support. Currently requires a forked version of
309 ALT-N's libdkim that I have put here:
310 http://duncanthrax.net/exim-experimental/
312 Note to Michael Haardt: I had to rename some vars in sieve.c. They
313 were called 'true' and it seems that C99 defines that as a reserved
314 keyword to be used with 'bool' variable types. That means you could
315 not include C99-style headers which use bools without triggering
316 build errors in sieve.c.
318 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
319 as mailq or other aliases. Changed the --help handling significantly
320 to do whats expected. exim_usage() emits usage/help information.
322 SC/01 Added the -bylocaldomain option to eximstats.
324 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
326 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
328 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
334 PH/01 Another patch from the Sieve maintainer.
336 PH/02 When an IPv6 address is converted to a string for single-key lookup
337 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
338 dots are used instead of colons so that keys in lsearch files need not
339 contain colons. This was done some time before quoting was made available
340 in lsearch files. However, iplsearch files do require colons in IPv6 keys
341 (notated using the quote facility) so as to distinguish them from IPv4
342 keys. This meant that lookups for IP addresses in host lists did not work
343 for iplsearch lookups.
345 This has been fixed by arranging for IPv6 addresses to be expressed with
346 colons if the lookup type is iplsearch. This is not incompatible, because
347 previously such lookups could never work.
349 The situation is now rather anomolous, since one *can* have colons in
350 ordinary lsearch keys. However, making the change in all cases is
351 incompatible and would probably break a number of configurations.
353 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
356 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
357 conversion specification without a maximum field width, thereby enabling
358 a rogue spamd server to cause a buffer overflow. While nobody in their
359 right mind would setup Exim to query an untrusted spamd server, an
360 attacker that gains access to a server running spamd could potentially
361 exploit this vulnerability to run arbitrary code as the Exim user.
363 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
364 $primary_hostname instead of what libspf2 thinks the hosts name is.
366 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
367 a directory entry by the name of the lookup key. Previously, if a
368 symlink pointed to a non-existing file or a file in a directory that
369 Exim lacked permissions to read, a lookup for a key matching that
370 symlink would fail. Now it is enough that a matching directory entry
371 exists, symlink or not. (Bugzilla 503.)
373 PH/03 The body_linecount and body_zerocount variables are now exported in the
376 PH/04 Added the $dnslist_matched variable.
378 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
379 This means they are set thereafter only if the connection becomes
382 PH/06 Added the client_condition to authenticators so that some can be skipped
383 by clients under certain conditions.
385 PH/07 The error message for a badly-placed control=no_multiline_responses left
386 "_responses" off the end of the name.
388 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
390 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
391 (without spaces) instead of just copying the configuration text.
393 PH/10 Added the /noupdate option to the ratelimit ACL condition.
395 PH/11 Added $max_received_linelength.
397 PH/12 Added +ignore_defer and +include_defer to host lists.
399 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
400 way in which PCRE > 7.0 is built.
402 PH/14 Implemented queue_only_load_latch.
404 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
405 MAIL command. The effect was to mangle the value on 64-bit systems.
407 PH/16 Another patch from the Sieve maintainer.
409 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
411 PH/18 If a system quota error occurred while trying to create the file for
412 a maildir delivery, the message "Mailbox is full" was not appended to the
413 bounce if the delivery eventually timed out. Change 4.67/27 below applied
414 only to a quota excession during the actual writing of the file.
416 PH/19 It seems that peer DN values may contain newlines (and other non-printing
417 characters?) which causes problems in log lines. The DN values are now
418 passed through string_printing() before being added to log lines.
420 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
421 and InterBase are left for another time.)
423 PH/21 Added message_body_newlines option.
425 PH/22 Guard against possible overflow in moan_check_errorcopy().
427 PH/23 POSIX allows open() to be a macro; guard against that.
429 PH/24 If the recipient of an error message contained an @ in the local part
430 (suitably quoted, of course), incorrect values were put in $domain and
431 $local_part during the evaluation of errors_copy.
437 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
438 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
441 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
442 issue a MAIL command.
444 PH/02 In an ACL statement such as
446 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
448 if a client was not listed at all, or was listed with a value other than
449 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
450 the condition was not true (as it should be), so access was not denied.
451 The bug was that the ! inversion was incorrectly passed on to the second
452 item. This has been fixed.
454 PH/03 Added additional dnslists conditions == and =& which are different from
455 = and & when the dns lookup returns more than one IP address.
457 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
458 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
460 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
461 FSYNC, which compiles an option called disable_fsync that allows for
462 bypassing fsync(). The documentation is heavily laced with warnings.
464 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
466 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
467 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
468 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
469 including adding "make clean"; (3) Added -fPIC when compiling the test
470 dynamically loaded module, to get rid of a warning.
472 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
473 message fails, move_frozen_messages = true and ignore_bounce_errors_after
474 = 0s. The bug is otherwise harmless.
476 PH/07 There was a bug in the dovecot authenticator such that the value of
477 $auth1 could be overwritten, and so not correctly preserved, after a
478 successful authentication. This usually meant that the value preserved by
479 the server_setid option was incorrect.
481 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
483 PH/09 Installed PCRE release 7.0.
485 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
486 run for batched SMTP input. It is now run at the start of every message
487 in the batch. While fixing this I discovered that the process information
488 (output by running exiwhat) was not always getting set for -bs and -bS
489 input. This is fixed, and it now also says "batched" for BSMTP.
491 PH/11 Added control=no_pipelining.
493 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
494 patch, slightly modified), and move the expansion of helo_data till after
495 the connection is made in the smtp transport (so it can use these
498 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
500 PH/14 Added log_selector = +pid.
502 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
504 PH/16 Add ${if forany and ${if forall.
506 PH/17 Added dsn_from option to vary the From: line in DSNs.
508 PH/18 Flush SMTP output before performing a callout, unless control =
509 no_callout_flush is set.
511 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
512 was true (the default) a successful delivery failed to delete the retry
513 item, thus causing premature timeout of the address. The bug is now
516 PH/20 Added hosts_avoid_pipelining to the smtp transport.
518 PH/21 Long custom messages for fakedefer and fakereject are now split up
519 into multiline reponses in the same way that messages for "deny" and
520 other ACL rejections are.
522 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
523 with slight modification.
525 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
526 draft, changing the syntax and factoring some duplicate code".
528 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
529 for deliveries of the second and subsequent messages over the same SMTP
532 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
533 ${reduce, with only minor "tidies".
535 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
537 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
538 expansion side effects.
540 PH/27 When a message times out after an over-quota error from an Exim-imposed
541 quota, the bounce message says "mailbox is full". This message was not
542 being given when it was a system quota that was exceeded. It now should
545 MH/03 Made $recipients available in local_scan(). local_scan() already has
546 better access to the recipient list through recipients_list[], but
547 $recipients can be useful in postmaster-provided expansion strings.
549 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
550 in the case of a MAIL command with additional options following the
551 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
552 were accidentally chopped off.
554 PH/29 SMTP synchronization checks are implemented when a command is read -
555 there is a check that no more input is waiting when there shouldn't be
556 any. However, for some commands, a delay in an ACL can mean that it is
557 some time before the response is written. In this time, more input might
558 arrive, invalidly. So now there are extra checks after an ACL has run for
559 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
560 pipelining has not been advertised.
562 PH/30 MH's patch to allow iscntrl() characters to be list separators.
564 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
565 returned in the SMTP response when smtp_return_error_details was false.
568 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
569 instead of the C I/O that was originally supplied, because problems were
572 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
573 Exim which did not show up earlier: it was assuming that a call to
574 SSL_CTX_set_info_callback() might give an error value. In fact, there is
575 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
576 was a macro that became an assignment, so it seemed to work. This has
577 changed to a proper function call with a void return, hence the compile
578 error. Exim's code has been fixed.
580 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
583 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
585 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
587 PH/37 If a message is not accepted after it has had an id assigned (e.g.
588 because it turns out to be too big or there is a timeout) there is no
589 "Completed" line in the log. When some messages of this type were
590 selected by exigrep, they were listed as "not completed". Others were
591 picked up by some special patterns. I have improved the selection
592 criteria to be more general.
594 PH/38 The host_find_failed option in the manualroute router can now be set
595 to "ignore", to completely ignore a host whose IP address cannot be
596 found. If all hosts are ignored, the behaviour is controlled by the new
597 host_all_ignored option.
599 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
600 homing or because of multiple MX records with /mx) generated more than
601 one IP address, and the following item turned out to be the local host,
602 all the secondary addresses of the first item were incorrectly removed
603 from the list, along with the local host and any following hosts (which
604 is what is supposed to happen).
606 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
607 whoever called Exim into the -H file. In the case of the daemon it was
608 behaving confusingly. When first started, it used values for whoever
609 started the daemon, but after a SIGHUP it used the Exim user (because it
610 calls itself on a restart). I have changed the code so that it now always
613 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
614 message are rejected with the same error (e.g. no authentication or bad
615 sender address), and a DATA command is nevertheless sent (as can happen
616 with PIPELINING or a stupid MUA), the error message that was given to the
617 RCPT commands is included in the rejection of the DATA command. This is
618 intended to be helpful for MUAs that show only the final error to their
621 PH/42 Another patch from the Sieve maintainer.
623 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
624 Eximstats - Fixed some broken HTML links and added missing column headers
626 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
627 columns for Rejects, Temp Rejects, Ham, and Spam rows.
629 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
631 PH/43 Yet another patch from the Sieve maintainer.
633 PH/44 I found a way to check for a TCP/IP connection going away before sending
634 the response to the final '.' that terminates a message, but only in the
635 case where the client has not sent further data following the '.'
636 (unfortunately, this is allowed). However, in many cases there won't be
637 any further data because there won't be any more messages to send. A call
638 to select() can be used: if it shows that the input is "ready", there is
639 either input waiting, or the socket has been closed. An attempt to read
640 the next input character can distinguish the two cases. Previously, Exim
641 would have sent an OK response which the client would never have see.
642 This could lead to message repetition. This fix should cure that, at
643 least in a lot of common cases.
645 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
646 advertised in response to EHLO.
652 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
653 fixed by 4.65/MH/01 (is this a record?) are fixed:
655 (i) An empty string was always treated as zero by the numeric comparison
656 operators. This behaviour has been restored.
658 (ii) It is documented that the numeric comparison operators always treat
659 their arguments as decimal numbers. This was broken in that numbers
660 starting with 0 were being interpreted as octal.
662 While fixing these problems I realized that there was another issue that
663 hadn't been noticed. Values of message_size_limit (both the global option
664 and the transport option) were treated as octal if they started with 0.
665 The documentation was vague. These values are now always treated as
666 decimal, and I will make that clear in the documentation.
672 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
673 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
676 MH/01 Don't check that the operands of numeric comparison operators are
677 integers when their expansion is in "skipping" mode (fixes bug
678 introduced by 4.64-PH/07).
680 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
681 child addresses, Exim now panics and dies. Previously, because the count
682 is held in a short int, deliveries were likely to be lost. As such a
683 large number of recipients for a single message is ridiculous
684 (performance will be very, very poor), I have chosen to impose a limit
685 rather than extend the field.
691 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
692 leftover -K file (the existence of which was triggered by #402).
693 While we were at it, introduced process PID as part of the -K
694 filename. This should rule out race conditions when creating
697 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
698 processing considerably. Previous code took too long for large mails,
699 triggering a timeout which in turn triggers #401.
701 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
702 in the DK code in transports.c. sendfile() is not really portable,
703 hence the _LINUX specificness.
705 TF/01 In the add_headers option to the mail command in an Exim filter,
706 there was a bug that Exim would claim a syntax error in any
707 header after the first one which had an odd number of characters
710 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
711 callout verification, Exim cached a "reject" for the entire domain. This
712 is correct for most verifications, but it is not correct for a recipient
713 verification with use_sender or use_postmaster set, because in that case
714 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
715 case of MAIL FROM:<> rejection from other early rejections (e.g.
716 rejection of HELO). When verifying a recipient using a non-null MAIL
717 address, the cache is ignored if it shows MAIL FROM:<> rejection.
718 Whatever the result of the callout, the value of the domain cache is
719 left unchanged (for any other kind of callout, getting as far as trying
720 RCPT means that the domain itself is ok).
722 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
725 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
726 manifest itself as EPIPE rather than ECONNECT. When tidying away a
727 session, the daemon ignores ECONNECT errors and logs others; it now
728 ignores EPIPE as well.
730 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
731 (quoted-printable decoding).
733 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
734 later the small subsequent patch to fix an introduced bug.
736 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
738 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
740 PH/08 An error is now given if message_size_limit is specified negative.
742 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
743 to be given (somewhat) arbitrary names.
745 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
748 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
749 miscellaneous code fixes
751 PH/10 Added the log_reject_target ACL modifier to specify where to log
754 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
755 hostname. This is wrong, because it relates to the incoming message (and
756 probably the interface on which it is arriving) and not to the outgoing
757 callout (which could be using a different interface). This has been
758 changed to use the value of the helo_data option from the smtp transport
759 instead - this is what is used when a message is actually being sent. If
760 there is no remote transport (possible with a router that sets up host
761 addresses), $smtp_active_hostname is used.
763 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
764 tweaks were necessary in order to get it to work (see also 21 below):
765 (a) The code assumed that strncpy() returns a negative number on buffer
766 overflow, which isn't the case. Replaced with Exim's string_format()
768 (b) There were several signed/unsigned issues. I just did the minimum
769 hacking in of casts. There is scope for a larger refactoring.
770 (c) The code used strcasecmp() which is not a standard C function.
771 Replaced with Exim's strcmpic() function.
772 (d) The code set only $1; it now sets $auth1 as well.
773 (e) A simple test gave the error "authentication client didn't specify
774 service in request". It would seem that Dovecot has changed its
775 interface. Fortunately there's a specification; I followed it and
776 changed what the client sends and it appears to be working now.
778 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
781 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
782 address A is aliased to B and C, where B exists and C does not. Without
783 -v the output is "A verified" because verification stops after a
784 successful redirection if more than one address is generated. However,
785 with -v the child addresses are also verified. Exim was outputting "A
786 failed to verify" and then showing the successful verification for C,
787 with its parentage. It now outputs "B failed to verify", showing B's
788 parentage before showing the successful verification of C.
790 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
791 look up a TXT record in a specific list after matching in a combined
794 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
795 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
796 they consult the DNS. I had assumed they would set it the way they
797 wanted; and indeed my experiments on Linux seem to show that in some
798 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
799 To be on the safe side, however, I have now made the interface to
800 host_find_byname() similar to host_find_bydns(), with an argument
801 containing the DNS resolver options. The host_find_byname() function now
802 sets these options at its start, just as host_find_bydns() does. The smtp
803 transport options dns_qualify_single and dns_search_parents are passed to
804 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
805 of host_find_byname() use the default settings of RES_DEFNAMES
806 (qualify_single) but not RES_DNSRCH (search_parents).
808 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
809 spool_read_header() do less string testing, by means of a preliminary
810 switch on the second character of optional "-foo" lines. (This is
811 overdue, caused by the large number of possibilities that now exist.
812 Originally there were few.) While I was there, I also converted the
813 str(n)cmp tests so they don't re-test the leading "-" and the first
814 character, in the hope this might squeeze out yet more improvement.
816 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
817 flag allowing group syntax was set by the header_syntax check but not
818 turned off, possible causing trouble later; (2) The flag was not being
819 set at all for the header_verify test, causing "group"-style headers to
820 be rejected. I have now set it in this case, and also caused header_
821 verify to ignore an empty address taken from a group. While doing this, I
822 came across some other cases where the code for allowing group syntax
823 while scanning a header line wasn't quite right (mostly, not resetting
824 the flag correctly in the right place). These bugs could have caused
825 trouble for malformed header lines. I hope it is now all correct.
827 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
828 with the "reply" argument non-NULL. The code, however (which originally
829 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
830 but it didn't always do it. This confused somebody who was copying the
831 code for some other use. I have removed all the tests.
833 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
834 feature that was used to support insecure browsers during the U.S. crypto
835 embargo. It requires special client support, and Exim is probably the
836 only MTA that supported it -- and would never use it because real RSA is
837 always available. This code has been removed, because it had the bad
838 effect of slowing Exim down by computing (never used) parameters for the
839 RSA_EXPORT functionality.
841 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
842 authenticator to fail if there's a tab character in the incoming data
843 (there should never be unless someone is messing about, as it's supposed
844 to be base64-encoded). Also added, on Timo's advice, the "secured" option
845 if the connection is using TLS or if the remote IP is the same as the
846 local IP, and the "valid-client-cert option" if a client certificate has
849 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
850 authenticators. This can be used for authorization after authentication
851 succeeds. (In the case of plaintext, it servers for both authentication
854 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
855 if any retry times were supplied.
857 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
858 connection, where there is no client IP address to check. In this
859 situation, the verify now always succeeds.
861 PH/25 Applied John Jetmore's -Mset patch.
863 PH/26 Added -bem to be like -Mset, but loading a message from a file.
865 PH/27 In a string expansion for a processed (not raw) header when multiple
866 headers of the same name were present, leading whitespace was being
867 removed from all of them, but trailing whitespace was being removed only
868 from the last one. Now trailing whitespace is removed from each header
869 before concatenation. Completely empty headers in a concatenation (as
872 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
873 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
875 PH/29 [Removed. This was a change that I later backed out, and forgot to
876 correct the ChangeLog entry (that I had efficiently created) before
877 committing the later change.]
879 PH/30 Exim was sometimes attempting to deliver messages that had suffered
880 address errors (4xx response to RCPT) over the same connection as other
881 messages routed to the same hosts. Such deliveries are always "forced",
882 so retry times are not inspected. This resulted in far too many retries
883 for the affected addresses. The effect occurred only when there were more
884 hosts than the hosts_max_try setting in the smtp transport when it had
885 the 4xx errors. Those hosts that it had tried were not added to the list
886 of hosts for which the message was waiting, so if all were tried, there
887 was no problem. Two fixes have been applied:
889 (i) If there are any address or message errors in an SMTP delivery, none
890 of the hosts (tried or untried) are now added to the list of hosts
891 for which the message is waiting, so the message should not be a
892 candidate for sending over the same connection that was used for a
893 successful delivery of some other message. This seems entirely
894 reasonable: after all the message is NOT "waiting for some host".
895 This is so "obvious" that I'm not sure why it wasn't done
896 previously. Hope I haven't missed anything, but it can't do any
897 harm, as the worst effect is to miss an optimization.
899 (ii) If, despite (i), such a delivery is accidentally attempted, the
900 routing retry time is respected, so at least it doesn't keep
901 hammering the server.
903 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
904 in ${readsocket because some servers need this prod.
906 PH/32 Added some extra debug output when updating a wait-xxx database.
908 PH/33 The hint "could be header name not terminated by colon", which has been
909 given for certain expansion errors for a long time, was not being given
910 for the ${if def:h_colon_omitted{... case.
912 PH/34 The spec says: "With one important exception, whenever a domain list is
913 being scanned, $domain contains the subject domain." There was at least
914 one case where this was not true.
916 PH/35 The error "getsockname() failed: connection reset by peer" was being
917 written to the panic log as well as the main log, but it isn't really
918 panic-worthy as it just means the connection died rather early on. I have
919 removed the panic log writing for the ECONNRESET error when getsockname()
922 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
923 runs only) independently of the message's sender address. This meant
924 that, if the 4xx error was in fact related to the sender, a different
925 message to the same recipient with a different sender could confuse
926 things. In particualar, this can happen when sending to a greylisting
927 server, but other circumstances could also provoke similar problems.
928 I have changed the default so that the retry time for these errors is now
929 based a combination of the sender and recipient addresses. This change
930 can be overridden by setting address_retry_include_sender=false in the
933 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
934 remote server are returned as part of bounce messages. This was not
935 happening for LMTP over a pipe (the lmtp transport), but now it is the
936 same for both kinds of LMTP.
938 PH/38 Despite being documented as not happening, Exim was rewriting addresses
939 in header lines that were in fact CNAMEs. This is no longer the case.
941 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
942 and queue runs started by the daemon processed all messages. This has
943 been fixed so that -R and -S can now usefully be given with -q<time>.
945 PH/40 Import PCRE release 6.7 (fixes some bugs).
947 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
949 PH/42 Give an error if -q is specified more than once.
951 PH/43 Renamed the variables $interface_address and $interface_port as
952 $received_ip_address and $received_port, to make it clear that these
953 values apply to message reception, and not to the outgoing interface when
954 a message is delivered. (The old names remain recognized, of course.)
956 PH/44 There was no timeout on the connect() call when using a Unix domain
957 socket in the ${readsocket expansion. There now is.
959 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
960 be meaningful with "accept".
962 SC/01 Eximstats V1.43
963 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
965 SC/02 Eximstats V1.44
966 Use a glob alias rather than an array ref in the generated
967 parser. This improves both readability and performance.
969 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
970 Collect SpamAssassin and rejection statistics.
971 Don't display local sender or destination tables unless
972 there is data to show.
973 Added average volumes into the top table text output.
975 SC/04 Eximstats V1.46
976 Collect data on the number of addresses (recipients)
977 as well as the number of messages.
979 SC/05 Eximstats V1.47
980 Added 'Message too big' to the list of mail rejection
981 reasons (thanks to Marco Gaiarin).
983 SC/06 Eximstats V1.48
984 Mainlog lines which have GMT offsets and are too short to
985 have a flag are now skipped.
987 SC/07 Eximstats V1.49 (Alain Williams)
988 Added the -emptyok flag.
990 SC/08 Eximstats V1.50
991 Fixes for obtaining the IP address from reject messages.
993 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
994 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
995 whitesspace changes from 4.64-PH/27
997 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1000 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1001 are found, allow negative numbers in numeric criteria)
1003 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1005 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1006 to match changes made in 4.64-PH/43
1008 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1010 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1011 "domain" is actually an IP address, and force a failure. This locks out
1012 those revolvers/nameservers that support "A-for-A" lookups, in
1013 contravention of the specifications.
1015 PH/48 When a host name was looked up from an IP address, and the subsequent
1016 forward lookup of the name timed out, the host name was left in
1017 $sender_host_name, contrary to the specification.
1019 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1020 restricted to single-key lookups, Exim was not diagnosing an error if
1021 * or *@ was used with a query-style lookup.
1023 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1025 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1026 long ago, but noone interested enough thought of it. Let's just say that
1027 the "1.1" means that there are some new functions that weren't there at
1028 some point in the past.
1030 PH/51 Error processing for expansion failure of helo_data from an smtp
1031 transport during callout processing was broken.
1033 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1034 tested/used via the -bh/-bhc/-bs options.
1036 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1037 bug, fixed in subsequent PCRE releases).
1039 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1040 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1042 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1048 SC/01 Use a glob alias rather than an array ref in eximstats generated
1049 parser. This improves both readability and performance.
1051 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1052 Don't display local sender or destination tables in eximstats unless
1053 there is data to show.
1054 Added average volumes into the eximstats top table text output.
1056 SC/03 Collect data on the number of addresses (recipients) as well
1057 as the number of messages in eximstats.
1059 TF/01 Correct an error in the documentation for the redirect router. Exim
1060 does (usually) call initgroups() when daemonizing.
1062 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1063 with consistent privilege compared to when running as a daemon.
1065 TF/03 Note in the spec that $authenticated_id is not set for local
1066 submissions from trusted users.
1068 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1069 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1071 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1072 by adding some example configuration directives to the default
1073 configuration file. A little bit of work is required to uncomment the
1074 directives and define how usernames and passwords are checked, but
1075 there is now a framework to start from.
1077 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1078 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1079 without this. I don't know how relevant this is to other LDAP libraries.
1081 PH/02 Add the verb name to the "unknown ACL verb" error.
1083 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1085 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1087 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1088 directory not expanded when it should be if an expanded home directory
1089 was set for the address (which is overridden by the transport).
1091 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1094 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1095 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1096 because it is too late at that time, and has no effect.
1098 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1099 security issue with \' (bugzilla #107). I could not use the
1100 PQescapeStringConn() function, because it needs a PGconn value as one of
1103 PH/08 When testing addresses using -bt, indicate those final addresses that
1104 are duplicates that would not cause an additional delivery. At least one
1105 person was confused, thinking that -bt output corresponded to deliveries.
1106 (Suppressing duplicates isn't a good idea as you lose the information
1107 about possibly different redirections that led to the duplicates.)
1109 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1110 systems where poll() doesn't work, in particular OS X.
1112 PH/10 Added more information to debugging output for retry time not reached.
1114 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1115 operations in malware.c.
1117 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1120 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1121 syslog_duplication set false, log lines that would normally be written
1122 both the the main log and to the reject log were not written to syslog at
1125 PH/14 In the default configuration, change the use of "message" in ACL warn
1126 statements to "add_header".
1128 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1129 not followed by a command (e.g. "seen endif").
1131 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1132 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1135 PH/17 Added extra conditions to the default value of delay_warning_condition
1139 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1140 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1141 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1144 The Auto-Submitted: and various List- headers are standardised, whereas I
1145 don't think Precedence: ever was.
1147 PH/18 Refactored debugging code in route_finduser() to show more information,
1148 in particular, the error code if getpwnam() issues one.
1150 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1151 This is apparently needed in addition to the PH/07 change above to avoid
1152 any possible encoding problems.
1154 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1155 but not after initializing Perl.
1157 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1158 output them only if debugging. By default they are written stderr,
1159 apparently, which is not desirable.
1161 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1164 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1167 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1169 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1170 authenticated or an ident call has been made. Suppress the default
1171 values for $authenticated_id and $authenticated_sender (but permit -oMai
1172 and -oMas) when testing with -bh.
1174 PH/24 Re-jigged the order of the tests in the default configuration so that the
1175 tests for valid domains and recipients precede the DNS black list and CSA
1176 tests, on the grounds that those ones are more expensive.
1178 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1179 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1180 This bug exists in every version of Exim that I still have, right back to
1183 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1184 However, an attempt to turn on case-sensitivity in a regex key by
1185 including (?-i) didn't work because the subject string was already
1186 lowercased, and the effects were non-intuitive. It turns out that a
1187 one-line patch can be used to allow (?-i) to work as expected.
1193 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1194 other effects) broke the use of negated acl sub-conditions.
1196 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1199 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1200 "Deny" causes Exim to reject the incoming connection with a 554 error.
1201 Unfortunately, if there is a major crisis, such as a disk failure,
1202 tcp-wrappers gives "deny", whereas what one would like would be some
1203 kind of temporary error. A kludge has been added to help with this.
1204 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1205 554 error is used if errno is still zero or contains ENOENT (which occurs
1206 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1209 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1211 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1212 errors. Otherwise a message that provokes a temporary error (when other
1213 messages do not) can cause a whole host to time out.
1215 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1216 addresses were routed directly to files or pipes from a redirect router.
1217 File deliveries just didn't batch; pipe deliveries might have suffered
1220 PH/06 A failure to get a lock for a hints database would erroneously always say
1221 "Failed to get write lock", even when it was really a read lock.
1223 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1224 of 0600. This has been changed to use the value of the lockfile_mode
1225 option (which defaults to 0600).
1227 PH/08 Applied small patch from the Sieve maintainer.
1229 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1230 folder from quota calculations, a direct delivery into this folder messed
1231 up the contents of the maildirsize file. This was because the regex was
1232 used only to exclude .Trash (or whatever) when the size of the mailbox
1233 was calculated. There was no check that a delivery was happening into an
1234 excluded directory. This bug has been fixed by ignoring all quota
1235 processing for deliveries into excluded directories.
1237 PH/10 Added the maildirfolder_create_regex option to appendfile.
1243 PH/01 The code for finding all the local interface addresses on a FreeBSD
1244 system running IPv6 was broken. This may well have applied to all BSD
1245 systems, as well as to others that have similar system calls. The broken
1246 code found IPv4 interfaces correctly, but gave incorrect values for the
1247 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1248 that it would not match correctly against @[] and not recognize the IPv6
1251 PH/02 The ipliteral router was not recognizing addresses of the form user@
1252 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1254 PH/03 Added disable_ipv6.
1256 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1257 decoded form, because it is most often used to construct To: headers
1258 lines in autoreplies, and the decoded form may well be syntactically
1259 invalid. However, $reply_address has leading white space removed, and all
1260 newlines turned into spaces so that the autoreply transport does not
1263 PH/05 If group was specified without a user on a router, and no group or user
1264 was specified on a transport, the group from the router was ignored.
1266 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1267 for visible compile-time settings that can be used to change these
1268 numbers, for those that want even more. Backwards compatibility with old
1269 spool files has been maintained. However, going back to a previous Exim
1270 release will lost any variables that are in spool files.
1272 PH/07 Two small changes when running in the test harness: increase delay when
1273 passing a TCP/IP connection to a new process, in case the original
1274 process has to generate a bounce, and remove special handling of
1275 127.0.0.2 (sic), which is no longer necessary.
1277 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1278 be the same on different OS.
1280 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1283 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1284 whether --show-vars was specified or not
1286 JJ/02 exipick: Added support for new ACL variable spool format introduced
1289 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1290 syntactically invalid From: or Reply-to: line, and a filter used this to
1291 generate an autoreply, and therefore failed to obtain an address for the
1292 autoreply, Exim could try to deliver to a non-existent relative file
1293 name, causing unrelated and misleading errors. What now happens is that
1294 it logs this as a hard delivery error, but does not attempt to create a
1297 PH/11 The exinext utility has a -C option for testing purposes, but although
1298 the given file was scanned by exinext itself; it wasn't being passed on
1299 when Exim was called.
1301 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1302 an end-of-file indication when reading a command response.
1304 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1305 compiled. In many other places in Exim, IPv6 addresses are always
1306 recognized, so I have changed this. It also means that IPv4 domain
1307 literals of the form [IPV4:n.n.n.n] are now always recognized.
1309 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1310 used if the router is not running as root, for example, when verifying at
1311 ACL time, or when using -bh. The debugging output from this situation was
1312 non-existent - all you got was a failure to exec. I have made two
1315 (a) Failures to set uid/gid, the current directory, or a process leader
1316 in a subprocess such as that created by queryprogram now generate
1317 suitable debugging ouput when -d is set.
1319 (b) The queryprogram router detects when it is not running as root,
1320 outputs suitable debugging information if -d is set, and then runs
1321 the subprocess without attempting to change uid/gid.
1323 PH/15 Minor change to Makefile for building test_host (undocumented testing
1326 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1327 additional section of a DNS packet that returns MX or SRV records.
1328 Instead, it always explicitly searches for A/AAAA records. This avoids
1329 major problems that occur when a DNS server includes only records of one
1330 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1331 fixed another bug: if SRV records were looked up and the corresponding
1332 address records were *not* found in the additional section, the port
1333 values from the SRV records were lost.
1335 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1336 using the correct key (the original address) when searching the retry
1337 rules in order to find which one to use for generating the retry hint.
1339 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1340 adding the default one. Similarly, if it contains a Reply-To: header, the
1341 errors_reply_to option, if set, is not used.
1343 PH/19 When calculating a retry time, Exim used to measure the "time since
1344 failure" by looking at the "first failed" field in the retry record. Now
1345 it does not use this if it is later than than the arrival time of the
1346 message. Instead it uses the arrival time. This makes for better
1347 behaviour in cases where some deliveries succeed, thus re-setting the
1348 "first failed" field. An example is a quota failure for a huge message
1349 when small messages continue to be delivered. Without this change, the
1350 "time since failure" will always be short, possible causing more frequent
1351 delivery attempts for the huge message than are intended.
1352 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1354 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1355 $1, $2, $3) because the numerical variables can be reset during some
1356 expansion items (e.g. "match"), thereby losing the authentication data.
1358 PH/21 Make -bV show the size of off_t variables so that the test suite can
1359 decide whether to run tests for quotas > 2G.
1361 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1362 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1363 filecount value is greater than 2G or if a quota value is greater than 2G
1364 on a system where the size of off_t is not greater than 4, a panic error
1367 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1368 never match. The debug and -bh output now contains an explicit error
1369 message indicating a malformed IPv4 address or mask.
1371 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1372 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1373 PH/23 above applies.
1375 PH/25 Do not write to syslog when running in the test harness. The only
1376 occasion when this arises is a failure to open the main or panic logs
1377 (for which there is an explicit test).
1379 PH/26 Added the /no_tell option to "control=freeze".
1381 PH/27 If a host name lookup failed very early in a connection, for example, if
1382 the IP address matched host_lookup and the reverse lookup yielded a name
1383 that did not have a forward lookup, an error message of the form "no IP
1384 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1385 could be logged. Now it outputs the IP address instead of "NULL".
1387 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1388 allows the sender and the authenticated sender to be set when
1389 submitting a message from within Exim. Since child_open_exim() is
1390 documented for local_scan(), the new function should be too.
1392 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1393 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1394 results in an empty string is now treated as unset.
1396 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1398 PH/31 Added sender_verify_fail logging option.
1400 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1401 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1402 tidied the source and removed it altogether.
1404 PH/33 When a queue run was abandoned because the load average was too high, a
1405 log line was always written; now it is written only if the queue_run log
1406 selector is set. In addition, the log line for abandonment now contains
1407 information about the queue run such as the pid. This is always present
1408 in "start" and "stop" lines but was omitted from the "abandon" line.
1410 PH/34 Omit spaces between a header name and the colon in the error message that
1411 is given when verify = headers_syntax fails (if there are lots of them,
1412 the message gets confusing).
1414 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1415 names, as there are now some PTR records that contain slashes. This check
1416 is only to protect against broken name servers that fall over on strange
1417 characters, so the fact that it applies to all lookups doesn't matter.
1419 PH/36 Now that the new test suite is complete, we can remove some of the
1420 special code in Exim that was needed for the old test suite. For example,
1421 sorting DNS records because real resolvers return them in an arbitrary
1422 order. The new test suite's fake resolver always returns records in the
1425 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1426 bounces) except when queue_only is set, to avoid logging races between
1427 the different processes.
1429 PH/38 Panic-die if .include specifies a non-absolute path.
1431 PH/39 A tweak to the "H" retry rule from its user.
1433 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1434 a label. They prevented compilation on older perls.
1436 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1437 a warning to be raised on newish perls.
1439 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1440 on queue. Changes to match documented behaviour of showing count of
1441 messages matching specified criteria.
1443 PH/40 Changed the default ident timeout from 30s to 5s.
1445 PH/41 Added support for the use of login_cap features, on those BSD systems
1446 that have them, for controlling the resources used by pipe deliveries.
1448 PH/42 The content-scanning code uses fopen() to create files in which to put
1449 message data. Previously it was not paying any attention to the mode of
1450 the files. Exim runs with umask(0) because the rest of the code creates
1451 files with open(), and sets the required mode explicitly. Thus, these
1452 files were ending up world-writeable. This was not a big issue, because,
1453 being within the spool directory, they were not world-accessible. I have
1454 created a function called modefopen, which takes an additional mode
1455 argument. It sets umask(777), creates the file, chmods it to the required
1456 mode, then resets the umask. All the relevant calls to fopen() in the
1457 content scanning code have been changed to use this function.
1459 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1460 to 24 hours. This avoids potential overflow problems when processing G
1461 and H retry rules. I suspect nobody ever tinkers with this value.
1463 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1465 PH/45 When the plaintext authenticator is running as a client, the server's
1466 challenges are checked to ensure they are valid base64 strings. By
1467 default, the authentication attempt is cancelled if an invalid string is
1468 received. Setting client_ignore_invalid_base64 true ignores these errors.
1469 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1470 they are received. Thus, the responses can be made to depend on the
1471 challenges. If an invalid string is ignored, an empty string is placed in
1474 PH/46 Messages that are created by the autoreply transport now contains a
1475 References: header, in accordance with RFCs 2822 and 3834.
1477 PH/47 Added authenticated_sender_force to the smtp transport.
1479 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1481 PH/49 Installed latest patch from the Sieve maintainer.
1483 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1484 was also set, the appendfile transport was unnecessarily scanning a
1485 directory of message files (e.g. for maildir delivery) to find the count
1486 of files (along with the size), even though it did not need this
1487 information. It now does the scan only if it needs to find either the
1488 size of the count of files.
1490 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1492 PH/52 Two bugs concerned with error handling when the smtp transport is
1495 (i) Exim was not creating retry information for temporary errors given
1496 for individual recipients after the DATA command when the smtp transport
1497 was used in LMTP mode. This meant that they could be retried too
1498 frequently, and not timed out correctly.
1500 (ii) Exim was setting the flag that allows error details to be returned
1501 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1502 recipients that were returned after the DATA command.
1504 PH/53 This is related to PH/52, but is more general: for any failing address,
1505 when detailed error information was permitted to be returned to the
1506 sender, but the error was temporary, then after the final timeout, only
1507 "retry timeout exceeded" was returned. Now it returns the full error as
1508 well as "retry timeout exceeded".
1510 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1511 do this, and (what is worse) MTAs that accept it.
1513 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1514 will now be deprecated.
1516 PH/56 New os.c-cygwin from the Cygwin maintainer.
1518 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1519 formats (previously only available in exim formats via -bpr, -bpru,
1520 and -bpra. Now also available in native and exiqgrep formats)
1522 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1523 with very large, slow to parse queues
1525 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1527 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1529 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1530 responses to authentication challenges, though it was showing the
1531 challenges; (ii) I've removed the CR characters from the debug output for
1534 PH/58 Allow for the insertion of a newline as well as a space when a string
1535 is turned into more than one encoded-word during RFC 2047 encoding. The
1536 Sieve code now uses this.
1538 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1539 data_4xx, lost_connection, tls_required.
1541 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1542 message was being sent as an SMTP response.
1544 PH/61 Add -l and -k options to exicyclog.
1546 PH/62 When verifying, if an address was redirected to one new address, so that
1547 verification continued, and the new address failed or deferred after
1548 having set something in $address_data, the value of $address_data was not
1549 passed back to the ACL. This was different to the case when no
1550 redirection occurred. The value is now passed back in both cases.
1552 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1553 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1554 use login_cap.h, so on its own it isn't the distinguishing feature. The
1555 new name refers directly to the setclassresources() function.
1557 PH/65 Added configuration files for NetBSD3.
1559 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1561 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1562 is preferred over IPv4.
1564 PH/68 The bounce_return_message and bounce_return_body options were not being
1565 honoured for bounces generated during the reception of non-SMTP messages.
1566 In particular, this applied to messages rejected by the ACL. This bug has
1567 been fixed. However, if bounce_return_message is true and bounce_return_
1568 body is false, the headers that are returned for a non-SMTP message
1569 include only those that have been read before the error was detected.
1570 (In the case of an ACL rejection, they have all been read.)
1572 PH/69 The HTML version of the specification is now built in a directory called
1573 spec_html instead of spec.html, because the latter looks like a path with
1574 a MIME-type, and this confuses some software.
1576 PH/70 Catch two compiler warnings in sieve.c.
1578 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1579 function verify_get_ident() calls ip_connect() to connect a socket, but
1580 if the "connect()" function timed out, ip_connect() used to close the
1581 socket. However, verify_get_ident() also closes the socket later, and in
1582 between Exim writes to the log, which may get opened at this point. When
1583 the socket was closed in ip_connect(), the log could get the same file
1584 descriptor number as the socket. This naturally causes chaos. The fix is
1585 not to close the socket in ip_connect(); the socket should be closed by
1586 the function that creates it. There was only one place in the code where
1587 this was missing, in the iplookup router, which I don't think anybody now
1588 uses, but I've fixed it anyway.
1590 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1591 well as to direct DNS lookups. Otherwise the handling of names in host
1592 lists is inconsistent and therefore confusing.
1598 PH/01 Two changes to the default runtime configuration:
1600 (1) Move the checks for relay_from_hosts and authenticated clients from
1601 after to before the (commented out) DNS black list checks.
1603 (2) Add control=submission to the relay_from_hosts and authenticated
1604 clients checks, on the grounds that messages accepted by these
1605 statements are most likely to be submissions.
1607 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1609 (1) Generate an error if the third argument for the ${prvs expansion is
1612 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1615 (3) Reset the variables that are obtained from the first argument of
1616 ${prvscheck and used in the second argument before leaving the code,
1617 because their memory is reclaimed, so using them afterwards may do
1620 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1621 one (it's much easier than Tom thought :-).
1623 (5) Because of (4), we can now allow for the use of $prvscheck_result
1624 inside the third argument.
1626 PH/03 For some reason, the default setting of PATH when running a command from
1627 a pipe transport was just "/usr/bin". I have changed it to
1630 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1631 anything to be listed in the output from -bV.
1633 PH/05 When a filter generated an autoreply, the entire To: header line was
1634 quoted in the delivery log line, like this:
1636 => >A.N.Other <ano@some.domain> <original@ddress> ...
1638 This has been changed so that it extracts the operative address. There
1639 may be more than one such address. If so, they are comma-separated, like
1642 => >ano@some.domain,ona@other.domain <original@ddress> ...
1644 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1645 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1646 not being looked up in the rDNS to get a host name, Exim was showing the
1647 IP address twice in Received: lines, even though the IP addresses were
1648 identical. For example:
1650 Received: from [1.2.3.4] (helo=[1.2.3.4])
1652 However, if the real host name was known, it was omitting the HELO data
1653 if it matched the actual IP address. This has been tidied up so that it
1654 doesn't show the same IP address twice.
1656 PH/07 When both +timestamp and +memory debugging was on, the value given by
1657 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1658 called by the debug printing, thereby overwriting the timestamp buffer.
1659 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1661 PH/08 When the original message was included in an autoreply transport, it
1662 always said "this is a copy of the message, including all the headers",
1663 even if body_only or headers_only was set. It now gives an appropriate
1666 PH/09 Applied a patch from the Sieve maintainer which:
1668 o fixes some comments
1669 o adds the (disabled) notify extension core
1670 o adds some debug output for the result of if/elsif tests
1671 o points to the current vacation draft in the documentation
1672 and documents the missing references header update
1676 o fixes a bug in processing the envelope test (when testing
1677 multiple envelope elements, the last element determinted the
1680 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1681 Electronic Mail") by including:
1683 Auto-submitted: auto-generated
1685 in the messages that it generates (bounce messages and others, such as
1686 warnings). In the case of bounce messages for non-SMTP mesages, there was
1687 also a typo: it was using "Auto_submitted" (underscore instead of
1688 hyphen). Since every message generated by Exim is necessarily in response
1689 to another message, thes have all been changed to:
1691 Auto-Submitted: auto-replied
1693 in accordance with these statements in the RFC:
1695 The auto-replied keyword:
1697 - SHOULD be used on messages sent in direct response to another
1698 message by an automatic process,
1700 - MUST NOT be used on manually-generated messages,
1702 - MAY be used on Delivery Status Notifications (DSNs) and Message
1703 Disposition Notifications (MDNs),
1705 - MUST NOT be used on messages generated by automatic or periodic
1706 processes, except for messages which are automatic responses to
1709 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1710 to the default Received: header definition.
1712 PH/12 Added log selector acl_warn_skipped (default on).
1714 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1715 variables because (a) they are in the wrong storage pool and (b) even if
1716 they were copied, it wouldn't work properly because of the caching.
1718 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1719 checking when decoding. Apparently there are clients that generate
1720 overlong encoded strings. Why am I not surprised?
1722 PH/15 If the first argument of "${if match_address" was not empty, but did not
1723 contain an "@" character, Exim crashed. Now it writes a panic log message
1724 and treats the condition as false.
1726 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1728 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1729 extension "envelope-auth". The code is finished and in agreement with
1730 other implementations, but there is no documentation so far and in fact,
1731 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1732 not changing the active code.
1734 Print executed "if" and "elsif" statements when debugging is used. This
1735 helps a great deal to understand what a filter does.
1737 Document more things not specified clearly in RFC3028. I had all this
1738 sorted out, when out of a sudden new issues came to my mind. Oops."
1740 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1743 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1744 canonical form (as documented). However, after a host name lookup from
1745 the IP address, check_host() was doing a simple string comparison with
1746 addresses acquired from the DNS when checking that the found name did
1747 have the original IP as one of its addresses. Since any found IPv6
1748 addresses are likely to be in abbreviated form, the comparison could
1749 fail. Luckily, there already exists a function for doing the comparison
1750 by converting both addresses to binary, so now that is used instead of
1751 the text comparison.
1753 PH/20 There was another similar case to PH/19, when a complete host name was
1754 given in a host list; looking up its IP address could give an abbreviated
1755 form, whereas the current host's name might or might not be abbreviated.
1756 The same fix has been applied.
1762 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1763 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1766 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1767 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1769 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1771 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1772 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1773 Also, for "Auto-Submitted", treat anything other than "no" as
1774 non-personal, in accordance with RFC 3834. (Previously it treated
1775 anything starting "auto-" as non-personal.)
1777 TF/01 The control=submission/name=... option had a problem with syntax
1778 errors if the name included a slash character. The /name= option
1779 now slurps the rest of the string, so it can include any characters
1780 but it must come last in the list of options (after /sender_retain
1783 PH/05 Some modifications to the interface to the fake nameserver for the new
1791 TK/01 Added the "success_on_redirect" address verification option. See
1792 NewStuff for rationale and an example.
1794 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1796 PH/02 Patch to exigrep to allow it to work on syslog lines.
1798 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1799 fread() to skip over the body file's header line, because in Cygwin the
1800 header line is locked and is inaccessible.
1802 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1803 co-exist for some time) to make it clear that it is the Exim ID that is
1804 referenced, not the Message-ID: header line.
1806 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1807 string_format() function, because snprintf() does not exist on all
1810 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1811 ${stat: expansion item.
1813 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1814 protocol synchronization error", to keep the pedants happy.
1816 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1817 well as for IRIX systems, when gcc is being used. See the host.c source
1820 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1822 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1825 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1826 transport and to the smtp transport in LMTP mode.
1828 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1830 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1831 run a filter in a subprocess. This could lead to confusion in subsequent
1832 lookups in the parent process. There should also be a search_tidyup() at
1833 the end of the subprocess.
1835 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1836 only if the host matched helo_try_verify_hosts, which caused the
1837 verification to occur when the EHLO/HELO command was issued. The ACL just
1838 tested the remembered result. Now, if a previous verification attempt has
1839 not happened, "verify = helo" does it there and then.
1841 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1843 TK/03 Fix log output including CR from clamd.
1845 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1846 exist provoked a memory error which could cause a segfault.
1848 PH/15 Installed PCRE 6.2
1850 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1852 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1853 of the problem. Specifically, suggested +O2 rather than +O1 for the
1856 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1858 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1859 record, the port was not correctly logged when the outgoing_port log
1860 selector was set (it logged the transort's default port).
1862 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1863 fallback_hosts, and "hosts" in the smtp transport.
1865 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1866 host errors such as "Connection refused".
1868 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1869 authentication with radiusclient 0.4.9:
1871 - Error returned from rc_read_config was caught wrongly
1872 - Username/password not passed on to radius server due to wrong length.
1874 The presumption is that some radiusclient API changes for 4.51/PH/17
1875 were not taken care of correctly. The code is still untested by me (my
1876 Linux distribution still has 0.3.2 of radiusclient), but it was
1877 contributed by a Radius user.
1879 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1880 expanding the "port" option of the smtp transport.
1882 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1883 while reading a MIME header. Thanks to Tom Hughes for a patch.
1885 PH/24 Include config.h inside local_scan.h so that configuration settings are
1888 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1889 that in an ACL for RCPT (for example), you can examine exactly what was
1892 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1893 commands, but it was not correctly comparing the address with the actual
1894 client host address. Thus, it would show the EHLO address in Received:
1895 header lines when this was not necessary.
1897 PH/27 Added the % operator to ${eval:}.
1899 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1900 it should be ignoring failures (because with -C, for example, it has lost
1901 privilege). It wasn't ignoring creation failures other than "already
1904 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1907 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1908 because an input line was too long, either on its own, or by virtue of
1909 too many continuations, the temporary file was not being removed, and the
1910 return code was incorrect.
1912 PH/31 Missing "BOOL" in function definition in filtertest.c.
1914 PH/32 Applied Sieve patches from the maintainer.
1916 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1918 PH/33 Added "verify = not_blind".
1920 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1921 Local/Makefile (with some defaults set). These are used in built scripts
1922 such as exicyclog, but they have never been used in the exim_install
1923 script (though there are many overriding facilities there). I have
1924 arranged that the exim_install script now takes note of these two
1927 PH/35 Installed configuration files for Dragonfly.
1929 PH/36 When a locally submitted message by a trusted user did not contain a
1930 From: header, and the sender address was obtained from -f or from an SMTP
1931 MAIL command, and the trusted user did not use -F to supply a sender
1932 name, $originator_name was incorrectly used when constructing a From:
1933 header. Furthermore, $originator_name was used for submission mode
1934 messages from external hosts without From: headers in a similar way,
1935 which is clearly wrong.
1937 PH/37 Added control=suppress_local_fixups.
1939 PH/38 When log_selector = +received_sender was set, and the addition of the
1940 sender made the log line's construction buffer exactly full, or one byte
1941 less than full, an overflow happened when the terminating "\n" was
1944 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1945 when the result of a list match is failure because a DNS lookup failed.
1947 PH/40 RM_COMMAND is now used in the building process.
1949 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1950 the "build-* directories that it finds.
1952 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1953 address in a domain literal was a prefix of an interface address.
1955 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1956 when verifying a sender address, unless rewrite_headers is false.
1958 PH/44 Wrote a long comment about why errors_to addresses are verified as
1959 recipients, not senders.
1961 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1962 the ratelimit ACL was added.
1964 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1966 PH/46 Added extra information about PostgreSQL errors to the error string.
1968 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1969 avoiding the need to install special zones in a real server. This is
1970 backwards compatible; if it can't find the fake resolver, it drops back.
1971 Thus, both old and new test suites can be run.
1973 TF/02 Added util/ratelimit.pl
1975 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1976 clock is set back in time.
1978 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1979 Candler <B.Candler@pobox.com>.
1981 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1982 OK for addresses that are the result of redirections.
1984 PH/48 A number of further additions for the benefit of the new test suite,
1985 including a fake gethostbyname() that interfaces to the fake DNS resolver
1988 TF/06 The fix for widen_domains has also been applied to qualify_single and
1989 search_parents which are the other dnslookup options that can cause
1992 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1995 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1997 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1998 service" if it encountered an MX record with an empty target hostname.
1999 The message is now "an MX or SRV record indicated no SMTP service".
2001 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2002 if the DNS of the sending site is misconfigured. This is quite a
2003 common situation. This change restores the behaviour of treating a
2004 helo verification defer as a failure.
2006 PH/51 If self=fail was set on a router, the bounce message did not include the
2007 actual error message.
2013 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2015 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2016 command itself ended in error, the underlying message about the transport
2017 filter timeout was being overwritten with the pipe command error. Now the
2018 underlying error message should be appended to the second error message.
2020 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2022 PH/02 Reduce the amount of output that "make" produces by default. Full output
2023 can still be requested.
2025 PH/03 The warning log line about a condition test deferring for a "warn" verb
2026 was being output only once per connection, rather than after each
2027 occurrence (because it was using the same function as for successful
2028 "warn" verbs). This seems wrong, so I have changed it.
2030 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2031 it should not have, which might have caused a crash in the right
2032 circumstances, but probably never did.
2034 PH/04 Installed a modified version of Tony Finch's patch to make submission
2035 mode fix the return path as well as the Sender: header line, and to
2036 add a /name= option so that you can make the user's friendly name appear
2039 TF/03 Added the control = fakedefer ACL modifier.
2041 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2042 Mark Lowes for thorough testing.
2044 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2046 TK/03 Merged latest SRS patch from Miles Wilton.
2048 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2049 with the definition in sysexits.h (which is #included earlier).
2050 Fortunately, Exim does not actually use EX_OK. The code used to try to
2051 preserve the sysexits.h value, by assumimg that macro definitions were
2052 scanned for macro replacements. I have been disabused of this notion,
2053 so now the code just undefines EX_OK before #including unistd.h.
2055 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2056 in the smtp transport. When a block could not be written in a single
2057 write() function, the timeout was being re-applied to each part-write.
2058 This seems wrong - if the receiver was accepting one byte at a time it
2059 would take for ever. The timeout is now adjusted when this happens. It
2060 doesn't have to be particularly precise.
2062 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2063 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2065 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2066 without a domain if the check to <postmaster@domain> fails.
2068 SC/01 Eximstats: added -xls and the ability to specify output files
2069 (patch written by Frank Heydlauf).
2071 SC/02 Eximstats: use FileHandles for outputing results.
2073 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2075 SC/04 Eximstats: fixed display of large numbers with -nvr option
2077 SC/05 Eximstats: fixed merging of reports with empty tables.
2079 SC/06 Eximstats: added the -include_original_destination flag
2081 SC/07 Eximstats: removed tabs and trailing whitespace.
2083 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2085 TK/06 MBOX spool code: Add real "From " MBOX separator line
2086 so the .eml file is really in mbox format (even though
2087 most programs do not really care). Patch from Alex Miller.
2089 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2090 The latter is generated from $received_to and is only set if the
2091 message has one envelope recipient. SA can use these headers,
2092 obviously out-of-the-box. Patch from Alex Miller.
2094 PH/08 The ${def test on a variable was returning false if the variable's
2095 value was "0", contrary to what the specification has always said!
2096 The result should be true unless the variable is empty.
2098 PH/09 The syntax error of a character other than { following "${if
2099 def:variable_name" (after optional whitespace) was not being diagnosed.
2100 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2101 accidental colon was present, for example, could give incorrect results.
2103 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2104 result is used (not including appendfile, where other changes are about
2107 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2108 This involved changing a lot of size variables from int to off_t. It
2109 should work with maildirs and everything.
2111 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2112 spamd dying while we are connected to it.
2114 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2117 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2118 function for BDB 4.3.
2120 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2122 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2123 the sources for PCRE have been re-arranged and more files are now
2126 PH/15 The code I had for printing potentially long long variables in PH/11
2127 above was not the best (it lost precision). The length of off_t variables
2128 is now inspected at build time, and an appropriate printing format (%ld
2129 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2130 to be "long long int" or "long int". This is needed for the internal
2131 formatting function string_vformat().
2133 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2134 the configuration file to be ":syslog", then the script "guesses" where
2135 the logs files are, rather than using the compiled in default. In our
2136 case the guess is not the same as the compiled default, so the script
2137 suddenly stopped working when I started to use syslog. The patch checks
2138 to see if log_file_path is "". If so, it attempts to read it from exim
2139 with no configuration file to get the compiled in version, before it
2140 falls back to the previous guessing code."
2142 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2143 implementing BATV in an Exim configuration. See NewStuff for the gory
2146 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2147 Makefile that are specific to HP-UX.
2149 PH/18 If the "use_postmaster" option was set for a recipient callout together
2150 with the "random" option, the postmaster address was used as the MAIL
2151 FROM address for the random test, but not for the subsequent recipient
2152 test. It is now used for both.
2154 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2155 patch removes a few documentation additions to RFC 3028, because the
2156 latest draft now contains them. It adds the new en;ascii-case comparator
2157 and a new error check for 8bit text in MIME parts. Comparator and
2158 require names are now matched exactly. I enabled the subaddress
2159 extension, but it is not well tested yet (read: it works for me)."
2161 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2162 rework some of the code of TK/09 above to avoid the hardwired use of
2163 "%lld" and "long long". Replaced the call to snprintf() with a call to
2166 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2167 records point to non-existent hosts", "retry timeout exceeded", and
2168 "retry time not reached for any host after a long failure period".
2170 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2171 experimental DomainKeys support:
2173 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2174 (2) On an error such as an illegally used "control", the wrong name for
2175 the control was given.
2177 These problems did NOT occur unless DomainKeys support was compiled.
2179 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2181 PH/24 Added ${if match_ip condition.
2183 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2184 databases so that it will be absolutely obvious if a crash occurs in the
2185 DB library. This is a regular occurrence (often caused by mis-matched
2188 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2189 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2190 picked up on a user's system that detects such things. There doesn't seem
2191 to be a gcc warning option for this - only an attribute that has to be
2192 put on the function's prototype. It seems that in Fedora Core 4 they have
2193 set this on a number of new functions. No doubt there will be more in due
2196 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2197 specify a transport. However, if an address that was verified by such a
2198 router was the subject of a callout, Exim crashed because it tried to
2199 read the rcpt_include_affixes from the non-existent transport. Now it
2200 just assumes that the setting of that option is false. This bug was
2201 introduced by 4.51/PH/31.
2203 PH/28 Changed -d+all to exclude +memory, because that information is very
2204 rarely of interest, but it makes the output a lot bigger. People tend to
2205 do -d+all out of habit.
2207 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2208 code in os-type was giving problems when libc.so lives in lib64, like on
2211 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2212 aren't the modern standard, and it seems that some systems' include files
2213 don't always have them. Exim was already checking for some of the newer
2214 ones like T_AAAA, and defining it itself. I've added checks for all the
2215 record types that Exim uses.
2217 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2218 not automatically generating a new one, as it is supposed to. This
2219 prevented TLS from working. If the file did exist, but contained invalid
2220 data, a new version was generated, as expected. It was only the case of a
2221 non-existent file that was broken.
2223 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2224 with a change in libdomainkeys > 0.64.
2226 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2227 from DNS. If the selector record carries the flag, it now has
2228 precedence over the domain-wide flag.
2230 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2232 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2233 the use of a port name that isn't defined in /etc/services) occurred, the
2234 message was deferred as in a normal delivery, and thus remained on the
2235 spool, instead of being failed because of the mua_wrapper setting. This
2236 is now fixed, and I tidied up some of the mua_wrapper messages at the
2239 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2240 the messages in a hash of arrays rather than using individual hashes.
2241 This is a bit cleaner and results in dramatic memory savings, albeit
2242 at a slight CPU cost.
2244 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2245 as requested by Marc Sherman.
2247 SC/10 Eximstats: added histograms for user specified patterns as requested
2250 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2252 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2253 fopen() in the content-scanning modules that did not already have it.
2259 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2260 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2262 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2264 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2266 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2267 to test Sieve filters that use "vacation".
2269 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2270 that changes the way the GnuTLS parameters are stored in the cache file.
2271 The new format can be generated externally. For backward compatibility,
2272 if the data in the cache doesn't make sense, Exim assumes it has read an
2273 old-format file, and it generates new data and writes a new file. This
2274 means that you can't go back to an older release without removing the
2277 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2278 work if there are any delivery delays because "one_time" forces the
2279 parent to be marked "delivered", so its unseen clone is never tried
2280 again. For this reason, Exim now forbids the simultaneous setting of
2283 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2284 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2285 entry if you want to know the details. The fix, however, neglected to
2286 consider the case when local delivery batching is involved. The test for
2287 "previously delivered" was not happening when checking to see if an
2288 address could be batched with a previous (undelivered) one; under
2289 certain circumstances this could lead to multiple deliveries to the same
2292 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2293 in its include files, and this causes problems building Exim.
2295 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2296 header_syntax) but Exim was just ignoring anything given after a slash.
2297 In particular, this caused confusion with an attempt to use "verify =
2298 reverse_host_lookup/defer_ok". An error is now given when options are
2299 supplied for verify items that do not have them. (Maybe reverse_host_
2300 lookup should have a defer_ok option, but that's a different point.)
2302 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2303 defined by RFC 821) to 2048, because there were problems with some AUTH
2304 commands, and RFC 1869 says the size should be increased for extended
2305 SMTP commands that take arguments.
2307 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2310 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2311 "unknown" error; now it says that the functionality isn't in the binary.
2313 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2314 an address' error message when a string expansion fails (syntax or
2315 whatever). Otherwise the password may appear in the log. Following change
2316 PH/42 below, there is no longer a chance of it appearing in a bounce
2319 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2321 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2322 of its addresses was ever tried. (Bugzilla bug #2.)
2324 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2325 the result incorrectly in the debug output. (It correctly added a newline
2326 to what was transported.)
2328 TF/01 Added $received_time.
2330 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2331 commented out examples of how to interface to a virus scanner and to
2332 SpamAssassin. Also added commented examples of av_scanner and
2333 spamd_address settings.
2335 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2336 and controls are allowed in which ACLs. There were a couple of minor
2337 errors. Some of the entries in the conditions table (which is a table of
2338 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2339 negation of where the condition IS allowed.
2341 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2343 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2344 header file does not have a version number, so I've had to invent a new
2345 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2346 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2347 radiusclient), but it was contributed by a Radius user.
2349 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2350 files or directories, for OpenSSL.
2352 PH/19 When an Exim process that is running as root has to create an Exim log
2353 file, it does so in a subprocess that runs as exim:exim so as to get the
2354 ownership right at creation (otherwise, other Exim processes might see
2355 the file with the wrong ownership). There was no test for failure of this
2356 fork() call, which would lead to the process getting stuck as it waited
2357 for a non-existent subprocess. Forks do occasionally fail when resources
2358 run out. I reviewed all the other calls to fork(); they all seem to check
2361 PH/20 When checking for unexpected SMTP input at connect time (before writing
2362 the banner), Exim was not dealing correctly with a non-positive return
2363 from the read() function. If the client had disconnected by this time,
2364 the result was a log entry for a synchronization error with an empty
2365 string after "input=" when read() returned zero. If read() returned -1
2366 (an event I could not check), uninitialized data bytes were printed.
2367 There were reports of junk text (parts of files, etc) appearing after
2370 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2372 PH/22 Added support for macro redefinition, and (re)definition in between
2373 driver and ACL definitions.
2375 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2376 forgetting to use the resulting value; it was using the unexpanded value.
2378 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2379 hadn't been configured. The fix is from Juergen Kreileder, who
2380 understands it better than I do:
2382 "Here's what I see happening with three configured cyrus_sasl
2383 authenticators configured (plain, login, cram-md5):
2385 On startup auth_cyrus_sasl_init() gets called for each of these.
2386 This means three calls to sasl_listmech() without a specified mech_list.
2387 => SASL tests which mechs of all available mechs actually work
2388 => three warnings about OTP not working
2389 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2391 With the patch, sasl_listmech() also gets called three times. But now
2392 SASL's mech_list option is set to the server_mech specified in the the
2393 authenticator. Or in other words, the answer from sasl_listmech()
2394 gets limited to just the mech you're testing for (which is different
2396 => the return list contains just 'plain' or 'login', 'cram-md5' or
2397 nothing depending on the value of ob->server_mech.
2399 I've just tested the patch: Authentication still works fine,
2400 unavailable mechs specified in the exim configuration are still
2401 caught, and the auth.log warnings about OTP are gone."
2403 PH/25 When debugging is enabled, the contents of the command line are added
2404 to the debugging output, even when log_selector=+arguments is not
2407 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2408 answer is "GNU", and only if the return is "GNU/something" is the answer
2411 PH/27 $acl_verify_message is now set immediately after the failure of a
2412 verification in an ACL, and so is available in subsequent modifiers. In
2413 particular, the message can be preserved by coding like this:
2415 warn !verify = sender
2416 set acl_m0 = $acl_verify_message
2418 Previously, $acl_verify_message was set only while expanding "message"
2419 and "log_message" when a very denied access.
2421 PH/28 Modified OS/os.c-Linux with
2423 -#ifndef OS_LOAD_AVERAGE
2424 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2426 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2427 nomenclature these days.)
2429 PH/29 Installed patch from the Sieve maintainer that adds the options
2430 sieve_useraddress and sieve_subaddress to the redirect router.
2432 PH/30 In these circumstances:
2433 . Two addresses routed to the same list of hosts;
2434 . First host does not offer TLS;
2435 . First host accepts first address;
2436 . First host gives temporary error to second address;
2437 . Second host offers TLS and a TLS session is established;
2438 . Second host accepts second address.
2439 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2440 and peerdn, if requested) that were in fact used only for the second
2443 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2444 attention to any local part prefix or suffix that was matched by the
2445 router that accepted the address. It now behaves in the same way as it
2446 does for delivery: the affixes are removed from the local part unless
2447 rcpt_include_affixes is set on the transport.
2449 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2450 timeout during the DATA phase of an incoming message.
2452 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2453 applied a patch sanctioned by the Sieve maintainer.
2455 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2456 the uid or gid is negative. A case of a negative gid caused this to be
2457 noticed. The fix allows for either to be negative.
2459 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2460 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2463 PH/36 The change PH/12 above was broken. Fixed it.
2465 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2466 the grounds that routing the same address twice would always produce the
2467 same answer. This might have been true once, but it is certainly no
2468 longer true now. Routing a child address may depend on the previous
2469 routing that produced that child. Some complicated redirection strategies
2470 went wrong when messages had multiple recipients, and made Exim's
2471 behaviour dependent on the order in which the addresses were given.
2473 I have moved the duplicate checking until after the routing is complete.
2474 Exim scans the addresses that are assigned to local and remote
2475 transports, and removes any duplicates. This means that more work will be
2476 done, as duplicates will always all be routed, but duplicates are
2477 presumably rare, so I don't expect this is of any significance.
2479 For deliveries to pipes, files, and autoreplies, the duplicate checking
2480 still happens during the routing process, since they are not going to be
2483 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2484 It corrects a timeout issue with spamd. This is Ian's comment: "The
2485 background is that sometimes spamd either never reads data from a
2486 connection it has accepted, or it never writes response data. The exiscan
2487 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2488 blindly assumes that writes won't block so it may never time out."
2490 PH/39 Allow G after quota size as well as K and M.
2492 PH/40 The value set for $authenticated_id in an authenticator may not contain
2493 binary zeroes or newlines because the value is written to log lines and
2494 to spool files. There was no check on this. Now the value is run through
2495 the string_printing() function so that such characters are converted to
2496 printable escape sequences.
2498 PH/41 $message_linecount is a new variable that contains the total number of
2499 lines in the message. Compare $body_linecount, which is the count for the
2502 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2503 bounce and delay warning messages, except in certain special cases, which
2506 (a) An SMTP error message from a remote host;
2507 (b) A message specified in a :fail: redirection;
2508 (c) A message specified in a "fail" command in a system filter;
2509 (d) A message specified in a FAIL return from the queryprogram router;
2510 (e) A message specified by the cannot_route_message router option.
2512 In these cases only, Exim does include the error details in bounce and
2513 warning messages. There are also a few cases where bland messages such
2514 as "unrouteable address" or "local delivery error" are given.
2516 PH/43 $value is now also set for the "else" part of a ${run expansion.
2518 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2519 being worked on, but at least Exim now implements the latest version to
2522 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2523 process to complete was treated as a delivery failure, a timeout while
2524 writing the message to the pipe was logged, but erroneously treated as a
2525 successful delivery. Such timeouts include transport filter timeouts. For
2526 consistency with the overall process timeout, these timeouts are now
2527 treated as errors, giving rise to delivery failures by default. However,
2528 there is now a new Boolean option for the pipe transport called
2529 timeout_defer, which, if set TRUE, converts the failures into defers for
2530 both kinds of timeout. A transport filter timeout is now identified in
2533 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2534 systems where "make" and "gmake" are different, calling "gmake" at top
2535 level broke things. I've arranged for the value of $(MAKE) to be passed
2536 from the Makefile to this script so that it can call the same version of
2540 A note about Exim versions 4.44 and 4.50
2541 ----------------------------------------
2543 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2544 changes of various kinds. As a consequence, a big documentation update was
2545 needed. This delayed the release for rather longer than seemed good, especially
2546 in the light of a couple of (minor) security issues. Therefore, the changes
2547 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2548 release. So 4.44 and 4.50 are in effect two different branches that both start
2551 I have left the 4.50 change log unchanged; it contains all the changes since
2552 4.43. The change log for 4.44 is below; many of its items are identical to
2553 those for 4.50. This seems to be the most sensible way to preserve the
2554 historical information.
2560 1. Minor wording change to the doc/README.SIEVE file.
2562 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2563 computation of the current number of files was incorrect.
2565 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2566 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2569 4. Give more explanation in the error message when the command for a transport
2570 filter fails to execute.
2572 5. There are several places where Exim runs a non-Exim command in a
2573 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2574 was being done only for the command run by the queryprogram router. It is
2575 now done for all such subprocesses. The other cases are: ${run, transport
2576 filters, and the commands run by the lmtp and pipe transports.
2578 6. Added CONFIGURE_GROUP build-time option.
2580 7. Some older OS have a limit of 256 on the maximum number of file
2581 descriptors. Exim was using setrlimit() to set 1000 as a large value
2582 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2583 systems. I've change it so that if it can't get 1000, it tries for 256.
2585 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2586 was an oversight, and furthermore, ever since the addition of extra
2587 controls (e.g. 4.43/32), the checks on when to allow different forms of
2588 "control" were broken. There should now be diagnostics for all cases when a
2589 control that does not make sense is encountered.
2591 9. Added the /retain_sender option to "control=submission".
2593 10. $recipients is now available in the predata ACL (oversight).
2595 11. Tidy the search cache before the fork to do a delivery from a message
2596 received from the command line. Otherwise the child will trigger a lookup
2597 failure and thereby defer the delivery if it tries to use (for example) a
2598 cached ldap connection that the parent has called unbind on.
2600 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2601 of $address_data from the recipient verification was clobbered by the
2602 sender verification.
2604 13. The value of address_data from a sender verification is now available in
2605 $sender_address_data in subsequent conditions in the ACL statement.
2607 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2609 15. Added a new option "connect=<time>" to callout options, to set a different
2612 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2613 was its contents. (It was OK if the option was not defined at all.)
2615 17. A "Completed" log line is now written for messages that are removed from
2616 the spool by the -Mrm option.
2618 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2619 information about exactly what failed.
2621 19. Added -dd to debug only the daemon process.
2623 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2624 handles timeouts, both on the server side and network timeouts. Renamed the
2625 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2627 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2628 It is now set to "smtps".
2630 22. $host_address is now set to the target address during the checking of
2631 ignore_target_hosts.
2633 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2634 being passed; this would have caused $sender_host_name to have been used if
2635 matching the list had actually called for a host name (not very likely,
2636 since this list is usually IP addresses). A host name is now passed as
2639 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2640 code that specifies a non-restarting handler (typically sigaction() in
2641 modern systems) in an attempt to fix a rare and obscure crash bug.
2643 25. Narrowed the window for a race in the daemon that could cause it to ignore
2644 SIGCHLD signals. This is not a major problem, because they are used only to
2645 wake it up if nothing else does.
2647 26. A malformed maildirsize file could cause Exim to calculate negative values
2648 for the mailbox size or file count. Odd effects could occur as a result.
2649 The maildirsize information is now recalculated if the size or filecount
2652 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2653 support for a long time. Removed HAVE_SYS_VFS_H.
2655 28. Installed the latest version of exipick from John Jetmore.
2657 29. In an address list, if the pattern was not a regular expression, an empty
2658 subject address (from a bounce message) matched only if the pattern was an
2659 empty string. Non-empty patterns were not even tested. This was the wrong
2660 because it is perfectly reasonable to use an empty address as part of a
2661 database query. An empty address is now tested by patterns that are
2662 lookups. However, all the other forms of pattern expect the subject to
2663 contain a local part and a domain, and therefore, for them, an empty
2664 address still always fails if the pattern is not itself empty.
2666 30. Exim went into a mad DNS loop when attempting to do a callout where the
2667 host was specified on an smtp transport, and looking it up yielded more
2668 than one IP address.
2670 31. Re-factored the code for checking spool and log partition space into a
2671 function that finds that data and another that does the check. The former
2672 is then used to implement four new variables: $spool_space, $log_space,
2673 $spool_inodes, and $log_inodes.
2675 32. The RFC2047 encoding function was originally intended for short strings
2676 such as real names; it was not keeping to the 75-character limit for
2677 encoded words that the RFC imposes. It now respects the limit, and
2678 generates multiple encoded words if necessary. To be on the safe side, I
2679 have increased the buffer size for the ${rfc2047: expansion operator from
2682 33. It is now permitted to omit both strings after an "if" condition; if the
2683 condition is true, the result is "true". As before, when the second string
2684 is omitted, a false condition yields an empty string. This makes it less
2685 cumbersome to write custom ACL and router conditions.
2687 34. Failure to deliver a bounce message always caused it to be frozen, even if
2688 there was an errors_to setting on the router. The errors_to setting is now
2691 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2692 canonical form (fully expanded) before being placed in
2693 $sender_host_address.
2695 36. The table in the code that translates DNS record types into text (T_A to
2696 "A" for instance) was missing entries for NS and CNAME. It is just possible
2697 that this could have caused confusion if both these types were looked up
2698 for the same domain, because the text type is used as part of Exim's
2699 per-process caching. But the chance of anyone hitting this buglet seems
2702 37. The dnsdb lookup has been extended in a number of ways.
2704 (1) There is a new type, "zns", which walks up the domain tree until it
2705 finds some nameserver records. It should be used with care.
2707 (2) There is a new type, "mxh", which is like "mx" except that it returns
2708 just the host names, not the priorities.
2710 (3) It is now possible to give a list of domains (or IP addresses) to be
2711 looked up. The behaviour when one of the lookups defers can be
2712 controlled by a keyword.
2714 (4) It is now possible to specify the separator character for use when
2715 multiple records are returned.
2717 38. The dnslists ACL condition has been extended: it is now possible to supply
2718 a list of IP addresses and/or domains to be looked up in a particular DNS
2721 39. Added log_selector=+queue_time_overall.
2723 40. When running the queue in the test harness, wait just a tad after forking a
2724 delivery process, to get repeatability of debugging output.
2726 41. Include certificate and key file names in error message when GnuTLS fails
2727 to set them up, because the GnuTLS error message doesn't include the name
2728 of the failing file when there is a problem reading it.
2730 42. Allow both -bf and -bF in the same test run.
2732 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2734 44. The "Exiscan patch" is now merged into the mainline Exim source.
2736 45. Sometimes the final signoff response after QUIT could fail to get
2737 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2738 < 0 before doing a fflush(). This bug looks as though it goes back to the
2739 introduction of TLS in release 3.20, but "sometimes" must have been rare
2740 because the tests only now provoked it.
2742 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2743 (this can affect the format of dates).
2745 47. exim_tidydb, when checking for the continued existence of a message for
2746 which it has found a message-specific retry record, was not finding
2747 messages that were in split spool directories. Consequently, it was
2748 deleting retry records that should have stayed in existence.
2750 48. Steve fixed some bugs in eximstats.
2752 49. The SPA authentication driver was not abandoning authentication and moving
2753 on to the next authenticator when an expansion was forced to fail,
2754 contradicting the general specification for all authenticators. Instead it
2755 was generating a temporary error. It now behaves as specified.
2757 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2758 (the order specifies the preference for clients). The order is now AES256,
2759 AES128, 3DES, ARCFOUR128.
2761 51. Small patch to Sieve code - explicitly set From: when generating an
2764 52. Exim crashed if a remote delivery caused a very long error message to be
2765 recorded - for instance if somebody sent an entire SpamAssassin report back
2766 as a large number of 550 error lines. This bug was coincidentally fixed by
2767 increasing the size of one of Exim's internal buffers (big_buffer) that
2768 happened as part of the Exiscan merge. However, to be on the safe side, I
2769 have made the code more robust (and fixed the comments that describe what
2772 53. Now that there can be additional text after "Completed" in log lines (if
2773 the queue_time_overall log selector is set), a one-byte patch to exigrep
2774 was needed to allow it to recognize "Completed" as not the last thing in
2777 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2778 patch that reportedly fixes this has been added. I am not expert enough to
2779 create a test for it. This is what the patch creator wrote:
2781 "I found a little strange behaviour of ldap code when working with
2782 Windows 2003 AD Domain, where users was placed in more than one
2783 Organization Units. When I tried to give exim partial DN, the exit code
2784 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2785 But simultaneously result of request was absolutely normal ldap result,
2786 so I produce this patch..."
2788 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2789 REFERENCE, so I have modified the code to exclude the patch when that macro
2792 55. Some experimental protocols are using DNS PTR records for new purposes. The
2793 keys for these records are domain names, not reversed IP addresses. The
2794 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2795 leaves it alone. Component reversal etc. now happens only for IP addresses.
2798 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2800 57. Double the size of the debug message buffer (to 2048) so that more of very
2801 long debug lines gets shown.
2803 58. The exicyclog utility now does better if the number of log files to keep
2804 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2806 59. Two changes related to the smtp_active_hostname option:
2808 (1) $smtp_active_hostname is now available as a variable.
2809 (2) The default for smtp_banner uses $smtp_active_hostname instead
2810 of $primary_hostname.
2812 60. The host_aton() function is supposed to be passed a string that is known
2813 to be a valid IP address. However, in the case of IPv6 addresses, it was
2814 not checking this. This is a hostage to fortune. Exim now panics and dies
2815 if the condition is not met. A case was found where this could be provoked
2816 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2817 components; fortuitously, this particular loophole had already been fixed
2818 by change 4.50/55 above.
2820 If there are any other similar loopholes, the new check in host_aton()
2821 itself should stop them being exploited. The report I received stated that
2822 data on the command line could provoke the exploit when Exim was running as
2823 exim, but did not say which command line option was involved. All I could
2824 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2825 running as the user.
2828 61. There was a buffer overflow vulnerability in the SPA authentication code
2829 (which came originally from the Samba project). I have added a test to the
2830 spa_base64_to_bits() function which I hope fixes it.
2833 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2834 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2836 63. The daemon start-up calls getloadavg() while still root for those OS that
2837 need the first call to be done as root, but it missed one case: when
2838 deliver_queue_load_max is set with deliver_drop_privilege. This is
2839 necessary for the benefit of the queue runner, because there is no re-exec
2840 when deliver_drop_privilege is set.
2842 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2843 This has been fixed.
2845 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2846 was in use, was not putting the data itself into the right store pool;
2847 consequently, it could be overwritten for a subsequent message in the same
2848 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2851 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2853 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2854 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2855 the function were treating the return as a boolean value, which happened to
2856 work because 0=false and not-0=true, but is not correct code.
2858 68. The host_aton() function was not handling scoped IPv6 addresses (those
2859 with, for example, "%eth0" on the end) correctly.
2861 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2862 negated items (that is, ~something) in unsigned ints. Some compilers
2863 apparently mutter when there is no cast.
2865 70. If an address verification called from an ACL failed, and did not produce a
2866 user-specific message (i.e. there was only a "system" message), nothing was
2867 put in $acl_verify_message. In this situation, it now puts the system
2870 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2871 session; change 4.31/43 added the unwanted input to the log line - except
2872 that it did not do this in the start of session case. It now does.
2874 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2875 This is wrong and can cause the other end to generate a synchronization
2876 error if it is another Exim or anything else that does the synchronization
2877 check. A QUIT command is no longer sent after a timeout.
2879 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2880 during host lookups.
2882 74. The defer_ok option of callout verification was not working if it was used
2883 when verifying addresses in header lines, that is, for this case:
2885 verify = header_sender/callout=defer_ok
2887 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2888 those file descriptors could be used for SMTP connections. If anything
2889 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2890 could be sent to the SMTP client, causing chaos. The daemon now opens
2891 stdin, stdout, and stderr to /dev/null when it puts itself into the
2894 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2895 log by default. The user can override this with suitable Perl magic.
2897 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2898 the log message when discard triggers, was not working for the DATA ACL or
2899 for the non-SMTP ACL.
2901 78. Error message wording change in sieve.c.
2903 79. If smtp_accept_max_per_host was set, the number of connections could be
2904 restricted to fewer than expected, because the daemon was trying to set up
2905 a new connection before checking whether the processes handling previous
2906 connections had finished. The check for completed processes is now done
2907 earlier. On busy systems, this bug wouldn't be noticed because something
2908 else would have woken the daemon, and it would have reaped the completed
2911 80. If a message was submitted locally by a user whose login name contained one
2912 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2913 It caused a spool format error. I have fixed the spool reading code. A
2914 related problem was that the "from" clause in the Received: line became
2915 illegal because of the space(s). It is now covered by ${quote_local_part.
2917 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2920 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2922 83. Patch from Sieve maintainer to fix unterminated string problem in
2923 "vacation" handling.
2925 84. Some minor changes to the Linux configuration files to help with other
2926 OS variants using glibc.
2928 85. One more patch for Sieve to update vacation handling to latest spec.
2931 ----------------------------------------------------
2932 See the note above about the 4.44 and 4.50 releases.
2933 ----------------------------------------------------
2939 1. Change 4.43/35 introduced a bug that caused file counts to be
2940 incorrectly computed when quota_filecount was set in an appendfile
2943 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2944 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2947 3. Give more explanation in the error message when the command for a transport
2948 filter fails to execute.
2950 4. There are several places where Exim runs a non-Exim command in a
2951 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2952 was being done only for the command run by the queryprogram router. It is
2953 now done for all such subprocesses. The other cases are: ${run, transport
2954 filters, and the commands run by the lmtp and pipe transports.
2956 5. Some older OS have a limit of 256 on the maximum number of file
2957 descriptors. Exim was using setrlimit() to set 1000 as a large value
2958 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2959 systems. I've change it so that if it can't get 1000, it tries for 256.
2961 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2962 was an oversight, and furthermore, ever since the addition of extra
2963 controls (e.g. 4.43/32), the checks on when to allow different forms of
2964 "control" were broken. There should now be diagnostics for all cases when a
2965 control that does not make sense is encountered.
2967 7. $recipients is now available in the predata ACL (oversight).
2969 8. Tidy the search cache before the fork to do a delivery from a message
2970 received from the command line. Otherwise the child will trigger a lookup
2971 failure and thereby defer the delivery if it tries to use (for example) a
2972 cached ldap connection that the parent has called unbind on.
2974 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2975 of $address_data from the recipient verification was clobbered by the
2976 sender verification.
2978 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2979 was its contents. (It was OK if the option was not defined at all.)
2981 11. A "Completed" log line is now written for messages that are removed from
2982 the spool by the -Mrm option.
2984 12. $host_address is now set to the target address during the checking of
2985 ignore_target_hosts.
2987 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2988 being passed; this would have caused $sender_host_name to have been used if
2989 matching the list had actually called for a host name (not very likely,
2990 since this list is usually IP addresses). A host name is now passed as
2993 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2994 code that specifies a non-restarting handler (typically sigaction() in
2995 modern systems) in an attempt to fix a rare and obscure crash bug.
2997 15. Narrowed the window for a race in the daemon that could cause it to ignore
2998 SIGCHLD signals. This is not a major problem, because they are used only to
2999 wake it up if nothing else does.
3001 16. A malformed maildirsize file could cause Exim to calculate negative values
3002 for the mailbox size or file count. Odd effects could occur as a result.
3003 The maildirsize information is now recalculated if the size or filecount
3006 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3007 support for a long time. Removed HAVE_SYS_VFS_H.
3009 18. Updated exipick to current release from John Jetmore.
3011 19. Allow an empty sender to be matched against a lookup in an address list.
3012 Previously the only cases considered were a regular expression, or an
3015 20. Exim went into a mad DNS lookup loop when doing a callout where the
3016 host was specified on the transport, if the DNS lookup yielded more than
3019 21. The RFC2047 encoding function was originally intended for short strings
3020 such as real names; it was not keeping to the 75-character limit for
3021 encoded words that the RFC imposes. It now respects the limit, and
3022 generates multiple encoded words if necessary. To be on the safe side, I
3023 have increased the buffer size for the ${rfc2047: expansion operator from
3026 22. Failure to deliver a bounce message always caused it to be frozen, even if
3027 there was an errors_to setting on the router. The errors_to setting is now
3030 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3031 canonical form (fully expanded) before being placed in
3032 $sender_host_address.
3034 24. Updated eximstats to version 1.33
3036 25. Include certificate and key file names in error message when GnuTLS fails
3037 to set them up, because the GnuTLS error message doesn't include the name
3038 of the failing file when there is a problem reading it.
3040 26. Expand error message when OpenSSL has problems setting up cert/key files.
3043 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3044 (this can affect the format of dates).
3046 28. exim_tidydb, when checking for the continued existence of a message for
3047 which it has found a message-specific retry record, was not finding
3048 messages that were in split spool directories. Consequently, it was
3049 deleting retry records that should have stayed in existence.
3051 29. eximstats updated to version 1.35
3052 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3053 1.35 - bugfix such that pie charts by volume are generated correctly
3055 30. The SPA authentication driver was not abandoning authentication and moving
3056 on to the next authenticator when an expansion was forced to fail,
3057 contradicting the general specification for all authenticators. Instead it
3058 was generating a temporary error. It now behaves as specified.
3060 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3061 (the order specifies the preference for clients). The order is now AES256,
3062 AES128, 3DES, ARCFOUR128.
3064 31. Small patch to Sieve code - explicitly set From: when generating an
3067 32. Exim crashed if a remote delivery caused a very long error message to be
3068 recorded - for instance if somebody sent an entire SpamAssassin report back
3069 as a large number of 550 error lines. This bug was coincidentally fixed by
3070 increasing the size of one of Exim's internal buffers (big_buffer) that
3071 happened as part of the Exiscan merge. However, to be on the safe side, I
3072 have made the code more robust (and fixed the comments that describe what
3075 33. Some experimental protocols are using DNS PTR records for new purposes. The
3076 keys for these records are domain names, not reversed IP addresses. The
3077 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3078 leaves it alone. Component reversal etc. now happens only for IP addresses.
3081 34. The host_aton() function is supposed to be passed a string that is known
3082 to be a valid IP address. However, in the case of IPv6 addresses, it was
3083 not checking this. This is a hostage to fortune. Exim now panics and dies
3084 if the condition is not met. A case was found where this could be provoked
3085 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3086 components; fortuitously, this particular loophole had already been fixed
3087 by change 4.50/55 or 4.44/33 above.
3089 If there are any other similar loopholes, the new check in host_aton()
3090 itself should stop them being exploited. The report I received stated that
3091 data on the command line could provoke the exploit when Exim was running as
3092 exim, but did not say which command line option was involved. All I could
3093 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3094 running as the user.
3097 35. There was a buffer overflow vulnerability in the SPA authentication code
3098 (which came originally from the Samba project). I have added a test to the
3099 spa_base64_to_bits() function which I hope fixes it.
3102 36. The daemon start-up calls getloadavg() while still root for those OS that
3103 need the first call to be done as root, but it missed one case: when
3104 deliver_queue_load_max is set with deliver_drop_privilege. This is
3105 necessary for the benefit of the queue runner, because there is no re-exec
3106 when deliver_drop_privilege is set.
3108 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3109 was in use, was not putting the data itself into the right store pool;
3110 consequently, it could be overwritten for a subsequent message in the same
3111 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3114 38. Sometimes the final signoff response after QUIT could fail to get
3115 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3116 < 0 before doing a fflush(). This bug looks as though it goes back to the
3117 introduction of TLS in release 3.20, but "sometimes" must have been rare
3118 because the tests only now provoked it.
3124 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3125 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3126 Now it returns an integer. A number of calls were still expecting a T/F
3127 return. Fortuitously, in all cases, the tests worked in OK situations,
3128 which is the norm. However, things would have gone wrong on any write
3129 failures on the smtp file descriptor. This function is used when sending
3130 messages over SMTP and also when doing verify callouts.
3132 2. When Exim is called to do synchronous delivery of a locally submitted
3133 message (the -odf or -odi options), it no longer closes stderr before doing
3136 3. Implemented the mua_wrapper option.
3138 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3140 5. Implemented the functions header_remove(), header_testname(),
3141 header_add_at_position(), and receive_remove_recipient(), and exported them
3144 6. If an ACL "warn" statement specified the addition of headers, Exim already
3145 inserted X-ACL-Warn: at the start if there was no header name. However, it
3146 was not making this test for the second and subsequent header lines if
3147 there were newlines in the string. This meant that an invalid header could
3148 be inserted if Exim was badly configured.
3150 7. Allow an ACL "warn" statement to add header lines at the start or after all
3151 the Received: headers, as well as at the end.
3153 8. Added the rcpt_4xx retry error code.
3155 9. Added postmaster_mailfrom=xxx to callout verification option.
3157 10. Added mailfrom=xxxx to the callout verification option, for verify=
3160 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3161 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3163 12. Inserted some casts to stop certain compilers warning when using pointer
3164 differences as field lengths or precisions in printf-type calls (mostly
3165 affecting debugging statements).
3167 13. Added optional readline() support for -be (dynamically loaded).
3169 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3170 same clock tick as a message's arrival, so that its received time was the
3171 same as the "first fail" time on the retry record, and that message
3172 remained on the queue past the ultimate address timeout, every queue runner
3173 would try a delivery (because it was past the ultimate address timeout) but
3174 after another failure, the ultimate address timeout, which should have then
3175 bounced the address, did not kick in. This was a "< instead of <=" error;
3176 in most cases the first failure would have been in the next clock tick
3177 after the received time, and all would be well.
3179 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3180 being recognized when the domain list was tested by the match_domain
3181 condition in an expansion string.
3183 16. Added the ${str2b64: operator.
3185 17. Exim was always calling setrlimit() to set a large limit for the number of
3186 processes, without checking whether the existing limit was already
3187 adequate. (It did check for the limit on file descriptors.) Furthermore,
3188 errors from getrlimit() and setrlimit() were being ignored. Now they are
3189 logged to the main and panic logs, but Exim does carry on, to try to do its
3190 job under whatever limits there are.
3192 18. Imported PCRE 5.0.
3194 19. Trivial typo in log message " temporarily refused connection" (the leading
3197 20. If the log selector return_path_on_delivery was set and an address was
3198 redirected to /dev/null, the delivery process crashed because it assumed
3199 that a return path would always be set for a "successful" delivery. In this
3200 case, the whole delivery is bypassed as an optimization, and therefore no
3203 21. Internal re-arrangement: the function for sending a challenge and reading
3204 a response while authentication was assuming a zero-terminated challenge
3205 string. It's now changed to take a pointer and a length, to allow for
3206 binary data in such strings.
3208 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3210 23. Exim was not respecting finduser_retries when seeking the login of the
3211 uid under which it was called; it was always trying 10 times. (The default
3212 setting of finduser_retries is zero.) Also, it was sleeping after the final
3213 failure, which is pointless.
3215 24. Implemented tls_on_connect_ports.
3217 25. Implemented acl_smtp_predata.
3219 26. If the domain in control=submission is set empty, Exim assumes that the
3220 authenticated id is a complete email address when it generates From: or
3221 Sender: header lines.
3223 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3224 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3225 chown and chgrp in /bin and hostname in /usr/bin.
3227 28. Exim was keeping the "process log" file open after each use, just as it
3228 does for the main log. This opens the possibility of it remaining open for
3229 long periods when the USR1 signal hits a daemon. Occasional processlog
3230 errors were reported, that could have been caused by this. Anyway, it seems
3231 much more sensible not to leave this file open at all, so that is what now
3234 29. The long-running daemon process does not normally write to the log once it
3235 has entered its main loop, and it closes the log before doing so. This is
3236 so that log files can straightforwardly be renamed and moved. However,
3237 there are a couple of unusual error situations where the daemon does write
3238 log entries, and I had neglected to close the log afterwards.
3240 30. The text of an SMTP error response that was received during a remote
3241 delivery was being truncated at 512 bytes. This is too short for some of
3242 the long messages that one sometimes sees. I've increased the limit to
3245 31. It is now possible to make retry rules that apply only when a message has a
3246 specific sender, in particular, an empty sender.
3248 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3249 it possible to be selective about when SMTP synchronization is enforced.
3251 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3253 32. Implemented hosts_connection_nolog.
3255 33. Added an ACL for QUIT.
3257 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3260 35. Added mailbox_size and mailbox_filecount to appendfile.
3262 36. Added control = no_multiline_responses to ACLs.
3264 37. There was a bug in the logic of the code that waits for the clock to tick
3265 in the case where the clock went backwards by a substantial amount such
3266 that the microsecond fraction of "now" was more than the microsecond
3267 fraction of "then" (but the whole seconds number was less).
3269 38. Added support for the libradius Radius client library this is found on
3270 FreeBSD (previously only the radiusclient library was supported).
3276 1. When certain lookups returned multiple values in the form name=value, the
3277 quoting of the values was not always being done properly. Specifically:
3278 (a) If the value started with a double quote, but contained no whitespace,
3280 (b) If the value contained whitespace other than a space character (i.e.
3281 tabs or newlines or carriage returns) it was not quoted.
3282 This fix has been applied to the mysql and pgsql lookups by writing a
3283 separate quoting function and calling it from the lookup code. The fix
3284 should probably also be applied to nisplus, ibase and oracle lookups, but
3285 since I cannot test any of those, I have not disturbed their existing code.
3287 2. A hit in the callout cache for a specific address caused a log line with no
3288 reason for rejecting RCPT. Now it says "Previous (cached) callout
3289 verification failure".
3291 3. There was an off-by-one bug in the queryprogram router. An over-long
3292 return line was truncated at 256 instead of 255 characters, thereby
3293 overflowing its buffer with the terminating zero. As well as fixing this, I
3294 have increased the buffer size to 1024 (and made a note to document this).
3296 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3297 when Exim is waiting for an SMTP response from a remote server, Exim
3298 restarts its select() call on the socket, thereby resetting its timeout.
3299 This is not a problem when such interrupts are rare. Somebody set up a cron
3300 job to run exiwhat every 2 minutes, which is less than the normal select()
3301 timeout (5 or 10 minutes). This meant that the select() timeout never
3302 kicked in because it was always reset. I have fixed this by comparing the
3303 time when an interrupt arrives with the time at the start of the first call
3304 to select(). If more time than the timeout has elapsed, the interrupt is
3305 treated as a timeout.
3307 5. Some internal re-factoring in preparation for the addition of Sieve
3308 extensions (by MH). In particular, the "personal" test is moved to a
3309 separate function, and given an option for scanning Cc: and Bcc: (which is
3310 not set for Exim filters).
3312 6. When Exim created an email address using the login of the caller as the
3313 local part (e.g. when creating a From: or Sender: header line), it was not
3314 quoting the local part when it contained special characters such as @.
3316 7. Installed new OpenBSD configuration files.
3318 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3319 try to make them clearer.
3321 9. Callout options, other than the timeout value, were being ignored when
3322 verifying sender addresses in header lines. For example, when using
3324 verify = header_sender/callout=no_cache
3326 the cache was (incorrectly) being used.
3328 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3329 only the Cygwin environment.
3331 11. When return_path_on_delivery was set as a log selector, if different remote
3332 addresses in the same message used different return paths and parallel
3333 remote delivery occurred, the wrong values would sometimes be logged.
3334 (Whenever a remote delivery process finished, the return path value from
3335 the most recently started remote delivery process was logged.)
3337 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3338 lines when AUTH and/or TLS are in use. This is the "received protocol"
3339 field. Exim used to use "asmtp" for authenticated SMTP, without any
3340 indication (in the protocol name) for TLS use. Now it follows the RFC and
3341 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3342 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3343 names appear in log lines as well as in Received: header lines.
3345 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3346 extensions, and comparison tests, and to fix some bugs.
3348 14. Changes to the "personal" filter test:
3350 (1) The test was buggy in that it was just doing the equivalent of
3351 "contains" tests on header lines. For example, if a user's address was
3352 anne@some.where, the "personal" test would incorrectly be true for
3354 To: susanne@some.where
3356 This test is now done by extracting each address from the header in turn,
3357 and checking the entire address. Other tests that are part of "personal"
3358 are now done using regular expressions (for example, to check local parts
3359 of addresses in From: header lines).
3361 (2) The list of non-personal local parts in From: addresses has been
3362 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3363 taken from the Sieve specification recommendations.
3365 (3) If the message contains any header line starting with "List-" it is
3366 treated as non-personal.
3368 (4) The test for "circular" in the Subject: header line has been removed
3369 because it now seems ill-conceived.
3371 15. Minor typos in src/EDITME comments corrected.
3373 16. Installed latest exipick from John Jetmore.
3375 17. If headers_add on a router specified a text string that was too long for
3376 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3377 of string_sprintf() is now avoided.
3379 18. $message_body_size was not set (it was always zero) when running the DATA
3380 ACL and the local_scan() function.
3382 19. For the "mail" command in an Exim filter, no default was being set for
3383 the once_repeat time, causing a random time value to be used if "once" was
3384 specified. (If the value happened to be <= 0, no repeat happened.) The
3385 default is now 0s, meaning "never repeat". The "vacation" command was OK
3386 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3387 (I found it when inspecting the code).
3389 20. There is now an overall timeout for performing a callout verification. It
3390 defaults to 4 times the callout timeout, which applies to individual SMTP
3391 commands during the callout. The overall timeout applies when there is more
3392 than one host that can be tried. The timeout is checked before trying the
3393 next host. This prevents very long delays if there are a large number of
3394 hosts and all are timing out (e.g. when the network connections are timing
3395 out). The value of the overall timeout can be changed by specifying an
3396 additional sub-option for "callout", called "maxwait". For example:
3398 verify = sender/callout=5s,maxwait=20s
3400 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3401 to the end before writing, but this should make it even safer).
3403 22. Exim was forgetting that it had advertised PIPELINING for the second and
3404 subsequent messages on an SMTP connection. It was also not resetting its
3405 memory on STARTTLS and an internal HELO.
3407 23. When Exim logs an SMTP synchronization error within a session, it now
3408 records whether PIPELINING has been advertised or not.
3410 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3411 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3412 rather than long int.
3414 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3416 26. Added the never_mail option to autoreply.
3422 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3423 crash if the getsockname() call failed; this can happen if a connection is
3424 closed very soon after it is established. The problem was simply in the
3425 order in which certain operations were done, causing Exim to try to write
3426 to the SMTP stream before it had set up the file descriptor. The bug has
3427 been fixed by making things happen in the correct order.
3433 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3434 before the connection was closed, thus losing the rejection response.
3436 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3437 some early Solaris releases, but causes trouble in current releases where
3438 socklen_t is defined.
3440 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3443 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3446 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3447 that releasing the top of it at the end releases what was used for sub-
3448 expansions (unless the block got too big). However, discard this block if
3449 the first thing is a variable or header, so that we can use its block when
3450 it is dynamic (useful for very large $message_headers, for example).
3452 6. Lookups now cache *every* query, not just the most recent. A new, separate
3453 store pool is used for this. It can be recovered when all lookup caches are
3454 flushed. Lookups now release memory at the end of their result strings.
3455 This has involved some general refactoring of the lookup sources.
3457 7. Some code has been added to the store_xxx() functions to reduce the amount
3458 of flapping under certain conditions.
3460 8. log_incoming_interface used to affect only the <= reception log lines. Now
3461 it causes the local interface and port to be added to several more SMTP log
3462 lines, for example "SMTP connection from", and rejection lines.
3464 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3466 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3468 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3469 could be overwritten at the end of the current message (or the start of a
3470 new message if it was set in a HELO ACL). The value is now preserved for
3471 the duration of the SMTP connection.
3473 12. If a transport had a headers_rewrite setting, and a matching header line
3474 contained an unqualified address, that address was qualified, even if it
3475 did not match any rewriting rules. The underlying bug was that the values
3476 of the flags that permit the existence of unqualified sender and recipient
3477 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3478 non-local messages, and by -bnq for local messages) were not being
3479 preserved with the message after it was received.
3481 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3482 "next input=" as part of the text comprising the host identity instead of
3483 the correct text. The code was using the same buffer for two different
3484 strings. However, depending on which order the printing function evaluated
3485 its arguments, the bug did not always show up. Under Linux, for example, my
3486 test suite worked just fine.
3488 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3489 doesn't work with some older versions of Perl. It has been changed to "my",
3490 which in any case is probably the better facility to use.
3492 15. A really picky compiler found some instances of statements for creating
3493 error messages that either had too many or two few arguments for the format
3496 16. The size of the buffer for calls to the DNS resolver has been increased
3497 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3498 for addresses that have a lot of PTR records. This alleviates a problem; it
3499 does not fully solve it.
3501 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3502 buffer now truncates the list and logs the incident, which is the same
3503 action as happens when Exim is looking up a host name and its aliases.
3504 Previously in this situation something unpredictable would happen;
3505 sometimes it was "internal error: store_reset failed".
3507 18. If a server dropped the connection unexpectedly when an Exim client was
3508 using GnuTLS and trying to read a response, the client delivery process
3509 crashed while trying to generate an error log message.
3511 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3512 string, for example:
3514 warn message = H1: something\nH2: something
3516 the text was added as a single header line from Exim's point of view
3517 though it ended up OK in the delivered message. However, searching for the
3518 second and subsequent header lines using $h_h2: did not work. This has been
3519 fixed. Similarly, if a system filter added multiple headers in this way,
3520 the routers could not see them.
3522 20. Expanded the error message when iplsearch is called with an invalid key to
3523 suggest using net-iplsearch in a host list.
3525 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3526 ACLs are no longer actually imposed (and a message to that effect is
3529 22. If a "gecos" field in a passwd entry contained escaped characters, in
3530 particular, if it contained a \" sequence, Exim got it wrong when building
3531 a From: or a Sender: header from that name. A second bug also caused
3532 incorrect handling when an unquoted " was present following a character
3533 that needed quoting.
3535 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3536 was not being matched caselessly.
3538 24. Arranged for all hyphens in the exim.8 source to be escaped with
3541 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3542 properly. Recipient callout cache records were still being keyed to include
3543 the sender, even when use_sender was set false. This led to far more
3544 callouts that were necessary. The sender is no longer included in the key
3545 when use_sender is false.
3547 26. Added "control = submission" modifier to ACLs.
3549 27. Added the ${base62d: operator to decode base 62 numbers.
3551 28. dnsdb lookups can now access SRV records.
3553 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3554 the configuration file.
3556 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3557 (-v) mode. This makes the output for a verbose queue run more intelligible.
3559 31. Added a use_postmaster feature to recipient callouts.
3561 32. Added the $body_zerocount variable, containing the number of binary zero
3562 bytes in the message body.
3564 33. The time of last modification of the "new" subdirectory is now used as the
3565 "mailbox time last read" when there is a quota error for a maildir
3568 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3570 35. Added +ignore_unknown as a special item in host lists.
3572 36. Code for decoding IPv6 addresses in host lists is now included, even if
3573 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3574 address was recognized as an IP address, but was then not correctly decoded
3575 into binary, causing unexpected and incorrect effects when compared with
3582 1. Very minor rewording of debugging text in manualroute to say "list of
3583 hosts" instead of "hostlist".
3585 2. If verify=header_syntax was set, and a header line with an unqualified
3586 address (no domain) and a large number of spaces between the end of the
3587 name and the colon was received, the reception process suffered a buffer
3588 overflow, and (when I tested it) crashed. This was caused by some obsolete
3589 code that should have been removed. The fix is to remove it!
3591 3. When running in the test harness, delay a bit after writing a bounce
3592 message to get a bit more predictability in the log output.
3594 4. Added a call to search_tidyup() just before forking a reception process. In
3595 theory, someone could use a lookup in the expansion of smtp_accept_max_
3596 per_host which, without the tidyup, could leave open a database connection.
3598 5. Added the variables $recipient_data and $sender_data which get set from a
3599 lookup success in an ACL "recipients" or "senders" condition, or a router
3600 "senders" option, similar to $domain_data and $local_part_data.
3602 6. Moved the writing of debug_print from before to after the "senders" test
3605 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3606 problems for message scanning, either using a data ACL, or using
3607 local_scan() because the Received: header was not generated till after they
3608 were called (in order to set the time as the time of reception completion).
3609 I have revised the way this works. The header is now generated after the
3610 body is received, but before the ACL or local_scan() are called. After they
3611 are run, the timestamp in the header is updated.
3617 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3618 before starting a queue runner without re-exec. This happened only when
3619 deliver_drop_privilege was set or when the Exim user was set to root. The
3620 effect of the bug was that timeouts during subsequent deliveries caused
3621 crashes instead of being properly handled. The handler is now left at its
3622 default (and expected) setting.
3624 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3625 message, again when deliver_drop_privilege is set or Exim is run as root.
3626 The bug described in (1) was not present in this case, but the tidying up
3627 of the other signals was missing. I have made the two cases consistent.
3629 3. The ignore_target_hosts setting on a manualroute router was being ignored
3630 for hosts that were looked up using the /MX notation.
3632 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3635 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3636 operated on the sender address. After changing the $sender_address to <>
3637 for the sender address verify, Exim was re-instated it as the original
3638 (before rewriting) address, but remembering that it had rewritten it, so it
3639 wasn't rewriting it again. This bug also had the effect of breaking the
3640 sender address verification caching when the sender address was rewritten.
3642 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3643 This has been changed so that if the ip literal address matches
3644 ignore_target_hosts, the router declines.
3646 7. Added expansion conditions match_domain, match_address, and match_local_
3647 part (NOT match_host).
3649 8. The placeholder for the Received: header didn't have a length field set.
3651 9. Added code to Exim itself and to exim_lock to test for a specific race
3652 condition that could lead to file corruption when using MBX delivery. The
3653 issue is with the lockfile that is created in /tmp. If this file is removed
3654 after a process has opened it but before that process has acquired a lock,
3655 there is the potential for a second process to recreate the file and also
3656 acquire a lock. This could lead to two Exim processes writing to the file
3657 at the same time. The added code performs the same test as UW imapd; it
3658 checks after acquiring the lock that its file descriptor still refers to
3659 the same named file.
3661 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3662 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3665 11. Added the smtp_active_hostname option. If used, this will typically be made
3666 to depend on the incoming interface address. Because $interface_address is
3667 not set up until the daemon has forked a reception process, error responses
3668 that can happen earlier (such as "too many connections") no longer contain
3671 12. If an expansion in a condition on a "warn" statement fails because a lookup
3672 defers, the "warn" statement is abandoned, and the next ACL statement is
3673 processed. Previously this caused the whole ACL to be aborted.
3675 13. Added the iplsearch lookup type.
3677 14. Added ident_timeout as a log selector.
3679 15. Added tls_certificate_verified as a log selector.
3681 16. Added a global option tls_require_ciphers (compare the smtp transport
3682 option of the same name). This controls incoming TLS connections.
3684 17. I finally figured out how to make tls_require_ciphers do a similar thing
3685 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3686 before starting the TLS session.
3688 18. Tabs are now shown as \t in -bP output.
3690 19. If the log selector return_path_on_delivery was set, Exim crashed when
3691 bouncing a message because it had too many Received: header lines.
3693 20. If two routers both had headers_remove settings, and the first one included
3694 a superfluous trailing colon, the final name in the first list and the
3695 first name in the second list were incorrectly joined into one item (with a
3696 colon in the middle).
3702 1. Added -C and -D options to the exinext utility, mainly to make it easier
3703 to include in the automated testing, but these could be helpful when
3704 multiple configurations are in use.
3706 2. The exinext utility was not formatting the output nicely when there was
3707 an alternate port involved in the retry record key, nor when there was a
3708 message id as well (for retries that were specific to a specific message
3709 and a specific host). It was also confused by IPv6 addresses, because of
3710 the additional colons they contain. I have fixed the IPv4 problem, and
3711 patched it up to do a reasonable job for IPv6.
3713 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3714 delivery, the log line now contains "pipelined" if PIPELINING was used.
3716 4. An SMTP transport process used to panic and die if the bind() call to set
3717 an explicit outgoing interface failed. This has been changed; it is now
3718 treated in the same way as a connect() failure.
3720 5. A reference to $sender_host_name in the part of a conditional expansion
3721 that was being skipped was still causing a DNS lookup. This no longer
3724 6. The def: expansion condition was not recognizing references to header lines
3725 that used bh_ and bheader_.
3727 7. Added the _cache feature to named lists.
3729 8. The code for checking quota_filecount in the appendfile transport was
3730 allowing one more file than it should have been.
3732 9. For compatibility with Sendmail, the command line option
3740 and sets the incoming protocol and host name (for trusted callers). The
3741 host name and its colon can be omitted when only the protocol is to be set.
3742 Note the Exim already has two private options, -pd and -ps, that refer to
3743 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3744 "s", but I don't think that's a major issue.
3746 10. A number of refactoring changes to the code, none of which should affect
3749 (a) The number of logging options was getting close to filling up the
3750 32-bit word that was used as a bit map. I have split them into two classes:
3751 those that are passed in the argument to log_write(), and those that are
3752 only ever tested independently outside of that function. These are now in
3753 separate 32-bit words, so there is plenty of room for expansion again.
3754 There is no change in the user interface or the logging behaviour.
3756 (b) When building, for example, log lines, the code previously used a
3757 macro that called string_cat() twice, in order to add two strings. This is
3758 not really sufficiently general. Furthermore, there was one instance where
3759 it was actually wrong because one of the argument was used twice, and in
3760 one call a function was used. (As it happened, calling the function twice
3761 did not affect the overall behaviour.) The macro has been replaced by a
3762 function that can join an arbitrary number of extra strings onto a growing
3765 (c) The code for expansion conditions now uses a table and a binary chop
3766 instead of a serial search (which was left over from when there were very
3767 few conditions). Also, it now recognizes conditions like "pam" even when
3768 the relevant support is not compiled in: a suitably worded error message is
3769 given if an attempt is made to use such a condition.
3771 11. Added ${time_interval:xxxxx}.
3773 12. A bug was causing one of the ddress fields not to be passed back correctly
3774 from remote delivery subprocesses. The field in question was not being
3775 subsequently used, so this caused to problems in practice.
3777 13. Added new log selectors queue_time and deliver_time.
3779 14. Might have fixed a bug in maildirsizefile handling that threw up
3780 "unexpected character" debug warnings, and recalculated the data
3781 unnecessarily. In any case, I expanded the warning message to give more
3784 15. Added the message "Restricted characters in address" to the statements in
3785 the default ACL that block characters like @ and % in local parts.
3787 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3788 Three changes have been made:
3790 (a) There was a serious bug; a negative response to MAIL caused the whole
3791 recipient domain to be cached as invalid, thereby blocking all messages
3792 to all local parts at the same domain, from all senders. This bug has
3793 been fixed. The domain is no longer cached after a negative response to
3794 MAIL if the sender used is not empty.
3796 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3799 (c) A new callout option, "use_sender" has been added for people who want
3800 the modified behaviour.
3806 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3809 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3810 indeed breaks things for older releases.
3812 3. Added additional logging to the case where there is a problem reading data
3813 from a filter that is running in a subprocess using a pipe, in order to
3814 try to track down a specific problem.
3816 4. Testing facility fudge: when running in the test harness and attempting
3817 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3818 getting "No route to host". Convert this to a timeout.
3820 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3823 6. Some OS don't have socklen_t but use size_t instead. This affects the
3824 fifth argument of getsockopt() amongst other things. This is now
3825 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3826 can be set for individual OS. I have set it for SunOS5, OSF1, and
3827 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3828 some earlier ones do not.
3830 7. Change 4.30/15 was not doing the test caselessly.
3832 8. The standard form for an IPv6 address literal was being rejected by address
3833 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3834 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3835 this, as well as the form without the "IPv6" on the front (but only when
3836 address literals are enabled, of course).
3838 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3840 10. Exim crashed if a message with an empty sender address specified by -f
3841 encountered a router with an errors_to setting. This could be provoked only
3842 by a command such as
3846 where an empty string was supplied; "<>" did not hit this bug.
3848 11. Installed PCRE release 4.5.
3850 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3851 remained set. It is now erased.
3853 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3854 times from message ids (which are base 36 rather than the normal 62).
3856 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3857 were being counted as actual protocol errors, and logged if the log
3858 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3859 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3860 and DATA following a set of rejected RCPTs do not count as protocol errors.
3861 In other words, Exim assumes they were pipelined, though this may not
3862 actually be the case. Of course, in all cases the client gets an
3863 appropriate error code.
3865 15. If a lookup fails in an ACL condition, a message about the failure may
3866 be available; it is used if testing the ACL cannot continue, because most
3867 such messages specify what the cause of the deferral is. However, some
3868 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3869 that caused an old message to be retained and used if a later statement
3870 caused a defer, replacing the real cause of the deferral.
3872 16. If an IP address had so many PTR records that the DNS lookup buffer
3873 was not large enough to hold them, Exim could crash while trying to process
3874 the truncated data. It now detects and logs this case.
3876 17. Further to 4.21/58, another change has been made: if (and only if) the
3877 first line of a message (the first header line) ends with CRLF, a bare LF
3878 in a subsequent header line has a space inserted after it, so as not to
3879 terminate the header.
3881 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3882 unnecessarily, used atoi() instead of strtol(), and didn't check the
3883 termination when getting file sizes from file names by regex.
3885 19. Completely re-implemented the support for maildirsize files, in the light
3886 of a number of problems with the previous contributed implementation
3887 (4.30/29). In particular:
3889 . If the quota is zero, the maildirsize file is maintained, but no quota is
3892 . If the maildir directory does not exist, it is created before any attempt
3893 to write a maildirsize file.
3895 . The quota value in the file is just a cache; if the quota is changed in
3896 the transport, the new value overrides.
3898 . A regular expression is available for excluding directories from the
3901 20. The autoreply transport checks the characters in options that define the
3902 message's headers; it allows continued headers, but it was checking with
3903 isspace() after an embedded newline instead of explicitly looking for a
3906 21. If all the "regular" hosts to which an address was routed had passed their
3907 expiry times, and had not reached their retry times, the address was
3908 bounced, even if fallback hosts were defined. Now Exim should go on to try
3911 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3912 equivalent code in the SMTP transport. Some hosts send humungous responses
3913 to HELO/EHLO, more than 1024 it seems.
3915 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3916 gives compiler warnings in some environments. I've now done it "properly",
3919 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3920 (because of problems with the built-in one) was declared to return uschar *
3921 instead of char *, causing compiler failure.
3923 25. Fixed a file descriptor leak when processing alias/forward files.
3925 26. Fixed a minor format string issue in dbfn.c.
3927 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3929 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3930 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3933 29. When a sender address is verified, it is cached, to save repeating the test
3934 when there is more than one recipient in a message. However, when the
3935 verification involves a callout, it is possible for different callout
3936 options to be set for different recipients. It is too complicated to keep
3937 track of this in the cache, so now Exim always runs a verification when a
3938 callout is required, relying on the callout cache for the optimization.
3939 The overhead is duplication of the address routing, but this should not be
3942 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3943 to be verified with callout=postmaster, and the main callout worked but the
3944 postmaster check failed, the verification correctly failed. However, if a
3945 subsequent RCPT command asked for sender verification *without* the
3946 postmaster check, incorrect caching caused this verification also to fail,
3949 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3950 it was not caching the DNS options (qualify_single, search_parents) that
3951 were used when the lookup failed. A subsequent lookup with different
3952 options therefore always gave the same answer, though there were cases
3953 where it should not have. (Example: a "domains = !$mx_any" option on a
3954 dnslookup router: the "domains" option is always processed without any
3955 widening, but the router might have qualify_single set.) Now Exim uses the
3956 cached value only when the same options are set.
3958 32. Added John Jetmore's "exipick" utility to the distribution.
3960 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3961 than a timeout (e.g. a certificate is required, and is not provided), an
3962 Exim server now closes the connection immediately. Previously it waited for
3963 the client to close - but if the client is SSL, it seems that they each
3964 wait for each other, leading to a delay before one of them times out.
3966 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3967 maintained 0.8.x compatibility because I don't think many are using it, and
3968 it is clearly obsolete.
3970 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3973 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3974 client certificate was expired. A simple patch fixes this, though I don't
3975 understand the full logic of why the verify callback is called multiple
3978 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3979 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3980 which causes problems with some clients (such as the Certicom SSL Plus
3981 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3982 disables the coutermeasure allowing Eudora to connect."
3984 38. Exim was not checking that a write() to a log file succeeded. This could
3985 lead to Bad Things if a log got too big, in particular if it hit a file
3986 size limit. Exim now panics and dies if it cannot write to a log file, just
3987 as it does if it cannot open a log file.
3989 39. Modified OS/Makefile-Linux so that it now contains
3991 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3993 The two -D definitions ensure that Exim is compiled with large file
3994 support, which makes it possible to handle log files that are bigger than
3997 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3998 instance) a domain was checked against a named list that involved a lookup,
3999 causing $domain_data to be set, then another domain was checked against the
4000 same list, then the first domain was re-checked, the value of $domain_data
4001 after the final check could be wrong. In particular, if the second check
4002 failed, it could be set empty. This bug probably also applied to
4005 41. The strip_trailing_dot option was not being applied to the address given
4006 with the -f command-line option.
4008 42. The code for reading a message's header from the spool was incrementing
4009 $received_count, but never initializing it. This meant that the value was
4010 incorrect (doubled) while delivering a message in the same process in which
4011 it was received. In the most common configuration of Exim, this never
4012 happens - a fresh exec is done - but it can happen when
4013 deliver_drop_privilege is set.
4015 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4016 it now includes up to 150 characters of the unexpected data in the log
4019 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4020 and building data strings. The size of both of these buffers was 10 000
4021 bytes - far larger than anybody would *ever* want, thought I. Needless to
4022 say, somebody hit the limit. I have increased the maximum line length to
4023 20 000 and the maximum data length of concatenated lines to 100 000. I have
4024 also fixed two bugs, because there was no checking on these buffers. Tsk,
4025 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4026 buffer is too small.
4028 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4029 lsearch lookups. Now it does.
4031 46. When parsing a route_list item in a manualroute router, a fixed-length
4032 buffer was used for the list of hosts. I made this 1024 bytes long,
4033 thinking that nobody would ever have a list of hosts that long. Wrong.
4034 Somebody had a whole pile of complicated expansion conditions, and the
4035 string was silently truncated, leading to an expansion error. It turns out
4036 that it is easier to change to an unlimited length (owing to other changes
4037 that have happened since this code was originally written) than to build
4038 structure for giving a limitation error. The length of the item that
4039 expands into the list of hosts is now unlimited.
4041 47. The lsearch lookup could not handle data where the length of text line was
4042 more than 4095 characters. Such lines were truncated, leading to shortened
4043 data being returned. It should now handle lines of any length.
4045 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4046 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4049 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4050 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4051 no longer happens in comment lines. A list of replacements is now placed
4052 at the head of all of the source files, except those whose only change is
4053 to replace PERL_COMMAND in the very first #! line.
4055 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4056 messages on the queue, with a bottom-up merge sort, using code contributed
4057 by Michael Haardt. This should make operations like -bp somewhat faster on
4058 large queues. It won't affect queue runners, except when queue_run_in_order
4061 51. Installed eximstats 1.31 in the distribution.
4063 52. Added support for SRV lookups to the dnslookup router.
4065 53. If an ACL referred to $message_body or $message_body_end, the value was not
4066 reset for any messages that followed in the same SMTP session.
4068 54. The store-handling optimization for building very long strings was not
4069 differentiating between the different store pools. I don't think this
4070 actually made any difference in practice, but I've tidied it.
4072 55. While running the routers to verify a sender address, $sender_address
4073 was still set to the sender address. This is wrong, because when routing to
4074 send a bounce to the sender, it would be empty. Therefore, I have changed
4075 it so that, while verifying a sender address, $sender_address is set to <>.
4076 (There is no change to what happens when verifying a recipient address.)
4078 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4079 target A or AAAA records (if not already returned) without resetting the
4080 qualify_single or search_parents options of the DNS resolver. These are
4081 inappropriate in this case because the targets of MX and SRV records must
4082 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4083 target that, when qualified, matched something in the local domain. These
4084 two options are now turned off when doing these lookups.
4086 57. It seems that at least some releases of Reiserfs (which does not have the
4087 concept of a fixed number of inodes) returns zero and not -1 for the
4088 number of available inodes. This interacted badly with check_spool_inodes,
4089 which assumed that -1 was the "no such thing" setting. What I have done is
4090 to check that the total number of inodes is greater than zero before doing
4091 the test of how many are available.
4093 58. When a "warn" ACL statement has a log_message modifier, the message is
4094 remembered, and not repeated. This is to avoid a lot of repetition when a
4095 message has many recipients that cause the same warning to be written.
4096 Howewer, Exim was preserving the list of already written lines for an
4097 entire SMTP session, which doesn't seem right. The memory is now reset if a
4098 new message is started.
4100 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4101 debugging output unless log_rewrite was also set.
4103 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4104 of (int)(handle) when we know that handle contains (void *)(-1).
4106 61. The Exim daemon panic-logs an error return when it closes the incoming
4107 connection. However "connection reset by peer" seems to be common, and
4108 isn't really an error worthy of noting specially, so that particular error
4111 62. When Exim is trying to find all the local interfaces, it used to panic and
4112 die if the ioctl to get the interface flags failed. However, it seems that
4113 on at least one OS (Solaris 9) it is possible to have an interface that is
4114 included in the list of interfaces, but for which you get a failure error
4115 for this call. This happens when the interface is not "plumbed" into a
4116 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4117 failure of the "get flags" call assumes that the interface is down.
4119 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4120 makes life easier for people who are doing arithmetic on fields extracted
4121 from dates, where you often get leading zeros that should not be
4122 interpreted as octal.
4124 64. Added qualify_domain to the redirect router, to override the global
4127 65. If a pathologically long header line contained very many addresses (the
4128 report of this problem mentioned 10 000) and each of them was rewritten,
4129 Exim could use up a very large amount of memory. (It kept on making new
4130 copies of the header line as it rewrote, and never released the old ones.)
4131 At the expense of a bit more processing, the header rewriting function has
4132 been changed so that it no longer eats memory in this way.
4134 66. The generation of the Received: header has been moved from the time that a
4135 message starts to be received, to the time that it finishes. The timestamp
4136 in the Received: header should now be very close to that of the <= log
4137 line. There are two side-effects of this change:
4139 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4140 logged header lines no longer include the local Received: line, because
4141 it has not yet been created. The same applies to a copy of the message
4142 that is returned to a non-SMTP sender when a message is rejected.
4144 (b) When a filter file is tested using -bf, no additional Received: header
4145 is added to the test message. After some thought, I decided that this
4148 This change does not affect the value of $received_for. It is still set
4149 after address rewriting, but before local_scan() is called.
4151 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4153 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4154 gave an unhelpful panic error message, and a defer error. I have managed to
4155 change this behaviour so that it now rejects any supplied certificate,
4156 which seems right, as the list of acceptable certificates is empty.
4158 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4159 gave an unhelpful defer error. I have not managed to make this reject any
4160 supplied certificates, but the error message it gives is "no certificate
4161 supplied", which is not helpful.
4163 70. exigrep's output now also includes lines that are not associated with any
4164 message, but which match the given pattern. Implemented by a patch from
4165 Martin Sluka, which also tidied up the Perl a bit.
4167 71. Recipient callout verification, like sender verification, was using <> in
4168 the MAIL FROM command. This isn't really the right thing, since the actual
4169 sender may affect whether the remote host accepts the recipient or not. I
4170 have changed it to use the actual sender in the callout; this means that
4171 the cache record is now keyed on a recipient/sender pair, not just the
4172 recipient address. There doesn't seem to be a real danger of callout loops,
4173 since a callout by the remote host to check the sender would use <>.
4174 [SEE ABOVE: changed after hitting problems.]
4176 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4177 temporary errors. However, in the case of such a code being given after
4178 the end of a data transmission (i.e. after ".") Exim was failing to write
4179 a retry record for the message. (Yes, there was some broken host that was
4180 actually sending 8xx at this point.)
4182 73. An unknown lookup type in a host list could cause Exim to panic-die when
4183 the list was checked. (An example that provoked this was putting <; in the
4184 middle of a list instead of at the start.) If this happened during a DATA
4185 ACL check, a -D file could be left lying around. This kind of configuration
4186 error no longer causes Exim to die; instead it causes a defer errror. The
4187 incident is still logged to the main and panic logs.
4189 74. Buglet left over from Exim 3 conversion. The message "too many messages
4190 in one connection" was written to the rejectlog but not the mainlog, except
4191 when address rewriting (yes!) was being logged.
4193 75. Added write_rejectlog option.
4195 76. When a system filter was run not as root (that is, when system_filter_user
4196 was set), the values of the $n variables were not being returned to the
4197 main process; thus, they were not subsequently available in the $sn
4200 77. Added +return_path_on_delivery log selector.
4202 78. A connection timeout was being treated differently from recipients deferred
4203 when testing hosts_max_try with a message that was older than the host's
4204 retry timeout. (The host should not be counted, thus allowing all hosts to
4205 be tried at least once before bouncing.) This may have been the cause of an
4206 occasionally reported bug whereby a message would remain on the queue
4207 longer than the retry timeout, but would be bounced if a delivery was
4208 forced. I say "may" because I never totally pinned down the problem;
4209 setting up timeout/retry tests is difficult. See also the next item.
4211 79. The ultimate address timeout was not being applied to errors that involved
4212 a combination of host plus message (for example, a timeout on a MAIL
4213 command). When an address resolved to a number of possible hosts, and they
4214 were not all tried for each delivery (e.g. because of hosts_max_try), a
4215 message could remain on the queue longer than the retry timeout.
4217 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4220 81. Fixed an obscure SMTP outgoing bug which required at least the following
4221 conditions: (a) there was another message waiting for the same server;
4222 (b) the server returned 5xx to all RCPT commands in the first message so
4223 that the message was not completed; (c) the server dropped the connection
4224 or gave a negative response to the RSET that Exim sends to abort the
4225 transaction. The observed case was a dropped connection after DATA that had
4226 been sent in pipelining mode. That is, the server had advertised PIPELINING
4227 but was not implementing it correctly. The effect of the bug was incorrect
4228 behaviour, such as trying another host, and this could lead to a crash.
4234 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4235 and daemon.c were passed as pointers to ints; they should have been
4236 pointers to socklen_t variables (which are typically unsigned ints).
4238 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4241 3. Fixed a really odd bug that affected only the testing scheme; patching a
4242 certain fixed string in the binary changed the value of another string that
4243 happened to be identical to the end of the original first string.
4245 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4246 name", it returns that address as the IP address. On some operating
4247 systems (e.g. Solaris), it also passes back the IP address string as the
4248 "host name". However, on others (e.g. Linux), it passes back an empty
4249 string. Exim wasn't checking for this, and was changing the host name to an
4250 empty string, assuming it had been canonicized.
4252 5. Although rare, it is permitted to have more than one PTR record for a given
4253 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4254 all the names associated with an address, because they do in Solaris.
4255 However, it seems that they do not in Linux for data that comes from the
4256 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4257 I found this out when I moved to a new Linux workstation and tried to run
4258 the Exim test suite.
4260 To get round this problem I have changed the code so that it now does its
4261 own call to the DNS to look up PTR records when searching for a host name.
4262 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4263 addresses that are only in /etc/hosts are still found.
4265 This behaviour is, however, controlled by an option called host_lookup_
4266 order, which defaults to "bydns:byaddr". If people want to use the other
4267 order, or indeed, just use one or the other means of lookup, they can
4268 specify it in this variable.
4270 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4271 some operating systems, this comes back from gethostbyaddr() as an empty
4272 string, and this is what Exim used to test for. However, it seems that in
4273 other systems, "." is yielded. Exim now tests for this case too.
4275 7. The values of check_spool_space and check_log_space are now held internally
4276 as a number of kilobytes instead of an absolute number of bytes. If a
4277 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4278 kilobyte. This means that much larger values can be stored.
4280 8. Exim monitor: an attempt to get the action menu when not actually pointing
4281 at a message produces an empty menu entitled "No message selected". This
4282 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4283 no entries in it ("Shell widget menu has zero width and/or height"). So I
4284 have added a single, blank menu entry in this case.
4286 9. Added ${quote_local_part.
4288 10. MIME decoding is now applied to the contents of Subject: header lines when
4291 11. Now that a reference to $sender_host_address automatically causes a reverse
4292 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4293 host lookup before query-style lookups in lists that might use this
4294 variable. This has therefore been abolished, and the "net-" prefix is no
4295 longer necessary for query-style lookups.
4297 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4298 have been a typo for LFLAGS, so it has been changed.
4300 13. The install script calls Exim with "-C /dev/null" in order to find the
4301 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4302 to be output. Howeve, since Exim outputs its version number before the
4303 error, it didn't break the script. It just looked ugly. I fixed this by
4304 always allowing "-C /dev/null" if the caller is root.
4306 14. Ignore overlarge ACL variable number when reading spool file - insurance
4307 against a later release with more variables having written the file.
4309 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4310 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4311 this, as well as the form without the "IPv6" on the front.
4313 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4314 OS/Makefile-Darwin file.
4316 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4317 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4319 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4320 variable. However, some OS use ints for inodes. I've added cast to long int
4321 to get rid of the compiler warning.
4323 19. I had forgotten to lock out "/../" in configuration file names when
4324 ALT_CONFIG_PREFIX was set.
4326 20. Routers used for verification do not need to specify transports. However,
4327 if such a router generated a host list, and callout was configured, Exim
4328 crashed, because it could not find a port number from the (non-existent)
4329 transport. It now assumes port 25 in this circumstance.
4331 21. Added the -t option to exigrep.
4333 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4334 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4335 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4336 them is set, LOOKUP_LSEARCH is forced.
4338 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4340 24. Added sender and host information to the "rejected by local_scan()" log
4341 line; previously there was no indication of these.
4343 25. Added .include_if_exists.
4345 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4346 Linux. It turns out that not all file systems support this. Apparently some
4347 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4348 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4349 not supported on the file descriptor, is now ignored when Exim is trying to
4350 sync a directory. This applies only to Linux.
4352 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4354 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4355 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4358 29. Added support for maildirsize files from supplied patch (modified a bit).
4360 30. The use of :fail: followed by an empty string could lead Exim to respond to
4361 sender verification failures with (e.g.):
4363 550 Verification failed for <xxx>
4364 550 Sender verify failed
4366 where the first response line was missing the '-' that indicates it is not
4367 the final line of the response.
4369 31. The loop for finding the name of the user that called Exim had a hardwired
4370 limit of 10; it now uses the value of finduser_retries, which is used for
4371 all other user lookups.
4373 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4376 33. Exim was neglecting to zero errno before one call of strtol() when
4377 expanding a string and expecting an integer value. On some systems this
4378 resulted in spurious "integer overflow" errors. Also, it was casting the
4379 result into an int without checking.
4381 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4382 did not work. The code looks as if it has *never* worked, though it appears
4383 to have been documented since at least releast 1.62. I have made it work.
4385 35. The "timeout_DNS" error in retry rules, also documented since at least
4386 1.62, also never worked. As it isn't clear exactly what this means, and
4387 clearly it isn't a major issue, I have abolished the feature by treating it
4388 as "timeout", and writing a warning to the main and panic logs.
4390 36. The display of retry rules for -brt wasn't always showing the error code
4393 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4394 timeout_connect_A, timeout_connect_MX.
4396 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4397 to the empty sender.
4399 39. The daemon was not analysing the content of -oX till after it had closed
4400 stderr and disconnected from the controlling terminal. This meant that any
4401 syntax errors were only noted on the panic log, and the return code from
4402 the command was 0. By re-arranging the code a little, I've made the
4403 decoding happen first, so such errors now appear on stderr, and the return
4404 code is 1. However, the actual setting up of the sockets still happens in
4405 the disconnected process, so errors there are still only recorded on the
4408 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4409 connections (as happens on some IP stacks) was logged at start up time as
4410 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4411 it from "IPv6 and IPv4", which means that two separate sockets are being
4414 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4415 says whether AF_INET or AF_INET6 was passed as an argument.
4417 42. Exiwhat output was messed up when time zones were included in log
4420 43. Exiwhat now gives more information about the daemon's listening ports,
4421 and whether -tls-on-connect was used.
4423 44. The "port" option of the smtp transport is now expanded.
4425 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4426 silently ignored. Now an error message is written to the main and panic
4429 46. There's a new ACL modifier called "logwrite" which writes to a log file
4430 as soon as it is encountered.
4432 47. Added $local_user_uid and $local_user_gid at routing time.
4434 48. Exim crashed when trying to verify a sender address that was being
4437 49. Exim was recognizing only a space character after ".include". It now also
4438 recognizes a tab character.
4440 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4441 extracting the relevant information from the specification. The man page no
4442 longer contains scrambled data for the -d option, and I've added a section
4443 at the front about calling Exim under different names.
4445 51. Added "extra_headers" argument to the "mail" command in filter files.
4447 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4450 53. Installed eximstats 1.29.
4452 54. Added transport_filter_timeout as a generic transport option.
4454 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4455 Cc: header lines. This was required by RFC 822, but it not required by RFC
4458 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4459 incoming messages that did not have them. Now it does so only if the
4460 message originates locally, that is, if there is no associated remote host
4461 address. When Resent- header lines are present, this applies to the Resent-
4462 lines rather than the non-Resent- lines.
4464 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4465 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4467 58. Messages for configuration errors now include the name of the main
4468 configuration file - useful now that there may be more than one file in a
4469 list (.included file names were always shown).
4471 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4472 for those rare installations that do not start the daemon as root or run it
4473 setuid root. I've cut out the call to initgroups() if the daemon is not
4476 60. The Exim user and group can now be bound into the binary as text strings
4477 that are looked up at the start of Exim's processing.
4479 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4481 62. Added $mailstore_basename variable.
4483 63. Installed patch to sieve.c from Michael Haardt.
4485 64. When Exim failed to open the panic log after failing to open the main log,
4486 the original message it was trying to log was written to stderr and debug
4487 output, but if they were not available (the usual case in production), it
4488 was lost. Now it is written to syslog before the two lines that record the
4489 failures to open the logs.
4491 65. Users' Exim filters run in subprocesses under the user's uid. It is
4492 possible for a "deliver" command or an alias in a "personal" command to
4493 provoke an address rewrite. If logging of address rewriting is configured,
4494 this fails because the process is not running as root or exim. There may be
4495 a better way of dealing with this, but for the moment (because 4.30 needs
4496 to be released), I have disabled address rewrite logging when running a
4497 filter in a non-root, non-exim process.
4503 1. The buildconfig auxiliary program wasn't quoting the value set for
4504 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4505 not defined. This bug was masked in 4.22 by the effect that was fixed in
4508 2. Some messages that were rejected after a message id was allocated were
4509 shown as "incomplete" by exigrep. It no longer does this for messages that
4510 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4512 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4513 have allow_domain_literals set, the ID did not get logged in the <= line.
4514 Domain literals are now always recognized in Message-ID: header lines.
4516 4. The first argument for a ${extract expansion item is the key name or field
4517 number. Leading and trailing spaces in this item were not being ignored,
4518 causing some misleading effects.
4520 5. When deliver_drop_privilege was set, single queue runner processes started
4521 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4522 same command in the process it spins off) were not dropping privilege.
4524 6. When the daemon running as "exim" started a queue runner, it always
4525 re-executed Exim in the spun-off process. This is a waste of effort when
4526 deliver_drop_privilege is set. The new process now just calls the
4527 queue-runner function directly.
4533 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4536 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4537 ignored. Though the use of -oP was forcing the writing of a pid file, it
4538 was always written to the default place.
4540 3. If the message "no IP address found for host xxxx" is generated during
4541 incoming verification, it is now followed by identification of the incoming
4542 connection (so you can more easily find what provoked it).
4544 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4546 5. Added some features to "harden" Exim a bit more against certain attacks:
4548 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4549 be put in Local/Makefile. This is like the never_users runtime option,
4550 but it cannot be overridden. The default setting is "root".
4552 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4553 prefix string with which any file named in a -C command line option
4556 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4557 is retained for -C and -D only if the caller of Exim is root. Without
4558 it, the exim user may also use -C and -D and retain privilege.
4560 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4561 command line option is disabled.
4563 6. Macro names set by the -D option must start with an upper case letter, just
4564 like macro names defined in the configuration file.
4566 7. Added "dereference=" facility to LDAP.
4568 8. Two instances of the typo "uknown" in the source files are fixed.
4570 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4571 the Configure-Makefile script screwed up while processing it.
4573 10. Incorporated PCRE 4.4.
4575 11. The SMTP synchronization check was not operating right at the start of an
4576 SMTP session. For example, it could not catch a HELO sent before the client
4577 waited for the greeting. There is now a check for outstanding input at the
4578 point when the greeting is written. Because of the duplex, asynchronous
4579 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4580 way, but not yet received, when the check is performed.
4582 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4583 on TCP/IP sockets, because this apparently causes some broken clients to
4586 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4587 unchanged) from the Cygwin maintainer.
4589 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4590 was supported instead of testing for mbx. Effectively a typo.
4592 15. The spa authenticator server code was not checking that the input it
4593 received was valid base64.
4595 16. The debug output line for the "set" modifier in ACLs was not showing the
4596 name of the variable that was being set.
4598 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4600 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4601 lookup on its own. Something else was needed to trigger the lookup. For
4602 example, a match in host_lookup or the need for a host name in a host list.
4603 Now, if $sender_host_name is referenced and the host name has not yet been
4604 looked up, a lookup is performed. If the lookup fails, the variable remains
4605 empty, and $host_lookup_failed is set to "1".
4607 19. Added "eqi" as a case-independent comparison operator.
4609 20. The saslauthd authentication condition could segfault if neither service
4610 nor realm was specified.
4612 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4613 error message that was logged was misleading, and incoming SMTP
4614 connections were dropped. The message is now more accurate, and temporary
4615 errors are given to SMTP connections.
4617 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4618 (except RSET) with a 421 error, until QUIT is received. However, it was
4619 failing to send a response to QUIT.
4621 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4622 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4623 ACL is now run after the helo_try_verify_hosts code.
4625 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4626 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4627 case-independent, so other case variants are also recognized). Apparently
4628 some systems use these upper case variants.
4630 25. If more than two messages were waiting for the same host, and a transport
4631 filter was specified for the transport, Exim sent two messages over the
4632 same TCP/IP connection, and then failed with "socket operation on non-
4633 socket" when it tried to send the third.
4635 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4637 27. The extern definition of crypt16() in expand.c was not being excluded when
4638 the OS had its own crypt16() function.
4640 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4641 as a preferred synonym for return_size_limit, both as an option and as an
4644 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4646 30. Changed the default configuration ACL to relax the local part checking rule
4647 for addresses that are not in any local domains. For these addresses,
4648 slashes and pipe symbols are allowed within local parts, but the sequence
4649 /../ is explicitly forbidden.
4651 31. SPA server authentication was not clearing the challenge buffer before
4654 32. log_message in a "warn" ACL statement was writing to the reject log as
4655 well as to the main log, which contradicts the documentation and doesn't
4656 seem right (because no rejection is happening). So I have stopped it.
4658 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4659 However, I am unable to do any testing of this.
4661 34. Fixed an infelicity in the appendfile transport. When checking directories
4662 for a mailbox, to see if any needed to be created, it was accidentally
4663 using path names with one or more superfluous leading slashes; tracing
4664 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4666 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4667 contents are added to the log line that is written for every discarded
4668 recipient. (Previously a log_message setting was ignored.)
4670 36. The ${quote: operator now quotes the string if it is empty.
4672 37. The install script runs exim in order to find its version number. If for
4673 some reason other than non-existence or emptiness, which it checks, it
4674 could not run './exim', it was installing it with an empty version number,
4675 i.e. as "exim-". This error state is now caught, and the installation is
4678 38. An argument was missing from the function that creates an error message
4679 when Exim fails to connect to the socket for saslauthd authentication.
4680 This could cause Exim to crash, or give a corrupted message.
4682 39. Added isip, isip4, and isip6 to ${if conditions.
4684 40. The ACL variables $acl_xx are now saved with the message, and can be
4685 accessed later in routers, transports, and filters.
4687 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4688 strings in the file are not string-expanded.
4690 42. If a MAIL command specified a SIZE value that was too large to fit into an
4691 int variable, the check against message_size_limit failed. Such values are
4692 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4693 day this will have to be increased, but I don't think I want to be around
4694 when emails are that large.
4701 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4702 iconv() is not standard in FreeBSD.
4704 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4705 IPv6 enabled. The observed symptom was a segmentation fault on return from
4706 the function os_common_find_running_interfaces() in src/os.c.
4708 3. In the check_special_case() function in daemon.c I had used "errno" as an
4709 argument name, which causes warnings on some systems. This was basically a
4710 typo, since it was named "eno" in the comments!
4712 4. The code that waits for the clock to tick (at a resolution of some fraction
4713 of a second) so as to ensure message-id uniqueness was always waiting for
4714 at least one whole tick, when it could have waited for less. [This is
4715 almost certainly not relevant at current processor speeds, where it is
4716 unlikely to ever wait at all. But we try to future-proof.]
4718 5. The function that sleeps for a time interval that includes fractions of a
4719 second contained a race. It did not block SIGALRM between setting the
4720 timer, and suspending (a couple of lines later). If the interval was short
4721 and the sigsuspend() was delayed until after it had expired, the suspension
4722 never ended. On busy systems this could lead to processes getting stuck for
4725 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4726 process, before it forks any delivery processes. The open lookup caching
4727 mechanism meant that the open file or database connection was passed into
4728 the delivery process. The problem was that delivery processes always tidy
4729 up cached lookup data. This could cause a problem for the next delivery
4730 process started by the queue runner, because the external queue runner
4731 process does not know about the closure. So the next delivery process
4732 still has data in the lookup cache. In the case of a file lookup, there was
4733 no problem because closing a file descriptor in a subprocess doesn't affect
4734 the parent. However, if the lookup was caching a connection to a database,
4735 the connection was closed, and the second delivery process was likely to
4736 see errors such as "PGSQL: query failed: server closed the connection
4737 unexpectedly". The problem has been fixed by closing all cached lookups
4738 in a queue runner before running a delivery process.
4740 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4741 seem to have the "const" qualifier which it has on other OS. I've
4744 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4745 *of the same type* (client or server) with the same public name that an
4746 error should be diagnosed.
4748 9. When Exim looked up a host name for an IP address, but failed to find the
4749 original IP address when looking up the host name (a safety check), it
4750 output the message "<ip address> does not match any IP for NULL", which was
4751 confusing, to say the least. The bug was that the host name should have
4752 appeared instead of "NULL".
4754 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4755 user that is built into the binary, and the -C or -D options is used, root
4756 privilege is dropped before the configuration file is read. In addition,
4757 logging is switched to stderr instead of the normal log files. If the
4758 configuration then re-defines the Exim user, the unprivileged environment
4759 is probably not what is expected, so Exim logs a panic warning message (but
4762 However, if deliver_drop_privilege is set, the unprivileged state may well
4763 be exactly what is intended, so the warning has been cut out in that case,
4764 and Exim is allowed to try to write to its normal log files.
4770 1. smtp_return_error_details was not giving details for temporary sender
4771 or receiver verification errors.
4773 2. Diagnose a configuration error if two authenticators have the same public
4776 3. Exim used not to create the message log file for a message until the first
4777 delivery attempt. This could be confusing when incoming messages were held
4778 for policy or load reasons. The message log file is now created at the time
4779 the message is received, and an initial "Received" line is written to it.
4781 4. The automatically generated man page for command line options had a minor
4782 bug that caused no ill effects; however, a more serious problem was that
4783 the procedure for building the man page automatically didn't always
4784 operate. Consequently, release 4.20 contains an out-of-date version. This
4785 shouldn't happen again.
4787 5. When building Exim with embedded Perl support, the script that builds the
4788 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4789 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4791 6. The freeze_tell option was not being used for messages that were frozen on
4792 arrival, either by an ACL or by local_scan().
4794 7. Added the smtp_incomplete_transaction log selector.
4796 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4797 was accepting AUTH without a new EHLO.
4799 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4800 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4803 10. Logging of TCP/IP connections (when configured) now happens in the main
4804 daemon process instead of the child process, so that the TCP/IP connection
4805 count is more accurate (but it can never be perfect).
4807 11. The use of "drop" in a nested ACL was not being handled correctly in the
4808 outer ACL. Now, if condition failure induced by the nested "drop" causes
4809 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4810 or "require"), the connection is dropped.
4812 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4813 that yield "discard" can now be used with an "accept" or a "discard" verb,
4814 but an error is generated for any others (because I can't see a useful way
4815 to define what should happen).
4817 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4818 are now processed in the same way as lines in the Exim configuration file.
4819 In particular, continuation lines are supported.
4821 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4823 15. Added -ti meaning -t -i.
4825 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4826 domains, and warn by logging if others are found.
4828 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4829 returned from GIFCONF when Exim is trying to find the list of interfaces on
4830 a host. The code in os.c has been modified to copy each ifreq to an aligned
4831 structure in all cases.
4833 Also, in some cases, the returned ifreq's were being copied to a 'struct
4834 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4835 means the last couple of bytes of an IPv6 address could be chopped if the
4836 ifreq contained only a normal sockaddr (14 bytes storage).
4838 18. Named domain lists were not supported in the hosts_treat_as_local option.
4839 An entry such as +xxxx was not recognized, and was treated as a literal
4842 19. Ensure that header lines added by a DATA ACL are included in the reject log
4843 if the ACL subsequently rejects the message.
4845 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4846 MD5 (which is deprecated).
4848 21. When testing a filter file using -bf, Exim was writing a message when it
4849 took the sender from a "From " line in the message, but it was not doing so
4850 when it took $return_path from a Return-Path: header line. It now does.
4852 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4853 with a valid header line field name (a series of printing characters
4854 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4856 23. Changed "disc" in the source to "disk" to conform to the documentation and
4857 the book and for uniformity.
4859 24. Ignore Sendmail's -Ooption=value command line item.
4861 25. When execve() failed while trying to run a command in a pipe transport,
4862 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4863 could be confused with a return value of 69 from the command itself. This
4864 has been changed to 127, the value the shell returns if it is asked to run
4865 a non-existent command. The wording for the related log line suggests a
4866 non-existent command as the problem.
4868 26. If received_header_text expands to an empty string, do not add a Received:
4869 header line to the message. (Well, it adds a token one on the spool, but
4870 marks it "old" so that it doesn't get used or transmitted.)
4872 27. Installed eximstats 1.28 (addition of -nt option).
4874 28. There was no check for failure on the call to getsockname() in the daemon
4875 code. This can fail if there is a shortage of resources on the system, with
4876 ENOMEM, for example. A temporary error is now given on failure.
4878 29. Contrary to the C standard, it seems that in some environments, the
4879 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4880 program. Exim now does this explicitly; it affects the formatting of
4881 timestamps using strftime().
4883 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4884 complaints. I've now initialized all the variables, to avoid this.
4886 32. Header lines added by a system filter were not being "seen" during
4887 transport-time rewrites.
4889 33. The info_callback() function passed to OpenSSL is set up with type void
4890 (*)(SSL *, int, int), as described somewhere. However, when calling the
4891 function (actually a macro) that sets it up, the type void(*)() is
4892 expected. I've put in a cast to prevent warnings from picky compilers.
4894 34. If a DNS black list lookup found a CNAME record, but there were no A
4895 records associated with the domain it pointed at, Exim crashed.
4897 35. If a DNS black list lookup returned more than one A record, Exim ignored
4898 all but the first. It now scans all returned addresses if a particular IP
4899 value is being sought. In this situation, the contents of the
4900 $dnslist_value variable are a list of all the addresses, separated by a
4903 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4904 to accept a host name and all its aliases if the forward lookup for any of
4905 them yielded the IP address of the incoming connection. Now it accepts only
4906 those names whose forward lookup yields the correct IP address. Any other
4907 names are discarded. This closes a loophole whereby a rogue DNS
4908 administrator could create reverse DNS records to break through a
4909 wildcarded host restriction in an ACL.
4911 37. If a user filter or a system filter that ran in a subprocess used any of
4912 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4913 the wrong values were passed to the pipe command ($thisaddress had the
4914 value of $0, $0 had the value of $1, etc). This bug was introduced by
4915 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4917 38. Improved the line breaking for long SMTP error messages from ACLs.
4918 Previously, if there was no break point between 40 and 75 characters, Exim
4919 left the rest of the message alone. Two changes have been made: (a) I've
4920 reduced the minimum length to 35 characters; (b) if it can't find a break
4921 point between 35 and 75 characters, it looks ahead and uses the first one
4922 that it finds. This may give the occasional overlong line, but at least the
4923 remaining text gets split now.
4925 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4926 file descriptors might be low, and that setting 1000 would always raise it.
4927 It turns out that in some environments, the limit is already over 1000 and
4928 that lowering it causes trouble. So now Exim takes care not to decrease it.
4930 40. When delivering a message, the value of $return_path is set to $sender_
4931 address at the start of routing (routers may change the value). By an
4932 oversight, this default was not being set up when an address was tested by
4933 -bt or -bv, which affected the outcome if any router or filter referred to
4936 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4937 log only when "message" or "log_message" are set. However, if one of the
4938 conditions was an address verification, or a call to a nested ACL, the
4939 messages generated by the underlying test were being passed through. This
4940 no longer happens. The underlying message is available in $acl_verify_
4941 message for both "message" and "log_message" expansions, so it can be
4942 passed through if needed.
4944 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4945 new expansion $bh_ to give the encoded byte string without charset
4946 translation. Translation happens only if iconv() is available; HAVE_ICONV
4947 indicates this at build time. HEADERS_CHARSET gives the charset to
4948 translate to; headers_charset can change it in the configuration, and
4949 "headers charset" can change it in an individual filter file.
4951 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4952 that creates RFC 2047 encoded "words" labels them as that charset instead
4953 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4954 expansion operator; (ii) when Exim creates a From: line for a local
4955 message; (iii) when a header line is rewritten to include a "phrase" part.
4957 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4958 buggy, causing it to skip the first lines of messages whose message ID
4959 ended in 'D'. This would not have bitten before Exim release 4.14, because
4960 message IDs were unlikely to end in 'D' before then. The effect was to have
4961 incorrect size information for certain domains.
4963 45. #include "config.h" was missing at the start of the crypt16.c module. This
4964 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4967 46. If there was a timeout during a "random" callout check, Exim treated it as
4968 a failure of the random address, and carried on sending RSET and the real
4969 address. If the delay was just some slowness somewhere, the response to the
4970 original RCPT would be taken as a response to RSET and so on, causing
4971 mayhem of various kinds.
4973 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4974 when I implemented it. It didn't allow for the fact that some option values
4975 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4976 the right test for positive values.
4978 48. Domain names in DNS records are case-independent. Exim always looks them up
4979 in lower case. Some resolvers return domain names in exactly the case they
4980 appear in the zone file, that is, they may contain uppercase letters. Not
4981 all resolvers do this - some return always lower case. Exim was treating a
4982 change of case by a resolver as a change of domain, similar to a widening
4983 of a domain abbreviation. This triggered its re-routing code and so it was
4984 trying to route what was effectively the same domain again. This normally
4985 caused routing to fail (because the router wouldn't handle the domain
4986 twice). Now Exim checks for this case specially, and just changes the
4987 casing of the domain that it ultimately uses when it transmits the message
4990 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4993 50. If a filter generated a file delivery with a non-absolute name (possible if
4994 no home directory exists for the router), the forbid_file option was not
4997 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4998 the existing equality matching.
5000 52. Exim was using ints instead of ino_t variables in some places where it was
5001 dealing with inode numbers.
5003 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5004 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5005 called TMPDIR, and if it finds it is different, it changes its value.
5007 54. The smtp_printf() function is now made available to local_scan() so
5008 additional output lines can be written before returning. There is also an
5009 smtp_fflush() function to enable the detection of a dropped connection.
5010 The variables smtp_input and smtp_batched_input are exported to
5013 55. Changed the default runtime configuration: the message "Unknown user"
5014 has been removed from the ACL, and instead placed on the localuser router,
5015 using the cannot_route_message feature. This means that any verification
5016 failures that generate their own messages won't get overridden. Similarly,
5017 the "Unrouteable address" message that was in the ACL for unverifiable
5018 relay addresses has also been removed.
5020 56. Added hosts_avoid_esmtp to the smtp transport.
5022 57. The exicyclog script was not checking for the esoteric option
5023 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5024 will work only if exicyclog is run under the appropriate euid.
5026 58. Following a discussion on the list, the rules by which Exim recognises line
5027 endings on incoming messages have been changed. The -dropcr and drop_cr
5028 options are now no-ops, retained only for backwards compatibility. The
5029 following line terminators are recognized: LF CRLF CR. However, special
5030 processing applies to CR:
5032 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5033 nor a local message in the state where . is a terminator.
5035 (ii) If a bare CR is encountered in a header line, an extra space is added
5036 after the line terminator so as not to end the header. The reasoning
5037 behind this is that bare CRs in header lines are most likely either
5038 to be mistakes, or people trying to play silly games.
5040 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5041 was being incorrectly given as 18 bytes larger than it should have been.
5042 This is a VOB (very old bug).
5044 60. This may never have affected anything current, but just in case it has:
5045 When the local host is found other than at the start of a list of hosts,
5046 the local host, those with the same MX, and any that follow, are discarded.
5047 When the list in question was part of a longer list of hosts, the following
5048 hosts (not currently being processed) were also being discarded. This no
5049 longer happens. I'm not sure if this situation could ever has previously
5052 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5055 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5056 header. This is something that is recommended in a new Internet Draft, and
5057 is something that is documented as being done by Sendmail. There are two
5058 possible values. For messages generated by the autoreply transport, Exim
5061 Auto-Submitted: auto-replied
5063 whereas for all other generated messages (e.g. bounces) it adds
5065 Auto-Submitted: auto-generated
5067 63. The "personal" condition in filters now includes a test for the
5068 Auto-Submitted: header. If it contains the string "auto-" the message it
5069 not considered personal.
5071 64. Added rcpt_include_affixes as a generic transport option.
5073 65. Added queue_only_override (default true).
5075 66. Added the syslog_duplication option.
5077 67. If what should have been the first header line of a message consisted of
5078 a space followed by a colon, Exim was mis-interpreting it as a header line.
5079 It isn't of course - it is syntactically invalid and should therefore be
5080 treated as the start of the message body. The misbehaviour could have
5081 caused a number of strange effects, including loss of data in subsequent
5082 header lines, and spool format errors.
5084 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5085 client host had authenticated. This control can now be exercised by an ACL
5086 for more flexibility.
5088 69. By default, callouts do not happen when testing with -bh. There is now a
5089 variant, -bhc, which does actually run the callout code, including
5090 consulting and updating the callout cache.
5092 70. Added support for saslauthd authentication, courtesy of Alexander
5095 71. If statvfs() failed on the spool or log directories while checking their
5096 size for availability, Exim confusingly gave the error "space shortage".
5097 Furthermore, in debugging mode it crashed with a floating point exception.
5098 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5099 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5100 serious problem, Exim now writes to the main and panic logs when this
5101 happens, with details of the failure. It then refuses to accept the
5102 incoming message, giving the message "spool directory problem" or "log
5103 directory problem" with a 421 code for SMTP messages.
5105 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5106 0, 1, and 2 exist, because some OS complain for execs without them (see
5107 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5108 descriptors. However, the code omitted to check that the open succeeded,
5109 causing mysterious errors if for some reason the permissions on /dev/null
5110 got screwed. Now Exim writes a message to the main and panic logs, and
5111 bombs out if it can't open /dev/null.
5113 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5114 interact so that it is all more flexible. It is supposed to remain
5115 backwards compatible. Also added extra_local_interfaces.
5117 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5118 to bomb out with an assertion failure - to the client this appears as a
5119 connection drop. This problem occurs in the part of the code that was taken
5120 from the Samba project. Fortunately, the assertion is in a very simple
5121 function, so I have fixed this by reproducing the function inline in the
5122 one place where it is called, and arranging for authentication to fail
5123 instead of killing the process with assert().
5125 75. The SPA client code was not working when the server requested OEM rather
5126 than Unicode encoding.
5128 76. Added code to make require_files with a specific uid setting more usable in
5129 the case where statting the file as root fails - usually a non-root-mounted
5130 NFS file system. When this happens and the failure is EACCES, Exim now
5131 forks a subprocess and does the per-uid checking as the relevant uid.
5133 77. Added process_log_path.
5135 78. If log_file_path was not explicitly set, a setting of check_log_space or
5136 check_log_inodes was ignored.
5138 79. If a space check for the spool or log partitions fails, the incident is now
5139 logged. Of course, in the latter case the data may get lost...
5141 80. Added the %p formatting code to string_format() so that it can be used to
5142 print addresses in debug_print(). Adjusted all the address printing in the
5143 debugging in store.c to use %p rather than %d.
5145 81. There was a concern that a line of code in smtp_in.c could overflow a
5146 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5147 initially expressed, the concern was not well-founded, because trailing
5148 spaces are removed early. However, if the trailing spaces were followed by
5149 a NULL, they did not get removed, so the overflow was possible. Two fixes
5152 (a) I re-wrote the offending code in a cleaner fashion.
5153 (b) If an incoming SMTP command contains a NULL character, it is rejected
5156 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5157 runs initgroups(), so that if the Exim user is in any additional groups,
5158 they will be used during message reception.
5164 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob