1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.624 2010/06/06 01:35:41 pdp 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_codedump 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 (I believe), leading to a segfault because of
28 an 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".
35 PP/09 Implemented reverse_ip expansion operator.
37 PP/10 Bugzilla 937: provide a "debug" ACL control.
39 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
45 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
46 $data_path, and $header_path variables; fixed documentation bugs and
49 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
50 exipick to access non-standard spools, including the "frozen" queue
53 NM/01 Bugzilla 965: Support mysql stored procedures.
54 Patch from Alain Williams
56 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
58 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
59 Patch from Andreas Metzler
61 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
62 Patch from Kirill Miazine
64 NM/05 Bugzilla 671: Added umask to procmail example.
66 JJ/03 installed exipick 20100323.0, fixing doc bug
68 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
69 directory. Notification and patch from Dan Rosenberg.
71 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
73 TK/02 Improve log output when DKIM signing operation fails.
75 MH/01 Treat the transport option dkim_domain as a colon separated
76 list, not as a single string, and sign the message with each element,
77 omitting multiple occurences of the same signer.
79 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
80 Bugzilla 985, 986. Patch by Simon Arlott
82 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
85 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
86 MBX locking. Notification from Dan Rosenberg.
92 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
94 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
96 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
98 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
100 NM/04 Bugzilla 915: Flag broken perl installation during build.
106 TK/01 Added patch by Johannes Berg that expands the main option
107 "spamd_address" if it starts with a dollar sign.
109 TK/02 Write list of recipients to X-Envelope-Sender header when building
110 the mbox-format spool file for content scanning (suggested by Jakob
113 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
114 (http://www.dcc-servers.net/) support via dccifd. Activated by
115 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
117 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
118 by Mark Daniel Reidel <mr@df.eu>.
120 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
121 When building exim an external PCRE library is now needed -
122 PCRE is a system library on the majority of modern systems.
123 See entry on PCRE_LIBS in EDITME file.
125 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
126 conversation. Added nologin parameter to request.
127 Patch contributed by Kirill Miazine.
129 TF/01 Do not log submission mode rewrites if they do not change the address.
131 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
133 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
134 log files in place. Contributed by Roberto Lima.
136 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
138 TF/03 Bugzilla 615: When checking the local_parts router precondition
139 after a local_part_suffix or local_part_prefix option, Exim now
140 does not use the address's named list lookup cache, since this
141 contains cached lookups for the whole local part.
143 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
144 Robert Millan. Documentation is in experimental-spec.txt.
146 TF/04 Bugzilla 668: Fix parallel build (make -j).
148 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
150 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
151 Patch provided by Jan Srzednicki.
153 TF/05 Leading white space used to be stripped from $spam_report which
154 wrecked the formatting. Now it is preserved.
156 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
157 that they are available at delivery time.
159 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
161 TF/08 TLS error reporting now respects the incoming_interface and
162 incoming_port log selectors.
164 TF/09 Produce a more useful error message if an SMTP transport's hosts
165 setting expands to an empty string.
167 NM/06 Bugzilla 744: EXPN did not work under TLS.
168 Patch provided by Phil Pennock.
170 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
171 Patch provided by Richard Godbee.
173 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
174 acl_smtp_notquit, added index entry.
176 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
177 Patch provided by Eugene Bujak.
179 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
180 accept(). Patch provided by Maxim Dounin.
182 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
183 Patch provided by Phil Pennock.
185 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
187 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
188 Patch provided by Brad "anomie" Jorsch.
190 NM/14 Bugzilla 622: Added timeout setting to transport filter.
191 Patch provided by Dean Brooks.
193 TK/05 Add native DKIM support (does not depend on external libraries).
195 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
196 Patch provided by Graeme Fowler.
198 NM/16 Bugzilla 851: Documentation example syntax fix.
200 NM/17 Changed NOTICE file to remove references to embedded PCRE.
202 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
205 NM/19 Bugzilla 745: TLS version reporting.
206 Patch provided by Phil Pennock.
208 NM/20 Bugzilla 167: bool: condition support.
209 Patch provided by Phil Pennock.
211 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
212 clients. Patch provided by Phil Pennock.
214 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
215 Patch provided by Brad "anomie" Jorsch.
217 NM/23 Bugzilla 687: Fix misparses in eximstats.
218 Patch provided by Heiko Schlittermann.
220 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
221 Patch provided by Heiko Schlittermann.
223 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
224 plus update to original patch.
226 NM/26 Bugzilla 799: Documentation correction for ratelimit.
228 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
229 Patch provided by David Brownlee.
231 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
233 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
235 NM/30 Bugzilla 888: TLS documentation bugfixes.
237 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
239 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
240 Unlike the original bugzilla I have changed all shell scripts in src tree.
242 NM/33 Bugzilla 898: Transport filter timeout fix.
243 Patch by Todd Rinaldo.
245 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
246 Patch by Serge Demonchaux.
248 NM/35 Bugzilla 39: Base64 decode bug fixes.
249 Patch by Jakob Hirsch.
251 NM/36 Bugzilla 909: Correct connect() call in dcc code.
253 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
255 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
257 NM/39 Bugzilla 911: Fixed MakeLinks build script.
263 TK/01 Add preliminary DKIM support. Currently requires a forked version of
264 ALT-N's libdkim that I have put here:
265 http://duncanthrax.net/exim-experimental/
267 Note to Michael Haardt: I had to rename some vars in sieve.c. They
268 were called 'true' and it seems that C99 defines that as a reserved
269 keyword to be used with 'bool' variable types. That means you could
270 not include C99-style headers which use bools without triggering
271 build errors in sieve.c.
273 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
274 as mailq or other aliases. Changed the --help handling significantly
275 to do whats expected. exim_usage() emits usage/help information.
277 SC/01 Added the -bylocaldomain option to eximstats.
279 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
281 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
283 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
289 PH/01 Another patch from the Sieve maintainer.
291 PH/02 When an IPv6 address is converted to a string for single-key lookup
292 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
293 dots are used instead of colons so that keys in lsearch files need not
294 contain colons. This was done some time before quoting was made available
295 in lsearch files. However, iplsearch files do require colons in IPv6 keys
296 (notated using the quote facility) so as to distinguish them from IPv4
297 keys. This meant that lookups for IP addresses in host lists did not work
298 for iplsearch lookups.
300 This has been fixed by arranging for IPv6 addresses to be expressed with
301 colons if the lookup type is iplsearch. This is not incompatible, because
302 previously such lookups could never work.
304 The situation is now rather anomolous, since one *can* have colons in
305 ordinary lsearch keys. However, making the change in all cases is
306 incompatible and would probably break a number of configurations.
308 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
311 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
312 conversion specification without a maximum field width, thereby enabling
313 a rogue spamd server to cause a buffer overflow. While nobody in their
314 right mind would setup Exim to query an untrusted spamd server, an
315 attacker that gains access to a server running spamd could potentially
316 exploit this vulnerability to run arbitrary code as the Exim user.
318 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
319 $primary_hostname instead of what libspf2 thinks the hosts name is.
321 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
322 a directory entry by the name of the lookup key. Previously, if a
323 symlink pointed to a non-existing file or a file in a directory that
324 Exim lacked permissions to read, a lookup for a key matching that
325 symlink would fail. Now it is enough that a matching directory entry
326 exists, symlink or not. (Bugzilla 503.)
328 PH/03 The body_linecount and body_zerocount variables are now exported in the
331 PH/04 Added the $dnslist_matched variable.
333 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
334 This means they are set thereafter only if the connection becomes
337 PH/06 Added the client_condition to authenticators so that some can be skipped
338 by clients under certain conditions.
340 PH/07 The error message for a badly-placed control=no_multiline_responses left
341 "_responses" off the end of the name.
343 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
345 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
346 (without spaces) instead of just copying the configuration text.
348 PH/10 Added the /noupdate option to the ratelimit ACL condition.
350 PH/11 Added $max_received_linelength.
352 PH/12 Added +ignore_defer and +include_defer to host lists.
354 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
355 way in which PCRE > 7.0 is built.
357 PH/14 Implemented queue_only_load_latch.
359 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
360 MAIL command. The effect was to mangle the value on 64-bit systems.
362 PH/16 Another patch from the Sieve maintainer.
364 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
366 PH/18 If a system quota error occurred while trying to create the file for
367 a maildir delivery, the message "Mailbox is full" was not appended to the
368 bounce if the delivery eventually timed out. Change 4.67/27 below applied
369 only to a quota excession during the actual writing of the file.
371 PH/19 It seems that peer DN values may contain newlines (and other non-printing
372 characters?) which causes problems in log lines. The DN values are now
373 passed through string_printing() before being added to log lines.
375 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
376 and InterBase are left for another time.)
378 PH/21 Added message_body_newlines option.
380 PH/22 Guard against possible overflow in moan_check_errorcopy().
382 PH/23 POSIX allows open() to be a macro; guard against that.
384 PH/24 If the recipient of an error message contained an @ in the local part
385 (suitably quoted, of course), incorrect values were put in $domain and
386 $local_part during the evaluation of errors_copy.
392 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
393 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
396 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
397 issue a MAIL command.
399 PH/02 In an ACL statement such as
401 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
403 if a client was not listed at all, or was listed with a value other than
404 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
405 the condition was not true (as it should be), so access was not denied.
406 The bug was that the ! inversion was incorrectly passed on to the second
407 item. This has been fixed.
409 PH/03 Added additional dnslists conditions == and =& which are different from
410 = and & when the dns lookup returns more than one IP address.
412 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
413 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
415 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
416 FSYNC, which compiles an option called disable_fsync that allows for
417 bypassing fsync(). The documentation is heavily laced with warnings.
419 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
421 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
422 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
423 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
424 including adding "make clean"; (3) Added -fPIC when compiling the test
425 dynamically loaded module, to get rid of a warning.
427 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
428 message fails, move_frozen_messages = true and ignore_bounce_errors_after
429 = 0s. The bug is otherwise harmless.
431 PH/07 There was a bug in the dovecot authenticator such that the value of
432 $auth1 could be overwritten, and so not correctly preserved, after a
433 successful authentication. This usually meant that the value preserved by
434 the server_setid option was incorrect.
436 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
438 PH/09 Installed PCRE release 7.0.
440 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
441 run for batched SMTP input. It is now run at the start of every message
442 in the batch. While fixing this I discovered that the process information
443 (output by running exiwhat) was not always getting set for -bs and -bS
444 input. This is fixed, and it now also says "batched" for BSMTP.
446 PH/11 Added control=no_pipelining.
448 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
449 patch, slightly modified), and move the expansion of helo_data till after
450 the connection is made in the smtp transport (so it can use these
453 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
455 PH/14 Added log_selector = +pid.
457 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
459 PH/16 Add ${if forany and ${if forall.
461 PH/17 Added dsn_from option to vary the From: line in DSNs.
463 PH/18 Flush SMTP output before performing a callout, unless control =
464 no_callout_flush is set.
466 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
467 was true (the default) a successful delivery failed to delete the retry
468 item, thus causing premature timeout of the address. The bug is now
471 PH/20 Added hosts_avoid_pipelining to the smtp transport.
473 PH/21 Long custom messages for fakedefer and fakereject are now split up
474 into multiline reponses in the same way that messages for "deny" and
475 other ACL rejections are.
477 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
478 with slight modification.
480 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
481 draft, changing the syntax and factoring some duplicate code".
483 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
484 for deliveries of the second and subsequent messages over the same SMTP
487 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
488 ${reduce, with only minor "tidies".
490 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
492 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
493 expansion side effects.
495 PH/27 When a message times out after an over-quota error from an Exim-imposed
496 quota, the bounce message says "mailbox is full". This message was not
497 being given when it was a system quota that was exceeded. It now should
500 MH/03 Made $recipients available in local_scan(). local_scan() already has
501 better access to the recipient list through recipients_list[], but
502 $recipients can be useful in postmaster-provided expansion strings.
504 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
505 in the case of a MAIL command with additional options following the
506 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
507 were accidentally chopped off.
509 PH/29 SMTP synchronization checks are implemented when a command is read -
510 there is a check that no more input is waiting when there shouldn't be
511 any. However, for some commands, a delay in an ACL can mean that it is
512 some time before the response is written. In this time, more input might
513 arrive, invalidly. So now there are extra checks after an ACL has run for
514 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
515 pipelining has not been advertised.
517 PH/30 MH's patch to allow iscntrl() characters to be list separators.
519 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
520 returned in the SMTP response when smtp_return_error_details was false.
523 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
524 instead of the C I/O that was originally supplied, because problems were
527 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
528 Exim which did not show up earlier: it was assuming that a call to
529 SSL_CTX_set_info_callback() might give an error value. In fact, there is
530 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
531 was a macro that became an assignment, so it seemed to work. This has
532 changed to a proper function call with a void return, hence the compile
533 error. Exim's code has been fixed.
535 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
538 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
540 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
542 PH/37 If a message is not accepted after it has had an id assigned (e.g.
543 because it turns out to be too big or there is a timeout) there is no
544 "Completed" line in the log. When some messages of this type were
545 selected by exigrep, they were listed as "not completed". Others were
546 picked up by some special patterns. I have improved the selection
547 criteria to be more general.
549 PH/38 The host_find_failed option in the manualroute router can now be set
550 to "ignore", to completely ignore a host whose IP address cannot be
551 found. If all hosts are ignored, the behaviour is controlled by the new
552 host_all_ignored option.
554 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
555 homing or because of multiple MX records with /mx) generated more than
556 one IP address, and the following item turned out to be the local host,
557 all the secondary addresses of the first item were incorrectly removed
558 from the list, along with the local host and any following hosts (which
559 is what is supposed to happen).
561 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
562 whoever called Exim into the -H file. In the case of the daemon it was
563 behaving confusingly. When first started, it used values for whoever
564 started the daemon, but after a SIGHUP it used the Exim user (because it
565 calls itself on a restart). I have changed the code so that it now always
568 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
569 message are rejected with the same error (e.g. no authentication or bad
570 sender address), and a DATA command is nevertheless sent (as can happen
571 with PIPELINING or a stupid MUA), the error message that was given to the
572 RCPT commands is included in the rejection of the DATA command. This is
573 intended to be helpful for MUAs that show only the final error to their
576 PH/42 Another patch from the Sieve maintainer.
578 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
579 Eximstats - Fixed some broken HTML links and added missing column headers
581 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
582 columns for Rejects, Temp Rejects, Ham, and Spam rows.
584 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
586 PH/43 Yet another patch from the Sieve maintainer.
588 PH/44 I found a way to check for a TCP/IP connection going away before sending
589 the response to the final '.' that terminates a message, but only in the
590 case where the client has not sent further data following the '.'
591 (unfortunately, this is allowed). However, in many cases there won't be
592 any further data because there won't be any more messages to send. A call
593 to select() can be used: if it shows that the input is "ready", there is
594 either input waiting, or the socket has been closed. An attempt to read
595 the next input character can distinguish the two cases. Previously, Exim
596 would have sent an OK response which the client would never have see.
597 This could lead to message repetition. This fix should cure that, at
598 least in a lot of common cases.
600 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
601 advertised in response to EHLO.
607 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
608 fixed by 4.65/MH/01 (is this a record?) are fixed:
610 (i) An empty string was always treated as zero by the numeric comparison
611 operators. This behaviour has been restored.
613 (ii) It is documented that the numeric comparison operators always treat
614 their arguments as decimal numbers. This was broken in that numbers
615 starting with 0 were being interpreted as octal.
617 While fixing these problems I realized that there was another issue that
618 hadn't been noticed. Values of message_size_limit (both the global option
619 and the transport option) were treated as octal if they started with 0.
620 The documentation was vague. These values are now always treated as
621 decimal, and I will make that clear in the documentation.
627 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
628 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
631 MH/01 Don't check that the operands of numeric comparison operators are
632 integers when their expansion is in "skipping" mode (fixes bug
633 introduced by 4.64-PH/07).
635 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
636 child addresses, Exim now panics and dies. Previously, because the count
637 is held in a short int, deliveries were likely to be lost. As such a
638 large number of recipients for a single message is ridiculous
639 (performance will be very, very poor), I have chosen to impose a limit
640 rather than extend the field.
646 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
647 leftover -K file (the existence of which was triggered by #402).
648 While we were at it, introduced process PID as part of the -K
649 filename. This should rule out race conditions when creating
652 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
653 processing considerably. Previous code took too long for large mails,
654 triggering a timeout which in turn triggers #401.
656 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
657 in the DK code in transports.c. sendfile() is not really portable,
658 hence the _LINUX specificness.
660 TF/01 In the add_headers option to the mail command in an Exim filter,
661 there was a bug that Exim would claim a syntax error in any
662 header after the first one which had an odd number of characters
665 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
666 callout verification, Exim cached a "reject" for the entire domain. This
667 is correct for most verifications, but it is not correct for a recipient
668 verification with use_sender or use_postmaster set, because in that case
669 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
670 case of MAIL FROM:<> rejection from other early rejections (e.g.
671 rejection of HELO). When verifying a recipient using a non-null MAIL
672 address, the cache is ignored if it shows MAIL FROM:<> rejection.
673 Whatever the result of the callout, the value of the domain cache is
674 left unchanged (for any other kind of callout, getting as far as trying
675 RCPT means that the domain itself is ok).
677 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
680 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
681 manifest itself as EPIPE rather than ECONNECT. When tidying away a
682 session, the daemon ignores ECONNECT errors and logs others; it now
683 ignores EPIPE as well.
685 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
686 (quoted-printable decoding).
688 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
689 later the small subsequent patch to fix an introduced bug.
691 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
693 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
695 PH/08 An error is now given if message_size_limit is specified negative.
697 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
698 to be given (somewhat) arbitrary names.
700 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
703 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
704 miscellaneous code fixes
706 PH/10 Added the log_reject_target ACL modifier to specify where to log
709 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
710 hostname. This is wrong, because it relates to the incoming message (and
711 probably the interface on which it is arriving) and not to the outgoing
712 callout (which could be using a different interface). This has been
713 changed to use the value of the helo_data option from the smtp transport
714 instead - this is what is used when a message is actually being sent. If
715 there is no remote transport (possible with a router that sets up host
716 addresses), $smtp_active_hostname is used.
718 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
719 tweaks were necessary in order to get it to work (see also 21 below):
720 (a) The code assumed that strncpy() returns a negative number on buffer
721 overflow, which isn't the case. Replaced with Exim's string_format()
723 (b) There were several signed/unsigned issues. I just did the minimum
724 hacking in of casts. There is scope for a larger refactoring.
725 (c) The code used strcasecmp() which is not a standard C function.
726 Replaced with Exim's strcmpic() function.
727 (d) The code set only $1; it now sets $auth1 as well.
728 (e) A simple test gave the error "authentication client didn't specify
729 service in request". It would seem that Dovecot has changed its
730 interface. Fortunately there's a specification; I followed it and
731 changed what the client sends and it appears to be working now.
733 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
736 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
737 address A is aliased to B and C, where B exists and C does not. Without
738 -v the output is "A verified" because verification stops after a
739 successful redirection if more than one address is generated. However,
740 with -v the child addresses are also verified. Exim was outputting "A
741 failed to verify" and then showing the successful verification for C,
742 with its parentage. It now outputs "B failed to verify", showing B's
743 parentage before showing the successful verification of C.
745 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
746 look up a TXT record in a specific list after matching in a combined
749 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
750 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
751 they consult the DNS. I had assumed they would set it the way they
752 wanted; and indeed my experiments on Linux seem to show that in some
753 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
754 To be on the safe side, however, I have now made the interface to
755 host_find_byname() similar to host_find_bydns(), with an argument
756 containing the DNS resolver options. The host_find_byname() function now
757 sets these options at its start, just as host_find_bydns() does. The smtp
758 transport options dns_qualify_single and dns_search_parents are passed to
759 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
760 of host_find_byname() use the default settings of RES_DEFNAMES
761 (qualify_single) but not RES_DNSRCH (search_parents).
763 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
764 spool_read_header() do less string testing, by means of a preliminary
765 switch on the second character of optional "-foo" lines. (This is
766 overdue, caused by the large number of possibilities that now exist.
767 Originally there were few.) While I was there, I also converted the
768 str(n)cmp tests so they don't re-test the leading "-" and the first
769 character, in the hope this might squeeze out yet more improvement.
771 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
772 flag allowing group syntax was set by the header_syntax check but not
773 turned off, possible causing trouble later; (2) The flag was not being
774 set at all for the header_verify test, causing "group"-style headers to
775 be rejected. I have now set it in this case, and also caused header_
776 verify to ignore an empty address taken from a group. While doing this, I
777 came across some other cases where the code for allowing group syntax
778 while scanning a header line wasn't quite right (mostly, not resetting
779 the flag correctly in the right place). These bugs could have caused
780 trouble for malformed header lines. I hope it is now all correct.
782 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
783 with the "reply" argument non-NULL. The code, however (which originally
784 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
785 but it didn't always do it. This confused somebody who was copying the
786 code for some other use. I have removed all the tests.
788 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
789 feature that was used to support insecure browsers during the U.S. crypto
790 embargo. It requires special client support, and Exim is probably the
791 only MTA that supported it -- and would never use it because real RSA is
792 always available. This code has been removed, because it had the bad
793 effect of slowing Exim down by computing (never used) parameters for the
794 RSA_EXPORT functionality.
796 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
797 authenticator to fail if there's a tab character in the incoming data
798 (there should never be unless someone is messing about, as it's supposed
799 to be base64-encoded). Also added, on Timo's advice, the "secured" option
800 if the connection is using TLS or if the remote IP is the same as the
801 local IP, and the "valid-client-cert option" if a client certificate has
804 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
805 authenticators. This can be used for authorization after authentication
806 succeeds. (In the case of plaintext, it servers for both authentication
809 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
810 if any retry times were supplied.
812 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
813 connection, where there is no client IP address to check. In this
814 situation, the verify now always succeeds.
816 PH/25 Applied John Jetmore's -Mset patch.
818 PH/26 Added -bem to be like -Mset, but loading a message from a file.
820 PH/27 In a string expansion for a processed (not raw) header when multiple
821 headers of the same name were present, leading whitespace was being
822 removed from all of them, but trailing whitespace was being removed only
823 from the last one. Now trailing whitespace is removed from each header
824 before concatenation. Completely empty headers in a concatenation (as
827 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
828 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
830 PH/29 [Removed. This was a change that I later backed out, and forgot to
831 correct the ChangeLog entry (that I had efficiently created) before
832 committing the later change.]
834 PH/30 Exim was sometimes attempting to deliver messages that had suffered
835 address errors (4xx response to RCPT) over the same connection as other
836 messages routed to the same hosts. Such deliveries are always "forced",
837 so retry times are not inspected. This resulted in far too many retries
838 for the affected addresses. The effect occurred only when there were more
839 hosts than the hosts_max_try setting in the smtp transport when it had
840 the 4xx errors. Those hosts that it had tried were not added to the list
841 of hosts for which the message was waiting, so if all were tried, there
842 was no problem. Two fixes have been applied:
844 (i) If there are any address or message errors in an SMTP delivery, none
845 of the hosts (tried or untried) are now added to the list of hosts
846 for which the message is waiting, so the message should not be a
847 candidate for sending over the same connection that was used for a
848 successful delivery of some other message. This seems entirely
849 reasonable: after all the message is NOT "waiting for some host".
850 This is so "obvious" that I'm not sure why it wasn't done
851 previously. Hope I haven't missed anything, but it can't do any
852 harm, as the worst effect is to miss an optimization.
854 (ii) If, despite (i), such a delivery is accidentally attempted, the
855 routing retry time is respected, so at least it doesn't keep
856 hammering the server.
858 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
859 in ${readsocket because some servers need this prod.
861 PH/32 Added some extra debug output when updating a wait-xxx database.
863 PH/33 The hint "could be header name not terminated by colon", which has been
864 given for certain expansion errors for a long time, was not being given
865 for the ${if def:h_colon_omitted{... case.
867 PH/34 The spec says: "With one important exception, whenever a domain list is
868 being scanned, $domain contains the subject domain." There was at least
869 one case where this was not true.
871 PH/35 The error "getsockname() failed: connection reset by peer" was being
872 written to the panic log as well as the main log, but it isn't really
873 panic-worthy as it just means the connection died rather early on. I have
874 removed the panic log writing for the ECONNRESET error when getsockname()
877 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
878 runs only) independently of the message's sender address. This meant
879 that, if the 4xx error was in fact related to the sender, a different
880 message to the same recipient with a different sender could confuse
881 things. In particualar, this can happen when sending to a greylisting
882 server, but other circumstances could also provoke similar problems.
883 I have changed the default so that the retry time for these errors is now
884 based a combination of the sender and recipient addresses. This change
885 can be overridden by setting address_retry_include_sender=false in the
888 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
889 remote server are returned as part of bounce messages. This was not
890 happening for LMTP over a pipe (the lmtp transport), but now it is the
891 same for both kinds of LMTP.
893 PH/38 Despite being documented as not happening, Exim was rewriting addresses
894 in header lines that were in fact CNAMEs. This is no longer the case.
896 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
897 and queue runs started by the daemon processed all messages. This has
898 been fixed so that -R and -S can now usefully be given with -q<time>.
900 PH/40 Import PCRE release 6.7 (fixes some bugs).
902 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
904 PH/42 Give an error if -q is specified more than once.
906 PH/43 Renamed the variables $interface_address and $interface_port as
907 $received_ip_address and $received_port, to make it clear that these
908 values apply to message reception, and not to the outgoing interface when
909 a message is delivered. (The old names remain recognized, of course.)
911 PH/44 There was no timeout on the connect() call when using a Unix domain
912 socket in the ${readsocket expansion. There now is.
914 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
915 be meaningful with "accept".
917 SC/01 Eximstats V1.43
918 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
920 SC/02 Eximstats V1.44
921 Use a glob alias rather than an array ref in the generated
922 parser. This improves both readability and performance.
924 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
925 Collect SpamAssassin and rejection statistics.
926 Don't display local sender or destination tables unless
927 there is data to show.
928 Added average volumes into the top table text output.
930 SC/04 Eximstats V1.46
931 Collect data on the number of addresses (recipients)
932 as well as the number of messages.
934 SC/05 Eximstats V1.47
935 Added 'Message too big' to the list of mail rejection
936 reasons (thanks to Marco Gaiarin).
938 SC/06 Eximstats V1.48
939 Mainlog lines which have GMT offsets and are too short to
940 have a flag are now skipped.
942 SC/07 Eximstats V1.49 (Alain Williams)
943 Added the -emptyok flag.
945 SC/08 Eximstats V1.50
946 Fixes for obtaining the IP address from reject messages.
948 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
949 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
950 whitesspace changes from 4.64-PH/27
952 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
955 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
956 are found, allow negative numbers in numeric criteria)
958 JJ/06 exipick.20061117.2, added new $message_body_missing variable
960 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
961 to match changes made in 4.64-PH/43
963 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
965 PH/47 Put in an explicit test for a DNS lookup of an address record where the
966 "domain" is actually an IP address, and force a failure. This locks out
967 those revolvers/nameservers that support "A-for-A" lookups, in
968 contravention of the specifications.
970 PH/48 When a host name was looked up from an IP address, and the subsequent
971 forward lookup of the name timed out, the host name was left in
972 $sender_host_name, contrary to the specification.
974 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
975 restricted to single-key lookups, Exim was not diagnosing an error if
976 * or *@ was used with a query-style lookup.
978 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
980 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
981 long ago, but noone interested enough thought of it. Let's just say that
982 the "1.1" means that there are some new functions that weren't there at
983 some point in the past.
985 PH/51 Error processing for expansion failure of helo_data from an smtp
986 transport during callout processing was broken.
988 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
989 tested/used via the -bh/-bhc/-bs options.
991 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
992 bug, fixed in subsequent PCRE releases).
994 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
995 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
997 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1003 SC/01 Use a glob alias rather than an array ref in eximstats generated
1004 parser. This improves both readability and performance.
1006 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1007 Don't display local sender or destination tables in eximstats unless
1008 there is data to show.
1009 Added average volumes into the eximstats top table text output.
1011 SC/03 Collect data on the number of addresses (recipients) as well
1012 as the number of messages in eximstats.
1014 TF/01 Correct an error in the documentation for the redirect router. Exim
1015 does (usually) call initgroups() when daemonizing.
1017 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1018 with consistent privilege compared to when running as a daemon.
1020 TF/03 Note in the spec that $authenticated_id is not set for local
1021 submissions from trusted users.
1023 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1024 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1026 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1027 by adding some example configuration directives to the default
1028 configuration file. A little bit of work is required to uncomment the
1029 directives and define how usernames and passwords are checked, but
1030 there is now a framework to start from.
1032 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1033 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1034 without this. I don't know how relevant this is to other LDAP libraries.
1036 PH/02 Add the verb name to the "unknown ACL verb" error.
1038 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1040 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1042 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1043 directory not expanded when it should be if an expanded home directory
1044 was set for the address (which is overridden by the transport).
1046 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1049 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1050 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1051 because it is too late at that time, and has no effect.
1053 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1054 security issue with \' (bugzilla #107). I could not use the
1055 PQescapeStringConn() function, because it needs a PGconn value as one of
1058 PH/08 When testing addresses using -bt, indicate those final addresses that
1059 are duplicates that would not cause an additional delivery. At least one
1060 person was confused, thinking that -bt output corresponded to deliveries.
1061 (Suppressing duplicates isn't a good idea as you lose the information
1062 about possibly different redirections that led to the duplicates.)
1064 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1065 systems where poll() doesn't work, in particular OS X.
1067 PH/10 Added more information to debugging output for retry time not reached.
1069 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1070 operations in malware.c.
1072 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1075 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1076 syslog_duplication set false, log lines that would normally be written
1077 both the the main log and to the reject log were not written to syslog at
1080 PH/14 In the default configuration, change the use of "message" in ACL warn
1081 statements to "add_header".
1083 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1084 not followed by a command (e.g. "seen endif").
1086 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1087 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1090 PH/17 Added extra conditions to the default value of delay_warning_condition
1094 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1095 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1096 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1099 The Auto-Submitted: and various List- headers are standardised, whereas I
1100 don't think Precedence: ever was.
1102 PH/18 Refactored debugging code in route_finduser() to show more information,
1103 in particular, the error code if getpwnam() issues one.
1105 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1106 This is apparently needed in addition to the PH/07 change above to avoid
1107 any possible encoding problems.
1109 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1110 but not after initializing Perl.
1112 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1113 output them only if debugging. By default they are written stderr,
1114 apparently, which is not desirable.
1116 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1119 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1122 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1124 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1125 authenticated or an ident call has been made. Suppress the default
1126 values for $authenticated_id and $authenticated_sender (but permit -oMai
1127 and -oMas) when testing with -bh.
1129 PH/24 Re-jigged the order of the tests in the default configuration so that the
1130 tests for valid domains and recipients precede the DNS black list and CSA
1131 tests, on the grounds that those ones are more expensive.
1133 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1134 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1135 This bug exists in every version of Exim that I still have, right back to
1138 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1139 However, an attempt to turn on case-sensitivity in a regex key by
1140 including (?-i) didn't work because the subject string was already
1141 lowercased, and the effects were non-intuitive. It turns out that a
1142 one-line patch can be used to allow (?-i) to work as expected.
1148 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1149 other effects) broke the use of negated acl sub-conditions.
1151 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1154 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1155 "Deny" causes Exim to reject the incoming connection with a 554 error.
1156 Unfortunately, if there is a major crisis, such as a disk failure,
1157 tcp-wrappers gives "deny", whereas what one would like would be some
1158 kind of temporary error. A kludge has been added to help with this.
1159 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1160 554 error is used if errno is still zero or contains ENOENT (which occurs
1161 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1164 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1166 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1167 errors. Otherwise a message that provokes a temporary error (when other
1168 messages do not) can cause a whole host to time out.
1170 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1171 addresses were routed directly to files or pipes from a redirect router.
1172 File deliveries just didn't batch; pipe deliveries might have suffered
1175 PH/06 A failure to get a lock for a hints database would erroneously always say
1176 "Failed to get write lock", even when it was really a read lock.
1178 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1179 of 0600. This has been changed to use the value of the lockfile_mode
1180 option (which defaults to 0600).
1182 PH/08 Applied small patch from the Sieve maintainer.
1184 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1185 folder from quota calculations, a direct delivery into this folder messed
1186 up the contents of the maildirsize file. This was because the regex was
1187 used only to exclude .Trash (or whatever) when the size of the mailbox
1188 was calculated. There was no check that a delivery was happening into an
1189 excluded directory. This bug has been fixed by ignoring all quota
1190 processing for deliveries into excluded directories.
1192 PH/10 Added the maildirfolder_create_regex option to appendfile.
1198 PH/01 The code for finding all the local interface addresses on a FreeBSD
1199 system running IPv6 was broken. This may well have applied to all BSD
1200 systems, as well as to others that have similar system calls. The broken
1201 code found IPv4 interfaces correctly, but gave incorrect values for the
1202 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1203 that it would not match correctly against @[] and not recognize the IPv6
1206 PH/02 The ipliteral router was not recognizing addresses of the form user@
1207 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1209 PH/03 Added disable_ipv6.
1211 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1212 decoded form, because it is most often used to construct To: headers
1213 lines in autoreplies, and the decoded form may well be syntactically
1214 invalid. However, $reply_address has leading white space removed, and all
1215 newlines turned into spaces so that the autoreply transport does not
1218 PH/05 If group was specified without a user on a router, and no group or user
1219 was specified on a transport, the group from the router was ignored.
1221 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1222 for visible compile-time settings that can be used to change these
1223 numbers, for those that want even more. Backwards compatibility with old
1224 spool files has been maintained. However, going back to a previous Exim
1225 release will lost any variables that are in spool files.
1227 PH/07 Two small changes when running in the test harness: increase delay when
1228 passing a TCP/IP connection to a new process, in case the original
1229 process has to generate a bounce, and remove special handling of
1230 127.0.0.2 (sic), which is no longer necessary.
1232 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1233 be the same on different OS.
1235 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1238 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1239 whether --show-vars was specified or not
1241 JJ/02 exipick: Added support for new ACL variable spool format introduced
1244 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1245 syntactically invalid From: or Reply-to: line, and a filter used this to
1246 generate an autoreply, and therefore failed to obtain an address for the
1247 autoreply, Exim could try to deliver to a non-existent relative file
1248 name, causing unrelated and misleading errors. What now happens is that
1249 it logs this as a hard delivery error, but does not attempt to create a
1252 PH/11 The exinext utility has a -C option for testing purposes, but although
1253 the given file was scanned by exinext itself; it wasn't being passed on
1254 when Exim was called.
1256 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1257 an end-of-file indication when reading a command response.
1259 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1260 compiled. In many other places in Exim, IPv6 addresses are always
1261 recognized, so I have changed this. It also means that IPv4 domain
1262 literals of the form [IPV4:n.n.n.n] are now always recognized.
1264 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1265 used if the router is not running as root, for example, when verifying at
1266 ACL time, or when using -bh. The debugging output from this situation was
1267 non-existent - all you got was a failure to exec. I have made two
1270 (a) Failures to set uid/gid, the current directory, or a process leader
1271 in a subprocess such as that created by queryprogram now generate
1272 suitable debugging ouput when -d is set.
1274 (b) The queryprogram router detects when it is not running as root,
1275 outputs suitable debugging information if -d is set, and then runs
1276 the subprocess without attempting to change uid/gid.
1278 PH/15 Minor change to Makefile for building test_host (undocumented testing
1281 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1282 additional section of a DNS packet that returns MX or SRV records.
1283 Instead, it always explicitly searches for A/AAAA records. This avoids
1284 major problems that occur when a DNS server includes only records of one
1285 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1286 fixed another bug: if SRV records were looked up and the corresponding
1287 address records were *not* found in the additional section, the port
1288 values from the SRV records were lost.
1290 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1291 using the correct key (the original address) when searching the retry
1292 rules in order to find which one to use for generating the retry hint.
1294 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1295 adding the default one. Similarly, if it contains a Reply-To: header, the
1296 errors_reply_to option, if set, is not used.
1298 PH/19 When calculating a retry time, Exim used to measure the "time since
1299 failure" by looking at the "first failed" field in the retry record. Now
1300 it does not use this if it is later than than the arrival time of the
1301 message. Instead it uses the arrival time. This makes for better
1302 behaviour in cases where some deliveries succeed, thus re-setting the
1303 "first failed" field. An example is a quota failure for a huge message
1304 when small messages continue to be delivered. Without this change, the
1305 "time since failure" will always be short, possible causing more frequent
1306 delivery attempts for the huge message than are intended.
1307 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1309 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1310 $1, $2, $3) because the numerical variables can be reset during some
1311 expansion items (e.g. "match"), thereby losing the authentication data.
1313 PH/21 Make -bV show the size of off_t variables so that the test suite can
1314 decide whether to run tests for quotas > 2G.
1316 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1317 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1318 filecount value is greater than 2G or if a quota value is greater than 2G
1319 on a system where the size of off_t is not greater than 4, a panic error
1322 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1323 never match. The debug and -bh output now contains an explicit error
1324 message indicating a malformed IPv4 address or mask.
1326 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1327 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1328 PH/23 above applies.
1330 PH/25 Do not write to syslog when running in the test harness. The only
1331 occasion when this arises is a failure to open the main or panic logs
1332 (for which there is an explicit test).
1334 PH/26 Added the /no_tell option to "control=freeze".
1336 PH/27 If a host name lookup failed very early in a connection, for example, if
1337 the IP address matched host_lookup and the reverse lookup yielded a name
1338 that did not have a forward lookup, an error message of the form "no IP
1339 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1340 could be logged. Now it outputs the IP address instead of "NULL".
1342 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1343 allows the sender and the authenticated sender to be set when
1344 submitting a message from within Exim. Since child_open_exim() is
1345 documented for local_scan(), the new function should be too.
1347 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1348 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1349 results in an empty string is now treated as unset.
1351 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1353 PH/31 Added sender_verify_fail logging option.
1355 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1356 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1357 tidied the source and removed it altogether.
1359 PH/33 When a queue run was abandoned because the load average was too high, a
1360 log line was always written; now it is written only if the queue_run log
1361 selector is set. In addition, the log line for abandonment now contains
1362 information about the queue run such as the pid. This is always present
1363 in "start" and "stop" lines but was omitted from the "abandon" line.
1365 PH/34 Omit spaces between a header name and the colon in the error message that
1366 is given when verify = headers_syntax fails (if there are lots of them,
1367 the message gets confusing).
1369 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1370 names, as there are now some PTR records that contain slashes. This check
1371 is only to protect against broken name servers that fall over on strange
1372 characters, so the fact that it applies to all lookups doesn't matter.
1374 PH/36 Now that the new test suite is complete, we can remove some of the
1375 special code in Exim that was needed for the old test suite. For example,
1376 sorting DNS records because real resolvers return them in an arbitrary
1377 order. The new test suite's fake resolver always returns records in the
1380 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1381 bounces) except when queue_only is set, to avoid logging races between
1382 the different processes.
1384 PH/38 Panic-die if .include specifies a non-absolute path.
1386 PH/39 A tweak to the "H" retry rule from its user.
1388 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1389 a label. They prevented compilation on older perls.
1391 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1392 a warning to be raised on newish perls.
1394 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1395 on queue. Changes to match documented behaviour of showing count of
1396 messages matching specified criteria.
1398 PH/40 Changed the default ident timeout from 30s to 5s.
1400 PH/41 Added support for the use of login_cap features, on those BSD systems
1401 that have them, for controlling the resources used by pipe deliveries.
1403 PH/42 The content-scanning code uses fopen() to create files in which to put
1404 message data. Previously it was not paying any attention to the mode of
1405 the files. Exim runs with umask(0) because the rest of the code creates
1406 files with open(), and sets the required mode explicitly. Thus, these
1407 files were ending up world-writeable. This was not a big issue, because,
1408 being within the spool directory, they were not world-accessible. I have
1409 created a function called modefopen, which takes an additional mode
1410 argument. It sets umask(777), creates the file, chmods it to the required
1411 mode, then resets the umask. All the relevant calls to fopen() in the
1412 content scanning code have been changed to use this function.
1414 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1415 to 24 hours. This avoids potential overflow problems when processing G
1416 and H retry rules. I suspect nobody ever tinkers with this value.
1418 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1420 PH/45 When the plaintext authenticator is running as a client, the server's
1421 challenges are checked to ensure they are valid base64 strings. By
1422 default, the authentication attempt is cancelled if an invalid string is
1423 received. Setting client_ignore_invalid_base64 true ignores these errors.
1424 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1425 they are received. Thus, the responses can be made to depend on the
1426 challenges. If an invalid string is ignored, an empty string is placed in
1429 PH/46 Messages that are created by the autoreply transport now contains a
1430 References: header, in accordance with RFCs 2822 and 3834.
1432 PH/47 Added authenticated_sender_force to the smtp transport.
1434 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1436 PH/49 Installed latest patch from the Sieve maintainer.
1438 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1439 was also set, the appendfile transport was unnecessarily scanning a
1440 directory of message files (e.g. for maildir delivery) to find the count
1441 of files (along with the size), even though it did not need this
1442 information. It now does the scan only if it needs to find either the
1443 size of the count of files.
1445 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1447 PH/52 Two bugs concerned with error handling when the smtp transport is
1450 (i) Exim was not creating retry information for temporary errors given
1451 for individual recipients after the DATA command when the smtp transport
1452 was used in LMTP mode. This meant that they could be retried too
1453 frequently, and not timed out correctly.
1455 (ii) Exim was setting the flag that allows error details to be returned
1456 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1457 recipients that were returned after the DATA command.
1459 PH/53 This is related to PH/52, but is more general: for any failing address,
1460 when detailed error information was permitted to be returned to the
1461 sender, but the error was temporary, then after the final timeout, only
1462 "retry timeout exceeded" was returned. Now it returns the full error as
1463 well as "retry timeout exceeded".
1465 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1466 do this, and (what is worse) MTAs that accept it.
1468 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1469 will now be deprecated.
1471 PH/56 New os.c-cygwin from the Cygwin maintainer.
1473 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1474 formats (previously only available in exim formats via -bpr, -bpru,
1475 and -bpra. Now also available in native and exiqgrep formats)
1477 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1478 with very large, slow to parse queues
1480 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1482 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1484 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1485 responses to authentication challenges, though it was showing the
1486 challenges; (ii) I've removed the CR characters from the debug output for
1489 PH/58 Allow for the insertion of a newline as well as a space when a string
1490 is turned into more than one encoded-word during RFC 2047 encoding. The
1491 Sieve code now uses this.
1493 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1494 data_4xx, lost_connection, tls_required.
1496 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1497 message was being sent as an SMTP response.
1499 PH/61 Add -l and -k options to exicyclog.
1501 PH/62 When verifying, if an address was redirected to one new address, so that
1502 verification continued, and the new address failed or deferred after
1503 having set something in $address_data, the value of $address_data was not
1504 passed back to the ACL. This was different to the case when no
1505 redirection occurred. The value is now passed back in both cases.
1507 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1508 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1509 use login_cap.h, so on its own it isn't the distinguishing feature. The
1510 new name refers directly to the setclassresources() function.
1512 PH/65 Added configuration files for NetBSD3.
1514 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1516 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1517 is preferred over IPv4.
1519 PH/68 The bounce_return_message and bounce_return_body options were not being
1520 honoured for bounces generated during the reception of non-SMTP messages.
1521 In particular, this applied to messages rejected by the ACL. This bug has
1522 been fixed. However, if bounce_return_message is true and bounce_return_
1523 body is false, the headers that are returned for a non-SMTP message
1524 include only those that have been read before the error was detected.
1525 (In the case of an ACL rejection, they have all been read.)
1527 PH/69 The HTML version of the specification is now built in a directory called
1528 spec_html instead of spec.html, because the latter looks like a path with
1529 a MIME-type, and this confuses some software.
1531 PH/70 Catch two compiler warnings in sieve.c.
1533 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1534 function verify_get_ident() calls ip_connect() to connect a socket, but
1535 if the "connect()" function timed out, ip_connect() used to close the
1536 socket. However, verify_get_ident() also closes the socket later, and in
1537 between Exim writes to the log, which may get opened at this point. When
1538 the socket was closed in ip_connect(), the log could get the same file
1539 descriptor number as the socket. This naturally causes chaos. The fix is
1540 not to close the socket in ip_connect(); the socket should be closed by
1541 the function that creates it. There was only one place in the code where
1542 this was missing, in the iplookup router, which I don't think anybody now
1543 uses, but I've fixed it anyway.
1545 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1546 well as to direct DNS lookups. Otherwise the handling of names in host
1547 lists is inconsistent and therefore confusing.
1553 PH/01 Two changes to the default runtime configuration:
1555 (1) Move the checks for relay_from_hosts and authenticated clients from
1556 after to before the (commented out) DNS black list checks.
1558 (2) Add control=submission to the relay_from_hosts and authenticated
1559 clients checks, on the grounds that messages accepted by these
1560 statements are most likely to be submissions.
1562 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1564 (1) Generate an error if the third argument for the ${prvs expansion is
1567 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1570 (3) Reset the variables that are obtained from the first argument of
1571 ${prvscheck and used in the second argument before leaving the code,
1572 because their memory is reclaimed, so using them afterwards may do
1575 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1576 one (it's much easier than Tom thought :-).
1578 (5) Because of (4), we can now allow for the use of $prvscheck_result
1579 inside the third argument.
1581 PH/03 For some reason, the default setting of PATH when running a command from
1582 a pipe transport was just "/usr/bin". I have changed it to
1585 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1586 anything to be listed in the output from -bV.
1588 PH/05 When a filter generated an autoreply, the entire To: header line was
1589 quoted in the delivery log line, like this:
1591 => >A.N.Other <ano@some.domain> <original@ddress> ...
1593 This has been changed so that it extracts the operative address. There
1594 may be more than one such address. If so, they are comma-separated, like
1597 => >ano@some.domain,ona@other.domain <original@ddress> ...
1599 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1600 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1601 not being looked up in the rDNS to get a host name, Exim was showing the
1602 IP address twice in Received: lines, even though the IP addresses were
1603 identical. For example:
1605 Received: from [1.2.3.4] (helo=[1.2.3.4])
1607 However, if the real host name was known, it was omitting the HELO data
1608 if it matched the actual IP address. This has been tidied up so that it
1609 doesn't show the same IP address twice.
1611 PH/07 When both +timestamp and +memory debugging was on, the value given by
1612 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1613 called by the debug printing, thereby overwriting the timestamp buffer.
1614 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1616 PH/08 When the original message was included in an autoreply transport, it
1617 always said "this is a copy of the message, including all the headers",
1618 even if body_only or headers_only was set. It now gives an appropriate
1621 PH/09 Applied a patch from the Sieve maintainer which:
1623 o fixes some comments
1624 o adds the (disabled) notify extension core
1625 o adds some debug output for the result of if/elsif tests
1626 o points to the current vacation draft in the documentation
1627 and documents the missing references header update
1631 o fixes a bug in processing the envelope test (when testing
1632 multiple envelope elements, the last element determinted the
1635 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1636 Electronic Mail") by including:
1638 Auto-submitted: auto-generated
1640 in the messages that it generates (bounce messages and others, such as
1641 warnings). In the case of bounce messages for non-SMTP mesages, there was
1642 also a typo: it was using "Auto_submitted" (underscore instead of
1643 hyphen). Since every message generated by Exim is necessarily in response
1644 to another message, thes have all been changed to:
1646 Auto-Submitted: auto-replied
1648 in accordance with these statements in the RFC:
1650 The auto-replied keyword:
1652 - SHOULD be used on messages sent in direct response to another
1653 message by an automatic process,
1655 - MUST NOT be used on manually-generated messages,
1657 - MAY be used on Delivery Status Notifications (DSNs) and Message
1658 Disposition Notifications (MDNs),
1660 - MUST NOT be used on messages generated by automatic or periodic
1661 processes, except for messages which are automatic responses to
1664 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1665 to the default Received: header definition.
1667 PH/12 Added log selector acl_warn_skipped (default on).
1669 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1670 variables because (a) they are in the wrong storage pool and (b) even if
1671 they were copied, it wouldn't work properly because of the caching.
1673 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1674 checking when decoding. Apparently there are clients that generate
1675 overlong encoded strings. Why am I not surprised?
1677 PH/15 If the first argument of "${if match_address" was not empty, but did not
1678 contain an "@" character, Exim crashed. Now it writes a panic log message
1679 and treats the condition as false.
1681 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1683 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1684 extension "envelope-auth". The code is finished and in agreement with
1685 other implementations, but there is no documentation so far and in fact,
1686 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1687 not changing the active code.
1689 Print executed "if" and "elsif" statements when debugging is used. This
1690 helps a great deal to understand what a filter does.
1692 Document more things not specified clearly in RFC3028. I had all this
1693 sorted out, when out of a sudden new issues came to my mind. Oops."
1695 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1698 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1699 canonical form (as documented). However, after a host name lookup from
1700 the IP address, check_host() was doing a simple string comparison with
1701 addresses acquired from the DNS when checking that the found name did
1702 have the original IP as one of its addresses. Since any found IPv6
1703 addresses are likely to be in abbreviated form, the comparison could
1704 fail. Luckily, there already exists a function for doing the comparison
1705 by converting both addresses to binary, so now that is used instead of
1706 the text comparison.
1708 PH/20 There was another similar case to PH/19, when a complete host name was
1709 given in a host list; looking up its IP address could give an abbreviated
1710 form, whereas the current host's name might or might not be abbreviated.
1711 The same fix has been applied.
1717 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1718 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1721 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1722 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1724 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1726 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1727 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1728 Also, for "Auto-Submitted", treat anything other than "no" as
1729 non-personal, in accordance with RFC 3834. (Previously it treated
1730 anything starting "auto-" as non-personal.)
1732 TF/01 The control=submission/name=... option had a problem with syntax
1733 errors if the name included a slash character. The /name= option
1734 now slurps the rest of the string, so it can include any characters
1735 but it must come last in the list of options (after /sender_retain
1738 PH/05 Some modifications to the interface to the fake nameserver for the new
1746 TK/01 Added the "success_on_redirect" address verification option. See
1747 NewStuff for rationale and an example.
1749 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1751 PH/02 Patch to exigrep to allow it to work on syslog lines.
1753 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1754 fread() to skip over the body file's header line, because in Cygwin the
1755 header line is locked and is inaccessible.
1757 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1758 co-exist for some time) to make it clear that it is the Exim ID that is
1759 referenced, not the Message-ID: header line.
1761 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1762 string_format() function, because snprintf() does not exist on all
1765 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1766 ${stat: expansion item.
1768 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1769 protocol synchronization error", to keep the pedants happy.
1771 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1772 well as for IRIX systems, when gcc is being used. See the host.c source
1775 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1777 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1780 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1781 transport and to the smtp transport in LMTP mode.
1783 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1785 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1786 run a filter in a subprocess. This could lead to confusion in subsequent
1787 lookups in the parent process. There should also be a search_tidyup() at
1788 the end of the subprocess.
1790 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1791 only if the host matched helo_try_verify_hosts, which caused the
1792 verification to occur when the EHLO/HELO command was issued. The ACL just
1793 tested the remembered result. Now, if a previous verification attempt has
1794 not happened, "verify = helo" does it there and then.
1796 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1798 TK/03 Fix log output including CR from clamd.
1800 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1801 exist provoked a memory error which could cause a segfault.
1803 PH/15 Installed PCRE 6.2
1805 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1807 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1808 of the problem. Specifically, suggested +O2 rather than +O1 for the
1811 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1813 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1814 record, the port was not correctly logged when the outgoing_port log
1815 selector was set (it logged the transort's default port).
1817 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1818 fallback_hosts, and "hosts" in the smtp transport.
1820 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1821 host errors such as "Connection refused".
1823 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1824 authentication with radiusclient 0.4.9:
1826 - Error returned from rc_read_config was caught wrongly
1827 - Username/password not passed on to radius server due to wrong length.
1829 The presumption is that some radiusclient API changes for 4.51/PH/17
1830 were not taken care of correctly. The code is still untested by me (my
1831 Linux distribution still has 0.3.2 of radiusclient), but it was
1832 contributed by a Radius user.
1834 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1835 expanding the "port" option of the smtp transport.
1837 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1838 while reading a MIME header. Thanks to Tom Hughes for a patch.
1840 PH/24 Include config.h inside local_scan.h so that configuration settings are
1843 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1844 that in an ACL for RCPT (for example), you can examine exactly what was
1847 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1848 commands, but it was not correctly comparing the address with the actual
1849 client host address. Thus, it would show the EHLO address in Received:
1850 header lines when this was not necessary.
1852 PH/27 Added the % operator to ${eval:}.
1854 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1855 it should be ignoring failures (because with -C, for example, it has lost
1856 privilege). It wasn't ignoring creation failures other than "already
1859 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1862 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1863 because an input line was too long, either on its own, or by virtue of
1864 too many continuations, the temporary file was not being removed, and the
1865 return code was incorrect.
1867 PH/31 Missing "BOOL" in function definition in filtertest.c.
1869 PH/32 Applied Sieve patches from the maintainer.
1871 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1873 PH/33 Added "verify = not_blind".
1875 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1876 Local/Makefile (with some defaults set). These are used in built scripts
1877 such as exicyclog, but they have never been used in the exim_install
1878 script (though there are many overriding facilities there). I have
1879 arranged that the exim_install script now takes note of these two
1882 PH/35 Installed configuration files for Dragonfly.
1884 PH/36 When a locally submitted message by a trusted user did not contain a
1885 From: header, and the sender address was obtained from -f or from an SMTP
1886 MAIL command, and the trusted user did not use -F to supply a sender
1887 name, $originator_name was incorrectly used when constructing a From:
1888 header. Furthermore, $originator_name was used for submission mode
1889 messages from external hosts without From: headers in a similar way,
1890 which is clearly wrong.
1892 PH/37 Added control=suppress_local_fixups.
1894 PH/38 When log_selector = +received_sender was set, and the addition of the
1895 sender made the log line's construction buffer exactly full, or one byte
1896 less than full, an overflow happened when the terminating "\n" was
1899 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1900 when the result of a list match is failure because a DNS lookup failed.
1902 PH/40 RM_COMMAND is now used in the building process.
1904 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1905 the "build-* directories that it finds.
1907 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1908 address in a domain literal was a prefix of an interface address.
1910 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1911 when verifying a sender address, unless rewrite_headers is false.
1913 PH/44 Wrote a long comment about why errors_to addresses are verified as
1914 recipients, not senders.
1916 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1917 the ratelimit ACL was added.
1919 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1921 PH/46 Added extra information about PostgreSQL errors to the error string.
1923 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1924 avoiding the need to install special zones in a real server. This is
1925 backwards compatible; if it can't find the fake resolver, it drops back.
1926 Thus, both old and new test suites can be run.
1928 TF/02 Added util/ratelimit.pl
1930 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1931 clock is set back in time.
1933 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1934 Candler <B.Candler@pobox.com>.
1936 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1937 OK for addresses that are the result of redirections.
1939 PH/48 A number of further additions for the benefit of the new test suite,
1940 including a fake gethostbyname() that interfaces to the fake DNS resolver
1943 TF/06 The fix for widen_domains has also been applied to qualify_single and
1944 search_parents which are the other dnslookup options that can cause
1947 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1950 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1952 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1953 service" if it encountered an MX record with an empty target hostname.
1954 The message is now "an MX or SRV record indicated no SMTP service".
1956 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1957 if the DNS of the sending site is misconfigured. This is quite a
1958 common situation. This change restores the behaviour of treating a
1959 helo verification defer as a failure.
1961 PH/51 If self=fail was set on a router, the bounce message did not include the
1962 actual error message.
1968 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1970 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1971 command itself ended in error, the underlying message about the transport
1972 filter timeout was being overwritten with the pipe command error. Now the
1973 underlying error message should be appended to the second error message.
1975 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1977 PH/02 Reduce the amount of output that "make" produces by default. Full output
1978 can still be requested.
1980 PH/03 The warning log line about a condition test deferring for a "warn" verb
1981 was being output only once per connection, rather than after each
1982 occurrence (because it was using the same function as for successful
1983 "warn" verbs). This seems wrong, so I have changed it.
1985 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1986 it should not have, which might have caused a crash in the right
1987 circumstances, but probably never did.
1989 PH/04 Installed a modified version of Tony Finch's patch to make submission
1990 mode fix the return path as well as the Sender: header line, and to
1991 add a /name= option so that you can make the user's friendly name appear
1994 TF/03 Added the control = fakedefer ACL modifier.
1996 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1997 Mark Lowes for thorough testing.
1999 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2001 TK/03 Merged latest SRS patch from Miles Wilton.
2003 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2004 with the definition in sysexits.h (which is #included earlier).
2005 Fortunately, Exim does not actually use EX_OK. The code used to try to
2006 preserve the sysexits.h value, by assumimg that macro definitions were
2007 scanned for macro replacements. I have been disabused of this notion,
2008 so now the code just undefines EX_OK before #including unistd.h.
2010 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2011 in the smtp transport. When a block could not be written in a single
2012 write() function, the timeout was being re-applied to each part-write.
2013 This seems wrong - if the receiver was accepting one byte at a time it
2014 would take for ever. The timeout is now adjusted when this happens. It
2015 doesn't have to be particularly precise.
2017 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2018 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2020 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2021 without a domain if the check to <postmaster@domain> fails.
2023 SC/01 Eximstats: added -xls and the ability to specify output files
2024 (patch written by Frank Heydlauf).
2026 SC/02 Eximstats: use FileHandles for outputing results.
2028 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2030 SC/04 Eximstats: fixed display of large numbers with -nvr option
2032 SC/05 Eximstats: fixed merging of reports with empty tables.
2034 SC/06 Eximstats: added the -include_original_destination flag
2036 SC/07 Eximstats: removed tabs and trailing whitespace.
2038 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2040 TK/06 MBOX spool code: Add real "From " MBOX separator line
2041 so the .eml file is really in mbox format (even though
2042 most programs do not really care). Patch from Alex Miller.
2044 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2045 The latter is generated from $received_to and is only set if the
2046 message has one envelope recipient. SA can use these headers,
2047 obviously out-of-the-box. Patch from Alex Miller.
2049 PH/08 The ${def test on a variable was returning false if the variable's
2050 value was "0", contrary to what the specification has always said!
2051 The result should be true unless the variable is empty.
2053 PH/09 The syntax error of a character other than { following "${if
2054 def:variable_name" (after optional whitespace) was not being diagnosed.
2055 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2056 accidental colon was present, for example, could give incorrect results.
2058 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2059 result is used (not including appendfile, where other changes are about
2062 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2063 This involved changing a lot of size variables from int to off_t. It
2064 should work with maildirs and everything.
2066 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2067 spamd dying while we are connected to it.
2069 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2072 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2073 function for BDB 4.3.
2075 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2077 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2078 the sources for PCRE have been re-arranged and more files are now
2081 PH/15 The code I had for printing potentially long long variables in PH/11
2082 above was not the best (it lost precision). The length of off_t variables
2083 is now inspected at build time, and an appropriate printing format (%ld
2084 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2085 to be "long long int" or "long int". This is needed for the internal
2086 formatting function string_vformat().
2088 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2089 the configuration file to be ":syslog", then the script "guesses" where
2090 the logs files are, rather than using the compiled in default. In our
2091 case the guess is not the same as the compiled default, so the script
2092 suddenly stopped working when I started to use syslog. The patch checks
2093 to see if log_file_path is "". If so, it attempts to read it from exim
2094 with no configuration file to get the compiled in version, before it
2095 falls back to the previous guessing code."
2097 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2098 implementing BATV in an Exim configuration. See NewStuff for the gory
2101 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2102 Makefile that are specific to HP-UX.
2104 PH/18 If the "use_postmaster" option was set for a recipient callout together
2105 with the "random" option, the postmaster address was used as the MAIL
2106 FROM address for the random test, but not for the subsequent recipient
2107 test. It is now used for both.
2109 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2110 patch removes a few documentation additions to RFC 3028, because the
2111 latest draft now contains them. It adds the new en;ascii-case comparator
2112 and a new error check for 8bit text in MIME parts. Comparator and
2113 require names are now matched exactly. I enabled the subaddress
2114 extension, but it is not well tested yet (read: it works for me)."
2116 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2117 rework some of the code of TK/09 above to avoid the hardwired use of
2118 "%lld" and "long long". Replaced the call to snprintf() with a call to
2121 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2122 records point to non-existent hosts", "retry timeout exceeded", and
2123 "retry time not reached for any host after a long failure period".
2125 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2126 experimental DomainKeys support:
2128 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2129 (2) On an error such as an illegally used "control", the wrong name for
2130 the control was given.
2132 These problems did NOT occur unless DomainKeys support was compiled.
2134 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2136 PH/24 Added ${if match_ip condition.
2138 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2139 databases so that it will be absolutely obvious if a crash occurs in the
2140 DB library. This is a regular occurrence (often caused by mis-matched
2143 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2144 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2145 picked up on a user's system that detects such things. There doesn't seem
2146 to be a gcc warning option for this - only an attribute that has to be
2147 put on the function's prototype. It seems that in Fedora Core 4 they have
2148 set this on a number of new functions. No doubt there will be more in due
2151 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2152 specify a transport. However, if an address that was verified by such a
2153 router was the subject of a callout, Exim crashed because it tried to
2154 read the rcpt_include_affixes from the non-existent transport. Now it
2155 just assumes that the setting of that option is false. This bug was
2156 introduced by 4.51/PH/31.
2158 PH/28 Changed -d+all to exclude +memory, because that information is very
2159 rarely of interest, but it makes the output a lot bigger. People tend to
2160 do -d+all out of habit.
2162 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2163 code in os-type was giving problems when libc.so lives in lib64, like on
2166 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2167 aren't the modern standard, and it seems that some systems' include files
2168 don't always have them. Exim was already checking for some of the newer
2169 ones like T_AAAA, and defining it itself. I've added checks for all the
2170 record types that Exim uses.
2172 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2173 not automatically generating a new one, as it is supposed to. This
2174 prevented TLS from working. If the file did exist, but contained invalid
2175 data, a new version was generated, as expected. It was only the case of a
2176 non-existent file that was broken.
2178 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2179 with a change in libdomainkeys > 0.64.
2181 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2182 from DNS. If the selector record carries the flag, it now has
2183 precedence over the domain-wide flag.
2185 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2187 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2188 the use of a port name that isn't defined in /etc/services) occurred, the
2189 message was deferred as in a normal delivery, and thus remained on the
2190 spool, instead of being failed because of the mua_wrapper setting. This
2191 is now fixed, and I tidied up some of the mua_wrapper messages at the
2194 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2195 the messages in a hash of arrays rather than using individual hashes.
2196 This is a bit cleaner and results in dramatic memory savings, albeit
2197 at a slight CPU cost.
2199 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2200 as requested by Marc Sherman.
2202 SC/10 Eximstats: added histograms for user specified patterns as requested
2205 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2207 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2208 fopen() in the content-scanning modules that did not already have it.
2214 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2215 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2217 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2219 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2221 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2222 to test Sieve filters that use "vacation".
2224 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2225 that changes the way the GnuTLS parameters are stored in the cache file.
2226 The new format can be generated externally. For backward compatibility,
2227 if the data in the cache doesn't make sense, Exim assumes it has read an
2228 old-format file, and it generates new data and writes a new file. This
2229 means that you can't go back to an older release without removing the
2232 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2233 work if there are any delivery delays because "one_time" forces the
2234 parent to be marked "delivered", so its unseen clone is never tried
2235 again. For this reason, Exim now forbids the simultaneous setting of
2238 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2239 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2240 entry if you want to know the details. The fix, however, neglected to
2241 consider the case when local delivery batching is involved. The test for
2242 "previously delivered" was not happening when checking to see if an
2243 address could be batched with a previous (undelivered) one; under
2244 certain circumstances this could lead to multiple deliveries to the same
2247 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2248 in its include files, and this causes problems building Exim.
2250 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2251 header_syntax) but Exim was just ignoring anything given after a slash.
2252 In particular, this caused confusion with an attempt to use "verify =
2253 reverse_host_lookup/defer_ok". An error is now given when options are
2254 supplied for verify items that do not have them. (Maybe reverse_host_
2255 lookup should have a defer_ok option, but that's a different point.)
2257 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2258 defined by RFC 821) to 2048, because there were problems with some AUTH
2259 commands, and RFC 1869 says the size should be increased for extended
2260 SMTP commands that take arguments.
2262 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2265 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2266 "unknown" error; now it says that the functionality isn't in the binary.
2268 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2269 an address' error message when a string expansion fails (syntax or
2270 whatever). Otherwise the password may appear in the log. Following change
2271 PH/42 below, there is no longer a chance of it appearing in a bounce
2274 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2276 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2277 of its addresses was ever tried. (Bugzilla bug #2.)
2279 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2280 the result incorrectly in the debug output. (It correctly added a newline
2281 to what was transported.)
2283 TF/01 Added $received_time.
2285 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2286 commented out examples of how to interface to a virus scanner and to
2287 SpamAssassin. Also added commented examples of av_scanner and
2288 spamd_address settings.
2290 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2291 and controls are allowed in which ACLs. There were a couple of minor
2292 errors. Some of the entries in the conditions table (which is a table of
2293 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2294 negation of where the condition IS allowed.
2296 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2298 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2299 header file does not have a version number, so I've had to invent a new
2300 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2301 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2302 radiusclient), but it was contributed by a Radius user.
2304 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2305 files or directories, for OpenSSL.
2307 PH/19 When an Exim process that is running as root has to create an Exim log
2308 file, it does so in a subprocess that runs as exim:exim so as to get the
2309 ownership right at creation (otherwise, other Exim processes might see
2310 the file with the wrong ownership). There was no test for failure of this
2311 fork() call, which would lead to the process getting stuck as it waited
2312 for a non-existent subprocess. Forks do occasionally fail when resources
2313 run out. I reviewed all the other calls to fork(); they all seem to check
2316 PH/20 When checking for unexpected SMTP input at connect time (before writing
2317 the banner), Exim was not dealing correctly with a non-positive return
2318 from the read() function. If the client had disconnected by this time,
2319 the result was a log entry for a synchronization error with an empty
2320 string after "input=" when read() returned zero. If read() returned -1
2321 (an event I could not check), uninitialized data bytes were printed.
2322 There were reports of junk text (parts of files, etc) appearing after
2325 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2327 PH/22 Added support for macro redefinition, and (re)definition in between
2328 driver and ACL definitions.
2330 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2331 forgetting to use the resulting value; it was using the unexpanded value.
2333 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2334 hadn't been configured. The fix is from Juergen Kreileder, who
2335 understands it better than I do:
2337 "Here's what I see happening with three configured cyrus_sasl
2338 authenticators configured (plain, login, cram-md5):
2340 On startup auth_cyrus_sasl_init() gets called for each of these.
2341 This means three calls to sasl_listmech() without a specified mech_list.
2342 => SASL tests which mechs of all available mechs actually work
2343 => three warnings about OTP not working
2344 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2346 With the patch, sasl_listmech() also gets called three times. But now
2347 SASL's mech_list option is set to the server_mech specified in the the
2348 authenticator. Or in other words, the answer from sasl_listmech()
2349 gets limited to just the mech you're testing for (which is different
2351 => the return list contains just 'plain' or 'login', 'cram-md5' or
2352 nothing depending on the value of ob->server_mech.
2354 I've just tested the patch: Authentication still works fine,
2355 unavailable mechs specified in the exim configuration are still
2356 caught, and the auth.log warnings about OTP are gone."
2358 PH/25 When debugging is enabled, the contents of the command line are added
2359 to the debugging output, even when log_selector=+arguments is not
2362 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2363 answer is "GNU", and only if the return is "GNU/something" is the answer
2366 PH/27 $acl_verify_message is now set immediately after the failure of a
2367 verification in an ACL, and so is available in subsequent modifiers. In
2368 particular, the message can be preserved by coding like this:
2370 warn !verify = sender
2371 set acl_m0 = $acl_verify_message
2373 Previously, $acl_verify_message was set only while expanding "message"
2374 and "log_message" when a very denied access.
2376 PH/28 Modified OS/os.c-Linux with
2378 -#ifndef OS_LOAD_AVERAGE
2379 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2381 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2382 nomenclature these days.)
2384 PH/29 Installed patch from the Sieve maintainer that adds the options
2385 sieve_useraddress and sieve_subaddress to the redirect router.
2387 PH/30 In these circumstances:
2388 . Two addresses routed to the same list of hosts;
2389 . First host does not offer TLS;
2390 . First host accepts first address;
2391 . First host gives temporary error to second address;
2392 . Second host offers TLS and a TLS session is established;
2393 . Second host accepts second address.
2394 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2395 and peerdn, if requested) that were in fact used only for the second
2398 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2399 attention to any local part prefix or suffix that was matched by the
2400 router that accepted the address. It now behaves in the same way as it
2401 does for delivery: the affixes are removed from the local part unless
2402 rcpt_include_affixes is set on the transport.
2404 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2405 timeout during the DATA phase of an incoming message.
2407 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2408 applied a patch sanctioned by the Sieve maintainer.
2410 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2411 the uid or gid is negative. A case of a negative gid caused this to be
2412 noticed. The fix allows for either to be negative.
2414 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2415 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2418 PH/36 The change PH/12 above was broken. Fixed it.
2420 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2421 the grounds that routing the same address twice would always produce the
2422 same answer. This might have been true once, but it is certainly no
2423 longer true now. Routing a child address may depend on the previous
2424 routing that produced that child. Some complicated redirection strategies
2425 went wrong when messages had multiple recipients, and made Exim's
2426 behaviour dependent on the order in which the addresses were given.
2428 I have moved the duplicate checking until after the routing is complete.
2429 Exim scans the addresses that are assigned to local and remote
2430 transports, and removes any duplicates. This means that more work will be
2431 done, as duplicates will always all be routed, but duplicates are
2432 presumably rare, so I don't expect this is of any significance.
2434 For deliveries to pipes, files, and autoreplies, the duplicate checking
2435 still happens during the routing process, since they are not going to be
2438 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2439 It corrects a timeout issue with spamd. This is Ian's comment: "The
2440 background is that sometimes spamd either never reads data from a
2441 connection it has accepted, or it never writes response data. The exiscan
2442 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2443 blindly assumes that writes won't block so it may never time out."
2445 PH/39 Allow G after quota size as well as K and M.
2447 PH/40 The value set for $authenticated_id in an authenticator may not contain
2448 binary zeroes or newlines because the value is written to log lines and
2449 to spool files. There was no check on this. Now the value is run through
2450 the string_printing() function so that such characters are converted to
2451 printable escape sequences.
2453 PH/41 $message_linecount is a new variable that contains the total number of
2454 lines in the message. Compare $body_linecount, which is the count for the
2457 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2458 bounce and delay warning messages, except in certain special cases, which
2461 (a) An SMTP error message from a remote host;
2462 (b) A message specified in a :fail: redirection;
2463 (c) A message specified in a "fail" command in a system filter;
2464 (d) A message specified in a FAIL return from the queryprogram router;
2465 (e) A message specified by the cannot_route_message router option.
2467 In these cases only, Exim does include the error details in bounce and
2468 warning messages. There are also a few cases where bland messages such
2469 as "unrouteable address" or "local delivery error" are given.
2471 PH/43 $value is now also set for the "else" part of a ${run expansion.
2473 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2474 being worked on, but at least Exim now implements the latest version to
2477 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2478 process to complete was treated as a delivery failure, a timeout while
2479 writing the message to the pipe was logged, but erroneously treated as a
2480 successful delivery. Such timeouts include transport filter timeouts. For
2481 consistency with the overall process timeout, these timeouts are now
2482 treated as errors, giving rise to delivery failures by default. However,
2483 there is now a new Boolean option for the pipe transport called
2484 timeout_defer, which, if set TRUE, converts the failures into defers for
2485 both kinds of timeout. A transport filter timeout is now identified in
2488 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2489 systems where "make" and "gmake" are different, calling "gmake" at top
2490 level broke things. I've arranged for the value of $(MAKE) to be passed
2491 from the Makefile to this script so that it can call the same version of
2495 A note about Exim versions 4.44 and 4.50
2496 ----------------------------------------
2498 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2499 changes of various kinds. As a consequence, a big documentation update was
2500 needed. This delayed the release for rather longer than seemed good, especially
2501 in the light of a couple of (minor) security issues. Therefore, the changes
2502 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2503 release. So 4.44 and 4.50 are in effect two different branches that both start
2506 I have left the 4.50 change log unchanged; it contains all the changes since
2507 4.43. The change log for 4.44 is below; many of its items are identical to
2508 those for 4.50. This seems to be the most sensible way to preserve the
2509 historical information.
2515 1. Minor wording change to the doc/README.SIEVE file.
2517 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2518 computation of the current number of files was incorrect.
2520 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2521 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2524 4. Give more explanation in the error message when the command for a transport
2525 filter fails to execute.
2527 5. There are several places where Exim runs a non-Exim command in a
2528 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2529 was being done only for the command run by the queryprogram router. It is
2530 now done for all such subprocesses. The other cases are: ${run, transport
2531 filters, and the commands run by the lmtp and pipe transports.
2533 6. Added CONFIGURE_GROUP build-time option.
2535 7. Some older OS have a limit of 256 on the maximum number of file
2536 descriptors. Exim was using setrlimit() to set 1000 as a large value
2537 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2538 systems. I've change it so that if it can't get 1000, it tries for 256.
2540 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2541 was an oversight, and furthermore, ever since the addition of extra
2542 controls (e.g. 4.43/32), the checks on when to allow different forms of
2543 "control" were broken. There should now be diagnostics for all cases when a
2544 control that does not make sense is encountered.
2546 9. Added the /retain_sender option to "control=submission".
2548 10. $recipients is now available in the predata ACL (oversight).
2550 11. Tidy the search cache before the fork to do a delivery from a message
2551 received from the command line. Otherwise the child will trigger a lookup
2552 failure and thereby defer the delivery if it tries to use (for example) a
2553 cached ldap connection that the parent has called unbind on.
2555 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2556 of $address_data from the recipient verification was clobbered by the
2557 sender verification.
2559 13. The value of address_data from a sender verification is now available in
2560 $sender_address_data in subsequent conditions in the ACL statement.
2562 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2564 15. Added a new option "connect=<time>" to callout options, to set a different
2567 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2568 was its contents. (It was OK if the option was not defined at all.)
2570 17. A "Completed" log line is now written for messages that are removed from
2571 the spool by the -Mrm option.
2573 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2574 information about exactly what failed.
2576 19. Added -dd to debug only the daemon process.
2578 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2579 handles timeouts, both on the server side and network timeouts. Renamed the
2580 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2582 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2583 It is now set to "smtps".
2585 22. $host_address is now set to the target address during the checking of
2586 ignore_target_hosts.
2588 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2589 being passed; this would have caused $sender_host_name to have been used if
2590 matching the list had actually called for a host name (not very likely,
2591 since this list is usually IP addresses). A host name is now passed as
2594 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2595 code that specifies a non-restarting handler (typically sigaction() in
2596 modern systems) in an attempt to fix a rare and obscure crash bug.
2598 25. Narrowed the window for a race in the daemon that could cause it to ignore
2599 SIGCHLD signals. This is not a major problem, because they are used only to
2600 wake it up if nothing else does.
2602 26. A malformed maildirsize file could cause Exim to calculate negative values
2603 for the mailbox size or file count. Odd effects could occur as a result.
2604 The maildirsize information is now recalculated if the size or filecount
2607 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2608 support for a long time. Removed HAVE_SYS_VFS_H.
2610 28. Installed the latest version of exipick from John Jetmore.
2612 29. In an address list, if the pattern was not a regular expression, an empty
2613 subject address (from a bounce message) matched only if the pattern was an
2614 empty string. Non-empty patterns were not even tested. This was the wrong
2615 because it is perfectly reasonable to use an empty address as part of a
2616 database query. An empty address is now tested by patterns that are
2617 lookups. However, all the other forms of pattern expect the subject to
2618 contain a local part and a domain, and therefore, for them, an empty
2619 address still always fails if the pattern is not itself empty.
2621 30. Exim went into a mad DNS loop when attempting to do a callout where the
2622 host was specified on an smtp transport, and looking it up yielded more
2623 than one IP address.
2625 31. Re-factored the code for checking spool and log partition space into a
2626 function that finds that data and another that does the check. The former
2627 is then used to implement four new variables: $spool_space, $log_space,
2628 $spool_inodes, and $log_inodes.
2630 32. The RFC2047 encoding function was originally intended for short strings
2631 such as real names; it was not keeping to the 75-character limit for
2632 encoded words that the RFC imposes. It now respects the limit, and
2633 generates multiple encoded words if necessary. To be on the safe side, I
2634 have increased the buffer size for the ${rfc2047: expansion operator from
2637 33. It is now permitted to omit both strings after an "if" condition; if the
2638 condition is true, the result is "true". As before, when the second string
2639 is omitted, a false condition yields an empty string. This makes it less
2640 cumbersome to write custom ACL and router conditions.
2642 34. Failure to deliver a bounce message always caused it to be frozen, even if
2643 there was an errors_to setting on the router. The errors_to setting is now
2646 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2647 canonical form (fully expanded) before being placed in
2648 $sender_host_address.
2650 36. The table in the code that translates DNS record types into text (T_A to
2651 "A" for instance) was missing entries for NS and CNAME. It is just possible
2652 that this could have caused confusion if both these types were looked up
2653 for the same domain, because the text type is used as part of Exim's
2654 per-process caching. But the chance of anyone hitting this buglet seems
2657 37. The dnsdb lookup has been extended in a number of ways.
2659 (1) There is a new type, "zns", which walks up the domain tree until it
2660 finds some nameserver records. It should be used with care.
2662 (2) There is a new type, "mxh", which is like "mx" except that it returns
2663 just the host names, not the priorities.
2665 (3) It is now possible to give a list of domains (or IP addresses) to be
2666 looked up. The behaviour when one of the lookups defers can be
2667 controlled by a keyword.
2669 (4) It is now possible to specify the separator character for use when
2670 multiple records are returned.
2672 38. The dnslists ACL condition has been extended: it is now possible to supply
2673 a list of IP addresses and/or domains to be looked up in a particular DNS
2676 39. Added log_selector=+queue_time_overall.
2678 40. When running the queue in the test harness, wait just a tad after forking a
2679 delivery process, to get repeatability of debugging output.
2681 41. Include certificate and key file names in error message when GnuTLS fails
2682 to set them up, because the GnuTLS error message doesn't include the name
2683 of the failing file when there is a problem reading it.
2685 42. Allow both -bf and -bF in the same test run.
2687 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2689 44. The "Exiscan patch" is now merged into the mainline Exim source.
2691 45. Sometimes the final signoff response after QUIT could fail to get
2692 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2693 < 0 before doing a fflush(). This bug looks as though it goes back to the
2694 introduction of TLS in release 3.20, but "sometimes" must have been rare
2695 because the tests only now provoked it.
2697 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2698 (this can affect the format of dates).
2700 47. exim_tidydb, when checking for the continued existence of a message for
2701 which it has found a message-specific retry record, was not finding
2702 messages that were in split spool directories. Consequently, it was
2703 deleting retry records that should have stayed in existence.
2705 48. Steve fixed some bugs in eximstats.
2707 49. The SPA authentication driver was not abandoning authentication and moving
2708 on to the next authenticator when an expansion was forced to fail,
2709 contradicting the general specification for all authenticators. Instead it
2710 was generating a temporary error. It now behaves as specified.
2712 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2713 (the order specifies the preference for clients). The order is now AES256,
2714 AES128, 3DES, ARCFOUR128.
2716 51. Small patch to Sieve code - explicitly set From: when generating an
2719 52. Exim crashed if a remote delivery caused a very long error message to be
2720 recorded - for instance if somebody sent an entire SpamAssassin report back
2721 as a large number of 550 error lines. This bug was coincidentally fixed by
2722 increasing the size of one of Exim's internal buffers (big_buffer) that
2723 happened as part of the Exiscan merge. However, to be on the safe side, I
2724 have made the code more robust (and fixed the comments that describe what
2727 53. Now that there can be additional text after "Completed" in log lines (if
2728 the queue_time_overall log selector is set), a one-byte patch to exigrep
2729 was needed to allow it to recognize "Completed" as not the last thing in
2732 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2733 patch that reportedly fixes this has been added. I am not expert enough to
2734 create a test for it. This is what the patch creator wrote:
2736 "I found a little strange behaviour of ldap code when working with
2737 Windows 2003 AD Domain, where users was placed in more than one
2738 Organization Units. When I tried to give exim partial DN, the exit code
2739 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2740 But simultaneously result of request was absolutely normal ldap result,
2741 so I produce this patch..."
2743 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2744 REFERENCE, so I have modified the code to exclude the patch when that macro
2747 55. Some experimental protocols are using DNS PTR records for new purposes. The
2748 keys for these records are domain names, not reversed IP addresses. The
2749 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2750 leaves it alone. Component reversal etc. now happens only for IP addresses.
2753 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2755 57. Double the size of the debug message buffer (to 2048) so that more of very
2756 long debug lines gets shown.
2758 58. The exicyclog utility now does better if the number of log files to keep
2759 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2761 59. Two changes related to the smtp_active_hostname option:
2763 (1) $smtp_active_hostname is now available as a variable.
2764 (2) The default for smtp_banner uses $smtp_active_hostname instead
2765 of $primary_hostname.
2767 60. The host_aton() function is supposed to be passed a string that is known
2768 to be a valid IP address. However, in the case of IPv6 addresses, it was
2769 not checking this. This is a hostage to fortune. Exim now panics and dies
2770 if the condition is not met. A case was found where this could be provoked
2771 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2772 components; fortuitously, this particular loophole had already been fixed
2773 by change 4.50/55 above.
2775 If there are any other similar loopholes, the new check in host_aton()
2776 itself should stop them being exploited. The report I received stated that
2777 data on the command line could provoke the exploit when Exim was running as
2778 exim, but did not say which command line option was involved. All I could
2779 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2780 running as the user.
2783 61. There was a buffer overflow vulnerability in the SPA authentication code
2784 (which came originally from the Samba project). I have added a test to the
2785 spa_base64_to_bits() function which I hope fixes it.
2788 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2789 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2791 63. The daemon start-up calls getloadavg() while still root for those OS that
2792 need the first call to be done as root, but it missed one case: when
2793 deliver_queue_load_max is set with deliver_drop_privilege. This is
2794 necessary for the benefit of the queue runner, because there is no re-exec
2795 when deliver_drop_privilege is set.
2797 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2798 This has been fixed.
2800 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2801 was in use, was not putting the data itself into the right store pool;
2802 consequently, it could be overwritten for a subsequent message in the same
2803 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2806 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2808 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2809 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2810 the function were treating the return as a boolean value, which happened to
2811 work because 0=false and not-0=true, but is not correct code.
2813 68. The host_aton() function was not handling scoped IPv6 addresses (those
2814 with, for example, "%eth0" on the end) correctly.
2816 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2817 negated items (that is, ~something) in unsigned ints. Some compilers
2818 apparently mutter when there is no cast.
2820 70. If an address verification called from an ACL failed, and did not produce a
2821 user-specific message (i.e. there was only a "system" message), nothing was
2822 put in $acl_verify_message. In this situation, it now puts the system
2825 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2826 session; change 4.31/43 added the unwanted input to the log line - except
2827 that it did not do this in the start of session case. It now does.
2829 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2830 This is wrong and can cause the other end to generate a synchronization
2831 error if it is another Exim or anything else that does the synchronization
2832 check. A QUIT command is no longer sent after a timeout.
2834 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2835 during host lookups.
2837 74. The defer_ok option of callout verification was not working if it was used
2838 when verifying addresses in header lines, that is, for this case:
2840 verify = header_sender/callout=defer_ok
2842 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2843 those file descriptors could be used for SMTP connections. If anything
2844 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2845 could be sent to the SMTP client, causing chaos. The daemon now opens
2846 stdin, stdout, and stderr to /dev/null when it puts itself into the
2849 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2850 log by default. The user can override this with suitable Perl magic.
2852 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2853 the log message when discard triggers, was not working for the DATA ACL or
2854 for the non-SMTP ACL.
2856 78. Error message wording change in sieve.c.
2858 79. If smtp_accept_max_per_host was set, the number of connections could be
2859 restricted to fewer than expected, because the daemon was trying to set up
2860 a new connection before checking whether the processes handling previous
2861 connections had finished. The check for completed processes is now done
2862 earlier. On busy systems, this bug wouldn't be noticed because something
2863 else would have woken the daemon, and it would have reaped the completed
2866 80. If a message was submitted locally by a user whose login name contained one
2867 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2868 It caused a spool format error. I have fixed the spool reading code. A
2869 related problem was that the "from" clause in the Received: line became
2870 illegal because of the space(s). It is now covered by ${quote_local_part.
2872 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2875 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2877 83. Patch from Sieve maintainer to fix unterminated string problem in
2878 "vacation" handling.
2880 84. Some minor changes to the Linux configuration files to help with other
2881 OS variants using glibc.
2883 85. One more patch for Sieve to update vacation handling to latest spec.
2886 ----------------------------------------------------
2887 See the note above about the 4.44 and 4.50 releases.
2888 ----------------------------------------------------
2894 1. Change 4.43/35 introduced a bug that caused file counts to be
2895 incorrectly computed when quota_filecount was set in an appendfile
2898 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2899 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2902 3. Give more explanation in the error message when the command for a transport
2903 filter fails to execute.
2905 4. There are several places where Exim runs a non-Exim command in a
2906 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2907 was being done only for the command run by the queryprogram router. It is
2908 now done for all such subprocesses. The other cases are: ${run, transport
2909 filters, and the commands run by the lmtp and pipe transports.
2911 5. Some older OS have a limit of 256 on the maximum number of file
2912 descriptors. Exim was using setrlimit() to set 1000 as a large value
2913 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2914 systems. I've change it so that if it can't get 1000, it tries for 256.
2916 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2917 was an oversight, and furthermore, ever since the addition of extra
2918 controls (e.g. 4.43/32), the checks on when to allow different forms of
2919 "control" were broken. There should now be diagnostics for all cases when a
2920 control that does not make sense is encountered.
2922 7. $recipients is now available in the predata ACL (oversight).
2924 8. Tidy the search cache before the fork to do a delivery from a message
2925 received from the command line. Otherwise the child will trigger a lookup
2926 failure and thereby defer the delivery if it tries to use (for example) a
2927 cached ldap connection that the parent has called unbind on.
2929 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2930 of $address_data from the recipient verification was clobbered by the
2931 sender verification.
2933 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2934 was its contents. (It was OK if the option was not defined at all.)
2936 11. A "Completed" log line is now written for messages that are removed from
2937 the spool by the -Mrm option.
2939 12. $host_address is now set to the target address during the checking of
2940 ignore_target_hosts.
2942 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2943 being passed; this would have caused $sender_host_name to have been used if
2944 matching the list had actually called for a host name (not very likely,
2945 since this list is usually IP addresses). A host name is now passed as
2948 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2949 code that specifies a non-restarting handler (typically sigaction() in
2950 modern systems) in an attempt to fix a rare and obscure crash bug.
2952 15. Narrowed the window for a race in the daemon that could cause it to ignore
2953 SIGCHLD signals. This is not a major problem, because they are used only to
2954 wake it up if nothing else does.
2956 16. A malformed maildirsize file could cause Exim to calculate negative values
2957 for the mailbox size or file count. Odd effects could occur as a result.
2958 The maildirsize information is now recalculated if the size or filecount
2961 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2962 support for a long time. Removed HAVE_SYS_VFS_H.
2964 18. Updated exipick to current release from John Jetmore.
2966 19. Allow an empty sender to be matched against a lookup in an address list.
2967 Previously the only cases considered were a regular expression, or an
2970 20. Exim went into a mad DNS lookup loop when doing a callout where the
2971 host was specified on the transport, if the DNS lookup yielded more than
2974 21. The RFC2047 encoding function was originally intended for short strings
2975 such as real names; it was not keeping to the 75-character limit for
2976 encoded words that the RFC imposes. It now respects the limit, and
2977 generates multiple encoded words if necessary. To be on the safe side, I
2978 have increased the buffer size for the ${rfc2047: expansion operator from
2981 22. Failure to deliver a bounce message always caused it to be frozen, even if
2982 there was an errors_to setting on the router. The errors_to setting is now
2985 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2986 canonical form (fully expanded) before being placed in
2987 $sender_host_address.
2989 24. Updated eximstats to version 1.33
2991 25. Include certificate and key file names in error message when GnuTLS fails
2992 to set them up, because the GnuTLS error message doesn't include the name
2993 of the failing file when there is a problem reading it.
2995 26. Expand error message when OpenSSL has problems setting up cert/key files.
2998 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2999 (this can affect the format of dates).
3001 28. exim_tidydb, when checking for the continued existence of a message for
3002 which it has found a message-specific retry record, was not finding
3003 messages that were in split spool directories. Consequently, it was
3004 deleting retry records that should have stayed in existence.
3006 29. eximstats updated to version 1.35
3007 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3008 1.35 - bugfix such that pie charts by volume are generated correctly
3010 30. The SPA authentication driver was not abandoning authentication and moving
3011 on to the next authenticator when an expansion was forced to fail,
3012 contradicting the general specification for all authenticators. Instead it
3013 was generating a temporary error. It now behaves as specified.
3015 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3016 (the order specifies the preference for clients). The order is now AES256,
3017 AES128, 3DES, ARCFOUR128.
3019 31. Small patch to Sieve code - explicitly set From: when generating an
3022 32. Exim crashed if a remote delivery caused a very long error message to be
3023 recorded - for instance if somebody sent an entire SpamAssassin report back
3024 as a large number of 550 error lines. This bug was coincidentally fixed by
3025 increasing the size of one of Exim's internal buffers (big_buffer) that
3026 happened as part of the Exiscan merge. However, to be on the safe side, I
3027 have made the code more robust (and fixed the comments that describe what
3030 33. Some experimental protocols are using DNS PTR records for new purposes. The
3031 keys for these records are domain names, not reversed IP addresses. The
3032 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3033 leaves it alone. Component reversal etc. now happens only for IP addresses.
3036 34. The host_aton() function is supposed to be passed a string that is known
3037 to be a valid IP address. However, in the case of IPv6 addresses, it was
3038 not checking this. This is a hostage to fortune. Exim now panics and dies
3039 if the condition is not met. A case was found where this could be provoked
3040 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3041 components; fortuitously, this particular loophole had already been fixed
3042 by change 4.50/55 or 4.44/33 above.
3044 If there are any other similar loopholes, the new check in host_aton()
3045 itself should stop them being exploited. The report I received stated that
3046 data on the command line could provoke the exploit when Exim was running as
3047 exim, but did not say which command line option was involved. All I could
3048 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3049 running as the user.
3052 35. There was a buffer overflow vulnerability in the SPA authentication code
3053 (which came originally from the Samba project). I have added a test to the
3054 spa_base64_to_bits() function which I hope fixes it.
3057 36. The daemon start-up calls getloadavg() while still root for those OS that
3058 need the first call to be done as root, but it missed one case: when
3059 deliver_queue_load_max is set with deliver_drop_privilege. This is
3060 necessary for the benefit of the queue runner, because there is no re-exec
3061 when deliver_drop_privilege is set.
3063 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3064 was in use, was not putting the data itself into the right store pool;
3065 consequently, it could be overwritten for a subsequent message in the same
3066 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3069 38. Sometimes the final signoff response after QUIT could fail to get
3070 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3071 < 0 before doing a fflush(). This bug looks as though it goes back to the
3072 introduction of TLS in release 3.20, but "sometimes" must have been rare
3073 because the tests only now provoked it.
3079 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3080 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3081 Now it returns an integer. A number of calls were still expecting a T/F
3082 return. Fortuitously, in all cases, the tests worked in OK situations,
3083 which is the norm. However, things would have gone wrong on any write
3084 failures on the smtp file descriptor. This function is used when sending
3085 messages over SMTP and also when doing verify callouts.
3087 2. When Exim is called to do synchronous delivery of a locally submitted
3088 message (the -odf or -odi options), it no longer closes stderr before doing
3091 3. Implemented the mua_wrapper option.
3093 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3095 5. Implemented the functions header_remove(), header_testname(),
3096 header_add_at_position(), and receive_remove_recipient(), and exported them
3099 6. If an ACL "warn" statement specified the addition of headers, Exim already
3100 inserted X-ACL-Warn: at the start if there was no header name. However, it
3101 was not making this test for the second and subsequent header lines if
3102 there were newlines in the string. This meant that an invalid header could
3103 be inserted if Exim was badly configured.
3105 7. Allow an ACL "warn" statement to add header lines at the start or after all
3106 the Received: headers, as well as at the end.
3108 8. Added the rcpt_4xx retry error code.
3110 9. Added postmaster_mailfrom=xxx to callout verification option.
3112 10. Added mailfrom=xxxx to the callout verification option, for verify=
3115 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3116 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3118 12. Inserted some casts to stop certain compilers warning when using pointer
3119 differences as field lengths or precisions in printf-type calls (mostly
3120 affecting debugging statements).
3122 13. Added optional readline() support for -be (dynamically loaded).
3124 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3125 same clock tick as a message's arrival, so that its received time was the
3126 same as the "first fail" time on the retry record, and that message
3127 remained on the queue past the ultimate address timeout, every queue runner
3128 would try a delivery (because it was past the ultimate address timeout) but
3129 after another failure, the ultimate address timeout, which should have then
3130 bounced the address, did not kick in. This was a "< instead of <=" error;
3131 in most cases the first failure would have been in the next clock tick
3132 after the received time, and all would be well.
3134 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3135 being recognized when the domain list was tested by the match_domain
3136 condition in an expansion string.
3138 16. Added the ${str2b64: operator.
3140 17. Exim was always calling setrlimit() to set a large limit for the number of
3141 processes, without checking whether the existing limit was already
3142 adequate. (It did check for the limit on file descriptors.) Furthermore,
3143 errors from getrlimit() and setrlimit() were being ignored. Now they are
3144 logged to the main and panic logs, but Exim does carry on, to try to do its
3145 job under whatever limits there are.
3147 18. Imported PCRE 5.0.
3149 19. Trivial typo in log message " temporarily refused connection" (the leading
3152 20. If the log selector return_path_on_delivery was set and an address was
3153 redirected to /dev/null, the delivery process crashed because it assumed
3154 that a return path would always be set for a "successful" delivery. In this
3155 case, the whole delivery is bypassed as an optimization, and therefore no
3158 21. Internal re-arrangement: the function for sending a challenge and reading
3159 a response while authentication was assuming a zero-terminated challenge
3160 string. It's now changed to take a pointer and a length, to allow for
3161 binary data in such strings.
3163 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3165 23. Exim was not respecting finduser_retries when seeking the login of the
3166 uid under which it was called; it was always trying 10 times. (The default
3167 setting of finduser_retries is zero.) Also, it was sleeping after the final
3168 failure, which is pointless.
3170 24. Implemented tls_on_connect_ports.
3172 25. Implemented acl_smtp_predata.
3174 26. If the domain in control=submission is set empty, Exim assumes that the
3175 authenticated id is a complete email address when it generates From: or
3176 Sender: header lines.
3178 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3179 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3180 chown and chgrp in /bin and hostname in /usr/bin.
3182 28. Exim was keeping the "process log" file open after each use, just as it
3183 does for the main log. This opens the possibility of it remaining open for
3184 long periods when the USR1 signal hits a daemon. Occasional processlog
3185 errors were reported, that could have been caused by this. Anyway, it seems
3186 much more sensible not to leave this file open at all, so that is what now
3189 29. The long-running daemon process does not normally write to the log once it
3190 has entered its main loop, and it closes the log before doing so. This is
3191 so that log files can straightforwardly be renamed and moved. However,
3192 there are a couple of unusual error situations where the daemon does write
3193 log entries, and I had neglected to close the log afterwards.
3195 30. The text of an SMTP error response that was received during a remote
3196 delivery was being truncated at 512 bytes. This is too short for some of
3197 the long messages that one sometimes sees. I've increased the limit to
3200 31. It is now possible to make retry rules that apply only when a message has a
3201 specific sender, in particular, an empty sender.
3203 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3204 it possible to be selective about when SMTP synchronization is enforced.
3206 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3208 32. Implemented hosts_connection_nolog.
3210 33. Added an ACL for QUIT.
3212 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3215 35. Added mailbox_size and mailbox_filecount to appendfile.
3217 36. Added control = no_multiline_responses to ACLs.
3219 37. There was a bug in the logic of the code that waits for the clock to tick
3220 in the case where the clock went backwards by a substantial amount such
3221 that the microsecond fraction of "now" was more than the microsecond
3222 fraction of "then" (but the whole seconds number was less).
3224 38. Added support for the libradius Radius client library this is found on
3225 FreeBSD (previously only the radiusclient library was supported).
3231 1. When certain lookups returned multiple values in the form name=value, the
3232 quoting of the values was not always being done properly. Specifically:
3233 (a) If the value started with a double quote, but contained no whitespace,
3235 (b) If the value contained whitespace other than a space character (i.e.
3236 tabs or newlines or carriage returns) it was not quoted.
3237 This fix has been applied to the mysql and pgsql lookups by writing a
3238 separate quoting function and calling it from the lookup code. The fix
3239 should probably also be applied to nisplus, ibase and oracle lookups, but
3240 since I cannot test any of those, I have not disturbed their existing code.
3242 2. A hit in the callout cache for a specific address caused a log line with no
3243 reason for rejecting RCPT. Now it says "Previous (cached) callout
3244 verification failure".
3246 3. There was an off-by-one bug in the queryprogram router. An over-long
3247 return line was truncated at 256 instead of 255 characters, thereby
3248 overflowing its buffer with the terminating zero. As well as fixing this, I
3249 have increased the buffer size to 1024 (and made a note to document this).
3251 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3252 when Exim is waiting for an SMTP response from a remote server, Exim
3253 restarts its select() call on the socket, thereby resetting its timeout.
3254 This is not a problem when such interrupts are rare. Somebody set up a cron
3255 job to run exiwhat every 2 minutes, which is less than the normal select()
3256 timeout (5 or 10 minutes). This meant that the select() timeout never
3257 kicked in because it was always reset. I have fixed this by comparing the
3258 time when an interrupt arrives with the time at the start of the first call
3259 to select(). If more time than the timeout has elapsed, the interrupt is
3260 treated as a timeout.
3262 5. Some internal re-factoring in preparation for the addition of Sieve
3263 extensions (by MH). In particular, the "personal" test is moved to a
3264 separate function, and given an option for scanning Cc: and Bcc: (which is
3265 not set for Exim filters).
3267 6. When Exim created an email address using the login of the caller as the
3268 local part (e.g. when creating a From: or Sender: header line), it was not
3269 quoting the local part when it contained special characters such as @.
3271 7. Installed new OpenBSD configuration files.
3273 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3274 try to make them clearer.
3276 9. Callout options, other than the timeout value, were being ignored when
3277 verifying sender addresses in header lines. For example, when using
3279 verify = header_sender/callout=no_cache
3281 the cache was (incorrectly) being used.
3283 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3284 only the Cygwin environment.
3286 11. When return_path_on_delivery was set as a log selector, if different remote
3287 addresses in the same message used different return paths and parallel
3288 remote delivery occurred, the wrong values would sometimes be logged.
3289 (Whenever a remote delivery process finished, the return path value from
3290 the most recently started remote delivery process was logged.)
3292 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3293 lines when AUTH and/or TLS are in use. This is the "received protocol"
3294 field. Exim used to use "asmtp" for authenticated SMTP, without any
3295 indication (in the protocol name) for TLS use. Now it follows the RFC and
3296 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3297 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3298 names appear in log lines as well as in Received: header lines.
3300 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3301 extensions, and comparison tests, and to fix some bugs.
3303 14. Changes to the "personal" filter test:
3305 (1) The test was buggy in that it was just doing the equivalent of
3306 "contains" tests on header lines. For example, if a user's address was
3307 anne@some.where, the "personal" test would incorrectly be true for
3309 To: susanne@some.where
3311 This test is now done by extracting each address from the header in turn,
3312 and checking the entire address. Other tests that are part of "personal"
3313 are now done using regular expressions (for example, to check local parts
3314 of addresses in From: header lines).
3316 (2) The list of non-personal local parts in From: addresses has been
3317 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3318 taken from the Sieve specification recommendations.
3320 (3) If the message contains any header line starting with "List-" it is
3321 treated as non-personal.
3323 (4) The test for "circular" in the Subject: header line has been removed
3324 because it now seems ill-conceived.
3326 15. Minor typos in src/EDITME comments corrected.
3328 16. Installed latest exipick from John Jetmore.
3330 17. If headers_add on a router specified a text string that was too long for
3331 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3332 of string_sprintf() is now avoided.
3334 18. $message_body_size was not set (it was always zero) when running the DATA
3335 ACL and the local_scan() function.
3337 19. For the "mail" command in an Exim filter, no default was being set for
3338 the once_repeat time, causing a random time value to be used if "once" was
3339 specified. (If the value happened to be <= 0, no repeat happened.) The
3340 default is now 0s, meaning "never repeat". The "vacation" command was OK
3341 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3342 (I found it when inspecting the code).
3344 20. There is now an overall timeout for performing a callout verification. It
3345 defaults to 4 times the callout timeout, which applies to individual SMTP
3346 commands during the callout. The overall timeout applies when there is more
3347 than one host that can be tried. The timeout is checked before trying the
3348 next host. This prevents very long delays if there are a large number of
3349 hosts and all are timing out (e.g. when the network connections are timing
3350 out). The value of the overall timeout can be changed by specifying an
3351 additional sub-option for "callout", called "maxwait". For example:
3353 verify = sender/callout=5s,maxwait=20s
3355 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3356 to the end before writing, but this should make it even safer).
3358 22. Exim was forgetting that it had advertised PIPELINING for the second and
3359 subsequent messages on an SMTP connection. It was also not resetting its
3360 memory on STARTTLS and an internal HELO.
3362 23. When Exim logs an SMTP synchronization error within a session, it now
3363 records whether PIPELINING has been advertised or not.
3365 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3366 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3367 rather than long int.
3369 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3371 26. Added the never_mail option to autoreply.
3377 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3378 crash if the getsockname() call failed; this can happen if a connection is
3379 closed very soon after it is established. The problem was simply in the
3380 order in which certain operations were done, causing Exim to try to write
3381 to the SMTP stream before it had set up the file descriptor. The bug has
3382 been fixed by making things happen in the correct order.
3388 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3389 before the connection was closed, thus losing the rejection response.
3391 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3392 some early Solaris releases, but causes trouble in current releases where
3393 socklen_t is defined.
3395 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3398 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3401 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3402 that releasing the top of it at the end releases what was used for sub-
3403 expansions (unless the block got too big). However, discard this block if
3404 the first thing is a variable or header, so that we can use its block when
3405 it is dynamic (useful for very large $message_headers, for example).
3407 6. Lookups now cache *every* query, not just the most recent. A new, separate
3408 store pool is used for this. It can be recovered when all lookup caches are
3409 flushed. Lookups now release memory at the end of their result strings.
3410 This has involved some general refactoring of the lookup sources.
3412 7. Some code has been added to the store_xxx() functions to reduce the amount
3413 of flapping under certain conditions.
3415 8. log_incoming_interface used to affect only the <= reception log lines. Now
3416 it causes the local interface and port to be added to several more SMTP log
3417 lines, for example "SMTP connection from", and rejection lines.
3419 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3421 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3423 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3424 could be overwritten at the end of the current message (or the start of a
3425 new message if it was set in a HELO ACL). The value is now preserved for
3426 the duration of the SMTP connection.
3428 12. If a transport had a headers_rewrite setting, and a matching header line
3429 contained an unqualified address, that address was qualified, even if it
3430 did not match any rewriting rules. The underlying bug was that the values
3431 of the flags that permit the existence of unqualified sender and recipient
3432 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3433 non-local messages, and by -bnq for local messages) were not being
3434 preserved with the message after it was received.
3436 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3437 "next input=" as part of the text comprising the host identity instead of
3438 the correct text. The code was using the same buffer for two different
3439 strings. However, depending on which order the printing function evaluated
3440 its arguments, the bug did not always show up. Under Linux, for example, my
3441 test suite worked just fine.
3443 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3444 doesn't work with some older versions of Perl. It has been changed to "my",
3445 which in any case is probably the better facility to use.
3447 15. A really picky compiler found some instances of statements for creating
3448 error messages that either had too many or two few arguments for the format
3451 16. The size of the buffer for calls to the DNS resolver has been increased
3452 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3453 for addresses that have a lot of PTR records. This alleviates a problem; it
3454 does not fully solve it.
3456 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3457 buffer now truncates the list and logs the incident, which is the same
3458 action as happens when Exim is looking up a host name and its aliases.
3459 Previously in this situation something unpredictable would happen;
3460 sometimes it was "internal error: store_reset failed".
3462 18. If a server dropped the connection unexpectedly when an Exim client was
3463 using GnuTLS and trying to read a response, the client delivery process
3464 crashed while trying to generate an error log message.
3466 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3467 string, for example:
3469 warn message = H1: something\nH2: something
3471 the text was added as a single header line from Exim's point of view
3472 though it ended up OK in the delivered message. However, searching for the
3473 second and subsequent header lines using $h_h2: did not work. This has been
3474 fixed. Similarly, if a system filter added multiple headers in this way,
3475 the routers could not see them.
3477 20. Expanded the error message when iplsearch is called with an invalid key to
3478 suggest using net-iplsearch in a host list.
3480 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3481 ACLs are no longer actually imposed (and a message to that effect is
3484 22. If a "gecos" field in a passwd entry contained escaped characters, in
3485 particular, if it contained a \" sequence, Exim got it wrong when building
3486 a From: or a Sender: header from that name. A second bug also caused
3487 incorrect handling when an unquoted " was present following a character
3488 that needed quoting.
3490 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3491 was not being matched caselessly.
3493 24. Arranged for all hyphens in the exim.8 source to be escaped with
3496 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3497 properly. Recipient callout cache records were still being keyed to include
3498 the sender, even when use_sender was set false. This led to far more
3499 callouts that were necessary. The sender is no longer included in the key
3500 when use_sender is false.
3502 26. Added "control = submission" modifier to ACLs.
3504 27. Added the ${base62d: operator to decode base 62 numbers.
3506 28. dnsdb lookups can now access SRV records.
3508 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3509 the configuration file.
3511 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3512 (-v) mode. This makes the output for a verbose queue run more intelligible.
3514 31. Added a use_postmaster feature to recipient callouts.
3516 32. Added the $body_zerocount variable, containing the number of binary zero
3517 bytes in the message body.
3519 33. The time of last modification of the "new" subdirectory is now used as the
3520 "mailbox time last read" when there is a quota error for a maildir
3523 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3525 35. Added +ignore_unknown as a special item in host lists.
3527 36. Code for decoding IPv6 addresses in host lists is now included, even if
3528 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3529 address was recognized as an IP address, but was then not correctly decoded
3530 into binary, causing unexpected and incorrect effects when compared with
3537 1. Very minor rewording of debugging text in manualroute to say "list of
3538 hosts" instead of "hostlist".
3540 2. If verify=header_syntax was set, and a header line with an unqualified
3541 address (no domain) and a large number of spaces between the end of the
3542 name and the colon was received, the reception process suffered a buffer
3543 overflow, and (when I tested it) crashed. This was caused by some obsolete
3544 code that should have been removed. The fix is to remove it!
3546 3. When running in the test harness, delay a bit after writing a bounce
3547 message to get a bit more predictability in the log output.
3549 4. Added a call to search_tidyup() just before forking a reception process. In
3550 theory, someone could use a lookup in the expansion of smtp_accept_max_
3551 per_host which, without the tidyup, could leave open a database connection.
3553 5. Added the variables $recipient_data and $sender_data which get set from a
3554 lookup success in an ACL "recipients" or "senders" condition, or a router
3555 "senders" option, similar to $domain_data and $local_part_data.
3557 6. Moved the writing of debug_print from before to after the "senders" test
3560 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3561 problems for message scanning, either using a data ACL, or using
3562 local_scan() because the Received: header was not generated till after they
3563 were called (in order to set the time as the time of reception completion).
3564 I have revised the way this works. The header is now generated after the
3565 body is received, but before the ACL or local_scan() are called. After they
3566 are run, the timestamp in the header is updated.
3572 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3573 before starting a queue runner without re-exec. This happened only when
3574 deliver_drop_privilege was set or when the Exim user was set to root. The
3575 effect of the bug was that timeouts during subsequent deliveries caused
3576 crashes instead of being properly handled. The handler is now left at its
3577 default (and expected) setting.
3579 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3580 message, again when deliver_drop_privilege is set or Exim is run as root.
3581 The bug described in (1) was not present in this case, but the tidying up
3582 of the other signals was missing. I have made the two cases consistent.
3584 3. The ignore_target_hosts setting on a manualroute router was being ignored
3585 for hosts that were looked up using the /MX notation.
3587 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3590 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3591 operated on the sender address. After changing the $sender_address to <>
3592 for the sender address verify, Exim was re-instated it as the original
3593 (before rewriting) address, but remembering that it had rewritten it, so it
3594 wasn't rewriting it again. This bug also had the effect of breaking the
3595 sender address verification caching when the sender address was rewritten.
3597 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3598 This has been changed so that if the ip literal address matches
3599 ignore_target_hosts, the router declines.
3601 7. Added expansion conditions match_domain, match_address, and match_local_
3602 part (NOT match_host).
3604 8. The placeholder for the Received: header didn't have a length field set.
3606 9. Added code to Exim itself and to exim_lock to test for a specific race
3607 condition that could lead to file corruption when using MBX delivery. The
3608 issue is with the lockfile that is created in /tmp. If this file is removed
3609 after a process has opened it but before that process has acquired a lock,
3610 there is the potential for a second process to recreate the file and also
3611 acquire a lock. This could lead to two Exim processes writing to the file
3612 at the same time. The added code performs the same test as UW imapd; it
3613 checks after acquiring the lock that its file descriptor still refers to
3614 the same named file.
3616 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3617 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3620 11. Added the smtp_active_hostname option. If used, this will typically be made
3621 to depend on the incoming interface address. Because $interface_address is
3622 not set up until the daemon has forked a reception process, error responses
3623 that can happen earlier (such as "too many connections") no longer contain
3626 12. If an expansion in a condition on a "warn" statement fails because a lookup
3627 defers, the "warn" statement is abandoned, and the next ACL statement is
3628 processed. Previously this caused the whole ACL to be aborted.
3630 13. Added the iplsearch lookup type.
3632 14. Added ident_timeout as a log selector.
3634 15. Added tls_certificate_verified as a log selector.
3636 16. Added a global option tls_require_ciphers (compare the smtp transport
3637 option of the same name). This controls incoming TLS connections.
3639 17. I finally figured out how to make tls_require_ciphers do a similar thing
3640 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3641 before starting the TLS session.
3643 18. Tabs are now shown as \t in -bP output.
3645 19. If the log selector return_path_on_delivery was set, Exim crashed when
3646 bouncing a message because it had too many Received: header lines.
3648 20. If two routers both had headers_remove settings, and the first one included
3649 a superfluous trailing colon, the final name in the first list and the
3650 first name in the second list were incorrectly joined into one item (with a
3651 colon in the middle).
3657 1. Added -C and -D options to the exinext utility, mainly to make it easier
3658 to include in the automated testing, but these could be helpful when
3659 multiple configurations are in use.
3661 2. The exinext utility was not formatting the output nicely when there was
3662 an alternate port involved in the retry record key, nor when there was a
3663 message id as well (for retries that were specific to a specific message
3664 and a specific host). It was also confused by IPv6 addresses, because of
3665 the additional colons they contain. I have fixed the IPv4 problem, and
3666 patched it up to do a reasonable job for IPv6.
3668 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3669 delivery, the log line now contains "pipelined" if PIPELINING was used.
3671 4. An SMTP transport process used to panic and die if the bind() call to set
3672 an explicit outgoing interface failed. This has been changed; it is now
3673 treated in the same way as a connect() failure.
3675 5. A reference to $sender_host_name in the part of a conditional expansion
3676 that was being skipped was still causing a DNS lookup. This no longer
3679 6. The def: expansion condition was not recognizing references to header lines
3680 that used bh_ and bheader_.
3682 7. Added the _cache feature to named lists.
3684 8. The code for checking quota_filecount in the appendfile transport was
3685 allowing one more file than it should have been.
3687 9. For compatibility with Sendmail, the command line option
3695 and sets the incoming protocol and host name (for trusted callers). The
3696 host name and its colon can be omitted when only the protocol is to be set.
3697 Note the Exim already has two private options, -pd and -ps, that refer to
3698 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3699 "s", but I don't think that's a major issue.
3701 10. A number of refactoring changes to the code, none of which should affect
3704 (a) The number of logging options was getting close to filling up the
3705 32-bit word that was used as a bit map. I have split them into two classes:
3706 those that are passed in the argument to log_write(), and those that are
3707 only ever tested independently outside of that function. These are now in
3708 separate 32-bit words, so there is plenty of room for expansion again.
3709 There is no change in the user interface or the logging behaviour.
3711 (b) When building, for example, log lines, the code previously used a
3712 macro that called string_cat() twice, in order to add two strings. This is
3713 not really sufficiently general. Furthermore, there was one instance where
3714 it was actually wrong because one of the argument was used twice, and in
3715 one call a function was used. (As it happened, calling the function twice
3716 did not affect the overall behaviour.) The macro has been replaced by a
3717 function that can join an arbitrary number of extra strings onto a growing
3720 (c) The code for expansion conditions now uses a table and a binary chop
3721 instead of a serial search (which was left over from when there were very
3722 few conditions). Also, it now recognizes conditions like "pam" even when
3723 the relevant support is not compiled in: a suitably worded error message is
3724 given if an attempt is made to use such a condition.
3726 11. Added ${time_interval:xxxxx}.
3728 12. A bug was causing one of the ddress fields not to be passed back correctly
3729 from remote delivery subprocesses. The field in question was not being
3730 subsequently used, so this caused to problems in practice.
3732 13. Added new log selectors queue_time and deliver_time.
3734 14. Might have fixed a bug in maildirsizefile handling that threw up
3735 "unexpected character" debug warnings, and recalculated the data
3736 unnecessarily. In any case, I expanded the warning message to give more
3739 15. Added the message "Restricted characters in address" to the statements in
3740 the default ACL that block characters like @ and % in local parts.
3742 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3743 Three changes have been made:
3745 (a) There was a serious bug; a negative response to MAIL caused the whole
3746 recipient domain to be cached as invalid, thereby blocking all messages
3747 to all local parts at the same domain, from all senders. This bug has
3748 been fixed. The domain is no longer cached after a negative response to
3749 MAIL if the sender used is not empty.
3751 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3754 (c) A new callout option, "use_sender" has been added for people who want
3755 the modified behaviour.
3761 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3764 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3765 indeed breaks things for older releases.
3767 3. Added additional logging to the case where there is a problem reading data
3768 from a filter that is running in a subprocess using a pipe, in order to
3769 try to track down a specific problem.
3771 4. Testing facility fudge: when running in the test harness and attempting
3772 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3773 getting "No route to host". Convert this to a timeout.
3775 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3778 6. Some OS don't have socklen_t but use size_t instead. This affects the
3779 fifth argument of getsockopt() amongst other things. This is now
3780 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3781 can be set for individual OS. I have set it for SunOS5, OSF1, and
3782 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3783 some earlier ones do not.
3785 7. Change 4.30/15 was not doing the test caselessly.
3787 8. The standard form for an IPv6 address literal was being rejected by address
3788 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3789 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3790 this, as well as the form without the "IPv6" on the front (but only when
3791 address literals are enabled, of course).
3793 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3795 10. Exim crashed if a message with an empty sender address specified by -f
3796 encountered a router with an errors_to setting. This could be provoked only
3797 by a command such as
3801 where an empty string was supplied; "<>" did not hit this bug.
3803 11. Installed PCRE release 4.5.
3805 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3806 remained set. It is now erased.
3808 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3809 times from message ids (which are base 36 rather than the normal 62).
3811 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3812 were being counted as actual protocol errors, and logged if the log
3813 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3814 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3815 and DATA following a set of rejected RCPTs do not count as protocol errors.
3816 In other words, Exim assumes they were pipelined, though this may not
3817 actually be the case. Of course, in all cases the client gets an
3818 appropriate error code.
3820 15. If a lookup fails in an ACL condition, a message about the failure may
3821 be available; it is used if testing the ACL cannot continue, because most
3822 such messages specify what the cause of the deferral is. However, some
3823 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3824 that caused an old message to be retained and used if a later statement
3825 caused a defer, replacing the real cause of the deferral.
3827 16. If an IP address had so many PTR records that the DNS lookup buffer
3828 was not large enough to hold them, Exim could crash while trying to process
3829 the truncated data. It now detects and logs this case.
3831 17. Further to 4.21/58, another change has been made: if (and only if) the
3832 first line of a message (the first header line) ends with CRLF, a bare LF
3833 in a subsequent header line has a space inserted after it, so as not to
3834 terminate the header.
3836 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3837 unnecessarily, used atoi() instead of strtol(), and didn't check the
3838 termination when getting file sizes from file names by regex.
3840 19. Completely re-implemented the support for maildirsize files, in the light
3841 of a number of problems with the previous contributed implementation
3842 (4.30/29). In particular:
3844 . If the quota is zero, the maildirsize file is maintained, but no quota is
3847 . If the maildir directory does not exist, it is created before any attempt
3848 to write a maildirsize file.
3850 . The quota value in the file is just a cache; if the quota is changed in
3851 the transport, the new value overrides.
3853 . A regular expression is available for excluding directories from the
3856 20. The autoreply transport checks the characters in options that define the
3857 message's headers; it allows continued headers, but it was checking with
3858 isspace() after an embedded newline instead of explicitly looking for a
3861 21. If all the "regular" hosts to which an address was routed had passed their
3862 expiry times, and had not reached their retry times, the address was
3863 bounced, even if fallback hosts were defined. Now Exim should go on to try
3866 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3867 equivalent code in the SMTP transport. Some hosts send humungous responses
3868 to HELO/EHLO, more than 1024 it seems.
3870 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3871 gives compiler warnings in some environments. I've now done it "properly",
3874 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3875 (because of problems with the built-in one) was declared to return uschar *
3876 instead of char *, causing compiler failure.
3878 25. Fixed a file descriptor leak when processing alias/forward files.
3880 26. Fixed a minor format string issue in dbfn.c.
3882 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3884 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3885 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3888 29. When a sender address is verified, it is cached, to save repeating the test
3889 when there is more than one recipient in a message. However, when the
3890 verification involves a callout, it is possible for different callout
3891 options to be set for different recipients. It is too complicated to keep
3892 track of this in the cache, so now Exim always runs a verification when a
3893 callout is required, relying on the callout cache for the optimization.
3894 The overhead is duplication of the address routing, but this should not be
3897 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3898 to be verified with callout=postmaster, and the main callout worked but the
3899 postmaster check failed, the verification correctly failed. However, if a
3900 subsequent RCPT command asked for sender verification *without* the
3901 postmaster check, incorrect caching caused this verification also to fail,
3904 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3905 it was not caching the DNS options (qualify_single, search_parents) that
3906 were used when the lookup failed. A subsequent lookup with different
3907 options therefore always gave the same answer, though there were cases
3908 where it should not have. (Example: a "domains = !$mx_any" option on a
3909 dnslookup router: the "domains" option is always processed without any
3910 widening, but the router might have qualify_single set.) Now Exim uses the
3911 cached value only when the same options are set.
3913 32. Added John Jetmore's "exipick" utility to the distribution.
3915 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3916 than a timeout (e.g. a certificate is required, and is not provided), an
3917 Exim server now closes the connection immediately. Previously it waited for
3918 the client to close - but if the client is SSL, it seems that they each
3919 wait for each other, leading to a delay before one of them times out.
3921 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3922 maintained 0.8.x compatibility because I don't think many are using it, and
3923 it is clearly obsolete.
3925 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3928 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3929 client certificate was expired. A simple patch fixes this, though I don't
3930 understand the full logic of why the verify callback is called multiple
3933 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3934 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3935 which causes problems with some clients (such as the Certicom SSL Plus
3936 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3937 disables the coutermeasure allowing Eudora to connect."
3939 38. Exim was not checking that a write() to a log file succeeded. This could
3940 lead to Bad Things if a log got too big, in particular if it hit a file
3941 size limit. Exim now panics and dies if it cannot write to a log file, just
3942 as it does if it cannot open a log file.
3944 39. Modified OS/Makefile-Linux so that it now contains
3946 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3948 The two -D definitions ensure that Exim is compiled with large file
3949 support, which makes it possible to handle log files that are bigger than
3952 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3953 instance) a domain was checked against a named list that involved a lookup,
3954 causing $domain_data to be set, then another domain was checked against the
3955 same list, then the first domain was re-checked, the value of $domain_data
3956 after the final check could be wrong. In particular, if the second check
3957 failed, it could be set empty. This bug probably also applied to
3960 41. The strip_trailing_dot option was not being applied to the address given
3961 with the -f command-line option.
3963 42. The code for reading a message's header from the spool was incrementing
3964 $received_count, but never initializing it. This meant that the value was
3965 incorrect (doubled) while delivering a message in the same process in which
3966 it was received. In the most common configuration of Exim, this never
3967 happens - a fresh exec is done - but it can happen when
3968 deliver_drop_privilege is set.
3970 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3971 it now includes up to 150 characters of the unexpected data in the log
3974 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3975 and building data strings. The size of both of these buffers was 10 000
3976 bytes - far larger than anybody would *ever* want, thought I. Needless to
3977 say, somebody hit the limit. I have increased the maximum line length to
3978 20 000 and the maximum data length of concatenated lines to 100 000. I have
3979 also fixed two bugs, because there was no checking on these buffers. Tsk,
3980 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3981 buffer is too small.
3983 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3984 lsearch lookups. Now it does.
3986 46. When parsing a route_list item in a manualroute router, a fixed-length
3987 buffer was used for the list of hosts. I made this 1024 bytes long,
3988 thinking that nobody would ever have a list of hosts that long. Wrong.
3989 Somebody had a whole pile of complicated expansion conditions, and the
3990 string was silently truncated, leading to an expansion error. It turns out
3991 that it is easier to change to an unlimited length (owing to other changes
3992 that have happened since this code was originally written) than to build
3993 structure for giving a limitation error. The length of the item that
3994 expands into the list of hosts is now unlimited.
3996 47. The lsearch lookup could not handle data where the length of text line was
3997 more than 4095 characters. Such lines were truncated, leading to shortened
3998 data being returned. It should now handle lines of any length.
4000 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4001 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4004 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4005 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4006 no longer happens in comment lines. A list of replacements is now placed
4007 at the head of all of the source files, except those whose only change is
4008 to replace PERL_COMMAND in the very first #! line.
4010 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4011 messages on the queue, with a bottom-up merge sort, using code contributed
4012 by Michael Haardt. This should make operations like -bp somewhat faster on
4013 large queues. It won't affect queue runners, except when queue_run_in_order
4016 51. Installed eximstats 1.31 in the distribution.
4018 52. Added support for SRV lookups to the dnslookup router.
4020 53. If an ACL referred to $message_body or $message_body_end, the value was not
4021 reset for any messages that followed in the same SMTP session.
4023 54. The store-handling optimization for building very long strings was not
4024 differentiating between the different store pools. I don't think this
4025 actually made any difference in practice, but I've tidied it.
4027 55. While running the routers to verify a sender address, $sender_address
4028 was still set to the sender address. This is wrong, because when routing to
4029 send a bounce to the sender, it would be empty. Therefore, I have changed
4030 it so that, while verifying a sender address, $sender_address is set to <>.
4031 (There is no change to what happens when verifying a recipient address.)
4033 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4034 target A or AAAA records (if not already returned) without resetting the
4035 qualify_single or search_parents options of the DNS resolver. These are
4036 inappropriate in this case because the targets of MX and SRV records must
4037 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4038 target that, when qualified, matched something in the local domain. These
4039 two options are now turned off when doing these lookups.
4041 57. It seems that at least some releases of Reiserfs (which does not have the
4042 concept of a fixed number of inodes) returns zero and not -1 for the
4043 number of available inodes. This interacted badly with check_spool_inodes,
4044 which assumed that -1 was the "no such thing" setting. What I have done is
4045 to check that the total number of inodes is greater than zero before doing
4046 the test of how many are available.
4048 58. When a "warn" ACL statement has a log_message modifier, the message is
4049 remembered, and not repeated. This is to avoid a lot of repetition when a
4050 message has many recipients that cause the same warning to be written.
4051 Howewer, Exim was preserving the list of already written lines for an
4052 entire SMTP session, which doesn't seem right. The memory is now reset if a
4053 new message is started.
4055 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4056 debugging output unless log_rewrite was also set.
4058 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4059 of (int)(handle) when we know that handle contains (void *)(-1).
4061 61. The Exim daemon panic-logs an error return when it closes the incoming
4062 connection. However "connection reset by peer" seems to be common, and
4063 isn't really an error worthy of noting specially, so that particular error
4066 62. When Exim is trying to find all the local interfaces, it used to panic and
4067 die if the ioctl to get the interface flags failed. However, it seems that
4068 on at least one OS (Solaris 9) it is possible to have an interface that is
4069 included in the list of interfaces, but for which you get a failure error
4070 for this call. This happens when the interface is not "plumbed" into a
4071 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4072 failure of the "get flags" call assumes that the interface is down.
4074 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4075 makes life easier for people who are doing arithmetic on fields extracted
4076 from dates, where you often get leading zeros that should not be
4077 interpreted as octal.
4079 64. Added qualify_domain to the redirect router, to override the global
4082 65. If a pathologically long header line contained very many addresses (the
4083 report of this problem mentioned 10 000) and each of them was rewritten,
4084 Exim could use up a very large amount of memory. (It kept on making new
4085 copies of the header line as it rewrote, and never released the old ones.)
4086 At the expense of a bit more processing, the header rewriting function has
4087 been changed so that it no longer eats memory in this way.
4089 66. The generation of the Received: header has been moved from the time that a
4090 message starts to be received, to the time that it finishes. The timestamp
4091 in the Received: header should now be very close to that of the <= log
4092 line. There are two side-effects of this change:
4094 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4095 logged header lines no longer include the local Received: line, because
4096 it has not yet been created. The same applies to a copy of the message
4097 that is returned to a non-SMTP sender when a message is rejected.
4099 (b) When a filter file is tested using -bf, no additional Received: header
4100 is added to the test message. After some thought, I decided that this
4103 This change does not affect the value of $received_for. It is still set
4104 after address rewriting, but before local_scan() is called.
4106 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4108 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4109 gave an unhelpful panic error message, and a defer error. I have managed to
4110 change this behaviour so that it now rejects any supplied certificate,
4111 which seems right, as the list of acceptable certificates is empty.
4113 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4114 gave an unhelpful defer error. I have not managed to make this reject any
4115 supplied certificates, but the error message it gives is "no certificate
4116 supplied", which is not helpful.
4118 70. exigrep's output now also includes lines that are not associated with any
4119 message, but which match the given pattern. Implemented by a patch from
4120 Martin Sluka, which also tidied up the Perl a bit.
4122 71. Recipient callout verification, like sender verification, was using <> in
4123 the MAIL FROM command. This isn't really the right thing, since the actual
4124 sender may affect whether the remote host accepts the recipient or not. I
4125 have changed it to use the actual sender in the callout; this means that
4126 the cache record is now keyed on a recipient/sender pair, not just the
4127 recipient address. There doesn't seem to be a real danger of callout loops,
4128 since a callout by the remote host to check the sender would use <>.
4129 [SEE ABOVE: changed after hitting problems.]
4131 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4132 temporary errors. However, in the case of such a code being given after
4133 the end of a data transmission (i.e. after ".") Exim was failing to write
4134 a retry record for the message. (Yes, there was some broken host that was
4135 actually sending 8xx at this point.)
4137 73. An unknown lookup type in a host list could cause Exim to panic-die when
4138 the list was checked. (An example that provoked this was putting <; in the
4139 middle of a list instead of at the start.) If this happened during a DATA
4140 ACL check, a -D file could be left lying around. This kind of configuration
4141 error no longer causes Exim to die; instead it causes a defer errror. The
4142 incident is still logged to the main and panic logs.
4144 74. Buglet left over from Exim 3 conversion. The message "too many messages
4145 in one connection" was written to the rejectlog but not the mainlog, except
4146 when address rewriting (yes!) was being logged.
4148 75. Added write_rejectlog option.
4150 76. When a system filter was run not as root (that is, when system_filter_user
4151 was set), the values of the $n variables were not being returned to the
4152 main process; thus, they were not subsequently available in the $sn
4155 77. Added +return_path_on_delivery log selector.
4157 78. A connection timeout was being treated differently from recipients deferred
4158 when testing hosts_max_try with a message that was older than the host's
4159 retry timeout. (The host should not be counted, thus allowing all hosts to
4160 be tried at least once before bouncing.) This may have been the cause of an
4161 occasionally reported bug whereby a message would remain on the queue
4162 longer than the retry timeout, but would be bounced if a delivery was
4163 forced. I say "may" because I never totally pinned down the problem;
4164 setting up timeout/retry tests is difficult. See also the next item.
4166 79. The ultimate address timeout was not being applied to errors that involved
4167 a combination of host plus message (for example, a timeout on a MAIL
4168 command). When an address resolved to a number of possible hosts, and they
4169 were not all tried for each delivery (e.g. because of hosts_max_try), a
4170 message could remain on the queue longer than the retry timeout.
4172 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4175 81. Fixed an obscure SMTP outgoing bug which required at least the following
4176 conditions: (a) there was another message waiting for the same server;
4177 (b) the server returned 5xx to all RCPT commands in the first message so
4178 that the message was not completed; (c) the server dropped the connection
4179 or gave a negative response to the RSET that Exim sends to abort the
4180 transaction. The observed case was a dropped connection after DATA that had
4181 been sent in pipelining mode. That is, the server had advertised PIPELINING
4182 but was not implementing it correctly. The effect of the bug was incorrect
4183 behaviour, such as trying another host, and this could lead to a crash.
4189 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4190 and daemon.c were passed as pointers to ints; they should have been
4191 pointers to socklen_t variables (which are typically unsigned ints).
4193 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4196 3. Fixed a really odd bug that affected only the testing scheme; patching a
4197 certain fixed string in the binary changed the value of another string that
4198 happened to be identical to the end of the original first string.
4200 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4201 name", it returns that address as the IP address. On some operating
4202 systems (e.g. Solaris), it also passes back the IP address string as the
4203 "host name". However, on others (e.g. Linux), it passes back an empty
4204 string. Exim wasn't checking for this, and was changing the host name to an
4205 empty string, assuming it had been canonicized.
4207 5. Although rare, it is permitted to have more than one PTR record for a given
4208 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4209 all the names associated with an address, because they do in Solaris.
4210 However, it seems that they do not in Linux for data that comes from the
4211 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4212 I found this out when I moved to a new Linux workstation and tried to run
4213 the Exim test suite.
4215 To get round this problem I have changed the code so that it now does its
4216 own call to the DNS to look up PTR records when searching for a host name.
4217 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4218 addresses that are only in /etc/hosts are still found.
4220 This behaviour is, however, controlled by an option called host_lookup_
4221 order, which defaults to "bydns:byaddr". If people want to use the other
4222 order, or indeed, just use one or the other means of lookup, they can
4223 specify it in this variable.
4225 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4226 some operating systems, this comes back from gethostbyaddr() as an empty
4227 string, and this is what Exim used to test for. However, it seems that in
4228 other systems, "." is yielded. Exim now tests for this case too.
4230 7. The values of check_spool_space and check_log_space are now held internally
4231 as a number of kilobytes instead of an absolute number of bytes. If a
4232 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4233 kilobyte. This means that much larger values can be stored.
4235 8. Exim monitor: an attempt to get the action menu when not actually pointing
4236 at a message produces an empty menu entitled "No message selected". This
4237 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4238 no entries in it ("Shell widget menu has zero width and/or height"). So I
4239 have added a single, blank menu entry in this case.
4241 9. Added ${quote_local_part.
4243 10. MIME decoding is now applied to the contents of Subject: header lines when
4246 11. Now that a reference to $sender_host_address automatically causes a reverse
4247 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4248 host lookup before query-style lookups in lists that might use this
4249 variable. This has therefore been abolished, and the "net-" prefix is no
4250 longer necessary for query-style lookups.
4252 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4253 have been a typo for LFLAGS, so it has been changed.
4255 13. The install script calls Exim with "-C /dev/null" in order to find the
4256 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4257 to be output. Howeve, since Exim outputs its version number before the
4258 error, it didn't break the script. It just looked ugly. I fixed this by
4259 always allowing "-C /dev/null" if the caller is root.
4261 14. Ignore overlarge ACL variable number when reading spool file - insurance
4262 against a later release with more variables having written the file.
4264 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4265 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4266 this, as well as the form without the "IPv6" on the front.
4268 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4269 OS/Makefile-Darwin file.
4271 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4272 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4274 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4275 variable. However, some OS use ints for inodes. I've added cast to long int
4276 to get rid of the compiler warning.
4278 19. I had forgotten to lock out "/../" in configuration file names when
4279 ALT_CONFIG_PREFIX was set.
4281 20. Routers used for verification do not need to specify transports. However,
4282 if such a router generated a host list, and callout was configured, Exim
4283 crashed, because it could not find a port number from the (non-existent)
4284 transport. It now assumes port 25 in this circumstance.
4286 21. Added the -t option to exigrep.
4288 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4289 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4290 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4291 them is set, LOOKUP_LSEARCH is forced.
4293 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4295 24. Added sender and host information to the "rejected by local_scan()" log
4296 line; previously there was no indication of these.
4298 25. Added .include_if_exists.
4300 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4301 Linux. It turns out that not all file systems support this. Apparently some
4302 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4303 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4304 not supported on the file descriptor, is now ignored when Exim is trying to
4305 sync a directory. This applies only to Linux.
4307 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4309 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4310 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4313 29. Added support for maildirsize files from supplied patch (modified a bit).
4315 30. The use of :fail: followed by an empty string could lead Exim to respond to
4316 sender verification failures with (e.g.):
4318 550 Verification failed for <xxx>
4319 550 Sender verify failed
4321 where the first response line was missing the '-' that indicates it is not
4322 the final line of the response.
4324 31. The loop for finding the name of the user that called Exim had a hardwired
4325 limit of 10; it now uses the value of finduser_retries, which is used for
4326 all other user lookups.
4328 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4331 33. Exim was neglecting to zero errno before one call of strtol() when
4332 expanding a string and expecting an integer value. On some systems this
4333 resulted in spurious "integer overflow" errors. Also, it was casting the
4334 result into an int without checking.
4336 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4337 did not work. The code looks as if it has *never* worked, though it appears
4338 to have been documented since at least releast 1.62. I have made it work.
4340 35. The "timeout_DNS" error in retry rules, also documented since at least
4341 1.62, also never worked. As it isn't clear exactly what this means, and
4342 clearly it isn't a major issue, I have abolished the feature by treating it
4343 as "timeout", and writing a warning to the main and panic logs.
4345 36. The display of retry rules for -brt wasn't always showing the error code
4348 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4349 timeout_connect_A, timeout_connect_MX.
4351 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4352 to the empty sender.
4354 39. The daemon was not analysing the content of -oX till after it had closed
4355 stderr and disconnected from the controlling terminal. This meant that any
4356 syntax errors were only noted on the panic log, and the return code from
4357 the command was 0. By re-arranging the code a little, I've made the
4358 decoding happen first, so such errors now appear on stderr, and the return
4359 code is 1. However, the actual setting up of the sockets still happens in
4360 the disconnected process, so errors there are still only recorded on the
4363 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4364 connections (as happens on some IP stacks) was logged at start up time as
4365 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4366 it from "IPv6 and IPv4", which means that two separate sockets are being
4369 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4370 says whether AF_INET or AF_INET6 was passed as an argument.
4372 42. Exiwhat output was messed up when time zones were included in log
4375 43. Exiwhat now gives more information about the daemon's listening ports,
4376 and whether -tls-on-connect was used.
4378 44. The "port" option of the smtp transport is now expanded.
4380 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4381 silently ignored. Now an error message is written to the main and panic
4384 46. There's a new ACL modifier called "logwrite" which writes to a log file
4385 as soon as it is encountered.
4387 47. Added $local_user_uid and $local_user_gid at routing time.
4389 48. Exim crashed when trying to verify a sender address that was being
4392 49. Exim was recognizing only a space character after ".include". It now also
4393 recognizes a tab character.
4395 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4396 extracting the relevant information from the specification. The man page no
4397 longer contains scrambled data for the -d option, and I've added a section
4398 at the front about calling Exim under different names.
4400 51. Added "extra_headers" argument to the "mail" command in filter files.
4402 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4405 53. Installed eximstats 1.29.
4407 54. Added transport_filter_timeout as a generic transport option.
4409 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4410 Cc: header lines. This was required by RFC 822, but it not required by RFC
4413 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4414 incoming messages that did not have them. Now it does so only if the
4415 message originates locally, that is, if there is no associated remote host
4416 address. When Resent- header lines are present, this applies to the Resent-
4417 lines rather than the non-Resent- lines.
4419 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4420 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4422 58. Messages for configuration errors now include the name of the main
4423 configuration file - useful now that there may be more than one file in a
4424 list (.included file names were always shown).
4426 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4427 for those rare installations that do not start the daemon as root or run it
4428 setuid root. I've cut out the call to initgroups() if the daemon is not
4431 60. The Exim user and group can now be bound into the binary as text strings
4432 that are looked up at the start of Exim's processing.
4434 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4436 62. Added $mailstore_basename variable.
4438 63. Installed patch to sieve.c from Michael Haardt.
4440 64. When Exim failed to open the panic log after failing to open the main log,
4441 the original message it was trying to log was written to stderr and debug
4442 output, but if they were not available (the usual case in production), it
4443 was lost. Now it is written to syslog before the two lines that record the
4444 failures to open the logs.
4446 65. Users' Exim filters run in subprocesses under the user's uid. It is
4447 possible for a "deliver" command or an alias in a "personal" command to
4448 provoke an address rewrite. If logging of address rewriting is configured,
4449 this fails because the process is not running as root or exim. There may be
4450 a better way of dealing with this, but for the moment (because 4.30 needs
4451 to be released), I have disabled address rewrite logging when running a
4452 filter in a non-root, non-exim process.
4458 1. The buildconfig auxiliary program wasn't quoting the value set for
4459 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4460 not defined. This bug was masked in 4.22 by the effect that was fixed in
4463 2. Some messages that were rejected after a message id was allocated were
4464 shown as "incomplete" by exigrep. It no longer does this for messages that
4465 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4467 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4468 have allow_domain_literals set, the ID did not get logged in the <= line.
4469 Domain literals are now always recognized in Message-ID: header lines.
4471 4. The first argument for a ${extract expansion item is the key name or field
4472 number. Leading and trailing spaces in this item were not being ignored,
4473 causing some misleading effects.
4475 5. When deliver_drop_privilege was set, single queue runner processes started
4476 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4477 same command in the process it spins off) were not dropping privilege.
4479 6. When the daemon running as "exim" started a queue runner, it always
4480 re-executed Exim in the spun-off process. This is a waste of effort when
4481 deliver_drop_privilege is set. The new process now just calls the
4482 queue-runner function directly.
4488 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4491 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4492 ignored. Though the use of -oP was forcing the writing of a pid file, it
4493 was always written to the default place.
4495 3. If the message "no IP address found for host xxxx" is generated during
4496 incoming verification, it is now followed by identification of the incoming
4497 connection (so you can more easily find what provoked it).
4499 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4501 5. Added some features to "harden" Exim a bit more against certain attacks:
4503 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4504 be put in Local/Makefile. This is like the never_users runtime option,
4505 but it cannot be overridden. The default setting is "root".
4507 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4508 prefix string with which any file named in a -C command line option
4511 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4512 is retained for -C and -D only if the caller of Exim is root. Without
4513 it, the exim user may also use -C and -D and retain privilege.
4515 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4516 command line option is disabled.
4518 6. Macro names set by the -D option must start with an upper case letter, just
4519 like macro names defined in the configuration file.
4521 7. Added "dereference=" facility to LDAP.
4523 8. Two instances of the typo "uknown" in the source files are fixed.
4525 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4526 the Configure-Makefile script screwed up while processing it.
4528 10. Incorporated PCRE 4.4.
4530 11. The SMTP synchronization check was not operating right at the start of an
4531 SMTP session. For example, it could not catch a HELO sent before the client
4532 waited for the greeting. There is now a check for outstanding input at the
4533 point when the greeting is written. Because of the duplex, asynchronous
4534 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4535 way, but not yet received, when the check is performed.
4537 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4538 on TCP/IP sockets, because this apparently causes some broken clients to
4541 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4542 unchanged) from the Cygwin maintainer.
4544 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4545 was supported instead of testing for mbx. Effectively a typo.
4547 15. The spa authenticator server code was not checking that the input it
4548 received was valid base64.
4550 16. The debug output line for the "set" modifier in ACLs was not showing the
4551 name of the variable that was being set.
4553 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4555 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4556 lookup on its own. Something else was needed to trigger the lookup. For
4557 example, a match in host_lookup or the need for a host name in a host list.
4558 Now, if $sender_host_name is referenced and the host name has not yet been
4559 looked up, a lookup is performed. If the lookup fails, the variable remains
4560 empty, and $host_lookup_failed is set to "1".
4562 19. Added "eqi" as a case-independent comparison operator.
4564 20. The saslauthd authentication condition could segfault if neither service
4565 nor realm was specified.
4567 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4568 error message that was logged was misleading, and incoming SMTP
4569 connections were dropped. The message is now more accurate, and temporary
4570 errors are given to SMTP connections.
4572 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4573 (except RSET) with a 421 error, until QUIT is received. However, it was
4574 failing to send a response to QUIT.
4576 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4577 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4578 ACL is now run after the helo_try_verify_hosts code.
4580 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4581 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4582 case-independent, so other case variants are also recognized). Apparently
4583 some systems use these upper case variants.
4585 25. If more than two messages were waiting for the same host, and a transport
4586 filter was specified for the transport, Exim sent two messages over the
4587 same TCP/IP connection, and then failed with "socket operation on non-
4588 socket" when it tried to send the third.
4590 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4592 27. The extern definition of crypt16() in expand.c was not being excluded when
4593 the OS had its own crypt16() function.
4595 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4596 as a preferred synonym for return_size_limit, both as an option and as an
4599 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4601 30. Changed the default configuration ACL to relax the local part checking rule
4602 for addresses that are not in any local domains. For these addresses,
4603 slashes and pipe symbols are allowed within local parts, but the sequence
4604 /../ is explicitly forbidden.
4606 31. SPA server authentication was not clearing the challenge buffer before
4609 32. log_message in a "warn" ACL statement was writing to the reject log as
4610 well as to the main log, which contradicts the documentation and doesn't
4611 seem right (because no rejection is happening). So I have stopped it.
4613 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4614 However, I am unable to do any testing of this.
4616 34. Fixed an infelicity in the appendfile transport. When checking directories
4617 for a mailbox, to see if any needed to be created, it was accidentally
4618 using path names with one or more superfluous leading slashes; tracing
4619 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4621 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4622 contents are added to the log line that is written for every discarded
4623 recipient. (Previously a log_message setting was ignored.)
4625 36. The ${quote: operator now quotes the string if it is empty.
4627 37. The install script runs exim in order to find its version number. If for
4628 some reason other than non-existence or emptiness, which it checks, it
4629 could not run './exim', it was installing it with an empty version number,
4630 i.e. as "exim-". This error state is now caught, and the installation is
4633 38. An argument was missing from the function that creates an error message
4634 when Exim fails to connect to the socket for saslauthd authentication.
4635 This could cause Exim to crash, or give a corrupted message.
4637 39. Added isip, isip4, and isip6 to ${if conditions.
4639 40. The ACL variables $acl_xx are now saved with the message, and can be
4640 accessed later in routers, transports, and filters.
4642 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4643 strings in the file are not string-expanded.
4645 42. If a MAIL command specified a SIZE value that was too large to fit into an
4646 int variable, the check against message_size_limit failed. Such values are
4647 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4648 day this will have to be increased, but I don't think I want to be around
4649 when emails are that large.
4656 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4657 iconv() is not standard in FreeBSD.
4659 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4660 IPv6 enabled. The observed symptom was a segmentation fault on return from
4661 the function os_common_find_running_interfaces() in src/os.c.
4663 3. In the check_special_case() function in daemon.c I had used "errno" as an
4664 argument name, which causes warnings on some systems. This was basically a
4665 typo, since it was named "eno" in the comments!
4667 4. The code that waits for the clock to tick (at a resolution of some fraction
4668 of a second) so as to ensure message-id uniqueness was always waiting for
4669 at least one whole tick, when it could have waited for less. [This is
4670 almost certainly not relevant at current processor speeds, where it is
4671 unlikely to ever wait at all. But we try to future-proof.]
4673 5. The function that sleeps for a time interval that includes fractions of a
4674 second contained a race. It did not block SIGALRM between setting the
4675 timer, and suspending (a couple of lines later). If the interval was short
4676 and the sigsuspend() was delayed until after it had expired, the suspension
4677 never ended. On busy systems this could lead to processes getting stuck for
4680 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4681 process, before it forks any delivery processes. The open lookup caching
4682 mechanism meant that the open file or database connection was passed into
4683 the delivery process. The problem was that delivery processes always tidy
4684 up cached lookup data. This could cause a problem for the next delivery
4685 process started by the queue runner, because the external queue runner
4686 process does not know about the closure. So the next delivery process
4687 still has data in the lookup cache. In the case of a file lookup, there was
4688 no problem because closing a file descriptor in a subprocess doesn't affect
4689 the parent. However, if the lookup was caching a connection to a database,
4690 the connection was closed, and the second delivery process was likely to
4691 see errors such as "PGSQL: query failed: server closed the connection
4692 unexpectedly". The problem has been fixed by closing all cached lookups
4693 in a queue runner before running a delivery process.
4695 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4696 seem to have the "const" qualifier which it has on other OS. I've
4699 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4700 *of the same type* (client or server) with the same public name that an
4701 error should be diagnosed.
4703 9. When Exim looked up a host name for an IP address, but failed to find the
4704 original IP address when looking up the host name (a safety check), it
4705 output the message "<ip address> does not match any IP for NULL", which was
4706 confusing, to say the least. The bug was that the host name should have
4707 appeared instead of "NULL".
4709 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4710 user that is built into the binary, and the -C or -D options is used, root
4711 privilege is dropped before the configuration file is read. In addition,
4712 logging is switched to stderr instead of the normal log files. If the
4713 configuration then re-defines the Exim user, the unprivileged environment
4714 is probably not what is expected, so Exim logs a panic warning message (but
4717 However, if deliver_drop_privilege is set, the unprivileged state may well
4718 be exactly what is intended, so the warning has been cut out in that case,
4719 and Exim is allowed to try to write to its normal log files.
4725 1. smtp_return_error_details was not giving details for temporary sender
4726 or receiver verification errors.
4728 2. Diagnose a configuration error if two authenticators have the same public
4731 3. Exim used not to create the message log file for a message until the first
4732 delivery attempt. This could be confusing when incoming messages were held
4733 for policy or load reasons. The message log file is now created at the time
4734 the message is received, and an initial "Received" line is written to it.
4736 4. The automatically generated man page for command line options had a minor
4737 bug that caused no ill effects; however, a more serious problem was that
4738 the procedure for building the man page automatically didn't always
4739 operate. Consequently, release 4.20 contains an out-of-date version. This
4740 shouldn't happen again.
4742 5. When building Exim with embedded Perl support, the script that builds the
4743 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4744 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4746 6. The freeze_tell option was not being used for messages that were frozen on
4747 arrival, either by an ACL or by local_scan().
4749 7. Added the smtp_incomplete_transaction log selector.
4751 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4752 was accepting AUTH without a new EHLO.
4754 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4755 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4758 10. Logging of TCP/IP connections (when configured) now happens in the main
4759 daemon process instead of the child process, so that the TCP/IP connection
4760 count is more accurate (but it can never be perfect).
4762 11. The use of "drop" in a nested ACL was not being handled correctly in the
4763 outer ACL. Now, if condition failure induced by the nested "drop" causes
4764 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4765 or "require"), the connection is dropped.
4767 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4768 that yield "discard" can now be used with an "accept" or a "discard" verb,
4769 but an error is generated for any others (because I can't see a useful way
4770 to define what should happen).
4772 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4773 are now processed in the same way as lines in the Exim configuration file.
4774 In particular, continuation lines are supported.
4776 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4778 15. Added -ti meaning -t -i.
4780 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4781 domains, and warn by logging if others are found.
4783 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4784 returned from GIFCONF when Exim is trying to find the list of interfaces on
4785 a host. The code in os.c has been modified to copy each ifreq to an aligned
4786 structure in all cases.
4788 Also, in some cases, the returned ifreq's were being copied to a 'struct
4789 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4790 means the last couple of bytes of an IPv6 address could be chopped if the
4791 ifreq contained only a normal sockaddr (14 bytes storage).
4793 18. Named domain lists were not supported in the hosts_treat_as_local option.
4794 An entry such as +xxxx was not recognized, and was treated as a literal
4797 19. Ensure that header lines added by a DATA ACL are included in the reject log
4798 if the ACL subsequently rejects the message.
4800 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4801 MD5 (which is deprecated).
4803 21. When testing a filter file using -bf, Exim was writing a message when it
4804 took the sender from a "From " line in the message, but it was not doing so
4805 when it took $return_path from a Return-Path: header line. It now does.
4807 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4808 with a valid header line field name (a series of printing characters
4809 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4811 23. Changed "disc" in the source to "disk" to conform to the documentation and
4812 the book and for uniformity.
4814 24. Ignore Sendmail's -Ooption=value command line item.
4816 25. When execve() failed while trying to run a command in a pipe transport,
4817 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4818 could be confused with a return value of 69 from the command itself. This
4819 has been changed to 127, the value the shell returns if it is asked to run
4820 a non-existent command. The wording for the related log line suggests a
4821 non-existent command as the problem.
4823 26. If received_header_text expands to an empty string, do not add a Received:
4824 header line to the message. (Well, it adds a token one on the spool, but
4825 marks it "old" so that it doesn't get used or transmitted.)
4827 27. Installed eximstats 1.28 (addition of -nt option).
4829 28. There was no check for failure on the call to getsockname() in the daemon
4830 code. This can fail if there is a shortage of resources on the system, with
4831 ENOMEM, for example. A temporary error is now given on failure.
4833 29. Contrary to the C standard, it seems that in some environments, the
4834 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4835 program. Exim now does this explicitly; it affects the formatting of
4836 timestamps using strftime().
4838 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4839 complaints. I've now initialized all the variables, to avoid this.
4841 32. Header lines added by a system filter were not being "seen" during
4842 transport-time rewrites.
4844 33. The info_callback() function passed to OpenSSL is set up with type void
4845 (*)(SSL *, int, int), as described somewhere. However, when calling the
4846 function (actually a macro) that sets it up, the type void(*)() is
4847 expected. I've put in a cast to prevent warnings from picky compilers.
4849 34. If a DNS black list lookup found a CNAME record, but there were no A
4850 records associated with the domain it pointed at, Exim crashed.
4852 35. If a DNS black list lookup returned more than one A record, Exim ignored
4853 all but the first. It now scans all returned addresses if a particular IP
4854 value is being sought. In this situation, the contents of the
4855 $dnslist_value variable are a list of all the addresses, separated by a
4858 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4859 to accept a host name and all its aliases if the forward lookup for any of
4860 them yielded the IP address of the incoming connection. Now it accepts only
4861 those names whose forward lookup yields the correct IP address. Any other
4862 names are discarded. This closes a loophole whereby a rogue DNS
4863 administrator could create reverse DNS records to break through a
4864 wildcarded host restriction in an ACL.
4866 37. If a user filter or a system filter that ran in a subprocess used any of
4867 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4868 the wrong values were passed to the pipe command ($thisaddress had the
4869 value of $0, $0 had the value of $1, etc). This bug was introduced by
4870 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4872 38. Improved the line breaking for long SMTP error messages from ACLs.
4873 Previously, if there was no break point between 40 and 75 characters, Exim
4874 left the rest of the message alone. Two changes have been made: (a) I've
4875 reduced the minimum length to 35 characters; (b) if it can't find a break
4876 point between 35 and 75 characters, it looks ahead and uses the first one
4877 that it finds. This may give the occasional overlong line, but at least the
4878 remaining text gets split now.
4880 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4881 file descriptors might be low, and that setting 1000 would always raise it.
4882 It turns out that in some environments, the limit is already over 1000 and
4883 that lowering it causes trouble. So now Exim takes care not to decrease it.
4885 40. When delivering a message, the value of $return_path is set to $sender_
4886 address at the start of routing (routers may change the value). By an
4887 oversight, this default was not being set up when an address was tested by
4888 -bt or -bv, which affected the outcome if any router or filter referred to
4891 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4892 log only when "message" or "log_message" are set. However, if one of the
4893 conditions was an address verification, or a call to a nested ACL, the
4894 messages generated by the underlying test were being passed through. This
4895 no longer happens. The underlying message is available in $acl_verify_
4896 message for both "message" and "log_message" expansions, so it can be
4897 passed through if needed.
4899 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4900 new expansion $bh_ to give the encoded byte string without charset
4901 translation. Translation happens only if iconv() is available; HAVE_ICONV
4902 indicates this at build time. HEADERS_CHARSET gives the charset to
4903 translate to; headers_charset can change it in the configuration, and
4904 "headers charset" can change it in an individual filter file.
4906 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4907 that creates RFC 2047 encoded "words" labels them as that charset instead
4908 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4909 expansion operator; (ii) when Exim creates a From: line for a local
4910 message; (iii) when a header line is rewritten to include a "phrase" part.
4912 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4913 buggy, causing it to skip the first lines of messages whose message ID
4914 ended in 'D'. This would not have bitten before Exim release 4.14, because
4915 message IDs were unlikely to end in 'D' before then. The effect was to have
4916 incorrect size information for certain domains.
4918 45. #include "config.h" was missing at the start of the crypt16.c module. This
4919 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4922 46. If there was a timeout during a "random" callout check, Exim treated it as
4923 a failure of the random address, and carried on sending RSET and the real
4924 address. If the delay was just some slowness somewhere, the response to the
4925 original RCPT would be taken as a response to RSET and so on, causing
4926 mayhem of various kinds.
4928 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4929 when I implemented it. It didn't allow for the fact that some option values
4930 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4931 the right test for positive values.
4933 48. Domain names in DNS records are case-independent. Exim always looks them up
4934 in lower case. Some resolvers return domain names in exactly the case they
4935 appear in the zone file, that is, they may contain uppercase letters. Not
4936 all resolvers do this - some return always lower case. Exim was treating a
4937 change of case by a resolver as a change of domain, similar to a widening
4938 of a domain abbreviation. This triggered its re-routing code and so it was
4939 trying to route what was effectively the same domain again. This normally
4940 caused routing to fail (because the router wouldn't handle the domain
4941 twice). Now Exim checks for this case specially, and just changes the
4942 casing of the domain that it ultimately uses when it transmits the message
4945 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4948 50. If a filter generated a file delivery with a non-absolute name (possible if
4949 no home directory exists for the router), the forbid_file option was not
4952 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4953 the existing equality matching.
4955 52. Exim was using ints instead of ino_t variables in some places where it was
4956 dealing with inode numbers.
4958 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4959 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4960 called TMPDIR, and if it finds it is different, it changes its value.
4962 54. The smtp_printf() function is now made available to local_scan() so
4963 additional output lines can be written before returning. There is also an
4964 smtp_fflush() function to enable the detection of a dropped connection.
4965 The variables smtp_input and smtp_batched_input are exported to
4968 55. Changed the default runtime configuration: the message "Unknown user"
4969 has been removed from the ACL, and instead placed on the localuser router,
4970 using the cannot_route_message feature. This means that any verification
4971 failures that generate their own messages won't get overridden. Similarly,
4972 the "Unrouteable address" message that was in the ACL for unverifiable
4973 relay addresses has also been removed.
4975 56. Added hosts_avoid_esmtp to the smtp transport.
4977 57. The exicyclog script was not checking for the esoteric option
4978 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4979 will work only if exicyclog is run under the appropriate euid.
4981 58. Following a discussion on the list, the rules by which Exim recognises line
4982 endings on incoming messages have been changed. The -dropcr and drop_cr
4983 options are now no-ops, retained only for backwards compatibility. The
4984 following line terminators are recognized: LF CRLF CR. However, special
4985 processing applies to CR:
4987 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4988 nor a local message in the state where . is a terminator.
4990 (ii) If a bare CR is encountered in a header line, an extra space is added
4991 after the line terminator so as not to end the header. The reasoning
4992 behind this is that bare CRs in header lines are most likely either
4993 to be mistakes, or people trying to play silly games.
4995 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4996 was being incorrectly given as 18 bytes larger than it should have been.
4997 This is a VOB (very old bug).
4999 60. This may never have affected anything current, but just in case it has:
5000 When the local host is found other than at the start of a list of hosts,
5001 the local host, those with the same MX, and any that follow, are discarded.
5002 When the list in question was part of a longer list of hosts, the following
5003 hosts (not currently being processed) were also being discarded. This no
5004 longer happens. I'm not sure if this situation could ever has previously
5007 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5010 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5011 header. This is something that is recommended in a new Internet Draft, and
5012 is something that is documented as being done by Sendmail. There are two
5013 possible values. For messages generated by the autoreply transport, Exim
5016 Auto-Submitted: auto-replied
5018 whereas for all other generated messages (e.g. bounces) it adds
5020 Auto-Submitted: auto-generated
5022 63. The "personal" condition in filters now includes a test for the
5023 Auto-Submitted: header. If it contains the string "auto-" the message it
5024 not considered personal.
5026 64. Added rcpt_include_affixes as a generic transport option.
5028 65. Added queue_only_override (default true).
5030 66. Added the syslog_duplication option.
5032 67. If what should have been the first header line of a message consisted of
5033 a space followed by a colon, Exim was mis-interpreting it as a header line.
5034 It isn't of course - it is syntactically invalid and should therefore be
5035 treated as the start of the message body. The misbehaviour could have
5036 caused a number of strange effects, including loss of data in subsequent
5037 header lines, and spool format errors.
5039 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5040 client host had authenticated. This control can now be exercised by an ACL
5041 for more flexibility.
5043 69. By default, callouts do not happen when testing with -bh. There is now a
5044 variant, -bhc, which does actually run the callout code, including
5045 consulting and updating the callout cache.
5047 70. Added support for saslauthd authentication, courtesy of Alexander
5050 71. If statvfs() failed on the spool or log directories while checking their
5051 size for availability, Exim confusingly gave the error "space shortage".
5052 Furthermore, in debugging mode it crashed with a floating point exception.
5053 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5054 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5055 serious problem, Exim now writes to the main and panic logs when this
5056 happens, with details of the failure. It then refuses to accept the
5057 incoming message, giving the message "spool directory problem" or "log
5058 directory problem" with a 421 code for SMTP messages.
5060 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5061 0, 1, and 2 exist, because some OS complain for execs without them (see
5062 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5063 descriptors. However, the code omitted to check that the open succeeded,
5064 causing mysterious errors if for some reason the permissions on /dev/null
5065 got screwed. Now Exim writes a message to the main and panic logs, and
5066 bombs out if it can't open /dev/null.
5068 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5069 interact so that it is all more flexible. It is supposed to remain
5070 backwards compatible. Also added extra_local_interfaces.
5072 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5073 to bomb out with an assertion failure - to the client this appears as a
5074 connection drop. This problem occurs in the part of the code that was taken
5075 from the Samba project. Fortunately, the assertion is in a very simple
5076 function, so I have fixed this by reproducing the function inline in the
5077 one place where it is called, and arranging for authentication to fail
5078 instead of killing the process with assert().
5080 75. The SPA client code was not working when the server requested OEM rather
5081 than Unicode encoding.
5083 76. Added code to make require_files with a specific uid setting more usable in
5084 the case where statting the file as root fails - usually a non-root-mounted
5085 NFS file system. When this happens and the failure is EACCES, Exim now
5086 forks a subprocess and does the per-uid checking as the relevant uid.
5088 77. Added process_log_path.
5090 78. If log_file_path was not explicitly set, a setting of check_log_space or
5091 check_log_inodes was ignored.
5093 79. If a space check for the spool or log partitions fails, the incident is now
5094 logged. Of course, in the latter case the data may get lost...
5096 80. Added the %p formatting code to string_format() so that it can be used to
5097 print addresses in debug_print(). Adjusted all the address printing in the
5098 debugging in store.c to use %p rather than %d.
5100 81. There was a concern that a line of code in smtp_in.c could overflow a
5101 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5102 initially expressed, the concern was not well-founded, because trailing
5103 spaces are removed early. However, if the trailing spaces were followed by
5104 a NULL, they did not get removed, so the overflow was possible. Two fixes
5107 (a) I re-wrote the offending code in a cleaner fashion.
5108 (b) If an incoming SMTP command contains a NULL character, it is rejected
5111 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5112 runs initgroups(), so that if the Exim user is in any additional groups,
5113 they will be used during message reception.
5119 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob