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
86 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
87 option (effectively making it always true).
89 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
90 files to be used while preserving root privileges.
96 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
97 $data_path, and $header_path variables; fixed documentation bugs and
100 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
101 exipick to access non-standard spools, including the "frozen" queue
104 NM/01 Bugzilla 965: Support mysql stored procedures.
105 Patch from Alain Williams
107 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
109 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
110 Patch from Andreas Metzler
112 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
113 Patch from Kirill Miazine
115 NM/05 Bugzilla 671: Added umask to procmail example.
117 JJ/03 installed exipick 20100323.0, fixing doc bug
119 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
120 directory. Notification and patch from Dan Rosenberg.
122 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
124 TK/02 Improve log output when DKIM signing operation fails.
126 MH/01 Treat the transport option dkim_domain as a colon separated
127 list, not as a single string, and sign the message with each element,
128 omitting multiple occurences of the same signer.
130 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
131 Bugzilla 985, 986. Patch by Simon Arlott
133 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
134 Patch by Simon Arlott
136 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
137 MBX locking. Notification from Dan Rosenberg.
143 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
145 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
147 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
149 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
151 NM/04 Bugzilla 915: Flag broken perl installation during build.
157 TK/01 Added patch by Johannes Berg that expands the main option
158 "spamd_address" if it starts with a dollar sign.
160 TK/02 Write list of recipients to X-Envelope-Sender header when building
161 the mbox-format spool file for content scanning (suggested by Jakob
164 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
165 (http://www.dcc-servers.net/) support via dccifd. Activated by
166 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
168 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
169 by Mark Daniel Reidel <mr@df.eu>.
171 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
172 When building exim an external PCRE library is now needed -
173 PCRE is a system library on the majority of modern systems.
174 See entry on PCRE_LIBS in EDITME file.
176 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
177 conversation. Added nologin parameter to request.
178 Patch contributed by Kirill Miazine.
180 TF/01 Do not log submission mode rewrites if they do not change the address.
182 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
184 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
185 log files in place. Contributed by Roberto Lima.
187 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
189 TF/03 Bugzilla 615: When checking the local_parts router precondition
190 after a local_part_suffix or local_part_prefix option, Exim now
191 does not use the address's named list lookup cache, since this
192 contains cached lookups for the whole local part.
194 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
195 Robert Millan. Documentation is in experimental-spec.txt.
197 TF/04 Bugzilla 668: Fix parallel build (make -j).
199 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
201 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
202 Patch provided by Jan Srzednicki.
204 TF/05 Leading white space used to be stripped from $spam_report which
205 wrecked the formatting. Now it is preserved.
207 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
208 that they are available at delivery time.
210 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
212 TF/08 TLS error reporting now respects the incoming_interface and
213 incoming_port log selectors.
215 TF/09 Produce a more useful error message if an SMTP transport's hosts
216 setting expands to an empty string.
218 NM/06 Bugzilla 744: EXPN did not work under TLS.
219 Patch provided by Phil Pennock.
221 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
222 Patch provided by Richard Godbee.
224 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
225 acl_smtp_notquit, added index entry.
227 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
228 Patch provided by Eugene Bujak.
230 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
231 accept(). Patch provided by Maxim Dounin.
233 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
234 Patch provided by Phil Pennock.
236 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
238 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
239 Patch provided by Brad "anomie" Jorsch.
241 NM/14 Bugzilla 622: Added timeout setting to transport filter.
242 Patch provided by Dean Brooks.
244 TK/05 Add native DKIM support (does not depend on external libraries).
246 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
247 Patch provided by Graeme Fowler.
249 NM/16 Bugzilla 851: Documentation example syntax fix.
251 NM/17 Changed NOTICE file to remove references to embedded PCRE.
253 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
256 NM/19 Bugzilla 745: TLS version reporting.
257 Patch provided by Phil Pennock.
259 NM/20 Bugzilla 167: bool: condition support.
260 Patch provided by Phil Pennock.
262 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
263 clients. Patch provided by Phil Pennock.
265 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
266 Patch provided by Brad "anomie" Jorsch.
268 NM/23 Bugzilla 687: Fix misparses in eximstats.
269 Patch provided by Heiko Schlittermann.
271 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
272 Patch provided by Heiko Schlittermann.
274 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
275 plus update to original patch.
277 NM/26 Bugzilla 799: Documentation correction for ratelimit.
279 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
280 Patch provided by David Brownlee.
282 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
284 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
286 NM/30 Bugzilla 888: TLS documentation bugfixes.
288 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
290 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
291 Unlike the original bugzilla I have changed all shell scripts in src tree.
293 NM/33 Bugzilla 898: Transport filter timeout fix.
294 Patch by Todd Rinaldo.
296 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
297 Patch by Serge Demonchaux.
299 NM/35 Bugzilla 39: Base64 decode bug fixes.
300 Patch by Jakob Hirsch.
302 NM/36 Bugzilla 909: Correct connect() call in dcc code.
304 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
306 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
308 NM/39 Bugzilla 911: Fixed MakeLinks build script.
314 TK/01 Add preliminary DKIM support. Currently requires a forked version of
315 ALT-N's libdkim that I have put here:
316 http://duncanthrax.net/exim-experimental/
318 Note to Michael Haardt: I had to rename some vars in sieve.c. They
319 were called 'true' and it seems that C99 defines that as a reserved
320 keyword to be used with 'bool' variable types. That means you could
321 not include C99-style headers which use bools without triggering
322 build errors in sieve.c.
324 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
325 as mailq or other aliases. Changed the --help handling significantly
326 to do whats expected. exim_usage() emits usage/help information.
328 SC/01 Added the -bylocaldomain option to eximstats.
330 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
332 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
334 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
340 PH/01 Another patch from the Sieve maintainer.
342 PH/02 When an IPv6 address is converted to a string for single-key lookup
343 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
344 dots are used instead of colons so that keys in lsearch files need not
345 contain colons. This was done some time before quoting was made available
346 in lsearch files. However, iplsearch files do require colons in IPv6 keys
347 (notated using the quote facility) so as to distinguish them from IPv4
348 keys. This meant that lookups for IP addresses in host lists did not work
349 for iplsearch lookups.
351 This has been fixed by arranging for IPv6 addresses to be expressed with
352 colons if the lookup type is iplsearch. This is not incompatible, because
353 previously such lookups could never work.
355 The situation is now rather anomolous, since one *can* have colons in
356 ordinary lsearch keys. However, making the change in all cases is
357 incompatible and would probably break a number of configurations.
359 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
362 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
363 conversion specification without a maximum field width, thereby enabling
364 a rogue spamd server to cause a buffer overflow. While nobody in their
365 right mind would setup Exim to query an untrusted spamd server, an
366 attacker that gains access to a server running spamd could potentially
367 exploit this vulnerability to run arbitrary code as the Exim user.
369 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
370 $primary_hostname instead of what libspf2 thinks the hosts name is.
372 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
373 a directory entry by the name of the lookup key. Previously, if a
374 symlink pointed to a non-existing file or a file in a directory that
375 Exim lacked permissions to read, a lookup for a key matching that
376 symlink would fail. Now it is enough that a matching directory entry
377 exists, symlink or not. (Bugzilla 503.)
379 PH/03 The body_linecount and body_zerocount variables are now exported in the
382 PH/04 Added the $dnslist_matched variable.
384 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
385 This means they are set thereafter only if the connection becomes
388 PH/06 Added the client_condition to authenticators so that some can be skipped
389 by clients under certain conditions.
391 PH/07 The error message for a badly-placed control=no_multiline_responses left
392 "_responses" off the end of the name.
394 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
396 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
397 (without spaces) instead of just copying the configuration text.
399 PH/10 Added the /noupdate option to the ratelimit ACL condition.
401 PH/11 Added $max_received_linelength.
403 PH/12 Added +ignore_defer and +include_defer to host lists.
405 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
406 way in which PCRE > 7.0 is built.
408 PH/14 Implemented queue_only_load_latch.
410 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
411 MAIL command. The effect was to mangle the value on 64-bit systems.
413 PH/16 Another patch from the Sieve maintainer.
415 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
417 PH/18 If a system quota error occurred while trying to create the file for
418 a maildir delivery, the message "Mailbox is full" was not appended to the
419 bounce if the delivery eventually timed out. Change 4.67/27 below applied
420 only to a quota excession during the actual writing of the file.
422 PH/19 It seems that peer DN values may contain newlines (and other non-printing
423 characters?) which causes problems in log lines. The DN values are now
424 passed through string_printing() before being added to log lines.
426 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
427 and InterBase are left for another time.)
429 PH/21 Added message_body_newlines option.
431 PH/22 Guard against possible overflow in moan_check_errorcopy().
433 PH/23 POSIX allows open() to be a macro; guard against that.
435 PH/24 If the recipient of an error message contained an @ in the local part
436 (suitably quoted, of course), incorrect values were put in $domain and
437 $local_part during the evaluation of errors_copy.
443 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
444 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
447 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
448 issue a MAIL command.
450 PH/02 In an ACL statement such as
452 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
454 if a client was not listed at all, or was listed with a value other than
455 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
456 the condition was not true (as it should be), so access was not denied.
457 The bug was that the ! inversion was incorrectly passed on to the second
458 item. This has been fixed.
460 PH/03 Added additional dnslists conditions == and =& which are different from
461 = and & when the dns lookup returns more than one IP address.
463 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
464 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
466 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
467 FSYNC, which compiles an option called disable_fsync that allows for
468 bypassing fsync(). The documentation is heavily laced with warnings.
470 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
472 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
473 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
474 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
475 including adding "make clean"; (3) Added -fPIC when compiling the test
476 dynamically loaded module, to get rid of a warning.
478 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
479 message fails, move_frozen_messages = true and ignore_bounce_errors_after
480 = 0s. The bug is otherwise harmless.
482 PH/07 There was a bug in the dovecot authenticator such that the value of
483 $auth1 could be overwritten, and so not correctly preserved, after a
484 successful authentication. This usually meant that the value preserved by
485 the server_setid option was incorrect.
487 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
489 PH/09 Installed PCRE release 7.0.
491 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
492 run for batched SMTP input. It is now run at the start of every message
493 in the batch. While fixing this I discovered that the process information
494 (output by running exiwhat) was not always getting set for -bs and -bS
495 input. This is fixed, and it now also says "batched" for BSMTP.
497 PH/11 Added control=no_pipelining.
499 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
500 patch, slightly modified), and move the expansion of helo_data till after
501 the connection is made in the smtp transport (so it can use these
504 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
506 PH/14 Added log_selector = +pid.
508 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
510 PH/16 Add ${if forany and ${if forall.
512 PH/17 Added dsn_from option to vary the From: line in DSNs.
514 PH/18 Flush SMTP output before performing a callout, unless control =
515 no_callout_flush is set.
517 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
518 was true (the default) a successful delivery failed to delete the retry
519 item, thus causing premature timeout of the address. The bug is now
522 PH/20 Added hosts_avoid_pipelining to the smtp transport.
524 PH/21 Long custom messages for fakedefer and fakereject are now split up
525 into multiline reponses in the same way that messages for "deny" and
526 other ACL rejections are.
528 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
529 with slight modification.
531 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
532 draft, changing the syntax and factoring some duplicate code".
534 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
535 for deliveries of the second and subsequent messages over the same SMTP
538 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
539 ${reduce, with only minor "tidies".
541 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
543 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
544 expansion side effects.
546 PH/27 When a message times out after an over-quota error from an Exim-imposed
547 quota, the bounce message says "mailbox is full". This message was not
548 being given when it was a system quota that was exceeded. It now should
551 MH/03 Made $recipients available in local_scan(). local_scan() already has
552 better access to the recipient list through recipients_list[], but
553 $recipients can be useful in postmaster-provided expansion strings.
555 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
556 in the case of a MAIL command with additional options following the
557 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
558 were accidentally chopped off.
560 PH/29 SMTP synchronization checks are implemented when a command is read -
561 there is a check that no more input is waiting when there shouldn't be
562 any. However, for some commands, a delay in an ACL can mean that it is
563 some time before the response is written. In this time, more input might
564 arrive, invalidly. So now there are extra checks after an ACL has run for
565 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
566 pipelining has not been advertised.
568 PH/30 MH's patch to allow iscntrl() characters to be list separators.
570 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
571 returned in the SMTP response when smtp_return_error_details was false.
574 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
575 instead of the C I/O that was originally supplied, because problems were
578 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
579 Exim which did not show up earlier: it was assuming that a call to
580 SSL_CTX_set_info_callback() might give an error value. In fact, there is
581 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
582 was a macro that became an assignment, so it seemed to work. This has
583 changed to a proper function call with a void return, hence the compile
584 error. Exim's code has been fixed.
586 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
589 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
591 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
593 PH/37 If a message is not accepted after it has had an id assigned (e.g.
594 because it turns out to be too big or there is a timeout) there is no
595 "Completed" line in the log. When some messages of this type were
596 selected by exigrep, they were listed as "not completed". Others were
597 picked up by some special patterns. I have improved the selection
598 criteria to be more general.
600 PH/38 The host_find_failed option in the manualroute router can now be set
601 to "ignore", to completely ignore a host whose IP address cannot be
602 found. If all hosts are ignored, the behaviour is controlled by the new
603 host_all_ignored option.
605 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
606 homing or because of multiple MX records with /mx) generated more than
607 one IP address, and the following item turned out to be the local host,
608 all the secondary addresses of the first item were incorrectly removed
609 from the list, along with the local host and any following hosts (which
610 is what is supposed to happen).
612 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
613 whoever called Exim into the -H file. In the case of the daemon it was
614 behaving confusingly. When first started, it used values for whoever
615 started the daemon, but after a SIGHUP it used the Exim user (because it
616 calls itself on a restart). I have changed the code so that it now always
619 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
620 message are rejected with the same error (e.g. no authentication or bad
621 sender address), and a DATA command is nevertheless sent (as can happen
622 with PIPELINING or a stupid MUA), the error message that was given to the
623 RCPT commands is included in the rejection of the DATA command. This is
624 intended to be helpful for MUAs that show only the final error to their
627 PH/42 Another patch from the Sieve maintainer.
629 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
630 Eximstats - Fixed some broken HTML links and added missing column headers
632 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
633 columns for Rejects, Temp Rejects, Ham, and Spam rows.
635 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
637 PH/43 Yet another patch from the Sieve maintainer.
639 PH/44 I found a way to check for a TCP/IP connection going away before sending
640 the response to the final '.' that terminates a message, but only in the
641 case where the client has not sent further data following the '.'
642 (unfortunately, this is allowed). However, in many cases there won't be
643 any further data because there won't be any more messages to send. A call
644 to select() can be used: if it shows that the input is "ready", there is
645 either input waiting, or the socket has been closed. An attempt to read
646 the next input character can distinguish the two cases. Previously, Exim
647 would have sent an OK response which the client would never have see.
648 This could lead to message repetition. This fix should cure that, at
649 least in a lot of common cases.
651 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
652 advertised in response to EHLO.
658 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
659 fixed by 4.65/MH/01 (is this a record?) are fixed:
661 (i) An empty string was always treated as zero by the numeric comparison
662 operators. This behaviour has been restored.
664 (ii) It is documented that the numeric comparison operators always treat
665 their arguments as decimal numbers. This was broken in that numbers
666 starting with 0 were being interpreted as octal.
668 While fixing these problems I realized that there was another issue that
669 hadn't been noticed. Values of message_size_limit (both the global option
670 and the transport option) were treated as octal if they started with 0.
671 The documentation was vague. These values are now always treated as
672 decimal, and I will make that clear in the documentation.
678 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
679 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
682 MH/01 Don't check that the operands of numeric comparison operators are
683 integers when their expansion is in "skipping" mode (fixes bug
684 introduced by 4.64-PH/07).
686 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
687 child addresses, Exim now panics and dies. Previously, because the count
688 is held in a short int, deliveries were likely to be lost. As such a
689 large number of recipients for a single message is ridiculous
690 (performance will be very, very poor), I have chosen to impose a limit
691 rather than extend the field.
697 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
698 leftover -K file (the existence of which was triggered by #402).
699 While we were at it, introduced process PID as part of the -K
700 filename. This should rule out race conditions when creating
703 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
704 processing considerably. Previous code took too long for large mails,
705 triggering a timeout which in turn triggers #401.
707 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
708 in the DK code in transports.c. sendfile() is not really portable,
709 hence the _LINUX specificness.
711 TF/01 In the add_headers option to the mail command in an Exim filter,
712 there was a bug that Exim would claim a syntax error in any
713 header after the first one which had an odd number of characters
716 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
717 callout verification, Exim cached a "reject" for the entire domain. This
718 is correct for most verifications, but it is not correct for a recipient
719 verification with use_sender or use_postmaster set, because in that case
720 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
721 case of MAIL FROM:<> rejection from other early rejections (e.g.
722 rejection of HELO). When verifying a recipient using a non-null MAIL
723 address, the cache is ignored if it shows MAIL FROM:<> rejection.
724 Whatever the result of the callout, the value of the domain cache is
725 left unchanged (for any other kind of callout, getting as far as trying
726 RCPT means that the domain itself is ok).
728 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
731 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
732 manifest itself as EPIPE rather than ECONNECT. When tidying away a
733 session, the daemon ignores ECONNECT errors and logs others; it now
734 ignores EPIPE as well.
736 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
737 (quoted-printable decoding).
739 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
740 later the small subsequent patch to fix an introduced bug.
742 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
744 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
746 PH/08 An error is now given if message_size_limit is specified negative.
748 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
749 to be given (somewhat) arbitrary names.
751 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
754 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
755 miscellaneous code fixes
757 PH/10 Added the log_reject_target ACL modifier to specify where to log
760 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
761 hostname. This is wrong, because it relates to the incoming message (and
762 probably the interface on which it is arriving) and not to the outgoing
763 callout (which could be using a different interface). This has been
764 changed to use the value of the helo_data option from the smtp transport
765 instead - this is what is used when a message is actually being sent. If
766 there is no remote transport (possible with a router that sets up host
767 addresses), $smtp_active_hostname is used.
769 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
770 tweaks were necessary in order to get it to work (see also 21 below):
771 (a) The code assumed that strncpy() returns a negative number on buffer
772 overflow, which isn't the case. Replaced with Exim's string_format()
774 (b) There were several signed/unsigned issues. I just did the minimum
775 hacking in of casts. There is scope for a larger refactoring.
776 (c) The code used strcasecmp() which is not a standard C function.
777 Replaced with Exim's strcmpic() function.
778 (d) The code set only $1; it now sets $auth1 as well.
779 (e) A simple test gave the error "authentication client didn't specify
780 service in request". It would seem that Dovecot has changed its
781 interface. Fortunately there's a specification; I followed it and
782 changed what the client sends and it appears to be working now.
784 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
787 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
788 address A is aliased to B and C, where B exists and C does not. Without
789 -v the output is "A verified" because verification stops after a
790 successful redirection if more than one address is generated. However,
791 with -v the child addresses are also verified. Exim was outputting "A
792 failed to verify" and then showing the successful verification for C,
793 with its parentage. It now outputs "B failed to verify", showing B's
794 parentage before showing the successful verification of C.
796 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
797 look up a TXT record in a specific list after matching in a combined
800 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
801 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
802 they consult the DNS. I had assumed they would set it the way they
803 wanted; and indeed my experiments on Linux seem to show that in some
804 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
805 To be on the safe side, however, I have now made the interface to
806 host_find_byname() similar to host_find_bydns(), with an argument
807 containing the DNS resolver options. The host_find_byname() function now
808 sets these options at its start, just as host_find_bydns() does. The smtp
809 transport options dns_qualify_single and dns_search_parents are passed to
810 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
811 of host_find_byname() use the default settings of RES_DEFNAMES
812 (qualify_single) but not RES_DNSRCH (search_parents).
814 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
815 spool_read_header() do less string testing, by means of a preliminary
816 switch on the second character of optional "-foo" lines. (This is
817 overdue, caused by the large number of possibilities that now exist.
818 Originally there were few.) While I was there, I also converted the
819 str(n)cmp tests so they don't re-test the leading "-" and the first
820 character, in the hope this might squeeze out yet more improvement.
822 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
823 flag allowing group syntax was set by the header_syntax check but not
824 turned off, possible causing trouble later; (2) The flag was not being
825 set at all for the header_verify test, causing "group"-style headers to
826 be rejected. I have now set it in this case, and also caused header_
827 verify to ignore an empty address taken from a group. While doing this, I
828 came across some other cases where the code for allowing group syntax
829 while scanning a header line wasn't quite right (mostly, not resetting
830 the flag correctly in the right place). These bugs could have caused
831 trouble for malformed header lines. I hope it is now all correct.
833 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
834 with the "reply" argument non-NULL. The code, however (which originally
835 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
836 but it didn't always do it. This confused somebody who was copying the
837 code for some other use. I have removed all the tests.
839 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
840 feature that was used to support insecure browsers during the U.S. crypto
841 embargo. It requires special client support, and Exim is probably the
842 only MTA that supported it -- and would never use it because real RSA is
843 always available. This code has been removed, because it had the bad
844 effect of slowing Exim down by computing (never used) parameters for the
845 RSA_EXPORT functionality.
847 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
848 authenticator to fail if there's a tab character in the incoming data
849 (there should never be unless someone is messing about, as it's supposed
850 to be base64-encoded). Also added, on Timo's advice, the "secured" option
851 if the connection is using TLS or if the remote IP is the same as the
852 local IP, and the "valid-client-cert option" if a client certificate has
855 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
856 authenticators. This can be used for authorization after authentication
857 succeeds. (In the case of plaintext, it servers for both authentication
860 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
861 if any retry times were supplied.
863 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
864 connection, where there is no client IP address to check. In this
865 situation, the verify now always succeeds.
867 PH/25 Applied John Jetmore's -Mset patch.
869 PH/26 Added -bem to be like -Mset, but loading a message from a file.
871 PH/27 In a string expansion for a processed (not raw) header when multiple
872 headers of the same name were present, leading whitespace was being
873 removed from all of them, but trailing whitespace was being removed only
874 from the last one. Now trailing whitespace is removed from each header
875 before concatenation. Completely empty headers in a concatenation (as
878 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
879 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
881 PH/29 [Removed. This was a change that I later backed out, and forgot to
882 correct the ChangeLog entry (that I had efficiently created) before
883 committing the later change.]
885 PH/30 Exim was sometimes attempting to deliver messages that had suffered
886 address errors (4xx response to RCPT) over the same connection as other
887 messages routed to the same hosts. Such deliveries are always "forced",
888 so retry times are not inspected. This resulted in far too many retries
889 for the affected addresses. The effect occurred only when there were more
890 hosts than the hosts_max_try setting in the smtp transport when it had
891 the 4xx errors. Those hosts that it had tried were not added to the list
892 of hosts for which the message was waiting, so if all were tried, there
893 was no problem. Two fixes have been applied:
895 (i) If there are any address or message errors in an SMTP delivery, none
896 of the hosts (tried or untried) are now added to the list of hosts
897 for which the message is waiting, so the message should not be a
898 candidate for sending over the same connection that was used for a
899 successful delivery of some other message. This seems entirely
900 reasonable: after all the message is NOT "waiting for some host".
901 This is so "obvious" that I'm not sure why it wasn't done
902 previously. Hope I haven't missed anything, but it can't do any
903 harm, as the worst effect is to miss an optimization.
905 (ii) If, despite (i), such a delivery is accidentally attempted, the
906 routing retry time is respected, so at least it doesn't keep
907 hammering the server.
909 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
910 in ${readsocket because some servers need this prod.
912 PH/32 Added some extra debug output when updating a wait-xxx database.
914 PH/33 The hint "could be header name not terminated by colon", which has been
915 given for certain expansion errors for a long time, was not being given
916 for the ${if def:h_colon_omitted{... case.
918 PH/34 The spec says: "With one important exception, whenever a domain list is
919 being scanned, $domain contains the subject domain." There was at least
920 one case where this was not true.
922 PH/35 The error "getsockname() failed: connection reset by peer" was being
923 written to the panic log as well as the main log, but it isn't really
924 panic-worthy as it just means the connection died rather early on. I have
925 removed the panic log writing for the ECONNRESET error when getsockname()
928 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
929 runs only) independently of the message's sender address. This meant
930 that, if the 4xx error was in fact related to the sender, a different
931 message to the same recipient with a different sender could confuse
932 things. In particualar, this can happen when sending to a greylisting
933 server, but other circumstances could also provoke similar problems.
934 I have changed the default so that the retry time for these errors is now
935 based a combination of the sender and recipient addresses. This change
936 can be overridden by setting address_retry_include_sender=false in the
939 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
940 remote server are returned as part of bounce messages. This was not
941 happening for LMTP over a pipe (the lmtp transport), but now it is the
942 same for both kinds of LMTP.
944 PH/38 Despite being documented as not happening, Exim was rewriting addresses
945 in header lines that were in fact CNAMEs. This is no longer the case.
947 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
948 and queue runs started by the daemon processed all messages. This has
949 been fixed so that -R and -S can now usefully be given with -q<time>.
951 PH/40 Import PCRE release 6.7 (fixes some bugs).
953 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
955 PH/42 Give an error if -q is specified more than once.
957 PH/43 Renamed the variables $interface_address and $interface_port as
958 $received_ip_address and $received_port, to make it clear that these
959 values apply to message reception, and not to the outgoing interface when
960 a message is delivered. (The old names remain recognized, of course.)
962 PH/44 There was no timeout on the connect() call when using a Unix domain
963 socket in the ${readsocket expansion. There now is.
965 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
966 be meaningful with "accept".
968 SC/01 Eximstats V1.43
969 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
971 SC/02 Eximstats V1.44
972 Use a glob alias rather than an array ref in the generated
973 parser. This improves both readability and performance.
975 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
976 Collect SpamAssassin and rejection statistics.
977 Don't display local sender or destination tables unless
978 there is data to show.
979 Added average volumes into the top table text output.
981 SC/04 Eximstats V1.46
982 Collect data on the number of addresses (recipients)
983 as well as the number of messages.
985 SC/05 Eximstats V1.47
986 Added 'Message too big' to the list of mail rejection
987 reasons (thanks to Marco Gaiarin).
989 SC/06 Eximstats V1.48
990 Mainlog lines which have GMT offsets and are too short to
991 have a flag are now skipped.
993 SC/07 Eximstats V1.49 (Alain Williams)
994 Added the -emptyok flag.
996 SC/08 Eximstats V1.50
997 Fixes for obtaining the IP address from reject messages.
999 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1000 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1001 whitesspace changes from 4.64-PH/27
1003 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1006 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1007 are found, allow negative numbers in numeric criteria)
1009 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1011 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1012 to match changes made in 4.64-PH/43
1014 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1016 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1017 "domain" is actually an IP address, and force a failure. This locks out
1018 those revolvers/nameservers that support "A-for-A" lookups, in
1019 contravention of the specifications.
1021 PH/48 When a host name was looked up from an IP address, and the subsequent
1022 forward lookup of the name timed out, the host name was left in
1023 $sender_host_name, contrary to the specification.
1025 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1026 restricted to single-key lookups, Exim was not diagnosing an error if
1027 * or *@ was used with a query-style lookup.
1029 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1031 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1032 long ago, but noone interested enough thought of it. Let's just say that
1033 the "1.1" means that there are some new functions that weren't there at
1034 some point in the past.
1036 PH/51 Error processing for expansion failure of helo_data from an smtp
1037 transport during callout processing was broken.
1039 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1040 tested/used via the -bh/-bhc/-bs options.
1042 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1043 bug, fixed in subsequent PCRE releases).
1045 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1046 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1048 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1054 SC/01 Use a glob alias rather than an array ref in eximstats generated
1055 parser. This improves both readability and performance.
1057 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1058 Don't display local sender or destination tables in eximstats unless
1059 there is data to show.
1060 Added average volumes into the eximstats top table text output.
1062 SC/03 Collect data on the number of addresses (recipients) as well
1063 as the number of messages in eximstats.
1065 TF/01 Correct an error in the documentation for the redirect router. Exim
1066 does (usually) call initgroups() when daemonizing.
1068 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1069 with consistent privilege compared to when running as a daemon.
1071 TF/03 Note in the spec that $authenticated_id is not set for local
1072 submissions from trusted users.
1074 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1075 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1077 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1078 by adding some example configuration directives to the default
1079 configuration file. A little bit of work is required to uncomment the
1080 directives and define how usernames and passwords are checked, but
1081 there is now a framework to start from.
1083 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1084 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1085 without this. I don't know how relevant this is to other LDAP libraries.
1087 PH/02 Add the verb name to the "unknown ACL verb" error.
1089 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1091 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1093 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1094 directory not expanded when it should be if an expanded home directory
1095 was set for the address (which is overridden by the transport).
1097 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1100 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1101 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1102 because it is too late at that time, and has no effect.
1104 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1105 security issue with \' (bugzilla #107). I could not use the
1106 PQescapeStringConn() function, because it needs a PGconn value as one of
1109 PH/08 When testing addresses using -bt, indicate those final addresses that
1110 are duplicates that would not cause an additional delivery. At least one
1111 person was confused, thinking that -bt output corresponded to deliveries.
1112 (Suppressing duplicates isn't a good idea as you lose the information
1113 about possibly different redirections that led to the duplicates.)
1115 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1116 systems where poll() doesn't work, in particular OS X.
1118 PH/10 Added more information to debugging output for retry time not reached.
1120 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1121 operations in malware.c.
1123 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1126 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1127 syslog_duplication set false, log lines that would normally be written
1128 both the the main log and to the reject log were not written to syslog at
1131 PH/14 In the default configuration, change the use of "message" in ACL warn
1132 statements to "add_header".
1134 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1135 not followed by a command (e.g. "seen endif").
1137 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1138 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1141 PH/17 Added extra conditions to the default value of delay_warning_condition
1145 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1146 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1147 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1150 The Auto-Submitted: and various List- headers are standardised, whereas I
1151 don't think Precedence: ever was.
1153 PH/18 Refactored debugging code in route_finduser() to show more information,
1154 in particular, the error code if getpwnam() issues one.
1156 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1157 This is apparently needed in addition to the PH/07 change above to avoid
1158 any possible encoding problems.
1160 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1161 but not after initializing Perl.
1163 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1164 output them only if debugging. By default they are written stderr,
1165 apparently, which is not desirable.
1167 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1170 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1173 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1175 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1176 authenticated or an ident call has been made. Suppress the default
1177 values for $authenticated_id and $authenticated_sender (but permit -oMai
1178 and -oMas) when testing with -bh.
1180 PH/24 Re-jigged the order of the tests in the default configuration so that the
1181 tests for valid domains and recipients precede the DNS black list and CSA
1182 tests, on the grounds that those ones are more expensive.
1184 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1185 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1186 This bug exists in every version of Exim that I still have, right back to
1189 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1190 However, an attempt to turn on case-sensitivity in a regex key by
1191 including (?-i) didn't work because the subject string was already
1192 lowercased, and the effects were non-intuitive. It turns out that a
1193 one-line patch can be used to allow (?-i) to work as expected.
1199 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1200 other effects) broke the use of negated acl sub-conditions.
1202 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1205 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1206 "Deny" causes Exim to reject the incoming connection with a 554 error.
1207 Unfortunately, if there is a major crisis, such as a disk failure,
1208 tcp-wrappers gives "deny", whereas what one would like would be some
1209 kind of temporary error. A kludge has been added to help with this.
1210 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1211 554 error is used if errno is still zero or contains ENOENT (which occurs
1212 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1215 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1217 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1218 errors. Otherwise a message that provokes a temporary error (when other
1219 messages do not) can cause a whole host to time out.
1221 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1222 addresses were routed directly to files or pipes from a redirect router.
1223 File deliveries just didn't batch; pipe deliveries might have suffered
1226 PH/06 A failure to get a lock for a hints database would erroneously always say
1227 "Failed to get write lock", even when it was really a read lock.
1229 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1230 of 0600. This has been changed to use the value of the lockfile_mode
1231 option (which defaults to 0600).
1233 PH/08 Applied small patch from the Sieve maintainer.
1235 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1236 folder from quota calculations, a direct delivery into this folder messed
1237 up the contents of the maildirsize file. This was because the regex was
1238 used only to exclude .Trash (or whatever) when the size of the mailbox
1239 was calculated. There was no check that a delivery was happening into an
1240 excluded directory. This bug has been fixed by ignoring all quota
1241 processing for deliveries into excluded directories.
1243 PH/10 Added the maildirfolder_create_regex option to appendfile.
1249 PH/01 The code for finding all the local interface addresses on a FreeBSD
1250 system running IPv6 was broken. This may well have applied to all BSD
1251 systems, as well as to others that have similar system calls. The broken
1252 code found IPv4 interfaces correctly, but gave incorrect values for the
1253 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1254 that it would not match correctly against @[] and not recognize the IPv6
1257 PH/02 The ipliteral router was not recognizing addresses of the form user@
1258 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1260 PH/03 Added disable_ipv6.
1262 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1263 decoded form, because it is most often used to construct To: headers
1264 lines in autoreplies, and the decoded form may well be syntactically
1265 invalid. However, $reply_address has leading white space removed, and all
1266 newlines turned into spaces so that the autoreply transport does not
1269 PH/05 If group was specified without a user on a router, and no group or user
1270 was specified on a transport, the group from the router was ignored.
1272 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1273 for visible compile-time settings that can be used to change these
1274 numbers, for those that want even more. Backwards compatibility with old
1275 spool files has been maintained. However, going back to a previous Exim
1276 release will lost any variables that are in spool files.
1278 PH/07 Two small changes when running in the test harness: increase delay when
1279 passing a TCP/IP connection to a new process, in case the original
1280 process has to generate a bounce, and remove special handling of
1281 127.0.0.2 (sic), which is no longer necessary.
1283 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1284 be the same on different OS.
1286 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1289 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1290 whether --show-vars was specified or not
1292 JJ/02 exipick: Added support for new ACL variable spool format introduced
1295 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1296 syntactically invalid From: or Reply-to: line, and a filter used this to
1297 generate an autoreply, and therefore failed to obtain an address for the
1298 autoreply, Exim could try to deliver to a non-existent relative file
1299 name, causing unrelated and misleading errors. What now happens is that
1300 it logs this as a hard delivery error, but does not attempt to create a
1303 PH/11 The exinext utility has a -C option for testing purposes, but although
1304 the given file was scanned by exinext itself; it wasn't being passed on
1305 when Exim was called.
1307 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1308 an end-of-file indication when reading a command response.
1310 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1311 compiled. In many other places in Exim, IPv6 addresses are always
1312 recognized, so I have changed this. It also means that IPv4 domain
1313 literals of the form [IPV4:n.n.n.n] are now always recognized.
1315 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1316 used if the router is not running as root, for example, when verifying at
1317 ACL time, or when using -bh. The debugging output from this situation was
1318 non-existent - all you got was a failure to exec. I have made two
1321 (a) Failures to set uid/gid, the current directory, or a process leader
1322 in a subprocess such as that created by queryprogram now generate
1323 suitable debugging ouput when -d is set.
1325 (b) The queryprogram router detects when it is not running as root,
1326 outputs suitable debugging information if -d is set, and then runs
1327 the subprocess without attempting to change uid/gid.
1329 PH/15 Minor change to Makefile for building test_host (undocumented testing
1332 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1333 additional section of a DNS packet that returns MX or SRV records.
1334 Instead, it always explicitly searches for A/AAAA records. This avoids
1335 major problems that occur when a DNS server includes only records of one
1336 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1337 fixed another bug: if SRV records were looked up and the corresponding
1338 address records were *not* found in the additional section, the port
1339 values from the SRV records were lost.
1341 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1342 using the correct key (the original address) when searching the retry
1343 rules in order to find which one to use for generating the retry hint.
1345 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1346 adding the default one. Similarly, if it contains a Reply-To: header, the
1347 errors_reply_to option, if set, is not used.
1349 PH/19 When calculating a retry time, Exim used to measure the "time since
1350 failure" by looking at the "first failed" field in the retry record. Now
1351 it does not use this if it is later than than the arrival time of the
1352 message. Instead it uses the arrival time. This makes for better
1353 behaviour in cases where some deliveries succeed, thus re-setting the
1354 "first failed" field. An example is a quota failure for a huge message
1355 when small messages continue to be delivered. Without this change, the
1356 "time since failure" will always be short, possible causing more frequent
1357 delivery attempts for the huge message than are intended.
1358 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1360 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1361 $1, $2, $3) because the numerical variables can be reset during some
1362 expansion items (e.g. "match"), thereby losing the authentication data.
1364 PH/21 Make -bV show the size of off_t variables so that the test suite can
1365 decide whether to run tests for quotas > 2G.
1367 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1368 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1369 filecount value is greater than 2G or if a quota value is greater than 2G
1370 on a system where the size of off_t is not greater than 4, a panic error
1373 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1374 never match. The debug and -bh output now contains an explicit error
1375 message indicating a malformed IPv4 address or mask.
1377 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1378 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1379 PH/23 above applies.
1381 PH/25 Do not write to syslog when running in the test harness. The only
1382 occasion when this arises is a failure to open the main or panic logs
1383 (for which there is an explicit test).
1385 PH/26 Added the /no_tell option to "control=freeze".
1387 PH/27 If a host name lookup failed very early in a connection, for example, if
1388 the IP address matched host_lookup and the reverse lookup yielded a name
1389 that did not have a forward lookup, an error message of the form "no IP
1390 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1391 could be logged. Now it outputs the IP address instead of "NULL".
1393 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1394 allows the sender and the authenticated sender to be set when
1395 submitting a message from within Exim. Since child_open_exim() is
1396 documented for local_scan(), the new function should be too.
1398 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1399 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1400 results in an empty string is now treated as unset.
1402 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1404 PH/31 Added sender_verify_fail logging option.
1406 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1407 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1408 tidied the source and removed it altogether.
1410 PH/33 When a queue run was abandoned because the load average was too high, a
1411 log line was always written; now it is written only if the queue_run log
1412 selector is set. In addition, the log line for abandonment now contains
1413 information about the queue run such as the pid. This is always present
1414 in "start" and "stop" lines but was omitted from the "abandon" line.
1416 PH/34 Omit spaces between a header name and the colon in the error message that
1417 is given when verify = headers_syntax fails (if there are lots of them,
1418 the message gets confusing).
1420 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1421 names, as there are now some PTR records that contain slashes. This check
1422 is only to protect against broken name servers that fall over on strange
1423 characters, so the fact that it applies to all lookups doesn't matter.
1425 PH/36 Now that the new test suite is complete, we can remove some of the
1426 special code in Exim that was needed for the old test suite. For example,
1427 sorting DNS records because real resolvers return them in an arbitrary
1428 order. The new test suite's fake resolver always returns records in the
1431 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1432 bounces) except when queue_only is set, to avoid logging races between
1433 the different processes.
1435 PH/38 Panic-die if .include specifies a non-absolute path.
1437 PH/39 A tweak to the "H" retry rule from its user.
1439 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1440 a label. They prevented compilation on older perls.
1442 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1443 a warning to be raised on newish perls.
1445 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1446 on queue. Changes to match documented behaviour of showing count of
1447 messages matching specified criteria.
1449 PH/40 Changed the default ident timeout from 30s to 5s.
1451 PH/41 Added support for the use of login_cap features, on those BSD systems
1452 that have them, for controlling the resources used by pipe deliveries.
1454 PH/42 The content-scanning code uses fopen() to create files in which to put
1455 message data. Previously it was not paying any attention to the mode of
1456 the files. Exim runs with umask(0) because the rest of the code creates
1457 files with open(), and sets the required mode explicitly. Thus, these
1458 files were ending up world-writeable. This was not a big issue, because,
1459 being within the spool directory, they were not world-accessible. I have
1460 created a function called modefopen, which takes an additional mode
1461 argument. It sets umask(777), creates the file, chmods it to the required
1462 mode, then resets the umask. All the relevant calls to fopen() in the
1463 content scanning code have been changed to use this function.
1465 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1466 to 24 hours. This avoids potential overflow problems when processing G
1467 and H retry rules. I suspect nobody ever tinkers with this value.
1469 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1471 PH/45 When the plaintext authenticator is running as a client, the server's
1472 challenges are checked to ensure they are valid base64 strings. By
1473 default, the authentication attempt is cancelled if an invalid string is
1474 received. Setting client_ignore_invalid_base64 true ignores these errors.
1475 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1476 they are received. Thus, the responses can be made to depend on the
1477 challenges. If an invalid string is ignored, an empty string is placed in
1480 PH/46 Messages that are created by the autoreply transport now contains a
1481 References: header, in accordance with RFCs 2822 and 3834.
1483 PH/47 Added authenticated_sender_force to the smtp transport.
1485 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1487 PH/49 Installed latest patch from the Sieve maintainer.
1489 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1490 was also set, the appendfile transport was unnecessarily scanning a
1491 directory of message files (e.g. for maildir delivery) to find the count
1492 of files (along with the size), even though it did not need this
1493 information. It now does the scan only if it needs to find either the
1494 size of the count of files.
1496 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1498 PH/52 Two bugs concerned with error handling when the smtp transport is
1501 (i) Exim was not creating retry information for temporary errors given
1502 for individual recipients after the DATA command when the smtp transport
1503 was used in LMTP mode. This meant that they could be retried too
1504 frequently, and not timed out correctly.
1506 (ii) Exim was setting the flag that allows error details to be returned
1507 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1508 recipients that were returned after the DATA command.
1510 PH/53 This is related to PH/52, but is more general: for any failing address,
1511 when detailed error information was permitted to be returned to the
1512 sender, but the error was temporary, then after the final timeout, only
1513 "retry timeout exceeded" was returned. Now it returns the full error as
1514 well as "retry timeout exceeded".
1516 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1517 do this, and (what is worse) MTAs that accept it.
1519 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1520 will now be deprecated.
1522 PH/56 New os.c-cygwin from the Cygwin maintainer.
1524 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1525 formats (previously only available in exim formats via -bpr, -bpru,
1526 and -bpra. Now also available in native and exiqgrep formats)
1528 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1529 with very large, slow to parse queues
1531 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1533 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1535 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1536 responses to authentication challenges, though it was showing the
1537 challenges; (ii) I've removed the CR characters from the debug output for
1540 PH/58 Allow for the insertion of a newline as well as a space when a string
1541 is turned into more than one encoded-word during RFC 2047 encoding. The
1542 Sieve code now uses this.
1544 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1545 data_4xx, lost_connection, tls_required.
1547 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1548 message was being sent as an SMTP response.
1550 PH/61 Add -l and -k options to exicyclog.
1552 PH/62 When verifying, if an address was redirected to one new address, so that
1553 verification continued, and the new address failed or deferred after
1554 having set something in $address_data, the value of $address_data was not
1555 passed back to the ACL. This was different to the case when no
1556 redirection occurred. The value is now passed back in both cases.
1558 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1559 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1560 use login_cap.h, so on its own it isn't the distinguishing feature. The
1561 new name refers directly to the setclassresources() function.
1563 PH/65 Added configuration files for NetBSD3.
1565 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1567 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1568 is preferred over IPv4.
1570 PH/68 The bounce_return_message and bounce_return_body options were not being
1571 honoured for bounces generated during the reception of non-SMTP messages.
1572 In particular, this applied to messages rejected by the ACL. This bug has
1573 been fixed. However, if bounce_return_message is true and bounce_return_
1574 body is false, the headers that are returned for a non-SMTP message
1575 include only those that have been read before the error was detected.
1576 (In the case of an ACL rejection, they have all been read.)
1578 PH/69 The HTML version of the specification is now built in a directory called
1579 spec_html instead of spec.html, because the latter looks like a path with
1580 a MIME-type, and this confuses some software.
1582 PH/70 Catch two compiler warnings in sieve.c.
1584 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1585 function verify_get_ident() calls ip_connect() to connect a socket, but
1586 if the "connect()" function timed out, ip_connect() used to close the
1587 socket. However, verify_get_ident() also closes the socket later, and in
1588 between Exim writes to the log, which may get opened at this point. When
1589 the socket was closed in ip_connect(), the log could get the same file
1590 descriptor number as the socket. This naturally causes chaos. The fix is
1591 not to close the socket in ip_connect(); the socket should be closed by
1592 the function that creates it. There was only one place in the code where
1593 this was missing, in the iplookup router, which I don't think anybody now
1594 uses, but I've fixed it anyway.
1596 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1597 well as to direct DNS lookups. Otherwise the handling of names in host
1598 lists is inconsistent and therefore confusing.
1604 PH/01 Two changes to the default runtime configuration:
1606 (1) Move the checks for relay_from_hosts and authenticated clients from
1607 after to before the (commented out) DNS black list checks.
1609 (2) Add control=submission to the relay_from_hosts and authenticated
1610 clients checks, on the grounds that messages accepted by these
1611 statements are most likely to be submissions.
1613 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1615 (1) Generate an error if the third argument for the ${prvs expansion is
1618 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1621 (3) Reset the variables that are obtained from the first argument of
1622 ${prvscheck and used in the second argument before leaving the code,
1623 because their memory is reclaimed, so using them afterwards may do
1626 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1627 one (it's much easier than Tom thought :-).
1629 (5) Because of (4), we can now allow for the use of $prvscheck_result
1630 inside the third argument.
1632 PH/03 For some reason, the default setting of PATH when running a command from
1633 a pipe transport was just "/usr/bin". I have changed it to
1636 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1637 anything to be listed in the output from -bV.
1639 PH/05 When a filter generated an autoreply, the entire To: header line was
1640 quoted in the delivery log line, like this:
1642 => >A.N.Other <ano@some.domain> <original@ddress> ...
1644 This has been changed so that it extracts the operative address. There
1645 may be more than one such address. If so, they are comma-separated, like
1648 => >ano@some.domain,ona@other.domain <original@ddress> ...
1650 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1651 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1652 not being looked up in the rDNS to get a host name, Exim was showing the
1653 IP address twice in Received: lines, even though the IP addresses were
1654 identical. For example:
1656 Received: from [1.2.3.4] (helo=[1.2.3.4])
1658 However, if the real host name was known, it was omitting the HELO data
1659 if it matched the actual IP address. This has been tidied up so that it
1660 doesn't show the same IP address twice.
1662 PH/07 When both +timestamp and +memory debugging was on, the value given by
1663 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1664 called by the debug printing, thereby overwriting the timestamp buffer.
1665 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1667 PH/08 When the original message was included in an autoreply transport, it
1668 always said "this is a copy of the message, including all the headers",
1669 even if body_only or headers_only was set. It now gives an appropriate
1672 PH/09 Applied a patch from the Sieve maintainer which:
1674 o fixes some comments
1675 o adds the (disabled) notify extension core
1676 o adds some debug output for the result of if/elsif tests
1677 o points to the current vacation draft in the documentation
1678 and documents the missing references header update
1682 o fixes a bug in processing the envelope test (when testing
1683 multiple envelope elements, the last element determinted the
1686 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1687 Electronic Mail") by including:
1689 Auto-submitted: auto-generated
1691 in the messages that it generates (bounce messages and others, such as
1692 warnings). In the case of bounce messages for non-SMTP mesages, there was
1693 also a typo: it was using "Auto_submitted" (underscore instead of
1694 hyphen). Since every message generated by Exim is necessarily in response
1695 to another message, thes have all been changed to:
1697 Auto-Submitted: auto-replied
1699 in accordance with these statements in the RFC:
1701 The auto-replied keyword:
1703 - SHOULD be used on messages sent in direct response to another
1704 message by an automatic process,
1706 - MUST NOT be used on manually-generated messages,
1708 - MAY be used on Delivery Status Notifications (DSNs) and Message
1709 Disposition Notifications (MDNs),
1711 - MUST NOT be used on messages generated by automatic or periodic
1712 processes, except for messages which are automatic responses to
1715 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1716 to the default Received: header definition.
1718 PH/12 Added log selector acl_warn_skipped (default on).
1720 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1721 variables because (a) they are in the wrong storage pool and (b) even if
1722 they were copied, it wouldn't work properly because of the caching.
1724 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1725 checking when decoding. Apparently there are clients that generate
1726 overlong encoded strings. Why am I not surprised?
1728 PH/15 If the first argument of "${if match_address" was not empty, but did not
1729 contain an "@" character, Exim crashed. Now it writes a panic log message
1730 and treats the condition as false.
1732 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1734 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1735 extension "envelope-auth". The code is finished and in agreement with
1736 other implementations, but there is no documentation so far and in fact,
1737 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1738 not changing the active code.
1740 Print executed "if" and "elsif" statements when debugging is used. This
1741 helps a great deal to understand what a filter does.
1743 Document more things not specified clearly in RFC3028. I had all this
1744 sorted out, when out of a sudden new issues came to my mind. Oops."
1746 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1749 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1750 canonical form (as documented). However, after a host name lookup from
1751 the IP address, check_host() was doing a simple string comparison with
1752 addresses acquired from the DNS when checking that the found name did
1753 have the original IP as one of its addresses. Since any found IPv6
1754 addresses are likely to be in abbreviated form, the comparison could
1755 fail. Luckily, there already exists a function for doing the comparison
1756 by converting both addresses to binary, so now that is used instead of
1757 the text comparison.
1759 PH/20 There was another similar case to PH/19, when a complete host name was
1760 given in a host list; looking up its IP address could give an abbreviated
1761 form, whereas the current host's name might or might not be abbreviated.
1762 The same fix has been applied.
1768 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1769 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1772 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1773 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1775 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1777 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1778 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1779 Also, for "Auto-Submitted", treat anything other than "no" as
1780 non-personal, in accordance with RFC 3834. (Previously it treated
1781 anything starting "auto-" as non-personal.)
1783 TF/01 The control=submission/name=... option had a problem with syntax
1784 errors if the name included a slash character. The /name= option
1785 now slurps the rest of the string, so it can include any characters
1786 but it must come last in the list of options (after /sender_retain
1789 PH/05 Some modifications to the interface to the fake nameserver for the new
1797 TK/01 Added the "success_on_redirect" address verification option. See
1798 NewStuff for rationale and an example.
1800 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1802 PH/02 Patch to exigrep to allow it to work on syslog lines.
1804 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1805 fread() to skip over the body file's header line, because in Cygwin the
1806 header line is locked and is inaccessible.
1808 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1809 co-exist for some time) to make it clear that it is the Exim ID that is
1810 referenced, not the Message-ID: header line.
1812 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1813 string_format() function, because snprintf() does not exist on all
1816 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1817 ${stat: expansion item.
1819 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1820 protocol synchronization error", to keep the pedants happy.
1822 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1823 well as for IRIX systems, when gcc is being used. See the host.c source
1826 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1828 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1831 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1832 transport and to the smtp transport in LMTP mode.
1834 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1836 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1837 run a filter in a subprocess. This could lead to confusion in subsequent
1838 lookups in the parent process. There should also be a search_tidyup() at
1839 the end of the subprocess.
1841 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1842 only if the host matched helo_try_verify_hosts, which caused the
1843 verification to occur when the EHLO/HELO command was issued. The ACL just
1844 tested the remembered result. Now, if a previous verification attempt has
1845 not happened, "verify = helo" does it there and then.
1847 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1849 TK/03 Fix log output including CR from clamd.
1851 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1852 exist provoked a memory error which could cause a segfault.
1854 PH/15 Installed PCRE 6.2
1856 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1858 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1859 of the problem. Specifically, suggested +O2 rather than +O1 for the
1862 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1864 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1865 record, the port was not correctly logged when the outgoing_port log
1866 selector was set (it logged the transort's default port).
1868 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1869 fallback_hosts, and "hosts" in the smtp transport.
1871 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1872 host errors such as "Connection refused".
1874 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1875 authentication with radiusclient 0.4.9:
1877 - Error returned from rc_read_config was caught wrongly
1878 - Username/password not passed on to radius server due to wrong length.
1880 The presumption is that some radiusclient API changes for 4.51/PH/17
1881 were not taken care of correctly. The code is still untested by me (my
1882 Linux distribution still has 0.3.2 of radiusclient), but it was
1883 contributed by a Radius user.
1885 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1886 expanding the "port" option of the smtp transport.
1888 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1889 while reading a MIME header. Thanks to Tom Hughes for a patch.
1891 PH/24 Include config.h inside local_scan.h so that configuration settings are
1894 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1895 that in an ACL for RCPT (for example), you can examine exactly what was
1898 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1899 commands, but it was not correctly comparing the address with the actual
1900 client host address. Thus, it would show the EHLO address in Received:
1901 header lines when this was not necessary.
1903 PH/27 Added the % operator to ${eval:}.
1905 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1906 it should be ignoring failures (because with -C, for example, it has lost
1907 privilege). It wasn't ignoring creation failures other than "already
1910 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1913 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1914 because an input line was too long, either on its own, or by virtue of
1915 too many continuations, the temporary file was not being removed, and the
1916 return code was incorrect.
1918 PH/31 Missing "BOOL" in function definition in filtertest.c.
1920 PH/32 Applied Sieve patches from the maintainer.
1922 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1924 PH/33 Added "verify = not_blind".
1926 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1927 Local/Makefile (with some defaults set). These are used in built scripts
1928 such as exicyclog, but they have never been used in the exim_install
1929 script (though there are many overriding facilities there). I have
1930 arranged that the exim_install script now takes note of these two
1933 PH/35 Installed configuration files for Dragonfly.
1935 PH/36 When a locally submitted message by a trusted user did not contain a
1936 From: header, and the sender address was obtained from -f or from an SMTP
1937 MAIL command, and the trusted user did not use -F to supply a sender
1938 name, $originator_name was incorrectly used when constructing a From:
1939 header. Furthermore, $originator_name was used for submission mode
1940 messages from external hosts without From: headers in a similar way,
1941 which is clearly wrong.
1943 PH/37 Added control=suppress_local_fixups.
1945 PH/38 When log_selector = +received_sender was set, and the addition of the
1946 sender made the log line's construction buffer exactly full, or one byte
1947 less than full, an overflow happened when the terminating "\n" was
1950 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1951 when the result of a list match is failure because a DNS lookup failed.
1953 PH/40 RM_COMMAND is now used in the building process.
1955 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1956 the "build-* directories that it finds.
1958 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1959 address in a domain literal was a prefix of an interface address.
1961 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1962 when verifying a sender address, unless rewrite_headers is false.
1964 PH/44 Wrote a long comment about why errors_to addresses are verified as
1965 recipients, not senders.
1967 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1968 the ratelimit ACL was added.
1970 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1972 PH/46 Added extra information about PostgreSQL errors to the error string.
1974 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1975 avoiding the need to install special zones in a real server. This is
1976 backwards compatible; if it can't find the fake resolver, it drops back.
1977 Thus, both old and new test suites can be run.
1979 TF/02 Added util/ratelimit.pl
1981 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1982 clock is set back in time.
1984 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1985 Candler <B.Candler@pobox.com>.
1987 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1988 OK for addresses that are the result of redirections.
1990 PH/48 A number of further additions for the benefit of the new test suite,
1991 including a fake gethostbyname() that interfaces to the fake DNS resolver
1994 TF/06 The fix for widen_domains has also been applied to qualify_single and
1995 search_parents which are the other dnslookup options that can cause
1998 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2001 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2003 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2004 service" if it encountered an MX record with an empty target hostname.
2005 The message is now "an MX or SRV record indicated no SMTP service".
2007 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2008 if the DNS of the sending site is misconfigured. This is quite a
2009 common situation. This change restores the behaviour of treating a
2010 helo verification defer as a failure.
2012 PH/51 If self=fail was set on a router, the bounce message did not include the
2013 actual error message.
2019 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2021 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2022 command itself ended in error, the underlying message about the transport
2023 filter timeout was being overwritten with the pipe command error. Now the
2024 underlying error message should be appended to the second error message.
2026 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2028 PH/02 Reduce the amount of output that "make" produces by default. Full output
2029 can still be requested.
2031 PH/03 The warning log line about a condition test deferring for a "warn" verb
2032 was being output only once per connection, rather than after each
2033 occurrence (because it was using the same function as for successful
2034 "warn" verbs). This seems wrong, so I have changed it.
2036 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2037 it should not have, which might have caused a crash in the right
2038 circumstances, but probably never did.
2040 PH/04 Installed a modified version of Tony Finch's patch to make submission
2041 mode fix the return path as well as the Sender: header line, and to
2042 add a /name= option so that you can make the user's friendly name appear
2045 TF/03 Added the control = fakedefer ACL modifier.
2047 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2048 Mark Lowes for thorough testing.
2050 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2052 TK/03 Merged latest SRS patch from Miles Wilton.
2054 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2055 with the definition in sysexits.h (which is #included earlier).
2056 Fortunately, Exim does not actually use EX_OK. The code used to try to
2057 preserve the sysexits.h value, by assumimg that macro definitions were
2058 scanned for macro replacements. I have been disabused of this notion,
2059 so now the code just undefines EX_OK before #including unistd.h.
2061 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2062 in the smtp transport. When a block could not be written in a single
2063 write() function, the timeout was being re-applied to each part-write.
2064 This seems wrong - if the receiver was accepting one byte at a time it
2065 would take for ever. The timeout is now adjusted when this happens. It
2066 doesn't have to be particularly precise.
2068 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2069 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2071 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2072 without a domain if the check to <postmaster@domain> fails.
2074 SC/01 Eximstats: added -xls and the ability to specify output files
2075 (patch written by Frank Heydlauf).
2077 SC/02 Eximstats: use FileHandles for outputing results.
2079 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2081 SC/04 Eximstats: fixed display of large numbers with -nvr option
2083 SC/05 Eximstats: fixed merging of reports with empty tables.
2085 SC/06 Eximstats: added the -include_original_destination flag
2087 SC/07 Eximstats: removed tabs and trailing whitespace.
2089 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2091 TK/06 MBOX spool code: Add real "From " MBOX separator line
2092 so the .eml file is really in mbox format (even though
2093 most programs do not really care). Patch from Alex Miller.
2095 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2096 The latter is generated from $received_to and is only set if the
2097 message has one envelope recipient. SA can use these headers,
2098 obviously out-of-the-box. Patch from Alex Miller.
2100 PH/08 The ${def test on a variable was returning false if the variable's
2101 value was "0", contrary to what the specification has always said!
2102 The result should be true unless the variable is empty.
2104 PH/09 The syntax error of a character other than { following "${if
2105 def:variable_name" (after optional whitespace) was not being diagnosed.
2106 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2107 accidental colon was present, for example, could give incorrect results.
2109 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2110 result is used (not including appendfile, where other changes are about
2113 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2114 This involved changing a lot of size variables from int to off_t. It
2115 should work with maildirs and everything.
2117 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2118 spamd dying while we are connected to it.
2120 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2123 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2124 function for BDB 4.3.
2126 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2128 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2129 the sources for PCRE have been re-arranged and more files are now
2132 PH/15 The code I had for printing potentially long long variables in PH/11
2133 above was not the best (it lost precision). The length of off_t variables
2134 is now inspected at build time, and an appropriate printing format (%ld
2135 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2136 to be "long long int" or "long int". This is needed for the internal
2137 formatting function string_vformat().
2139 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2140 the configuration file to be ":syslog", then the script "guesses" where
2141 the logs files are, rather than using the compiled in default. In our
2142 case the guess is not the same as the compiled default, so the script
2143 suddenly stopped working when I started to use syslog. The patch checks
2144 to see if log_file_path is "". If so, it attempts to read it from exim
2145 with no configuration file to get the compiled in version, before it
2146 falls back to the previous guessing code."
2148 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2149 implementing BATV in an Exim configuration. See NewStuff for the gory
2152 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2153 Makefile that are specific to HP-UX.
2155 PH/18 If the "use_postmaster" option was set for a recipient callout together
2156 with the "random" option, the postmaster address was used as the MAIL
2157 FROM address for the random test, but not for the subsequent recipient
2158 test. It is now used for both.
2160 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2161 patch removes a few documentation additions to RFC 3028, because the
2162 latest draft now contains them. It adds the new en;ascii-case comparator
2163 and a new error check for 8bit text in MIME parts. Comparator and
2164 require names are now matched exactly. I enabled the subaddress
2165 extension, but it is not well tested yet (read: it works for me)."
2167 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2168 rework some of the code of TK/09 above to avoid the hardwired use of
2169 "%lld" and "long long". Replaced the call to snprintf() with a call to
2172 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2173 records point to non-existent hosts", "retry timeout exceeded", and
2174 "retry time not reached for any host after a long failure period".
2176 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2177 experimental DomainKeys support:
2179 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2180 (2) On an error such as an illegally used "control", the wrong name for
2181 the control was given.
2183 These problems did NOT occur unless DomainKeys support was compiled.
2185 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2187 PH/24 Added ${if match_ip condition.
2189 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2190 databases so that it will be absolutely obvious if a crash occurs in the
2191 DB library. This is a regular occurrence (often caused by mis-matched
2194 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2195 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2196 picked up on a user's system that detects such things. There doesn't seem
2197 to be a gcc warning option for this - only an attribute that has to be
2198 put on the function's prototype. It seems that in Fedora Core 4 they have
2199 set this on a number of new functions. No doubt there will be more in due
2202 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2203 specify a transport. However, if an address that was verified by such a
2204 router was the subject of a callout, Exim crashed because it tried to
2205 read the rcpt_include_affixes from the non-existent transport. Now it
2206 just assumes that the setting of that option is false. This bug was
2207 introduced by 4.51/PH/31.
2209 PH/28 Changed -d+all to exclude +memory, because that information is very
2210 rarely of interest, but it makes the output a lot bigger. People tend to
2211 do -d+all out of habit.
2213 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2214 code in os-type was giving problems when libc.so lives in lib64, like on
2217 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2218 aren't the modern standard, and it seems that some systems' include files
2219 don't always have them. Exim was already checking for some of the newer
2220 ones like T_AAAA, and defining it itself. I've added checks for all the
2221 record types that Exim uses.
2223 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2224 not automatically generating a new one, as it is supposed to. This
2225 prevented TLS from working. If the file did exist, but contained invalid
2226 data, a new version was generated, as expected. It was only the case of a
2227 non-existent file that was broken.
2229 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2230 with a change in libdomainkeys > 0.64.
2232 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2233 from DNS. If the selector record carries the flag, it now has
2234 precedence over the domain-wide flag.
2236 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2238 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2239 the use of a port name that isn't defined in /etc/services) occurred, the
2240 message was deferred as in a normal delivery, and thus remained on the
2241 spool, instead of being failed because of the mua_wrapper setting. This
2242 is now fixed, and I tidied up some of the mua_wrapper messages at the
2245 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2246 the messages in a hash of arrays rather than using individual hashes.
2247 This is a bit cleaner and results in dramatic memory savings, albeit
2248 at a slight CPU cost.
2250 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2251 as requested by Marc Sherman.
2253 SC/10 Eximstats: added histograms for user specified patterns as requested
2256 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2258 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2259 fopen() in the content-scanning modules that did not already have it.
2265 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2266 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2268 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2270 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2272 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2273 to test Sieve filters that use "vacation".
2275 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2276 that changes the way the GnuTLS parameters are stored in the cache file.
2277 The new format can be generated externally. For backward compatibility,
2278 if the data in the cache doesn't make sense, Exim assumes it has read an
2279 old-format file, and it generates new data and writes a new file. This
2280 means that you can't go back to an older release without removing the
2283 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2284 work if there are any delivery delays because "one_time" forces the
2285 parent to be marked "delivered", so its unseen clone is never tried
2286 again. For this reason, Exim now forbids the simultaneous setting of
2289 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2290 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2291 entry if you want to know the details. The fix, however, neglected to
2292 consider the case when local delivery batching is involved. The test for
2293 "previously delivered" was not happening when checking to see if an
2294 address could be batched with a previous (undelivered) one; under
2295 certain circumstances this could lead to multiple deliveries to the same
2298 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2299 in its include files, and this causes problems building Exim.
2301 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2302 header_syntax) but Exim was just ignoring anything given after a slash.
2303 In particular, this caused confusion with an attempt to use "verify =
2304 reverse_host_lookup/defer_ok". An error is now given when options are
2305 supplied for verify items that do not have them. (Maybe reverse_host_
2306 lookup should have a defer_ok option, but that's a different point.)
2308 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2309 defined by RFC 821) to 2048, because there were problems with some AUTH
2310 commands, and RFC 1869 says the size should be increased for extended
2311 SMTP commands that take arguments.
2313 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2316 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2317 "unknown" error; now it says that the functionality isn't in the binary.
2319 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2320 an address' error message when a string expansion fails (syntax or
2321 whatever). Otherwise the password may appear in the log. Following change
2322 PH/42 below, there is no longer a chance of it appearing in a bounce
2325 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2327 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2328 of its addresses was ever tried. (Bugzilla bug #2.)
2330 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2331 the result incorrectly in the debug output. (It correctly added a newline
2332 to what was transported.)
2334 TF/01 Added $received_time.
2336 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2337 commented out examples of how to interface to a virus scanner and to
2338 SpamAssassin. Also added commented examples of av_scanner and
2339 spamd_address settings.
2341 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2342 and controls are allowed in which ACLs. There were a couple of minor
2343 errors. Some of the entries in the conditions table (which is a table of
2344 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2345 negation of where the condition IS allowed.
2347 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2349 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2350 header file does not have a version number, so I've had to invent a new
2351 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2352 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2353 radiusclient), but it was contributed by a Radius user.
2355 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2356 files or directories, for OpenSSL.
2358 PH/19 When an Exim process that is running as root has to create an Exim log
2359 file, it does so in a subprocess that runs as exim:exim so as to get the
2360 ownership right at creation (otherwise, other Exim processes might see
2361 the file with the wrong ownership). There was no test for failure of this
2362 fork() call, which would lead to the process getting stuck as it waited
2363 for a non-existent subprocess. Forks do occasionally fail when resources
2364 run out. I reviewed all the other calls to fork(); they all seem to check
2367 PH/20 When checking for unexpected SMTP input at connect time (before writing
2368 the banner), Exim was not dealing correctly with a non-positive return
2369 from the read() function. If the client had disconnected by this time,
2370 the result was a log entry for a synchronization error with an empty
2371 string after "input=" when read() returned zero. If read() returned -1
2372 (an event I could not check), uninitialized data bytes were printed.
2373 There were reports of junk text (parts of files, etc) appearing after
2376 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2378 PH/22 Added support for macro redefinition, and (re)definition in between
2379 driver and ACL definitions.
2381 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2382 forgetting to use the resulting value; it was using the unexpanded value.
2384 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2385 hadn't been configured. The fix is from Juergen Kreileder, who
2386 understands it better than I do:
2388 "Here's what I see happening with three configured cyrus_sasl
2389 authenticators configured (plain, login, cram-md5):
2391 On startup auth_cyrus_sasl_init() gets called for each of these.
2392 This means three calls to sasl_listmech() without a specified mech_list.
2393 => SASL tests which mechs of all available mechs actually work
2394 => three warnings about OTP not working
2395 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2397 With the patch, sasl_listmech() also gets called three times. But now
2398 SASL's mech_list option is set to the server_mech specified in the the
2399 authenticator. Or in other words, the answer from sasl_listmech()
2400 gets limited to just the mech you're testing for (which is different
2402 => the return list contains just 'plain' or 'login', 'cram-md5' or
2403 nothing depending on the value of ob->server_mech.
2405 I've just tested the patch: Authentication still works fine,
2406 unavailable mechs specified in the exim configuration are still
2407 caught, and the auth.log warnings about OTP are gone."
2409 PH/25 When debugging is enabled, the contents of the command line are added
2410 to the debugging output, even when log_selector=+arguments is not
2413 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2414 answer is "GNU", and only if the return is "GNU/something" is the answer
2417 PH/27 $acl_verify_message is now set immediately after the failure of a
2418 verification in an ACL, and so is available in subsequent modifiers. In
2419 particular, the message can be preserved by coding like this:
2421 warn !verify = sender
2422 set acl_m0 = $acl_verify_message
2424 Previously, $acl_verify_message was set only while expanding "message"
2425 and "log_message" when a very denied access.
2427 PH/28 Modified OS/os.c-Linux with
2429 -#ifndef OS_LOAD_AVERAGE
2430 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2432 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2433 nomenclature these days.)
2435 PH/29 Installed patch from the Sieve maintainer that adds the options
2436 sieve_useraddress and sieve_subaddress to the redirect router.
2438 PH/30 In these circumstances:
2439 . Two addresses routed to the same list of hosts;
2440 . First host does not offer TLS;
2441 . First host accepts first address;
2442 . First host gives temporary error to second address;
2443 . Second host offers TLS and a TLS session is established;
2444 . Second host accepts second address.
2445 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2446 and peerdn, if requested) that were in fact used only for the second
2449 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2450 attention to any local part prefix or suffix that was matched by the
2451 router that accepted the address. It now behaves in the same way as it
2452 does for delivery: the affixes are removed from the local part unless
2453 rcpt_include_affixes is set on the transport.
2455 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2456 timeout during the DATA phase of an incoming message.
2458 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2459 applied a patch sanctioned by the Sieve maintainer.
2461 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2462 the uid or gid is negative. A case of a negative gid caused this to be
2463 noticed. The fix allows for either to be negative.
2465 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2466 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2469 PH/36 The change PH/12 above was broken. Fixed it.
2471 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2472 the grounds that routing the same address twice would always produce the
2473 same answer. This might have been true once, but it is certainly no
2474 longer true now. Routing a child address may depend on the previous
2475 routing that produced that child. Some complicated redirection strategies
2476 went wrong when messages had multiple recipients, and made Exim's
2477 behaviour dependent on the order in which the addresses were given.
2479 I have moved the duplicate checking until after the routing is complete.
2480 Exim scans the addresses that are assigned to local and remote
2481 transports, and removes any duplicates. This means that more work will be
2482 done, as duplicates will always all be routed, but duplicates are
2483 presumably rare, so I don't expect this is of any significance.
2485 For deliveries to pipes, files, and autoreplies, the duplicate checking
2486 still happens during the routing process, since they are not going to be
2489 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2490 It corrects a timeout issue with spamd. This is Ian's comment: "The
2491 background is that sometimes spamd either never reads data from a
2492 connection it has accepted, or it never writes response data. The exiscan
2493 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2494 blindly assumes that writes won't block so it may never time out."
2496 PH/39 Allow G after quota size as well as K and M.
2498 PH/40 The value set for $authenticated_id in an authenticator may not contain
2499 binary zeroes or newlines because the value is written to log lines and
2500 to spool files. There was no check on this. Now the value is run through
2501 the string_printing() function so that such characters are converted to
2502 printable escape sequences.
2504 PH/41 $message_linecount is a new variable that contains the total number of
2505 lines in the message. Compare $body_linecount, which is the count for the
2508 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2509 bounce and delay warning messages, except in certain special cases, which
2512 (a) An SMTP error message from a remote host;
2513 (b) A message specified in a :fail: redirection;
2514 (c) A message specified in a "fail" command in a system filter;
2515 (d) A message specified in a FAIL return from the queryprogram router;
2516 (e) A message specified by the cannot_route_message router option.
2518 In these cases only, Exim does include the error details in bounce and
2519 warning messages. There are also a few cases where bland messages such
2520 as "unrouteable address" or "local delivery error" are given.
2522 PH/43 $value is now also set for the "else" part of a ${run expansion.
2524 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2525 being worked on, but at least Exim now implements the latest version to
2528 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2529 process to complete was treated as a delivery failure, a timeout while
2530 writing the message to the pipe was logged, but erroneously treated as a
2531 successful delivery. Such timeouts include transport filter timeouts. For
2532 consistency with the overall process timeout, these timeouts are now
2533 treated as errors, giving rise to delivery failures by default. However,
2534 there is now a new Boolean option for the pipe transport called
2535 timeout_defer, which, if set TRUE, converts the failures into defers for
2536 both kinds of timeout. A transport filter timeout is now identified in
2539 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2540 systems where "make" and "gmake" are different, calling "gmake" at top
2541 level broke things. I've arranged for the value of $(MAKE) to be passed
2542 from the Makefile to this script so that it can call the same version of
2546 A note about Exim versions 4.44 and 4.50
2547 ----------------------------------------
2549 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2550 changes of various kinds. As a consequence, a big documentation update was
2551 needed. This delayed the release for rather longer than seemed good, especially
2552 in the light of a couple of (minor) security issues. Therefore, the changes
2553 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2554 release. So 4.44 and 4.50 are in effect two different branches that both start
2557 I have left the 4.50 change log unchanged; it contains all the changes since
2558 4.43. The change log for 4.44 is below; many of its items are identical to
2559 those for 4.50. This seems to be the most sensible way to preserve the
2560 historical information.
2566 1. Minor wording change to the doc/README.SIEVE file.
2568 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2569 computation of the current number of files was incorrect.
2571 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2572 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2575 4. Give more explanation in the error message when the command for a transport
2576 filter fails to execute.
2578 5. There are several places where Exim runs a non-Exim command in a
2579 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2580 was being done only for the command run by the queryprogram router. It is
2581 now done for all such subprocesses. The other cases are: ${run, transport
2582 filters, and the commands run by the lmtp and pipe transports.
2584 6. Added CONFIGURE_GROUP build-time option.
2586 7. Some older OS have a limit of 256 on the maximum number of file
2587 descriptors. Exim was using setrlimit() to set 1000 as a large value
2588 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2589 systems. I've change it so that if it can't get 1000, it tries for 256.
2591 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2592 was an oversight, and furthermore, ever since the addition of extra
2593 controls (e.g. 4.43/32), the checks on when to allow different forms of
2594 "control" were broken. There should now be diagnostics for all cases when a
2595 control that does not make sense is encountered.
2597 9. Added the /retain_sender option to "control=submission".
2599 10. $recipients is now available in the predata ACL (oversight).
2601 11. Tidy the search cache before the fork to do a delivery from a message
2602 received from the command line. Otherwise the child will trigger a lookup
2603 failure and thereby defer the delivery if it tries to use (for example) a
2604 cached ldap connection that the parent has called unbind on.
2606 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2607 of $address_data from the recipient verification was clobbered by the
2608 sender verification.
2610 13. The value of address_data from a sender verification is now available in
2611 $sender_address_data in subsequent conditions in the ACL statement.
2613 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2615 15. Added a new option "connect=<time>" to callout options, to set a different
2618 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2619 was its contents. (It was OK if the option was not defined at all.)
2621 17. A "Completed" log line is now written for messages that are removed from
2622 the spool by the -Mrm option.
2624 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2625 information about exactly what failed.
2627 19. Added -dd to debug only the daemon process.
2629 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2630 handles timeouts, both on the server side and network timeouts. Renamed the
2631 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2633 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2634 It is now set to "smtps".
2636 22. $host_address is now set to the target address during the checking of
2637 ignore_target_hosts.
2639 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2640 being passed; this would have caused $sender_host_name to have been used if
2641 matching the list had actually called for a host name (not very likely,
2642 since this list is usually IP addresses). A host name is now passed as
2645 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2646 code that specifies a non-restarting handler (typically sigaction() in
2647 modern systems) in an attempt to fix a rare and obscure crash bug.
2649 25. Narrowed the window for a race in the daemon that could cause it to ignore
2650 SIGCHLD signals. This is not a major problem, because they are used only to
2651 wake it up if nothing else does.
2653 26. A malformed maildirsize file could cause Exim to calculate negative values
2654 for the mailbox size or file count. Odd effects could occur as a result.
2655 The maildirsize information is now recalculated if the size or filecount
2658 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2659 support for a long time. Removed HAVE_SYS_VFS_H.
2661 28. Installed the latest version of exipick from John Jetmore.
2663 29. In an address list, if the pattern was not a regular expression, an empty
2664 subject address (from a bounce message) matched only if the pattern was an
2665 empty string. Non-empty patterns were not even tested. This was the wrong
2666 because it is perfectly reasonable to use an empty address as part of a
2667 database query. An empty address is now tested by patterns that are
2668 lookups. However, all the other forms of pattern expect the subject to
2669 contain a local part and a domain, and therefore, for them, an empty
2670 address still always fails if the pattern is not itself empty.
2672 30. Exim went into a mad DNS loop when attempting to do a callout where the
2673 host was specified on an smtp transport, and looking it up yielded more
2674 than one IP address.
2676 31. Re-factored the code for checking spool and log partition space into a
2677 function that finds that data and another that does the check. The former
2678 is then used to implement four new variables: $spool_space, $log_space,
2679 $spool_inodes, and $log_inodes.
2681 32. The RFC2047 encoding function was originally intended for short strings
2682 such as real names; it was not keeping to the 75-character limit for
2683 encoded words that the RFC imposes. It now respects the limit, and
2684 generates multiple encoded words if necessary. To be on the safe side, I
2685 have increased the buffer size for the ${rfc2047: expansion operator from
2688 33. It is now permitted to omit both strings after an "if" condition; if the
2689 condition is true, the result is "true". As before, when the second string
2690 is omitted, a false condition yields an empty string. This makes it less
2691 cumbersome to write custom ACL and router conditions.
2693 34. Failure to deliver a bounce message always caused it to be frozen, even if
2694 there was an errors_to setting on the router. The errors_to setting is now
2697 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2698 canonical form (fully expanded) before being placed in
2699 $sender_host_address.
2701 36. The table in the code that translates DNS record types into text (T_A to
2702 "A" for instance) was missing entries for NS and CNAME. It is just possible
2703 that this could have caused confusion if both these types were looked up
2704 for the same domain, because the text type is used as part of Exim's
2705 per-process caching. But the chance of anyone hitting this buglet seems
2708 37. The dnsdb lookup has been extended in a number of ways.
2710 (1) There is a new type, "zns", which walks up the domain tree until it
2711 finds some nameserver records. It should be used with care.
2713 (2) There is a new type, "mxh", which is like "mx" except that it returns
2714 just the host names, not the priorities.
2716 (3) It is now possible to give a list of domains (or IP addresses) to be
2717 looked up. The behaviour when one of the lookups defers can be
2718 controlled by a keyword.
2720 (4) It is now possible to specify the separator character for use when
2721 multiple records are returned.
2723 38. The dnslists ACL condition has been extended: it is now possible to supply
2724 a list of IP addresses and/or domains to be looked up in a particular DNS
2727 39. Added log_selector=+queue_time_overall.
2729 40. When running the queue in the test harness, wait just a tad after forking a
2730 delivery process, to get repeatability of debugging output.
2732 41. Include certificate and key file names in error message when GnuTLS fails
2733 to set them up, because the GnuTLS error message doesn't include the name
2734 of the failing file when there is a problem reading it.
2736 42. Allow both -bf and -bF in the same test run.
2738 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2740 44. The "Exiscan patch" is now merged into the mainline Exim source.
2742 45. Sometimes the final signoff response after QUIT could fail to get
2743 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2744 < 0 before doing a fflush(). This bug looks as though it goes back to the
2745 introduction of TLS in release 3.20, but "sometimes" must have been rare
2746 because the tests only now provoked it.
2748 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2749 (this can affect the format of dates).
2751 47. exim_tidydb, when checking for the continued existence of a message for
2752 which it has found a message-specific retry record, was not finding
2753 messages that were in split spool directories. Consequently, it was
2754 deleting retry records that should have stayed in existence.
2756 48. Steve fixed some bugs in eximstats.
2758 49. The SPA authentication driver was not abandoning authentication and moving
2759 on to the next authenticator when an expansion was forced to fail,
2760 contradicting the general specification for all authenticators. Instead it
2761 was generating a temporary error. It now behaves as specified.
2763 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2764 (the order specifies the preference for clients). The order is now AES256,
2765 AES128, 3DES, ARCFOUR128.
2767 51. Small patch to Sieve code - explicitly set From: when generating an
2770 52. Exim crashed if a remote delivery caused a very long error message to be
2771 recorded - for instance if somebody sent an entire SpamAssassin report back
2772 as a large number of 550 error lines. This bug was coincidentally fixed by
2773 increasing the size of one of Exim's internal buffers (big_buffer) that
2774 happened as part of the Exiscan merge. However, to be on the safe side, I
2775 have made the code more robust (and fixed the comments that describe what
2778 53. Now that there can be additional text after "Completed" in log lines (if
2779 the queue_time_overall log selector is set), a one-byte patch to exigrep
2780 was needed to allow it to recognize "Completed" as not the last thing in
2783 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2784 patch that reportedly fixes this has been added. I am not expert enough to
2785 create a test for it. This is what the patch creator wrote:
2787 "I found a little strange behaviour of ldap code when working with
2788 Windows 2003 AD Domain, where users was placed in more than one
2789 Organization Units. When I tried to give exim partial DN, the exit code
2790 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2791 But simultaneously result of request was absolutely normal ldap result,
2792 so I produce this patch..."
2794 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2795 REFERENCE, so I have modified the code to exclude the patch when that macro
2798 55. Some experimental protocols are using DNS PTR records for new purposes. The
2799 keys for these records are domain names, not reversed IP addresses. The
2800 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2801 leaves it alone. Component reversal etc. now happens only for IP addresses.
2804 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2806 57. Double the size of the debug message buffer (to 2048) so that more of very
2807 long debug lines gets shown.
2809 58. The exicyclog utility now does better if the number of log files to keep
2810 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2812 59. Two changes related to the smtp_active_hostname option:
2814 (1) $smtp_active_hostname is now available as a variable.
2815 (2) The default for smtp_banner uses $smtp_active_hostname instead
2816 of $primary_hostname.
2818 60. The host_aton() function is supposed to be passed a string that is known
2819 to be a valid IP address. However, in the case of IPv6 addresses, it was
2820 not checking this. This is a hostage to fortune. Exim now panics and dies
2821 if the condition is not met. A case was found where this could be provoked
2822 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2823 components; fortuitously, this particular loophole had already been fixed
2824 by change 4.50/55 above.
2826 If there are any other similar loopholes, the new check in host_aton()
2827 itself should stop them being exploited. The report I received stated that
2828 data on the command line could provoke the exploit when Exim was running as
2829 exim, but did not say which command line option was involved. All I could
2830 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2831 running as the user.
2834 61. There was a buffer overflow vulnerability in the SPA authentication code
2835 (which came originally from the Samba project). I have added a test to the
2836 spa_base64_to_bits() function which I hope fixes it.
2839 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2840 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2842 63. The daemon start-up calls getloadavg() while still root for those OS that
2843 need the first call to be done as root, but it missed one case: when
2844 deliver_queue_load_max is set with deliver_drop_privilege. This is
2845 necessary for the benefit of the queue runner, because there is no re-exec
2846 when deliver_drop_privilege is set.
2848 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2849 This has been fixed.
2851 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2852 was in use, was not putting the data itself into the right store pool;
2853 consequently, it could be overwritten for a subsequent message in the same
2854 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2857 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2859 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2860 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2861 the function were treating the return as a boolean value, which happened to
2862 work because 0=false and not-0=true, but is not correct code.
2864 68. The host_aton() function was not handling scoped IPv6 addresses (those
2865 with, for example, "%eth0" on the end) correctly.
2867 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2868 negated items (that is, ~something) in unsigned ints. Some compilers
2869 apparently mutter when there is no cast.
2871 70. If an address verification called from an ACL failed, and did not produce a
2872 user-specific message (i.e. there was only a "system" message), nothing was
2873 put in $acl_verify_message. In this situation, it now puts the system
2876 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2877 session; change 4.31/43 added the unwanted input to the log line - except
2878 that it did not do this in the start of session case. It now does.
2880 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2881 This is wrong and can cause the other end to generate a synchronization
2882 error if it is another Exim or anything else that does the synchronization
2883 check. A QUIT command is no longer sent after a timeout.
2885 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2886 during host lookups.
2888 74. The defer_ok option of callout verification was not working if it was used
2889 when verifying addresses in header lines, that is, for this case:
2891 verify = header_sender/callout=defer_ok
2893 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2894 those file descriptors could be used for SMTP connections. If anything
2895 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2896 could be sent to the SMTP client, causing chaos. The daemon now opens
2897 stdin, stdout, and stderr to /dev/null when it puts itself into the
2900 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2901 log by default. The user can override this with suitable Perl magic.
2903 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2904 the log message when discard triggers, was not working for the DATA ACL or
2905 for the non-SMTP ACL.
2907 78. Error message wording change in sieve.c.
2909 79. If smtp_accept_max_per_host was set, the number of connections could be
2910 restricted to fewer than expected, because the daemon was trying to set up
2911 a new connection before checking whether the processes handling previous
2912 connections had finished. The check for completed processes is now done
2913 earlier. On busy systems, this bug wouldn't be noticed because something
2914 else would have woken the daemon, and it would have reaped the completed
2917 80. If a message was submitted locally by a user whose login name contained one
2918 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2919 It caused a spool format error. I have fixed the spool reading code. A
2920 related problem was that the "from" clause in the Received: line became
2921 illegal because of the space(s). It is now covered by ${quote_local_part.
2923 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2926 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2928 83. Patch from Sieve maintainer to fix unterminated string problem in
2929 "vacation" handling.
2931 84. Some minor changes to the Linux configuration files to help with other
2932 OS variants using glibc.
2934 85. One more patch for Sieve to update vacation handling to latest spec.
2937 ----------------------------------------------------
2938 See the note above about the 4.44 and 4.50 releases.
2939 ----------------------------------------------------
2945 1. Change 4.43/35 introduced a bug that caused file counts to be
2946 incorrectly computed when quota_filecount was set in an appendfile
2949 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2950 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2953 3. Give more explanation in the error message when the command for a transport
2954 filter fails to execute.
2956 4. There are several places where Exim runs a non-Exim command in a
2957 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2958 was being done only for the command run by the queryprogram router. It is
2959 now done for all such subprocesses. The other cases are: ${run, transport
2960 filters, and the commands run by the lmtp and pipe transports.
2962 5. Some older OS have a limit of 256 on the maximum number of file
2963 descriptors. Exim was using setrlimit() to set 1000 as a large value
2964 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2965 systems. I've change it so that if it can't get 1000, it tries for 256.
2967 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2968 was an oversight, and furthermore, ever since the addition of extra
2969 controls (e.g. 4.43/32), the checks on when to allow different forms of
2970 "control" were broken. There should now be diagnostics for all cases when a
2971 control that does not make sense is encountered.
2973 7. $recipients is now available in the predata ACL (oversight).
2975 8. Tidy the search cache before the fork to do a delivery from a message
2976 received from the command line. Otherwise the child will trigger a lookup
2977 failure and thereby defer the delivery if it tries to use (for example) a
2978 cached ldap connection that the parent has called unbind on.
2980 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2981 of $address_data from the recipient verification was clobbered by the
2982 sender verification.
2984 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2985 was its contents. (It was OK if the option was not defined at all.)
2987 11. A "Completed" log line is now written for messages that are removed from
2988 the spool by the -Mrm option.
2990 12. $host_address is now set to the target address during the checking of
2991 ignore_target_hosts.
2993 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2994 being passed; this would have caused $sender_host_name to have been used if
2995 matching the list had actually called for a host name (not very likely,
2996 since this list is usually IP addresses). A host name is now passed as
2999 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3000 code that specifies a non-restarting handler (typically sigaction() in
3001 modern systems) in an attempt to fix a rare and obscure crash bug.
3003 15. Narrowed the window for a race in the daemon that could cause it to ignore
3004 SIGCHLD signals. This is not a major problem, because they are used only to
3005 wake it up if nothing else does.
3007 16. A malformed maildirsize file could cause Exim to calculate negative values
3008 for the mailbox size or file count. Odd effects could occur as a result.
3009 The maildirsize information is now recalculated if the size or filecount
3012 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3013 support for a long time. Removed HAVE_SYS_VFS_H.
3015 18. Updated exipick to current release from John Jetmore.
3017 19. Allow an empty sender to be matched against a lookup in an address list.
3018 Previously the only cases considered were a regular expression, or an
3021 20. Exim went into a mad DNS lookup loop when doing a callout where the
3022 host was specified on the transport, if the DNS lookup yielded more than
3025 21. The RFC2047 encoding function was originally intended for short strings
3026 such as real names; it was not keeping to the 75-character limit for
3027 encoded words that the RFC imposes. It now respects the limit, and
3028 generates multiple encoded words if necessary. To be on the safe side, I
3029 have increased the buffer size for the ${rfc2047: expansion operator from
3032 22. Failure to deliver a bounce message always caused it to be frozen, even if
3033 there was an errors_to setting on the router. The errors_to setting is now
3036 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3037 canonical form (fully expanded) before being placed in
3038 $sender_host_address.
3040 24. Updated eximstats to version 1.33
3042 25. Include certificate and key file names in error message when GnuTLS fails
3043 to set them up, because the GnuTLS error message doesn't include the name
3044 of the failing file when there is a problem reading it.
3046 26. Expand error message when OpenSSL has problems setting up cert/key files.
3049 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3050 (this can affect the format of dates).
3052 28. exim_tidydb, when checking for the continued existence of a message for
3053 which it has found a message-specific retry record, was not finding
3054 messages that were in split spool directories. Consequently, it was
3055 deleting retry records that should have stayed in existence.
3057 29. eximstats updated to version 1.35
3058 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3059 1.35 - bugfix such that pie charts by volume are generated correctly
3061 30. The SPA authentication driver was not abandoning authentication and moving
3062 on to the next authenticator when an expansion was forced to fail,
3063 contradicting the general specification for all authenticators. Instead it
3064 was generating a temporary error. It now behaves as specified.
3066 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3067 (the order specifies the preference for clients). The order is now AES256,
3068 AES128, 3DES, ARCFOUR128.
3070 31. Small patch to Sieve code - explicitly set From: when generating an
3073 32. Exim crashed if a remote delivery caused a very long error message to be
3074 recorded - for instance if somebody sent an entire SpamAssassin report back
3075 as a large number of 550 error lines. This bug was coincidentally fixed by
3076 increasing the size of one of Exim's internal buffers (big_buffer) that
3077 happened as part of the Exiscan merge. However, to be on the safe side, I
3078 have made the code more robust (and fixed the comments that describe what
3081 33. Some experimental protocols are using DNS PTR records for new purposes. The
3082 keys for these records are domain names, not reversed IP addresses. The
3083 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3084 leaves it alone. Component reversal etc. now happens only for IP addresses.
3087 34. The host_aton() function is supposed to be passed a string that is known
3088 to be a valid IP address. However, in the case of IPv6 addresses, it was
3089 not checking this. This is a hostage to fortune. Exim now panics and dies
3090 if the condition is not met. A case was found where this could be provoked
3091 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3092 components; fortuitously, this particular loophole had already been fixed
3093 by change 4.50/55 or 4.44/33 above.
3095 If there are any other similar loopholes, the new check in host_aton()
3096 itself should stop them being exploited. The report I received stated that
3097 data on the command line could provoke the exploit when Exim was running as
3098 exim, but did not say which command line option was involved. All I could
3099 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3100 running as the user.
3103 35. There was a buffer overflow vulnerability in the SPA authentication code
3104 (which came originally from the Samba project). I have added a test to the
3105 spa_base64_to_bits() function which I hope fixes it.
3108 36. The daemon start-up calls getloadavg() while still root for those OS that
3109 need the first call to be done as root, but it missed one case: when
3110 deliver_queue_load_max is set with deliver_drop_privilege. This is
3111 necessary for the benefit of the queue runner, because there is no re-exec
3112 when deliver_drop_privilege is set.
3114 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3115 was in use, was not putting the data itself into the right store pool;
3116 consequently, it could be overwritten for a subsequent message in the same
3117 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3120 38. Sometimes the final signoff response after QUIT could fail to get
3121 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3122 < 0 before doing a fflush(). This bug looks as though it goes back to the
3123 introduction of TLS in release 3.20, but "sometimes" must have been rare
3124 because the tests only now provoked it.
3130 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3131 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3132 Now it returns an integer. A number of calls were still expecting a T/F
3133 return. Fortuitously, in all cases, the tests worked in OK situations,
3134 which is the norm. However, things would have gone wrong on any write
3135 failures on the smtp file descriptor. This function is used when sending
3136 messages over SMTP and also when doing verify callouts.
3138 2. When Exim is called to do synchronous delivery of a locally submitted
3139 message (the -odf or -odi options), it no longer closes stderr before doing
3142 3. Implemented the mua_wrapper option.
3144 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3146 5. Implemented the functions header_remove(), header_testname(),
3147 header_add_at_position(), and receive_remove_recipient(), and exported them
3150 6. If an ACL "warn" statement specified the addition of headers, Exim already
3151 inserted X-ACL-Warn: at the start if there was no header name. However, it
3152 was not making this test for the second and subsequent header lines if
3153 there were newlines in the string. This meant that an invalid header could
3154 be inserted if Exim was badly configured.
3156 7. Allow an ACL "warn" statement to add header lines at the start or after all
3157 the Received: headers, as well as at the end.
3159 8. Added the rcpt_4xx retry error code.
3161 9. Added postmaster_mailfrom=xxx to callout verification option.
3163 10. Added mailfrom=xxxx to the callout verification option, for verify=
3166 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3167 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3169 12. Inserted some casts to stop certain compilers warning when using pointer
3170 differences as field lengths or precisions in printf-type calls (mostly
3171 affecting debugging statements).
3173 13. Added optional readline() support for -be (dynamically loaded).
3175 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3176 same clock tick as a message's arrival, so that its received time was the
3177 same as the "first fail" time on the retry record, and that message
3178 remained on the queue past the ultimate address timeout, every queue runner
3179 would try a delivery (because it was past the ultimate address timeout) but
3180 after another failure, the ultimate address timeout, which should have then
3181 bounced the address, did not kick in. This was a "< instead of <=" error;
3182 in most cases the first failure would have been in the next clock tick
3183 after the received time, and all would be well.
3185 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3186 being recognized when the domain list was tested by the match_domain
3187 condition in an expansion string.
3189 16. Added the ${str2b64: operator.
3191 17. Exim was always calling setrlimit() to set a large limit for the number of
3192 processes, without checking whether the existing limit was already
3193 adequate. (It did check for the limit on file descriptors.) Furthermore,
3194 errors from getrlimit() and setrlimit() were being ignored. Now they are
3195 logged to the main and panic logs, but Exim does carry on, to try to do its
3196 job under whatever limits there are.
3198 18. Imported PCRE 5.0.
3200 19. Trivial typo in log message " temporarily refused connection" (the leading
3203 20. If the log selector return_path_on_delivery was set and an address was
3204 redirected to /dev/null, the delivery process crashed because it assumed
3205 that a return path would always be set for a "successful" delivery. In this
3206 case, the whole delivery is bypassed as an optimization, and therefore no
3209 21. Internal re-arrangement: the function for sending a challenge and reading
3210 a response while authentication was assuming a zero-terminated challenge
3211 string. It's now changed to take a pointer and a length, to allow for
3212 binary data in such strings.
3214 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3216 23. Exim was not respecting finduser_retries when seeking the login of the
3217 uid under which it was called; it was always trying 10 times. (The default
3218 setting of finduser_retries is zero.) Also, it was sleeping after the final
3219 failure, which is pointless.
3221 24. Implemented tls_on_connect_ports.
3223 25. Implemented acl_smtp_predata.
3225 26. If the domain in control=submission is set empty, Exim assumes that the
3226 authenticated id is a complete email address when it generates From: or
3227 Sender: header lines.
3229 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3230 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3231 chown and chgrp in /bin and hostname in /usr/bin.
3233 28. Exim was keeping the "process log" file open after each use, just as it
3234 does for the main log. This opens the possibility of it remaining open for
3235 long periods when the USR1 signal hits a daemon. Occasional processlog
3236 errors were reported, that could have been caused by this. Anyway, it seems
3237 much more sensible not to leave this file open at all, so that is what now
3240 29. The long-running daemon process does not normally write to the log once it
3241 has entered its main loop, and it closes the log before doing so. This is
3242 so that log files can straightforwardly be renamed and moved. However,
3243 there are a couple of unusual error situations where the daemon does write
3244 log entries, and I had neglected to close the log afterwards.
3246 30. The text of an SMTP error response that was received during a remote
3247 delivery was being truncated at 512 bytes. This is too short for some of
3248 the long messages that one sometimes sees. I've increased the limit to
3251 31. It is now possible to make retry rules that apply only when a message has a
3252 specific sender, in particular, an empty sender.
3254 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3255 it possible to be selective about when SMTP synchronization is enforced.
3257 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3259 32. Implemented hosts_connection_nolog.
3261 33. Added an ACL for QUIT.
3263 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3266 35. Added mailbox_size and mailbox_filecount to appendfile.
3268 36. Added control = no_multiline_responses to ACLs.
3270 37. There was a bug in the logic of the code that waits for the clock to tick
3271 in the case where the clock went backwards by a substantial amount such
3272 that the microsecond fraction of "now" was more than the microsecond
3273 fraction of "then" (but the whole seconds number was less).
3275 38. Added support for the libradius Radius client library this is found on
3276 FreeBSD (previously only the radiusclient library was supported).
3282 1. When certain lookups returned multiple values in the form name=value, the
3283 quoting of the values was not always being done properly. Specifically:
3284 (a) If the value started with a double quote, but contained no whitespace,
3286 (b) If the value contained whitespace other than a space character (i.e.
3287 tabs or newlines or carriage returns) it was not quoted.
3288 This fix has been applied to the mysql and pgsql lookups by writing a
3289 separate quoting function and calling it from the lookup code. The fix
3290 should probably also be applied to nisplus, ibase and oracle lookups, but
3291 since I cannot test any of those, I have not disturbed their existing code.
3293 2. A hit in the callout cache for a specific address caused a log line with no
3294 reason for rejecting RCPT. Now it says "Previous (cached) callout
3295 verification failure".
3297 3. There was an off-by-one bug in the queryprogram router. An over-long
3298 return line was truncated at 256 instead of 255 characters, thereby
3299 overflowing its buffer with the terminating zero. As well as fixing this, I
3300 have increased the buffer size to 1024 (and made a note to document this).
3302 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3303 when Exim is waiting for an SMTP response from a remote server, Exim
3304 restarts its select() call on the socket, thereby resetting its timeout.
3305 This is not a problem when such interrupts are rare. Somebody set up a cron
3306 job to run exiwhat every 2 minutes, which is less than the normal select()
3307 timeout (5 or 10 minutes). This meant that the select() timeout never
3308 kicked in because it was always reset. I have fixed this by comparing the
3309 time when an interrupt arrives with the time at the start of the first call
3310 to select(). If more time than the timeout has elapsed, the interrupt is
3311 treated as a timeout.
3313 5. Some internal re-factoring in preparation for the addition of Sieve
3314 extensions (by MH). In particular, the "personal" test is moved to a
3315 separate function, and given an option for scanning Cc: and Bcc: (which is
3316 not set for Exim filters).
3318 6. When Exim created an email address using the login of the caller as the
3319 local part (e.g. when creating a From: or Sender: header line), it was not
3320 quoting the local part when it contained special characters such as @.
3322 7. Installed new OpenBSD configuration files.
3324 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3325 try to make them clearer.
3327 9. Callout options, other than the timeout value, were being ignored when
3328 verifying sender addresses in header lines. For example, when using
3330 verify = header_sender/callout=no_cache
3332 the cache was (incorrectly) being used.
3334 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3335 only the Cygwin environment.
3337 11. When return_path_on_delivery was set as a log selector, if different remote
3338 addresses in the same message used different return paths and parallel
3339 remote delivery occurred, the wrong values would sometimes be logged.
3340 (Whenever a remote delivery process finished, the return path value from
3341 the most recently started remote delivery process was logged.)
3343 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3344 lines when AUTH and/or TLS are in use. This is the "received protocol"
3345 field. Exim used to use "asmtp" for authenticated SMTP, without any
3346 indication (in the protocol name) for TLS use. Now it follows the RFC and
3347 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3348 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3349 names appear in log lines as well as in Received: header lines.
3351 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3352 extensions, and comparison tests, and to fix some bugs.
3354 14. Changes to the "personal" filter test:
3356 (1) The test was buggy in that it was just doing the equivalent of
3357 "contains" tests on header lines. For example, if a user's address was
3358 anne@some.where, the "personal" test would incorrectly be true for
3360 To: susanne@some.where
3362 This test is now done by extracting each address from the header in turn,
3363 and checking the entire address. Other tests that are part of "personal"
3364 are now done using regular expressions (for example, to check local parts
3365 of addresses in From: header lines).
3367 (2) The list of non-personal local parts in From: addresses has been
3368 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3369 taken from the Sieve specification recommendations.
3371 (3) If the message contains any header line starting with "List-" it is
3372 treated as non-personal.
3374 (4) The test for "circular" in the Subject: header line has been removed
3375 because it now seems ill-conceived.
3377 15. Minor typos in src/EDITME comments corrected.
3379 16. Installed latest exipick from John Jetmore.
3381 17. If headers_add on a router specified a text string that was too long for
3382 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3383 of string_sprintf() is now avoided.
3385 18. $message_body_size was not set (it was always zero) when running the DATA
3386 ACL and the local_scan() function.
3388 19. For the "mail" command in an Exim filter, no default was being set for
3389 the once_repeat time, causing a random time value to be used if "once" was
3390 specified. (If the value happened to be <= 0, no repeat happened.) The
3391 default is now 0s, meaning "never repeat". The "vacation" command was OK
3392 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3393 (I found it when inspecting the code).
3395 20. There is now an overall timeout for performing a callout verification. It
3396 defaults to 4 times the callout timeout, which applies to individual SMTP
3397 commands during the callout. The overall timeout applies when there is more
3398 than one host that can be tried. The timeout is checked before trying the
3399 next host. This prevents very long delays if there are a large number of
3400 hosts and all are timing out (e.g. when the network connections are timing
3401 out). The value of the overall timeout can be changed by specifying an
3402 additional sub-option for "callout", called "maxwait". For example:
3404 verify = sender/callout=5s,maxwait=20s
3406 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3407 to the end before writing, but this should make it even safer).
3409 22. Exim was forgetting that it had advertised PIPELINING for the second and
3410 subsequent messages on an SMTP connection. It was also not resetting its
3411 memory on STARTTLS and an internal HELO.
3413 23. When Exim logs an SMTP synchronization error within a session, it now
3414 records whether PIPELINING has been advertised or not.
3416 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3417 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3418 rather than long int.
3420 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3422 26. Added the never_mail option to autoreply.
3428 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3429 crash if the getsockname() call failed; this can happen if a connection is
3430 closed very soon after it is established. The problem was simply in the
3431 order in which certain operations were done, causing Exim to try to write
3432 to the SMTP stream before it had set up the file descriptor. The bug has
3433 been fixed by making things happen in the correct order.
3439 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3440 before the connection was closed, thus losing the rejection response.
3442 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3443 some early Solaris releases, but causes trouble in current releases where
3444 socklen_t is defined.
3446 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3449 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3452 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3453 that releasing the top of it at the end releases what was used for sub-
3454 expansions (unless the block got too big). However, discard this block if
3455 the first thing is a variable or header, so that we can use its block when
3456 it is dynamic (useful for very large $message_headers, for example).
3458 6. Lookups now cache *every* query, not just the most recent. A new, separate
3459 store pool is used for this. It can be recovered when all lookup caches are
3460 flushed. Lookups now release memory at the end of their result strings.
3461 This has involved some general refactoring of the lookup sources.
3463 7. Some code has been added to the store_xxx() functions to reduce the amount
3464 of flapping under certain conditions.
3466 8. log_incoming_interface used to affect only the <= reception log lines. Now
3467 it causes the local interface and port to be added to several more SMTP log
3468 lines, for example "SMTP connection from", and rejection lines.
3470 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3472 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3474 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3475 could be overwritten at the end of the current message (or the start of a
3476 new message if it was set in a HELO ACL). The value is now preserved for
3477 the duration of the SMTP connection.
3479 12. If a transport had a headers_rewrite setting, and a matching header line
3480 contained an unqualified address, that address was qualified, even if it
3481 did not match any rewriting rules. The underlying bug was that the values
3482 of the flags that permit the existence of unqualified sender and recipient
3483 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3484 non-local messages, and by -bnq for local messages) were not being
3485 preserved with the message after it was received.
3487 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3488 "next input=" as part of the text comprising the host identity instead of
3489 the correct text. The code was using the same buffer for two different
3490 strings. However, depending on which order the printing function evaluated
3491 its arguments, the bug did not always show up. Under Linux, for example, my
3492 test suite worked just fine.
3494 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3495 doesn't work with some older versions of Perl. It has been changed to "my",
3496 which in any case is probably the better facility to use.
3498 15. A really picky compiler found some instances of statements for creating
3499 error messages that either had too many or two few arguments for the format
3502 16. The size of the buffer for calls to the DNS resolver has been increased
3503 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3504 for addresses that have a lot of PTR records. This alleviates a problem; it
3505 does not fully solve it.
3507 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3508 buffer now truncates the list and logs the incident, which is the same
3509 action as happens when Exim is looking up a host name and its aliases.
3510 Previously in this situation something unpredictable would happen;
3511 sometimes it was "internal error: store_reset failed".
3513 18. If a server dropped the connection unexpectedly when an Exim client was
3514 using GnuTLS and trying to read a response, the client delivery process
3515 crashed while trying to generate an error log message.
3517 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3518 string, for example:
3520 warn message = H1: something\nH2: something
3522 the text was added as a single header line from Exim's point of view
3523 though it ended up OK in the delivered message. However, searching for the
3524 second and subsequent header lines using $h_h2: did not work. This has been
3525 fixed. Similarly, if a system filter added multiple headers in this way,
3526 the routers could not see them.
3528 20. Expanded the error message when iplsearch is called with an invalid key to
3529 suggest using net-iplsearch in a host list.
3531 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3532 ACLs are no longer actually imposed (and a message to that effect is
3535 22. If a "gecos" field in a passwd entry contained escaped characters, in
3536 particular, if it contained a \" sequence, Exim got it wrong when building
3537 a From: or a Sender: header from that name. A second bug also caused
3538 incorrect handling when an unquoted " was present following a character
3539 that needed quoting.
3541 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3542 was not being matched caselessly.
3544 24. Arranged for all hyphens in the exim.8 source to be escaped with
3547 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3548 properly. Recipient callout cache records were still being keyed to include
3549 the sender, even when use_sender was set false. This led to far more
3550 callouts that were necessary. The sender is no longer included in the key
3551 when use_sender is false.
3553 26. Added "control = submission" modifier to ACLs.
3555 27. Added the ${base62d: operator to decode base 62 numbers.
3557 28. dnsdb lookups can now access SRV records.
3559 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3560 the configuration file.
3562 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3563 (-v) mode. This makes the output for a verbose queue run more intelligible.
3565 31. Added a use_postmaster feature to recipient callouts.
3567 32. Added the $body_zerocount variable, containing the number of binary zero
3568 bytes in the message body.
3570 33. The time of last modification of the "new" subdirectory is now used as the
3571 "mailbox time last read" when there is a quota error for a maildir
3574 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3576 35. Added +ignore_unknown as a special item in host lists.
3578 36. Code for decoding IPv6 addresses in host lists is now included, even if
3579 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3580 address was recognized as an IP address, but was then not correctly decoded
3581 into binary, causing unexpected and incorrect effects when compared with
3588 1. Very minor rewording of debugging text in manualroute to say "list of
3589 hosts" instead of "hostlist".
3591 2. If verify=header_syntax was set, and a header line with an unqualified
3592 address (no domain) and a large number of spaces between the end of the
3593 name and the colon was received, the reception process suffered a buffer
3594 overflow, and (when I tested it) crashed. This was caused by some obsolete
3595 code that should have been removed. The fix is to remove it!
3597 3. When running in the test harness, delay a bit after writing a bounce
3598 message to get a bit more predictability in the log output.
3600 4. Added a call to search_tidyup() just before forking a reception process. In
3601 theory, someone could use a lookup in the expansion of smtp_accept_max_
3602 per_host which, without the tidyup, could leave open a database connection.
3604 5. Added the variables $recipient_data and $sender_data which get set from a
3605 lookup success in an ACL "recipients" or "senders" condition, or a router
3606 "senders" option, similar to $domain_data and $local_part_data.
3608 6. Moved the writing of debug_print from before to after the "senders" test
3611 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3612 problems for message scanning, either using a data ACL, or using
3613 local_scan() because the Received: header was not generated till after they
3614 were called (in order to set the time as the time of reception completion).
3615 I have revised the way this works. The header is now generated after the
3616 body is received, but before the ACL or local_scan() are called. After they
3617 are run, the timestamp in the header is updated.
3623 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3624 before starting a queue runner without re-exec. This happened only when
3625 deliver_drop_privilege was set or when the Exim user was set to root. The
3626 effect of the bug was that timeouts during subsequent deliveries caused
3627 crashes instead of being properly handled. The handler is now left at its
3628 default (and expected) setting.
3630 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3631 message, again when deliver_drop_privilege is set or Exim is run as root.
3632 The bug described in (1) was not present in this case, but the tidying up
3633 of the other signals was missing. I have made the two cases consistent.
3635 3. The ignore_target_hosts setting on a manualroute router was being ignored
3636 for hosts that were looked up using the /MX notation.
3638 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3641 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3642 operated on the sender address. After changing the $sender_address to <>
3643 for the sender address verify, Exim was re-instated it as the original
3644 (before rewriting) address, but remembering that it had rewritten it, so it
3645 wasn't rewriting it again. This bug also had the effect of breaking the
3646 sender address verification caching when the sender address was rewritten.
3648 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3649 This has been changed so that if the ip literal address matches
3650 ignore_target_hosts, the router declines.
3652 7. Added expansion conditions match_domain, match_address, and match_local_
3653 part (NOT match_host).
3655 8. The placeholder for the Received: header didn't have a length field set.
3657 9. Added code to Exim itself and to exim_lock to test for a specific race
3658 condition that could lead to file corruption when using MBX delivery. The
3659 issue is with the lockfile that is created in /tmp. If this file is removed
3660 after a process has opened it but before that process has acquired a lock,
3661 there is the potential for a second process to recreate the file and also
3662 acquire a lock. This could lead to two Exim processes writing to the file
3663 at the same time. The added code performs the same test as UW imapd; it
3664 checks after acquiring the lock that its file descriptor still refers to
3665 the same named file.
3667 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3668 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3671 11. Added the smtp_active_hostname option. If used, this will typically be made
3672 to depend on the incoming interface address. Because $interface_address is
3673 not set up until the daemon has forked a reception process, error responses
3674 that can happen earlier (such as "too many connections") no longer contain
3677 12. If an expansion in a condition on a "warn" statement fails because a lookup
3678 defers, the "warn" statement is abandoned, and the next ACL statement is
3679 processed. Previously this caused the whole ACL to be aborted.
3681 13. Added the iplsearch lookup type.
3683 14. Added ident_timeout as a log selector.
3685 15. Added tls_certificate_verified as a log selector.
3687 16. Added a global option tls_require_ciphers (compare the smtp transport
3688 option of the same name). This controls incoming TLS connections.
3690 17. I finally figured out how to make tls_require_ciphers do a similar thing
3691 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3692 before starting the TLS session.
3694 18. Tabs are now shown as \t in -bP output.
3696 19. If the log selector return_path_on_delivery was set, Exim crashed when
3697 bouncing a message because it had too many Received: header lines.
3699 20. If two routers both had headers_remove settings, and the first one included
3700 a superfluous trailing colon, the final name in the first list and the
3701 first name in the second list were incorrectly joined into one item (with a
3702 colon in the middle).
3708 1. Added -C and -D options to the exinext utility, mainly to make it easier
3709 to include in the automated testing, but these could be helpful when
3710 multiple configurations are in use.
3712 2. The exinext utility was not formatting the output nicely when there was
3713 an alternate port involved in the retry record key, nor when there was a
3714 message id as well (for retries that were specific to a specific message
3715 and a specific host). It was also confused by IPv6 addresses, because of
3716 the additional colons they contain. I have fixed the IPv4 problem, and
3717 patched it up to do a reasonable job for IPv6.
3719 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3720 delivery, the log line now contains "pipelined" if PIPELINING was used.
3722 4. An SMTP transport process used to panic and die if the bind() call to set
3723 an explicit outgoing interface failed. This has been changed; it is now
3724 treated in the same way as a connect() failure.
3726 5. A reference to $sender_host_name in the part of a conditional expansion
3727 that was being skipped was still causing a DNS lookup. This no longer
3730 6. The def: expansion condition was not recognizing references to header lines
3731 that used bh_ and bheader_.
3733 7. Added the _cache feature to named lists.
3735 8. The code for checking quota_filecount in the appendfile transport was
3736 allowing one more file than it should have been.
3738 9. For compatibility with Sendmail, the command line option
3746 and sets the incoming protocol and host name (for trusted callers). The
3747 host name and its colon can be omitted when only the protocol is to be set.
3748 Note the Exim already has two private options, -pd and -ps, that refer to
3749 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3750 "s", but I don't think that's a major issue.
3752 10. A number of refactoring changes to the code, none of which should affect
3755 (a) The number of logging options was getting close to filling up the
3756 32-bit word that was used as a bit map. I have split them into two classes:
3757 those that are passed in the argument to log_write(), and those that are
3758 only ever tested independently outside of that function. These are now in
3759 separate 32-bit words, so there is plenty of room for expansion again.
3760 There is no change in the user interface or the logging behaviour.
3762 (b) When building, for example, log lines, the code previously used a
3763 macro that called string_cat() twice, in order to add two strings. This is
3764 not really sufficiently general. Furthermore, there was one instance where
3765 it was actually wrong because one of the argument was used twice, and in
3766 one call a function was used. (As it happened, calling the function twice
3767 did not affect the overall behaviour.) The macro has been replaced by a
3768 function that can join an arbitrary number of extra strings onto a growing
3771 (c) The code for expansion conditions now uses a table and a binary chop
3772 instead of a serial search (which was left over from when there were very
3773 few conditions). Also, it now recognizes conditions like "pam" even when
3774 the relevant support is not compiled in: a suitably worded error message is
3775 given if an attempt is made to use such a condition.
3777 11. Added ${time_interval:xxxxx}.
3779 12. A bug was causing one of the ddress fields not to be passed back correctly
3780 from remote delivery subprocesses. The field in question was not being
3781 subsequently used, so this caused to problems in practice.
3783 13. Added new log selectors queue_time and deliver_time.
3785 14. Might have fixed a bug in maildirsizefile handling that threw up
3786 "unexpected character" debug warnings, and recalculated the data
3787 unnecessarily. In any case, I expanded the warning message to give more
3790 15. Added the message "Restricted characters in address" to the statements in
3791 the default ACL that block characters like @ and % in local parts.
3793 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3794 Three changes have been made:
3796 (a) There was a serious bug; a negative response to MAIL caused the whole
3797 recipient domain to be cached as invalid, thereby blocking all messages
3798 to all local parts at the same domain, from all senders. This bug has
3799 been fixed. The domain is no longer cached after a negative response to
3800 MAIL if the sender used is not empty.
3802 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3805 (c) A new callout option, "use_sender" has been added for people who want
3806 the modified behaviour.
3812 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3815 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3816 indeed breaks things for older releases.
3818 3. Added additional logging to the case where there is a problem reading data
3819 from a filter that is running in a subprocess using a pipe, in order to
3820 try to track down a specific problem.
3822 4. Testing facility fudge: when running in the test harness and attempting
3823 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3824 getting "No route to host". Convert this to a timeout.
3826 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3829 6. Some OS don't have socklen_t but use size_t instead. This affects the
3830 fifth argument of getsockopt() amongst other things. This is now
3831 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3832 can be set for individual OS. I have set it for SunOS5, OSF1, and
3833 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3834 some earlier ones do not.
3836 7. Change 4.30/15 was not doing the test caselessly.
3838 8. The standard form for an IPv6 address literal was being rejected by address
3839 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3840 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3841 this, as well as the form without the "IPv6" on the front (but only when
3842 address literals are enabled, of course).
3844 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3846 10. Exim crashed if a message with an empty sender address specified by -f
3847 encountered a router with an errors_to setting. This could be provoked only
3848 by a command such as
3852 where an empty string was supplied; "<>" did not hit this bug.
3854 11. Installed PCRE release 4.5.
3856 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3857 remained set. It is now erased.
3859 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3860 times from message ids (which are base 36 rather than the normal 62).
3862 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3863 were being counted as actual protocol errors, and logged if the log
3864 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3865 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3866 and DATA following a set of rejected RCPTs do not count as protocol errors.
3867 In other words, Exim assumes they were pipelined, though this may not
3868 actually be the case. Of course, in all cases the client gets an
3869 appropriate error code.
3871 15. If a lookup fails in an ACL condition, a message about the failure may
3872 be available; it is used if testing the ACL cannot continue, because most
3873 such messages specify what the cause of the deferral is. However, some
3874 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3875 that caused an old message to be retained and used if a later statement
3876 caused a defer, replacing the real cause of the deferral.
3878 16. If an IP address had so many PTR records that the DNS lookup buffer
3879 was not large enough to hold them, Exim could crash while trying to process
3880 the truncated data. It now detects and logs this case.
3882 17. Further to 4.21/58, another change has been made: if (and only if) the
3883 first line of a message (the first header line) ends with CRLF, a bare LF
3884 in a subsequent header line has a space inserted after it, so as not to
3885 terminate the header.
3887 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3888 unnecessarily, used atoi() instead of strtol(), and didn't check the
3889 termination when getting file sizes from file names by regex.
3891 19. Completely re-implemented the support for maildirsize files, in the light
3892 of a number of problems with the previous contributed implementation
3893 (4.30/29). In particular:
3895 . If the quota is zero, the maildirsize file is maintained, but no quota is
3898 . If the maildir directory does not exist, it is created before any attempt
3899 to write a maildirsize file.
3901 . The quota value in the file is just a cache; if the quota is changed in
3902 the transport, the new value overrides.
3904 . A regular expression is available for excluding directories from the
3907 20. The autoreply transport checks the characters in options that define the
3908 message's headers; it allows continued headers, but it was checking with
3909 isspace() after an embedded newline instead of explicitly looking for a
3912 21. If all the "regular" hosts to which an address was routed had passed their
3913 expiry times, and had not reached their retry times, the address was
3914 bounced, even if fallback hosts were defined. Now Exim should go on to try
3917 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3918 equivalent code in the SMTP transport. Some hosts send humungous responses
3919 to HELO/EHLO, more than 1024 it seems.
3921 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3922 gives compiler warnings in some environments. I've now done it "properly",
3925 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3926 (because of problems with the built-in one) was declared to return uschar *
3927 instead of char *, causing compiler failure.
3929 25. Fixed a file descriptor leak when processing alias/forward files.
3931 26. Fixed a minor format string issue in dbfn.c.
3933 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3935 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3936 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3939 29. When a sender address is verified, it is cached, to save repeating the test
3940 when there is more than one recipient in a message. However, when the
3941 verification involves a callout, it is possible for different callout
3942 options to be set for different recipients. It is too complicated to keep
3943 track of this in the cache, so now Exim always runs a verification when a
3944 callout is required, relying on the callout cache for the optimization.
3945 The overhead is duplication of the address routing, but this should not be
3948 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3949 to be verified with callout=postmaster, and the main callout worked but the
3950 postmaster check failed, the verification correctly failed. However, if a
3951 subsequent RCPT command asked for sender verification *without* the
3952 postmaster check, incorrect caching caused this verification also to fail,
3955 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3956 it was not caching the DNS options (qualify_single, search_parents) that
3957 were used when the lookup failed. A subsequent lookup with different
3958 options therefore always gave the same answer, though there were cases
3959 where it should not have. (Example: a "domains = !$mx_any" option on a
3960 dnslookup router: the "domains" option is always processed without any
3961 widening, but the router might have qualify_single set.) Now Exim uses the
3962 cached value only when the same options are set.
3964 32. Added John Jetmore's "exipick" utility to the distribution.
3966 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3967 than a timeout (e.g. a certificate is required, and is not provided), an
3968 Exim server now closes the connection immediately. Previously it waited for
3969 the client to close - but if the client is SSL, it seems that they each
3970 wait for each other, leading to a delay before one of them times out.
3972 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3973 maintained 0.8.x compatibility because I don't think many are using it, and
3974 it is clearly obsolete.
3976 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3979 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3980 client certificate was expired. A simple patch fixes this, though I don't
3981 understand the full logic of why the verify callback is called multiple
3984 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3985 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3986 which causes problems with some clients (such as the Certicom SSL Plus
3987 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3988 disables the coutermeasure allowing Eudora to connect."
3990 38. Exim was not checking that a write() to a log file succeeded. This could
3991 lead to Bad Things if a log got too big, in particular if it hit a file
3992 size limit. Exim now panics and dies if it cannot write to a log file, just
3993 as it does if it cannot open a log file.
3995 39. Modified OS/Makefile-Linux so that it now contains
3997 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3999 The two -D definitions ensure that Exim is compiled with large file
4000 support, which makes it possible to handle log files that are bigger than
4003 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4004 instance) a domain was checked against a named list that involved a lookup,
4005 causing $domain_data to be set, then another domain was checked against the
4006 same list, then the first domain was re-checked, the value of $domain_data
4007 after the final check could be wrong. In particular, if the second check
4008 failed, it could be set empty. This bug probably also applied to
4011 41. The strip_trailing_dot option was not being applied to the address given
4012 with the -f command-line option.
4014 42. The code for reading a message's header from the spool was incrementing
4015 $received_count, but never initializing it. This meant that the value was
4016 incorrect (doubled) while delivering a message in the same process in which
4017 it was received. In the most common configuration of Exim, this never
4018 happens - a fresh exec is done - but it can happen when
4019 deliver_drop_privilege is set.
4021 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4022 it now includes up to 150 characters of the unexpected data in the log
4025 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4026 and building data strings. The size of both of these buffers was 10 000
4027 bytes - far larger than anybody would *ever* want, thought I. Needless to
4028 say, somebody hit the limit. I have increased the maximum line length to
4029 20 000 and the maximum data length of concatenated lines to 100 000. I have
4030 also fixed two bugs, because there was no checking on these buffers. Tsk,
4031 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4032 buffer is too small.
4034 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4035 lsearch lookups. Now it does.
4037 46. When parsing a route_list item in a manualroute router, a fixed-length
4038 buffer was used for the list of hosts. I made this 1024 bytes long,
4039 thinking that nobody would ever have a list of hosts that long. Wrong.
4040 Somebody had a whole pile of complicated expansion conditions, and the
4041 string was silently truncated, leading to an expansion error. It turns out
4042 that it is easier to change to an unlimited length (owing to other changes
4043 that have happened since this code was originally written) than to build
4044 structure for giving a limitation error. The length of the item that
4045 expands into the list of hosts is now unlimited.
4047 47. The lsearch lookup could not handle data where the length of text line was
4048 more than 4095 characters. Such lines were truncated, leading to shortened
4049 data being returned. It should now handle lines of any length.
4051 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4052 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4055 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4056 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4057 no longer happens in comment lines. A list of replacements is now placed
4058 at the head of all of the source files, except those whose only change is
4059 to replace PERL_COMMAND in the very first #! line.
4061 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4062 messages on the queue, with a bottom-up merge sort, using code contributed
4063 by Michael Haardt. This should make operations like -bp somewhat faster on
4064 large queues. It won't affect queue runners, except when queue_run_in_order
4067 51. Installed eximstats 1.31 in the distribution.
4069 52. Added support for SRV lookups to the dnslookup router.
4071 53. If an ACL referred to $message_body or $message_body_end, the value was not
4072 reset for any messages that followed in the same SMTP session.
4074 54. The store-handling optimization for building very long strings was not
4075 differentiating between the different store pools. I don't think this
4076 actually made any difference in practice, but I've tidied it.
4078 55. While running the routers to verify a sender address, $sender_address
4079 was still set to the sender address. This is wrong, because when routing to
4080 send a bounce to the sender, it would be empty. Therefore, I have changed
4081 it so that, while verifying a sender address, $sender_address is set to <>.
4082 (There is no change to what happens when verifying a recipient address.)
4084 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4085 target A or AAAA records (if not already returned) without resetting the
4086 qualify_single or search_parents options of the DNS resolver. These are
4087 inappropriate in this case because the targets of MX and SRV records must
4088 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4089 target that, when qualified, matched something in the local domain. These
4090 two options are now turned off when doing these lookups.
4092 57. It seems that at least some releases of Reiserfs (which does not have the
4093 concept of a fixed number of inodes) returns zero and not -1 for the
4094 number of available inodes. This interacted badly with check_spool_inodes,
4095 which assumed that -1 was the "no such thing" setting. What I have done is
4096 to check that the total number of inodes is greater than zero before doing
4097 the test of how many are available.
4099 58. When a "warn" ACL statement has a log_message modifier, the message is
4100 remembered, and not repeated. This is to avoid a lot of repetition when a
4101 message has many recipients that cause the same warning to be written.
4102 Howewer, Exim was preserving the list of already written lines for an
4103 entire SMTP session, which doesn't seem right. The memory is now reset if a
4104 new message is started.
4106 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4107 debugging output unless log_rewrite was also set.
4109 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4110 of (int)(handle) when we know that handle contains (void *)(-1).
4112 61. The Exim daemon panic-logs an error return when it closes the incoming
4113 connection. However "connection reset by peer" seems to be common, and
4114 isn't really an error worthy of noting specially, so that particular error
4117 62. When Exim is trying to find all the local interfaces, it used to panic and
4118 die if the ioctl to get the interface flags failed. However, it seems that
4119 on at least one OS (Solaris 9) it is possible to have an interface that is
4120 included in the list of interfaces, but for which you get a failure error
4121 for this call. This happens when the interface is not "plumbed" into a
4122 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4123 failure of the "get flags" call assumes that the interface is down.
4125 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4126 makes life easier for people who are doing arithmetic on fields extracted
4127 from dates, where you often get leading zeros that should not be
4128 interpreted as octal.
4130 64. Added qualify_domain to the redirect router, to override the global
4133 65. If a pathologically long header line contained very many addresses (the
4134 report of this problem mentioned 10 000) and each of them was rewritten,
4135 Exim could use up a very large amount of memory. (It kept on making new
4136 copies of the header line as it rewrote, and never released the old ones.)
4137 At the expense of a bit more processing, the header rewriting function has
4138 been changed so that it no longer eats memory in this way.
4140 66. The generation of the Received: header has been moved from the time that a
4141 message starts to be received, to the time that it finishes. The timestamp
4142 in the Received: header should now be very close to that of the <= log
4143 line. There are two side-effects of this change:
4145 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4146 logged header lines no longer include the local Received: line, because
4147 it has not yet been created. The same applies to a copy of the message
4148 that is returned to a non-SMTP sender when a message is rejected.
4150 (b) When a filter file is tested using -bf, no additional Received: header
4151 is added to the test message. After some thought, I decided that this
4154 This change does not affect the value of $received_for. It is still set
4155 after address rewriting, but before local_scan() is called.
4157 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4159 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4160 gave an unhelpful panic error message, and a defer error. I have managed to
4161 change this behaviour so that it now rejects any supplied certificate,
4162 which seems right, as the list of acceptable certificates is empty.
4164 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4165 gave an unhelpful defer error. I have not managed to make this reject any
4166 supplied certificates, but the error message it gives is "no certificate
4167 supplied", which is not helpful.
4169 70. exigrep's output now also includes lines that are not associated with any
4170 message, but which match the given pattern. Implemented by a patch from
4171 Martin Sluka, which also tidied up the Perl a bit.
4173 71. Recipient callout verification, like sender verification, was using <> in
4174 the MAIL FROM command. This isn't really the right thing, since the actual
4175 sender may affect whether the remote host accepts the recipient or not. I
4176 have changed it to use the actual sender in the callout; this means that
4177 the cache record is now keyed on a recipient/sender pair, not just the
4178 recipient address. There doesn't seem to be a real danger of callout loops,
4179 since a callout by the remote host to check the sender would use <>.
4180 [SEE ABOVE: changed after hitting problems.]
4182 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4183 temporary errors. However, in the case of such a code being given after
4184 the end of a data transmission (i.e. after ".") Exim was failing to write
4185 a retry record for the message. (Yes, there was some broken host that was
4186 actually sending 8xx at this point.)
4188 73. An unknown lookup type in a host list could cause Exim to panic-die when
4189 the list was checked. (An example that provoked this was putting <; in the
4190 middle of a list instead of at the start.) If this happened during a DATA
4191 ACL check, a -D file could be left lying around. This kind of configuration
4192 error no longer causes Exim to die; instead it causes a defer errror. The
4193 incident is still logged to the main and panic logs.
4195 74. Buglet left over from Exim 3 conversion. The message "too many messages
4196 in one connection" was written to the rejectlog but not the mainlog, except
4197 when address rewriting (yes!) was being logged.
4199 75. Added write_rejectlog option.
4201 76. When a system filter was run not as root (that is, when system_filter_user
4202 was set), the values of the $n variables were not being returned to the
4203 main process; thus, they were not subsequently available in the $sn
4206 77. Added +return_path_on_delivery log selector.
4208 78. A connection timeout was being treated differently from recipients deferred
4209 when testing hosts_max_try with a message that was older than the host's
4210 retry timeout. (The host should not be counted, thus allowing all hosts to
4211 be tried at least once before bouncing.) This may have been the cause of an
4212 occasionally reported bug whereby a message would remain on the queue
4213 longer than the retry timeout, but would be bounced if a delivery was
4214 forced. I say "may" because I never totally pinned down the problem;
4215 setting up timeout/retry tests is difficult. See also the next item.
4217 79. The ultimate address timeout was not being applied to errors that involved
4218 a combination of host plus message (for example, a timeout on a MAIL
4219 command). When an address resolved to a number of possible hosts, and they
4220 were not all tried for each delivery (e.g. because of hosts_max_try), a
4221 message could remain on the queue longer than the retry timeout.
4223 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4226 81. Fixed an obscure SMTP outgoing bug which required at least the following
4227 conditions: (a) there was another message waiting for the same server;
4228 (b) the server returned 5xx to all RCPT commands in the first message so
4229 that the message was not completed; (c) the server dropped the connection
4230 or gave a negative response to the RSET that Exim sends to abort the
4231 transaction. The observed case was a dropped connection after DATA that had
4232 been sent in pipelining mode. That is, the server had advertised PIPELINING
4233 but was not implementing it correctly. The effect of the bug was incorrect
4234 behaviour, such as trying another host, and this could lead to a crash.
4240 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4241 and daemon.c were passed as pointers to ints; they should have been
4242 pointers to socklen_t variables (which are typically unsigned ints).
4244 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4247 3. Fixed a really odd bug that affected only the testing scheme; patching a
4248 certain fixed string in the binary changed the value of another string that
4249 happened to be identical to the end of the original first string.
4251 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4252 name", it returns that address as the IP address. On some operating
4253 systems (e.g. Solaris), it also passes back the IP address string as the
4254 "host name". However, on others (e.g. Linux), it passes back an empty
4255 string. Exim wasn't checking for this, and was changing the host name to an
4256 empty string, assuming it had been canonicized.
4258 5. Although rare, it is permitted to have more than one PTR record for a given
4259 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4260 all the names associated with an address, because they do in Solaris.
4261 However, it seems that they do not in Linux for data that comes from the
4262 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4263 I found this out when I moved to a new Linux workstation and tried to run
4264 the Exim test suite.
4266 To get round this problem I have changed the code so that it now does its
4267 own call to the DNS to look up PTR records when searching for a host name.
4268 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4269 addresses that are only in /etc/hosts are still found.
4271 This behaviour is, however, controlled by an option called host_lookup_
4272 order, which defaults to "bydns:byaddr". If people want to use the other
4273 order, or indeed, just use one or the other means of lookup, they can
4274 specify it in this variable.
4276 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4277 some operating systems, this comes back from gethostbyaddr() as an empty
4278 string, and this is what Exim used to test for. However, it seems that in
4279 other systems, "." is yielded. Exim now tests for this case too.
4281 7. The values of check_spool_space and check_log_space are now held internally
4282 as a number of kilobytes instead of an absolute number of bytes. If a
4283 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4284 kilobyte. This means that much larger values can be stored.
4286 8. Exim monitor: an attempt to get the action menu when not actually pointing
4287 at a message produces an empty menu entitled "No message selected". This
4288 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4289 no entries in it ("Shell widget menu has zero width and/or height"). So I
4290 have added a single, blank menu entry in this case.
4292 9. Added ${quote_local_part.
4294 10. MIME decoding is now applied to the contents of Subject: header lines when
4297 11. Now that a reference to $sender_host_address automatically causes a reverse
4298 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4299 host lookup before query-style lookups in lists that might use this
4300 variable. This has therefore been abolished, and the "net-" prefix is no
4301 longer necessary for query-style lookups.
4303 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4304 have been a typo for LFLAGS, so it has been changed.
4306 13. The install script calls Exim with "-C /dev/null" in order to find the
4307 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4308 to be output. Howeve, since Exim outputs its version number before the
4309 error, it didn't break the script. It just looked ugly. I fixed this by
4310 always allowing "-C /dev/null" if the caller is root.
4312 14. Ignore overlarge ACL variable number when reading spool file - insurance
4313 against a later release with more variables having written the file.
4315 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4316 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4317 this, as well as the form without the "IPv6" on the front.
4319 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4320 OS/Makefile-Darwin file.
4322 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4323 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4325 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4326 variable. However, some OS use ints for inodes. I've added cast to long int
4327 to get rid of the compiler warning.
4329 19. I had forgotten to lock out "/../" in configuration file names when
4330 ALT_CONFIG_PREFIX was set.
4332 20. Routers used for verification do not need to specify transports. However,
4333 if such a router generated a host list, and callout was configured, Exim
4334 crashed, because it could not find a port number from the (non-existent)
4335 transport. It now assumes port 25 in this circumstance.
4337 21. Added the -t option to exigrep.
4339 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4340 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4341 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4342 them is set, LOOKUP_LSEARCH is forced.
4344 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4346 24. Added sender and host information to the "rejected by local_scan()" log
4347 line; previously there was no indication of these.
4349 25. Added .include_if_exists.
4351 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4352 Linux. It turns out that not all file systems support this. Apparently some
4353 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4354 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4355 not supported on the file descriptor, is now ignored when Exim is trying to
4356 sync a directory. This applies only to Linux.
4358 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4360 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4361 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4364 29. Added support for maildirsize files from supplied patch (modified a bit).
4366 30. The use of :fail: followed by an empty string could lead Exim to respond to
4367 sender verification failures with (e.g.):
4369 550 Verification failed for <xxx>
4370 550 Sender verify failed
4372 where the first response line was missing the '-' that indicates it is not
4373 the final line of the response.
4375 31. The loop for finding the name of the user that called Exim had a hardwired
4376 limit of 10; it now uses the value of finduser_retries, which is used for
4377 all other user lookups.
4379 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4382 33. Exim was neglecting to zero errno before one call of strtol() when
4383 expanding a string and expecting an integer value. On some systems this
4384 resulted in spurious "integer overflow" errors. Also, it was casting the
4385 result into an int without checking.
4387 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4388 did not work. The code looks as if it has *never* worked, though it appears
4389 to have been documented since at least releast 1.62. I have made it work.
4391 35. The "timeout_DNS" error in retry rules, also documented since at least
4392 1.62, also never worked. As it isn't clear exactly what this means, and
4393 clearly it isn't a major issue, I have abolished the feature by treating it
4394 as "timeout", and writing a warning to the main and panic logs.
4396 36. The display of retry rules for -brt wasn't always showing the error code
4399 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4400 timeout_connect_A, timeout_connect_MX.
4402 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4403 to the empty sender.
4405 39. The daemon was not analysing the content of -oX till after it had closed
4406 stderr and disconnected from the controlling terminal. This meant that any
4407 syntax errors were only noted on the panic log, and the return code from
4408 the command was 0. By re-arranging the code a little, I've made the
4409 decoding happen first, so such errors now appear on stderr, and the return
4410 code is 1. However, the actual setting up of the sockets still happens in
4411 the disconnected process, so errors there are still only recorded on the
4414 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4415 connections (as happens on some IP stacks) was logged at start up time as
4416 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4417 it from "IPv6 and IPv4", which means that two separate sockets are being
4420 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4421 says whether AF_INET or AF_INET6 was passed as an argument.
4423 42. Exiwhat output was messed up when time zones were included in log
4426 43. Exiwhat now gives more information about the daemon's listening ports,
4427 and whether -tls-on-connect was used.
4429 44. The "port" option of the smtp transport is now expanded.
4431 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4432 silently ignored. Now an error message is written to the main and panic
4435 46. There's a new ACL modifier called "logwrite" which writes to a log file
4436 as soon as it is encountered.
4438 47. Added $local_user_uid and $local_user_gid at routing time.
4440 48. Exim crashed when trying to verify a sender address that was being
4443 49. Exim was recognizing only a space character after ".include". It now also
4444 recognizes a tab character.
4446 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4447 extracting the relevant information from the specification. The man page no
4448 longer contains scrambled data for the -d option, and I've added a section
4449 at the front about calling Exim under different names.
4451 51. Added "extra_headers" argument to the "mail" command in filter files.
4453 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4456 53. Installed eximstats 1.29.
4458 54. Added transport_filter_timeout as a generic transport option.
4460 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4461 Cc: header lines. This was required by RFC 822, but it not required by RFC
4464 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4465 incoming messages that did not have them. Now it does so only if the
4466 message originates locally, that is, if there is no associated remote host
4467 address. When Resent- header lines are present, this applies to the Resent-
4468 lines rather than the non-Resent- lines.
4470 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4471 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4473 58. Messages for configuration errors now include the name of the main
4474 configuration file - useful now that there may be more than one file in a
4475 list (.included file names were always shown).
4477 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4478 for those rare installations that do not start the daemon as root or run it
4479 setuid root. I've cut out the call to initgroups() if the daemon is not
4482 60. The Exim user and group can now be bound into the binary as text strings
4483 that are looked up at the start of Exim's processing.
4485 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4487 62. Added $mailstore_basename variable.
4489 63. Installed patch to sieve.c from Michael Haardt.
4491 64. When Exim failed to open the panic log after failing to open the main log,
4492 the original message it was trying to log was written to stderr and debug
4493 output, but if they were not available (the usual case in production), it
4494 was lost. Now it is written to syslog before the two lines that record the
4495 failures to open the logs.
4497 65. Users' Exim filters run in subprocesses under the user's uid. It is
4498 possible for a "deliver" command or an alias in a "personal" command to
4499 provoke an address rewrite. If logging of address rewriting is configured,
4500 this fails because the process is not running as root or exim. There may be
4501 a better way of dealing with this, but for the moment (because 4.30 needs
4502 to be released), I have disabled address rewrite logging when running a
4503 filter in a non-root, non-exim process.
4509 1. The buildconfig auxiliary program wasn't quoting the value set for
4510 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4511 not defined. This bug was masked in 4.22 by the effect that was fixed in
4514 2. Some messages that were rejected after a message id was allocated were
4515 shown as "incomplete" by exigrep. It no longer does this for messages that
4516 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4518 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4519 have allow_domain_literals set, the ID did not get logged in the <= line.
4520 Domain literals are now always recognized in Message-ID: header lines.
4522 4. The first argument for a ${extract expansion item is the key name or field
4523 number. Leading and trailing spaces in this item were not being ignored,
4524 causing some misleading effects.
4526 5. When deliver_drop_privilege was set, single queue runner processes started
4527 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4528 same command in the process it spins off) were not dropping privilege.
4530 6. When the daemon running as "exim" started a queue runner, it always
4531 re-executed Exim in the spun-off process. This is a waste of effort when
4532 deliver_drop_privilege is set. The new process now just calls the
4533 queue-runner function directly.
4539 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4542 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4543 ignored. Though the use of -oP was forcing the writing of a pid file, it
4544 was always written to the default place.
4546 3. If the message "no IP address found for host xxxx" is generated during
4547 incoming verification, it is now followed by identification of the incoming
4548 connection (so you can more easily find what provoked it).
4550 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4552 5. Added some features to "harden" Exim a bit more against certain attacks:
4554 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4555 be put in Local/Makefile. This is like the never_users runtime option,
4556 but it cannot be overridden. The default setting is "root".
4558 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4559 prefix string with which any file named in a -C command line option
4562 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4563 is retained for -C and -D only if the caller of Exim is root. Without
4564 it, the exim user may also use -C and -D and retain privilege.
4566 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4567 command line option is disabled.
4569 6. Macro names set by the -D option must start with an upper case letter, just
4570 like macro names defined in the configuration file.
4572 7. Added "dereference=" facility to LDAP.
4574 8. Two instances of the typo "uknown" in the source files are fixed.
4576 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4577 the Configure-Makefile script screwed up while processing it.
4579 10. Incorporated PCRE 4.4.
4581 11. The SMTP synchronization check was not operating right at the start of an
4582 SMTP session. For example, it could not catch a HELO sent before the client
4583 waited for the greeting. There is now a check for outstanding input at the
4584 point when the greeting is written. Because of the duplex, asynchronous
4585 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4586 way, but not yet received, when the check is performed.
4588 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4589 on TCP/IP sockets, because this apparently causes some broken clients to
4592 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4593 unchanged) from the Cygwin maintainer.
4595 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4596 was supported instead of testing for mbx. Effectively a typo.
4598 15. The spa authenticator server code was not checking that the input it
4599 received was valid base64.
4601 16. The debug output line for the "set" modifier in ACLs was not showing the
4602 name of the variable that was being set.
4604 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4606 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4607 lookup on its own. Something else was needed to trigger the lookup. For
4608 example, a match in host_lookup or the need for a host name in a host list.
4609 Now, if $sender_host_name is referenced and the host name has not yet been
4610 looked up, a lookup is performed. If the lookup fails, the variable remains
4611 empty, and $host_lookup_failed is set to "1".
4613 19. Added "eqi" as a case-independent comparison operator.
4615 20. The saslauthd authentication condition could segfault if neither service
4616 nor realm was specified.
4618 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4619 error message that was logged was misleading, and incoming SMTP
4620 connections were dropped. The message is now more accurate, and temporary
4621 errors are given to SMTP connections.
4623 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4624 (except RSET) with a 421 error, until QUIT is received. However, it was
4625 failing to send a response to QUIT.
4627 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4628 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4629 ACL is now run after the helo_try_verify_hosts code.
4631 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4632 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4633 case-independent, so other case variants are also recognized). Apparently
4634 some systems use these upper case variants.
4636 25. If more than two messages were waiting for the same host, and a transport
4637 filter was specified for the transport, Exim sent two messages over the
4638 same TCP/IP connection, and then failed with "socket operation on non-
4639 socket" when it tried to send the third.
4641 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4643 27. The extern definition of crypt16() in expand.c was not being excluded when
4644 the OS had its own crypt16() function.
4646 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4647 as a preferred synonym for return_size_limit, both as an option and as an
4650 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4652 30. Changed the default configuration ACL to relax the local part checking rule
4653 for addresses that are not in any local domains. For these addresses,
4654 slashes and pipe symbols are allowed within local parts, but the sequence
4655 /../ is explicitly forbidden.
4657 31. SPA server authentication was not clearing the challenge buffer before
4660 32. log_message in a "warn" ACL statement was writing to the reject log as
4661 well as to the main log, which contradicts the documentation and doesn't
4662 seem right (because no rejection is happening). So I have stopped it.
4664 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4665 However, I am unable to do any testing of this.
4667 34. Fixed an infelicity in the appendfile transport. When checking directories
4668 for a mailbox, to see if any needed to be created, it was accidentally
4669 using path names with one or more superfluous leading slashes; tracing
4670 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4672 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4673 contents are added to the log line that is written for every discarded
4674 recipient. (Previously a log_message setting was ignored.)
4676 36. The ${quote: operator now quotes the string if it is empty.
4678 37. The install script runs exim in order to find its version number. If for
4679 some reason other than non-existence or emptiness, which it checks, it
4680 could not run './exim', it was installing it with an empty version number,
4681 i.e. as "exim-". This error state is now caught, and the installation is
4684 38. An argument was missing from the function that creates an error message
4685 when Exim fails to connect to the socket for saslauthd authentication.
4686 This could cause Exim to crash, or give a corrupted message.
4688 39. Added isip, isip4, and isip6 to ${if conditions.
4690 40. The ACL variables $acl_xx are now saved with the message, and can be
4691 accessed later in routers, transports, and filters.
4693 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4694 strings in the file are not string-expanded.
4696 42. If a MAIL command specified a SIZE value that was too large to fit into an
4697 int variable, the check against message_size_limit failed. Such values are
4698 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4699 day this will have to be increased, but I don't think I want to be around
4700 when emails are that large.
4707 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4708 iconv() is not standard in FreeBSD.
4710 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4711 IPv6 enabled. The observed symptom was a segmentation fault on return from
4712 the function os_common_find_running_interfaces() in src/os.c.
4714 3. In the check_special_case() function in daemon.c I had used "errno" as an
4715 argument name, which causes warnings on some systems. This was basically a
4716 typo, since it was named "eno" in the comments!
4718 4. The code that waits for the clock to tick (at a resolution of some fraction
4719 of a second) so as to ensure message-id uniqueness was always waiting for
4720 at least one whole tick, when it could have waited for less. [This is
4721 almost certainly not relevant at current processor speeds, where it is
4722 unlikely to ever wait at all. But we try to future-proof.]
4724 5. The function that sleeps for a time interval that includes fractions of a
4725 second contained a race. It did not block SIGALRM between setting the
4726 timer, and suspending (a couple of lines later). If the interval was short
4727 and the sigsuspend() was delayed until after it had expired, the suspension
4728 never ended. On busy systems this could lead to processes getting stuck for
4731 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4732 process, before it forks any delivery processes. The open lookup caching
4733 mechanism meant that the open file or database connection was passed into
4734 the delivery process. The problem was that delivery processes always tidy
4735 up cached lookup data. This could cause a problem for the next delivery
4736 process started by the queue runner, because the external queue runner
4737 process does not know about the closure. So the next delivery process
4738 still has data in the lookup cache. In the case of a file lookup, there was
4739 no problem because closing a file descriptor in a subprocess doesn't affect
4740 the parent. However, if the lookup was caching a connection to a database,
4741 the connection was closed, and the second delivery process was likely to
4742 see errors such as "PGSQL: query failed: server closed the connection
4743 unexpectedly". The problem has been fixed by closing all cached lookups
4744 in a queue runner before running a delivery process.
4746 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4747 seem to have the "const" qualifier which it has on other OS. I've
4750 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4751 *of the same type* (client or server) with the same public name that an
4752 error should be diagnosed.
4754 9. When Exim looked up a host name for an IP address, but failed to find the
4755 original IP address when looking up the host name (a safety check), it
4756 output the message "<ip address> does not match any IP for NULL", which was
4757 confusing, to say the least. The bug was that the host name should have
4758 appeared instead of "NULL".
4760 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4761 user that is built into the binary, and the -C or -D options is used, root
4762 privilege is dropped before the configuration file is read. In addition,
4763 logging is switched to stderr instead of the normal log files. If the
4764 configuration then re-defines the Exim user, the unprivileged environment
4765 is probably not what is expected, so Exim logs a panic warning message (but
4768 However, if deliver_drop_privilege is set, the unprivileged state may well
4769 be exactly what is intended, so the warning has been cut out in that case,
4770 and Exim is allowed to try to write to its normal log files.
4776 1. smtp_return_error_details was not giving details for temporary sender
4777 or receiver verification errors.
4779 2. Diagnose a configuration error if two authenticators have the same public
4782 3. Exim used not to create the message log file for a message until the first
4783 delivery attempt. This could be confusing when incoming messages were held
4784 for policy or load reasons. The message log file is now created at the time
4785 the message is received, and an initial "Received" line is written to it.
4787 4. The automatically generated man page for command line options had a minor
4788 bug that caused no ill effects; however, a more serious problem was that
4789 the procedure for building the man page automatically didn't always
4790 operate. Consequently, release 4.20 contains an out-of-date version. This
4791 shouldn't happen again.
4793 5. When building Exim with embedded Perl support, the script that builds the
4794 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4795 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4797 6. The freeze_tell option was not being used for messages that were frozen on
4798 arrival, either by an ACL or by local_scan().
4800 7. Added the smtp_incomplete_transaction log selector.
4802 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4803 was accepting AUTH without a new EHLO.
4805 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4806 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4809 10. Logging of TCP/IP connections (when configured) now happens in the main
4810 daemon process instead of the child process, so that the TCP/IP connection
4811 count is more accurate (but it can never be perfect).
4813 11. The use of "drop" in a nested ACL was not being handled correctly in the
4814 outer ACL. Now, if condition failure induced by the nested "drop" causes
4815 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4816 or "require"), the connection is dropped.
4818 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4819 that yield "discard" can now be used with an "accept" or a "discard" verb,
4820 but an error is generated for any others (because I can't see a useful way
4821 to define what should happen).
4823 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4824 are now processed in the same way as lines in the Exim configuration file.
4825 In particular, continuation lines are supported.
4827 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4829 15. Added -ti meaning -t -i.
4831 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4832 domains, and warn by logging if others are found.
4834 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4835 returned from GIFCONF when Exim is trying to find the list of interfaces on
4836 a host. The code in os.c has been modified to copy each ifreq to an aligned
4837 structure in all cases.
4839 Also, in some cases, the returned ifreq's were being copied to a 'struct
4840 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4841 means the last couple of bytes of an IPv6 address could be chopped if the
4842 ifreq contained only a normal sockaddr (14 bytes storage).
4844 18. Named domain lists were not supported in the hosts_treat_as_local option.
4845 An entry such as +xxxx was not recognized, and was treated as a literal
4848 19. Ensure that header lines added by a DATA ACL are included in the reject log
4849 if the ACL subsequently rejects the message.
4851 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4852 MD5 (which is deprecated).
4854 21. When testing a filter file using -bf, Exim was writing a message when it
4855 took the sender from a "From " line in the message, but it was not doing so
4856 when it took $return_path from a Return-Path: header line. It now does.
4858 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4859 with a valid header line field name (a series of printing characters
4860 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4862 23. Changed "disc" in the source to "disk" to conform to the documentation and
4863 the book and for uniformity.
4865 24. Ignore Sendmail's -Ooption=value command line item.
4867 25. When execve() failed while trying to run a command in a pipe transport,
4868 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4869 could be confused with a return value of 69 from the command itself. This
4870 has been changed to 127, the value the shell returns if it is asked to run
4871 a non-existent command. The wording for the related log line suggests a
4872 non-existent command as the problem.
4874 26. If received_header_text expands to an empty string, do not add a Received:
4875 header line to the message. (Well, it adds a token one on the spool, but
4876 marks it "old" so that it doesn't get used or transmitted.)
4878 27. Installed eximstats 1.28 (addition of -nt option).
4880 28. There was no check for failure on the call to getsockname() in the daemon
4881 code. This can fail if there is a shortage of resources on the system, with
4882 ENOMEM, for example. A temporary error is now given on failure.
4884 29. Contrary to the C standard, it seems that in some environments, the
4885 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4886 program. Exim now does this explicitly; it affects the formatting of
4887 timestamps using strftime().
4889 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4890 complaints. I've now initialized all the variables, to avoid this.
4892 32. Header lines added by a system filter were not being "seen" during
4893 transport-time rewrites.
4895 33. The info_callback() function passed to OpenSSL is set up with type void
4896 (*)(SSL *, int, int), as described somewhere. However, when calling the
4897 function (actually a macro) that sets it up, the type void(*)() is
4898 expected. I've put in a cast to prevent warnings from picky compilers.
4900 34. If a DNS black list lookup found a CNAME record, but there were no A
4901 records associated with the domain it pointed at, Exim crashed.
4903 35. If a DNS black list lookup returned more than one A record, Exim ignored
4904 all but the first. It now scans all returned addresses if a particular IP
4905 value is being sought. In this situation, the contents of the
4906 $dnslist_value variable are a list of all the addresses, separated by a
4909 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4910 to accept a host name and all its aliases if the forward lookup for any of
4911 them yielded the IP address of the incoming connection. Now it accepts only
4912 those names whose forward lookup yields the correct IP address. Any other
4913 names are discarded. This closes a loophole whereby a rogue DNS
4914 administrator could create reverse DNS records to break through a
4915 wildcarded host restriction in an ACL.
4917 37. If a user filter or a system filter that ran in a subprocess used any of
4918 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4919 the wrong values were passed to the pipe command ($thisaddress had the
4920 value of $0, $0 had the value of $1, etc). This bug was introduced by
4921 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4923 38. Improved the line breaking for long SMTP error messages from ACLs.
4924 Previously, if there was no break point between 40 and 75 characters, Exim
4925 left the rest of the message alone. Two changes have been made: (a) I've
4926 reduced the minimum length to 35 characters; (b) if it can't find a break
4927 point between 35 and 75 characters, it looks ahead and uses the first one
4928 that it finds. This may give the occasional overlong line, but at least the
4929 remaining text gets split now.
4931 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4932 file descriptors might be low, and that setting 1000 would always raise it.
4933 It turns out that in some environments, the limit is already over 1000 and
4934 that lowering it causes trouble. So now Exim takes care not to decrease it.
4936 40. When delivering a message, the value of $return_path is set to $sender_
4937 address at the start of routing (routers may change the value). By an
4938 oversight, this default was not being set up when an address was tested by
4939 -bt or -bv, which affected the outcome if any router or filter referred to
4942 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4943 log only when "message" or "log_message" are set. However, if one of the
4944 conditions was an address verification, or a call to a nested ACL, the
4945 messages generated by the underlying test were being passed through. This
4946 no longer happens. The underlying message is available in $acl_verify_
4947 message for both "message" and "log_message" expansions, so it can be
4948 passed through if needed.
4950 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4951 new expansion $bh_ to give the encoded byte string without charset
4952 translation. Translation happens only if iconv() is available; HAVE_ICONV
4953 indicates this at build time. HEADERS_CHARSET gives the charset to
4954 translate to; headers_charset can change it in the configuration, and
4955 "headers charset" can change it in an individual filter file.
4957 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4958 that creates RFC 2047 encoded "words" labels them as that charset instead
4959 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4960 expansion operator; (ii) when Exim creates a From: line for a local
4961 message; (iii) when a header line is rewritten to include a "phrase" part.
4963 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4964 buggy, causing it to skip the first lines of messages whose message ID
4965 ended in 'D'. This would not have bitten before Exim release 4.14, because
4966 message IDs were unlikely to end in 'D' before then. The effect was to have
4967 incorrect size information for certain domains.
4969 45. #include "config.h" was missing at the start of the crypt16.c module. This
4970 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4973 46. If there was a timeout during a "random" callout check, Exim treated it as
4974 a failure of the random address, and carried on sending RSET and the real
4975 address. If the delay was just some slowness somewhere, the response to the
4976 original RCPT would be taken as a response to RSET and so on, causing
4977 mayhem of various kinds.
4979 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4980 when I implemented it. It didn't allow for the fact that some option values
4981 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4982 the right test for positive values.
4984 48. Domain names in DNS records are case-independent. Exim always looks them up
4985 in lower case. Some resolvers return domain names in exactly the case they
4986 appear in the zone file, that is, they may contain uppercase letters. Not
4987 all resolvers do this - some return always lower case. Exim was treating a
4988 change of case by a resolver as a change of domain, similar to a widening
4989 of a domain abbreviation. This triggered its re-routing code and so it was
4990 trying to route what was effectively the same domain again. This normally
4991 caused routing to fail (because the router wouldn't handle the domain
4992 twice). Now Exim checks for this case specially, and just changes the
4993 casing of the domain that it ultimately uses when it transmits the message
4996 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4999 50. If a filter generated a file delivery with a non-absolute name (possible if
5000 no home directory exists for the router), the forbid_file option was not
5003 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5004 the existing equality matching.
5006 52. Exim was using ints instead of ino_t variables in some places where it was
5007 dealing with inode numbers.
5009 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5010 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5011 called TMPDIR, and if it finds it is different, it changes its value.
5013 54. The smtp_printf() function is now made available to local_scan() so
5014 additional output lines can be written before returning. There is also an
5015 smtp_fflush() function to enable the detection of a dropped connection.
5016 The variables smtp_input and smtp_batched_input are exported to
5019 55. Changed the default runtime configuration: the message "Unknown user"
5020 has been removed from the ACL, and instead placed on the localuser router,
5021 using the cannot_route_message feature. This means that any verification
5022 failures that generate their own messages won't get overridden. Similarly,
5023 the "Unrouteable address" message that was in the ACL for unverifiable
5024 relay addresses has also been removed.
5026 56. Added hosts_avoid_esmtp to the smtp transport.
5028 57. The exicyclog script was not checking for the esoteric option
5029 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5030 will work only if exicyclog is run under the appropriate euid.
5032 58. Following a discussion on the list, the rules by which Exim recognises line
5033 endings on incoming messages have been changed. The -dropcr and drop_cr
5034 options are now no-ops, retained only for backwards compatibility. The
5035 following line terminators are recognized: LF CRLF CR. However, special
5036 processing applies to CR:
5038 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5039 nor a local message in the state where . is a terminator.
5041 (ii) If a bare CR is encountered in a header line, an extra space is added
5042 after the line terminator so as not to end the header. The reasoning
5043 behind this is that bare CRs in header lines are most likely either
5044 to be mistakes, or people trying to play silly games.
5046 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5047 was being incorrectly given as 18 bytes larger than it should have been.
5048 This is a VOB (very old bug).
5050 60. This may never have affected anything current, but just in case it has:
5051 When the local host is found other than at the start of a list of hosts,
5052 the local host, those with the same MX, and any that follow, are discarded.
5053 When the list in question was part of a longer list of hosts, the following
5054 hosts (not currently being processed) were also being discarded. This no
5055 longer happens. I'm not sure if this situation could ever has previously
5058 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5061 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5062 header. This is something that is recommended in a new Internet Draft, and
5063 is something that is documented as being done by Sendmail. There are two
5064 possible values. For messages generated by the autoreply transport, Exim
5067 Auto-Submitted: auto-replied
5069 whereas for all other generated messages (e.g. bounces) it adds
5071 Auto-Submitted: auto-generated
5073 63. The "personal" condition in filters now includes a test for the
5074 Auto-Submitted: header. If it contains the string "auto-" the message it
5075 not considered personal.
5077 64. Added rcpt_include_affixes as a generic transport option.
5079 65. Added queue_only_override (default true).
5081 66. Added the syslog_duplication option.
5083 67. If what should have been the first header line of a message consisted of
5084 a space followed by a colon, Exim was mis-interpreting it as a header line.
5085 It isn't of course - it is syntactically invalid and should therefore be
5086 treated as the start of the message body. The misbehaviour could have
5087 caused a number of strange effects, including loss of data in subsequent
5088 header lines, and spool format errors.
5090 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5091 client host had authenticated. This control can now be exercised by an ACL
5092 for more flexibility.
5094 69. By default, callouts do not happen when testing with -bh. There is now a
5095 variant, -bhc, which does actually run the callout code, including
5096 consulting and updating the callout cache.
5098 70. Added support for saslauthd authentication, courtesy of Alexander
5101 71. If statvfs() failed on the spool or log directories while checking their
5102 size for availability, Exim confusingly gave the error "space shortage".
5103 Furthermore, in debugging mode it crashed with a floating point exception.
5104 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5105 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5106 serious problem, Exim now writes to the main and panic logs when this
5107 happens, with details of the failure. It then refuses to accept the
5108 incoming message, giving the message "spool directory problem" or "log
5109 directory problem" with a 421 code for SMTP messages.
5111 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5112 0, 1, and 2 exist, because some OS complain for execs without them (see
5113 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5114 descriptors. However, the code omitted to check that the open succeeded,
5115 causing mysterious errors if for some reason the permissions on /dev/null
5116 got screwed. Now Exim writes a message to the main and panic logs, and
5117 bombs out if it can't open /dev/null.
5119 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5120 interact so that it is all more flexible. It is supposed to remain
5121 backwards compatible. Also added extra_local_interfaces.
5123 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5124 to bomb out with an assertion failure - to the client this appears as a
5125 connection drop. This problem occurs in the part of the code that was taken
5126 from the Samba project. Fortunately, the assertion is in a very simple
5127 function, so I have fixed this by reproducing the function inline in the
5128 one place where it is called, and arranging for authentication to fail
5129 instead of killing the process with assert().
5131 75. The SPA client code was not working when the server requested OEM rather
5132 than Unicode encoding.
5134 76. Added code to make require_files with a specific uid setting more usable in
5135 the case where statting the file as root fails - usually a non-root-mounted
5136 NFS file system. When this happens and the failure is EACCES, Exim now
5137 forks a subprocess and does the per-uid checking as the relevant uid.
5139 77. Added process_log_path.
5141 78. If log_file_path was not explicitly set, a setting of check_log_space or
5142 check_log_inodes was ignored.
5144 79. If a space check for the spool or log partitions fails, the incident is now
5145 logged. Of course, in the latter case the data may get lost...
5147 80. Added the %p formatting code to string_format() so that it can be used to
5148 print addresses in debug_print(). Adjusted all the address printing in the
5149 debugging in store.c to use %p rather than %d.
5151 81. There was a concern that a line of code in smtp_in.c could overflow a
5152 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5153 initially expressed, the concern was not well-founded, because trailing
5154 spaces are removed early. However, if the trailing spaces were followed by
5155 a NULL, they did not get removed, so the overflow was possible. Two fixes
5158 (a) I re-wrote the offending code in a cleaner fashion.
5159 (b) If an incoming SMTP command contains a NULL character, it is rejected
5162 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5163 runs initgroups(), so that if the Exim user is in any additional groups,
5164 they will be used during message reception.
5170 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob