1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.625 2010/06/06 02:08:50 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.
41 PP/12 Bugzilla 973: Implement --version.
47 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
48 $data_path, and $header_path variables; fixed documentation bugs and
51 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
52 exipick to access non-standard spools, including the "frozen" queue
55 NM/01 Bugzilla 965: Support mysql stored procedures.
56 Patch from Alain Williams
58 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
60 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
61 Patch from Andreas Metzler
63 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
64 Patch from Kirill Miazine
66 NM/05 Bugzilla 671: Added umask to procmail example.
68 JJ/03 installed exipick 20100323.0, fixing doc bug
70 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
71 directory. Notification and patch from Dan Rosenberg.
73 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
75 TK/02 Improve log output when DKIM signing operation fails.
77 MH/01 Treat the transport option dkim_domain as a colon separated
78 list, not as a single string, and sign the message with each element,
79 omitting multiple occurences of the same signer.
81 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
82 Bugzilla 985, 986. Patch by Simon Arlott
84 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
87 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
88 MBX locking. Notification from Dan Rosenberg.
94 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
96 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
98 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
100 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
102 NM/04 Bugzilla 915: Flag broken perl installation during build.
108 TK/01 Added patch by Johannes Berg that expands the main option
109 "spamd_address" if it starts with a dollar sign.
111 TK/02 Write list of recipients to X-Envelope-Sender header when building
112 the mbox-format spool file for content scanning (suggested by Jakob
115 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
116 (http://www.dcc-servers.net/) support via dccifd. Activated by
117 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
119 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
120 by Mark Daniel Reidel <mr@df.eu>.
122 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
123 When building exim an external PCRE library is now needed -
124 PCRE is a system library on the majority of modern systems.
125 See entry on PCRE_LIBS in EDITME file.
127 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
128 conversation. Added nologin parameter to request.
129 Patch contributed by Kirill Miazine.
131 TF/01 Do not log submission mode rewrites if they do not change the address.
133 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
135 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
136 log files in place. Contributed by Roberto Lima.
138 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
140 TF/03 Bugzilla 615: When checking the local_parts router precondition
141 after a local_part_suffix or local_part_prefix option, Exim now
142 does not use the address's named list lookup cache, since this
143 contains cached lookups for the whole local part.
145 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
146 Robert Millan. Documentation is in experimental-spec.txt.
148 TF/04 Bugzilla 668: Fix parallel build (make -j).
150 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
152 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
153 Patch provided by Jan Srzednicki.
155 TF/05 Leading white space used to be stripped from $spam_report which
156 wrecked the formatting. Now it is preserved.
158 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
159 that they are available at delivery time.
161 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
163 TF/08 TLS error reporting now respects the incoming_interface and
164 incoming_port log selectors.
166 TF/09 Produce a more useful error message if an SMTP transport's hosts
167 setting expands to an empty string.
169 NM/06 Bugzilla 744: EXPN did not work under TLS.
170 Patch provided by Phil Pennock.
172 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
173 Patch provided by Richard Godbee.
175 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
176 acl_smtp_notquit, added index entry.
178 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
179 Patch provided by Eugene Bujak.
181 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
182 accept(). Patch provided by Maxim Dounin.
184 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
185 Patch provided by Phil Pennock.
187 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
189 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
190 Patch provided by Brad "anomie" Jorsch.
192 NM/14 Bugzilla 622: Added timeout setting to transport filter.
193 Patch provided by Dean Brooks.
195 TK/05 Add native DKIM support (does not depend on external libraries).
197 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
198 Patch provided by Graeme Fowler.
200 NM/16 Bugzilla 851: Documentation example syntax fix.
202 NM/17 Changed NOTICE file to remove references to embedded PCRE.
204 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
207 NM/19 Bugzilla 745: TLS version reporting.
208 Patch provided by Phil Pennock.
210 NM/20 Bugzilla 167: bool: condition support.
211 Patch provided by Phil Pennock.
213 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
214 clients. Patch provided by Phil Pennock.
216 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
217 Patch provided by Brad "anomie" Jorsch.
219 NM/23 Bugzilla 687: Fix misparses in eximstats.
220 Patch provided by Heiko Schlittermann.
222 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
223 Patch provided by Heiko Schlittermann.
225 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
226 plus update to original patch.
228 NM/26 Bugzilla 799: Documentation correction for ratelimit.
230 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
231 Patch provided by David Brownlee.
233 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
235 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
237 NM/30 Bugzilla 888: TLS documentation bugfixes.
239 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
241 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
242 Unlike the original bugzilla I have changed all shell scripts in src tree.
244 NM/33 Bugzilla 898: Transport filter timeout fix.
245 Patch by Todd Rinaldo.
247 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
248 Patch by Serge Demonchaux.
250 NM/35 Bugzilla 39: Base64 decode bug fixes.
251 Patch by Jakob Hirsch.
253 NM/36 Bugzilla 909: Correct connect() call in dcc code.
255 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
257 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
259 NM/39 Bugzilla 911: Fixed MakeLinks build script.
265 TK/01 Add preliminary DKIM support. Currently requires a forked version of
266 ALT-N's libdkim that I have put here:
267 http://duncanthrax.net/exim-experimental/
269 Note to Michael Haardt: I had to rename some vars in sieve.c. They
270 were called 'true' and it seems that C99 defines that as a reserved
271 keyword to be used with 'bool' variable types. That means you could
272 not include C99-style headers which use bools without triggering
273 build errors in sieve.c.
275 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
276 as mailq or other aliases. Changed the --help handling significantly
277 to do whats expected. exim_usage() emits usage/help information.
279 SC/01 Added the -bylocaldomain option to eximstats.
281 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
283 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
285 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
291 PH/01 Another patch from the Sieve maintainer.
293 PH/02 When an IPv6 address is converted to a string for single-key lookup
294 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
295 dots are used instead of colons so that keys in lsearch files need not
296 contain colons. This was done some time before quoting was made available
297 in lsearch files. However, iplsearch files do require colons in IPv6 keys
298 (notated using the quote facility) so as to distinguish them from IPv4
299 keys. This meant that lookups for IP addresses in host lists did not work
300 for iplsearch lookups.
302 This has been fixed by arranging for IPv6 addresses to be expressed with
303 colons if the lookup type is iplsearch. This is not incompatible, because
304 previously such lookups could never work.
306 The situation is now rather anomolous, since one *can* have colons in
307 ordinary lsearch keys. However, making the change in all cases is
308 incompatible and would probably break a number of configurations.
310 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
313 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
314 conversion specification without a maximum field width, thereby enabling
315 a rogue spamd server to cause a buffer overflow. While nobody in their
316 right mind would setup Exim to query an untrusted spamd server, an
317 attacker that gains access to a server running spamd could potentially
318 exploit this vulnerability to run arbitrary code as the Exim user.
320 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
321 $primary_hostname instead of what libspf2 thinks the hosts name is.
323 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
324 a directory entry by the name of the lookup key. Previously, if a
325 symlink pointed to a non-existing file or a file in a directory that
326 Exim lacked permissions to read, a lookup for a key matching that
327 symlink would fail. Now it is enough that a matching directory entry
328 exists, symlink or not. (Bugzilla 503.)
330 PH/03 The body_linecount and body_zerocount variables are now exported in the
333 PH/04 Added the $dnslist_matched variable.
335 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
336 This means they are set thereafter only if the connection becomes
339 PH/06 Added the client_condition to authenticators so that some can be skipped
340 by clients under certain conditions.
342 PH/07 The error message for a badly-placed control=no_multiline_responses left
343 "_responses" off the end of the name.
345 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
347 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
348 (without spaces) instead of just copying the configuration text.
350 PH/10 Added the /noupdate option to the ratelimit ACL condition.
352 PH/11 Added $max_received_linelength.
354 PH/12 Added +ignore_defer and +include_defer to host lists.
356 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
357 way in which PCRE > 7.0 is built.
359 PH/14 Implemented queue_only_load_latch.
361 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
362 MAIL command. The effect was to mangle the value on 64-bit systems.
364 PH/16 Another patch from the Sieve maintainer.
366 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
368 PH/18 If a system quota error occurred while trying to create the file for
369 a maildir delivery, the message "Mailbox is full" was not appended to the
370 bounce if the delivery eventually timed out. Change 4.67/27 below applied
371 only to a quota excession during the actual writing of the file.
373 PH/19 It seems that peer DN values may contain newlines (and other non-printing
374 characters?) which causes problems in log lines. The DN values are now
375 passed through string_printing() before being added to log lines.
377 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
378 and InterBase are left for another time.)
380 PH/21 Added message_body_newlines option.
382 PH/22 Guard against possible overflow in moan_check_errorcopy().
384 PH/23 POSIX allows open() to be a macro; guard against that.
386 PH/24 If the recipient of an error message contained an @ in the local part
387 (suitably quoted, of course), incorrect values were put in $domain and
388 $local_part during the evaluation of errors_copy.
394 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
395 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
398 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
399 issue a MAIL command.
401 PH/02 In an ACL statement such as
403 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
405 if a client was not listed at all, or was listed with a value other than
406 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
407 the condition was not true (as it should be), so access was not denied.
408 The bug was that the ! inversion was incorrectly passed on to the second
409 item. This has been fixed.
411 PH/03 Added additional dnslists conditions == and =& which are different from
412 = and & when the dns lookup returns more than one IP address.
414 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
415 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
417 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
418 FSYNC, which compiles an option called disable_fsync that allows for
419 bypassing fsync(). The documentation is heavily laced with warnings.
421 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
423 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
424 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
425 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
426 including adding "make clean"; (3) Added -fPIC when compiling the test
427 dynamically loaded module, to get rid of a warning.
429 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
430 message fails, move_frozen_messages = true and ignore_bounce_errors_after
431 = 0s. The bug is otherwise harmless.
433 PH/07 There was a bug in the dovecot authenticator such that the value of
434 $auth1 could be overwritten, and so not correctly preserved, after a
435 successful authentication. This usually meant that the value preserved by
436 the server_setid option was incorrect.
438 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
440 PH/09 Installed PCRE release 7.0.
442 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
443 run for batched SMTP input. It is now run at the start of every message
444 in the batch. While fixing this I discovered that the process information
445 (output by running exiwhat) was not always getting set for -bs and -bS
446 input. This is fixed, and it now also says "batched" for BSMTP.
448 PH/11 Added control=no_pipelining.
450 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
451 patch, slightly modified), and move the expansion of helo_data till after
452 the connection is made in the smtp transport (so it can use these
455 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
457 PH/14 Added log_selector = +pid.
459 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
461 PH/16 Add ${if forany and ${if forall.
463 PH/17 Added dsn_from option to vary the From: line in DSNs.
465 PH/18 Flush SMTP output before performing a callout, unless control =
466 no_callout_flush is set.
468 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
469 was true (the default) a successful delivery failed to delete the retry
470 item, thus causing premature timeout of the address. The bug is now
473 PH/20 Added hosts_avoid_pipelining to the smtp transport.
475 PH/21 Long custom messages for fakedefer and fakereject are now split up
476 into multiline reponses in the same way that messages for "deny" and
477 other ACL rejections are.
479 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
480 with slight modification.
482 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
483 draft, changing the syntax and factoring some duplicate code".
485 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
486 for deliveries of the second and subsequent messages over the same SMTP
489 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
490 ${reduce, with only minor "tidies".
492 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
494 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
495 expansion side effects.
497 PH/27 When a message times out after an over-quota error from an Exim-imposed
498 quota, the bounce message says "mailbox is full". This message was not
499 being given when it was a system quota that was exceeded. It now should
502 MH/03 Made $recipients available in local_scan(). local_scan() already has
503 better access to the recipient list through recipients_list[], but
504 $recipients can be useful in postmaster-provided expansion strings.
506 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
507 in the case of a MAIL command with additional options following the
508 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
509 were accidentally chopped off.
511 PH/29 SMTP synchronization checks are implemented when a command is read -
512 there is a check that no more input is waiting when there shouldn't be
513 any. However, for some commands, a delay in an ACL can mean that it is
514 some time before the response is written. In this time, more input might
515 arrive, invalidly. So now there are extra checks after an ACL has run for
516 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
517 pipelining has not been advertised.
519 PH/30 MH's patch to allow iscntrl() characters to be list separators.
521 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
522 returned in the SMTP response when smtp_return_error_details was false.
525 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
526 instead of the C I/O that was originally supplied, because problems were
529 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
530 Exim which did not show up earlier: it was assuming that a call to
531 SSL_CTX_set_info_callback() might give an error value. In fact, there is
532 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
533 was a macro that became an assignment, so it seemed to work. This has
534 changed to a proper function call with a void return, hence the compile
535 error. Exim's code has been fixed.
537 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
540 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
542 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
544 PH/37 If a message is not accepted after it has had an id assigned (e.g.
545 because it turns out to be too big or there is a timeout) there is no
546 "Completed" line in the log. When some messages of this type were
547 selected by exigrep, they were listed as "not completed". Others were
548 picked up by some special patterns. I have improved the selection
549 criteria to be more general.
551 PH/38 The host_find_failed option in the manualroute router can now be set
552 to "ignore", to completely ignore a host whose IP address cannot be
553 found. If all hosts are ignored, the behaviour is controlled by the new
554 host_all_ignored option.
556 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
557 homing or because of multiple MX records with /mx) generated more than
558 one IP address, and the following item turned out to be the local host,
559 all the secondary addresses of the first item were incorrectly removed
560 from the list, along with the local host and any following hosts (which
561 is what is supposed to happen).
563 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
564 whoever called Exim into the -H file. In the case of the daemon it was
565 behaving confusingly. When first started, it used values for whoever
566 started the daemon, but after a SIGHUP it used the Exim user (because it
567 calls itself on a restart). I have changed the code so that it now always
570 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
571 message are rejected with the same error (e.g. no authentication or bad
572 sender address), and a DATA command is nevertheless sent (as can happen
573 with PIPELINING or a stupid MUA), the error message that was given to the
574 RCPT commands is included in the rejection of the DATA command. This is
575 intended to be helpful for MUAs that show only the final error to their
578 PH/42 Another patch from the Sieve maintainer.
580 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
581 Eximstats - Fixed some broken HTML links and added missing column headers
583 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
584 columns for Rejects, Temp Rejects, Ham, and Spam rows.
586 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
588 PH/43 Yet another patch from the Sieve maintainer.
590 PH/44 I found a way to check for a TCP/IP connection going away before sending
591 the response to the final '.' that terminates a message, but only in the
592 case where the client has not sent further data following the '.'
593 (unfortunately, this is allowed). However, in many cases there won't be
594 any further data because there won't be any more messages to send. A call
595 to select() can be used: if it shows that the input is "ready", there is
596 either input waiting, or the socket has been closed. An attempt to read
597 the next input character can distinguish the two cases. Previously, Exim
598 would have sent an OK response which the client would never have see.
599 This could lead to message repetition. This fix should cure that, at
600 least in a lot of common cases.
602 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
603 advertised in response to EHLO.
609 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
610 fixed by 4.65/MH/01 (is this a record?) are fixed:
612 (i) An empty string was always treated as zero by the numeric comparison
613 operators. This behaviour has been restored.
615 (ii) It is documented that the numeric comparison operators always treat
616 their arguments as decimal numbers. This was broken in that numbers
617 starting with 0 were being interpreted as octal.
619 While fixing these problems I realized that there was another issue that
620 hadn't been noticed. Values of message_size_limit (both the global option
621 and the transport option) were treated as octal if they started with 0.
622 The documentation was vague. These values are now always treated as
623 decimal, and I will make that clear in the documentation.
629 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
630 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
633 MH/01 Don't check that the operands of numeric comparison operators are
634 integers when their expansion is in "skipping" mode (fixes bug
635 introduced by 4.64-PH/07).
637 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
638 child addresses, Exim now panics and dies. Previously, because the count
639 is held in a short int, deliveries were likely to be lost. As such a
640 large number of recipients for a single message is ridiculous
641 (performance will be very, very poor), I have chosen to impose a limit
642 rather than extend the field.
648 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
649 leftover -K file (the existence of which was triggered by #402).
650 While we were at it, introduced process PID as part of the -K
651 filename. This should rule out race conditions when creating
654 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
655 processing considerably. Previous code took too long for large mails,
656 triggering a timeout which in turn triggers #401.
658 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
659 in the DK code in transports.c. sendfile() is not really portable,
660 hence the _LINUX specificness.
662 TF/01 In the add_headers option to the mail command in an Exim filter,
663 there was a bug that Exim would claim a syntax error in any
664 header after the first one which had an odd number of characters
667 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
668 callout verification, Exim cached a "reject" for the entire domain. This
669 is correct for most verifications, but it is not correct for a recipient
670 verification with use_sender or use_postmaster set, because in that case
671 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
672 case of MAIL FROM:<> rejection from other early rejections (e.g.
673 rejection of HELO). When verifying a recipient using a non-null MAIL
674 address, the cache is ignored if it shows MAIL FROM:<> rejection.
675 Whatever the result of the callout, the value of the domain cache is
676 left unchanged (for any other kind of callout, getting as far as trying
677 RCPT means that the domain itself is ok).
679 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
682 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
683 manifest itself as EPIPE rather than ECONNECT. When tidying away a
684 session, the daemon ignores ECONNECT errors and logs others; it now
685 ignores EPIPE as well.
687 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
688 (quoted-printable decoding).
690 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
691 later the small subsequent patch to fix an introduced bug.
693 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
695 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
697 PH/08 An error is now given if message_size_limit is specified negative.
699 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
700 to be given (somewhat) arbitrary names.
702 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
705 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
706 miscellaneous code fixes
708 PH/10 Added the log_reject_target ACL modifier to specify where to log
711 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
712 hostname. This is wrong, because it relates to the incoming message (and
713 probably the interface on which it is arriving) and not to the outgoing
714 callout (which could be using a different interface). This has been
715 changed to use the value of the helo_data option from the smtp transport
716 instead - this is what is used when a message is actually being sent. If
717 there is no remote transport (possible with a router that sets up host
718 addresses), $smtp_active_hostname is used.
720 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
721 tweaks were necessary in order to get it to work (see also 21 below):
722 (a) The code assumed that strncpy() returns a negative number on buffer
723 overflow, which isn't the case. Replaced with Exim's string_format()
725 (b) There were several signed/unsigned issues. I just did the minimum
726 hacking in of casts. There is scope for a larger refactoring.
727 (c) The code used strcasecmp() which is not a standard C function.
728 Replaced with Exim's strcmpic() function.
729 (d) The code set only $1; it now sets $auth1 as well.
730 (e) A simple test gave the error "authentication client didn't specify
731 service in request". It would seem that Dovecot has changed its
732 interface. Fortunately there's a specification; I followed it and
733 changed what the client sends and it appears to be working now.
735 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
738 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
739 address A is aliased to B and C, where B exists and C does not. Without
740 -v the output is "A verified" because verification stops after a
741 successful redirection if more than one address is generated. However,
742 with -v the child addresses are also verified. Exim was outputting "A
743 failed to verify" and then showing the successful verification for C,
744 with its parentage. It now outputs "B failed to verify", showing B's
745 parentage before showing the successful verification of C.
747 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
748 look up a TXT record in a specific list after matching in a combined
751 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
752 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
753 they consult the DNS. I had assumed they would set it the way they
754 wanted; and indeed my experiments on Linux seem to show that in some
755 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
756 To be on the safe side, however, I have now made the interface to
757 host_find_byname() similar to host_find_bydns(), with an argument
758 containing the DNS resolver options. The host_find_byname() function now
759 sets these options at its start, just as host_find_bydns() does. The smtp
760 transport options dns_qualify_single and dns_search_parents are passed to
761 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
762 of host_find_byname() use the default settings of RES_DEFNAMES
763 (qualify_single) but not RES_DNSRCH (search_parents).
765 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
766 spool_read_header() do less string testing, by means of a preliminary
767 switch on the second character of optional "-foo" lines. (This is
768 overdue, caused by the large number of possibilities that now exist.
769 Originally there were few.) While I was there, I also converted the
770 str(n)cmp tests so they don't re-test the leading "-" and the first
771 character, in the hope this might squeeze out yet more improvement.
773 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
774 flag allowing group syntax was set by the header_syntax check but not
775 turned off, possible causing trouble later; (2) The flag was not being
776 set at all for the header_verify test, causing "group"-style headers to
777 be rejected. I have now set it in this case, and also caused header_
778 verify to ignore an empty address taken from a group. While doing this, I
779 came across some other cases where the code for allowing group syntax
780 while scanning a header line wasn't quite right (mostly, not resetting
781 the flag correctly in the right place). These bugs could have caused
782 trouble for malformed header lines. I hope it is now all correct.
784 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
785 with the "reply" argument non-NULL. The code, however (which originally
786 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
787 but it didn't always do it. This confused somebody who was copying the
788 code for some other use. I have removed all the tests.
790 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
791 feature that was used to support insecure browsers during the U.S. crypto
792 embargo. It requires special client support, and Exim is probably the
793 only MTA that supported it -- and would never use it because real RSA is
794 always available. This code has been removed, because it had the bad
795 effect of slowing Exim down by computing (never used) parameters for the
796 RSA_EXPORT functionality.
798 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
799 authenticator to fail if there's a tab character in the incoming data
800 (there should never be unless someone is messing about, as it's supposed
801 to be base64-encoded). Also added, on Timo's advice, the "secured" option
802 if the connection is using TLS or if the remote IP is the same as the
803 local IP, and the "valid-client-cert option" if a client certificate has
806 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
807 authenticators. This can be used for authorization after authentication
808 succeeds. (In the case of plaintext, it servers for both authentication
811 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
812 if any retry times were supplied.
814 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
815 connection, where there is no client IP address to check. In this
816 situation, the verify now always succeeds.
818 PH/25 Applied John Jetmore's -Mset patch.
820 PH/26 Added -bem to be like -Mset, but loading a message from a file.
822 PH/27 In a string expansion for a processed (not raw) header when multiple
823 headers of the same name were present, leading whitespace was being
824 removed from all of them, but trailing whitespace was being removed only
825 from the last one. Now trailing whitespace is removed from each header
826 before concatenation. Completely empty headers in a concatenation (as
829 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
830 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
832 PH/29 [Removed. This was a change that I later backed out, and forgot to
833 correct the ChangeLog entry (that I had efficiently created) before
834 committing the later change.]
836 PH/30 Exim was sometimes attempting to deliver messages that had suffered
837 address errors (4xx response to RCPT) over the same connection as other
838 messages routed to the same hosts. Such deliveries are always "forced",
839 so retry times are not inspected. This resulted in far too many retries
840 for the affected addresses. The effect occurred only when there were more
841 hosts than the hosts_max_try setting in the smtp transport when it had
842 the 4xx errors. Those hosts that it had tried were not added to the list
843 of hosts for which the message was waiting, so if all were tried, there
844 was no problem. Two fixes have been applied:
846 (i) If there are any address or message errors in an SMTP delivery, none
847 of the hosts (tried or untried) are now added to the list of hosts
848 for which the message is waiting, so the message should not be a
849 candidate for sending over the same connection that was used for a
850 successful delivery of some other message. This seems entirely
851 reasonable: after all the message is NOT "waiting for some host".
852 This is so "obvious" that I'm not sure why it wasn't done
853 previously. Hope I haven't missed anything, but it can't do any
854 harm, as the worst effect is to miss an optimization.
856 (ii) If, despite (i), such a delivery is accidentally attempted, the
857 routing retry time is respected, so at least it doesn't keep
858 hammering the server.
860 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
861 in ${readsocket because some servers need this prod.
863 PH/32 Added some extra debug output when updating a wait-xxx database.
865 PH/33 The hint "could be header name not terminated by colon", which has been
866 given for certain expansion errors for a long time, was not being given
867 for the ${if def:h_colon_omitted{... case.
869 PH/34 The spec says: "With one important exception, whenever a domain list is
870 being scanned, $domain contains the subject domain." There was at least
871 one case where this was not true.
873 PH/35 The error "getsockname() failed: connection reset by peer" was being
874 written to the panic log as well as the main log, but it isn't really
875 panic-worthy as it just means the connection died rather early on. I have
876 removed the panic log writing for the ECONNRESET error when getsockname()
879 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
880 runs only) independently of the message's sender address. This meant
881 that, if the 4xx error was in fact related to the sender, a different
882 message to the same recipient with a different sender could confuse
883 things. In particualar, this can happen when sending to a greylisting
884 server, but other circumstances could also provoke similar problems.
885 I have changed the default so that the retry time for these errors is now
886 based a combination of the sender and recipient addresses. This change
887 can be overridden by setting address_retry_include_sender=false in the
890 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
891 remote server are returned as part of bounce messages. This was not
892 happening for LMTP over a pipe (the lmtp transport), but now it is the
893 same for both kinds of LMTP.
895 PH/38 Despite being documented as not happening, Exim was rewriting addresses
896 in header lines that were in fact CNAMEs. This is no longer the case.
898 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
899 and queue runs started by the daemon processed all messages. This has
900 been fixed so that -R and -S can now usefully be given with -q<time>.
902 PH/40 Import PCRE release 6.7 (fixes some bugs).
904 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
906 PH/42 Give an error if -q is specified more than once.
908 PH/43 Renamed the variables $interface_address and $interface_port as
909 $received_ip_address and $received_port, to make it clear that these
910 values apply to message reception, and not to the outgoing interface when
911 a message is delivered. (The old names remain recognized, of course.)
913 PH/44 There was no timeout on the connect() call when using a Unix domain
914 socket in the ${readsocket expansion. There now is.
916 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
917 be meaningful with "accept".
919 SC/01 Eximstats V1.43
920 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
922 SC/02 Eximstats V1.44
923 Use a glob alias rather than an array ref in the generated
924 parser. This improves both readability and performance.
926 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
927 Collect SpamAssassin and rejection statistics.
928 Don't display local sender or destination tables unless
929 there is data to show.
930 Added average volumes into the top table text output.
932 SC/04 Eximstats V1.46
933 Collect data on the number of addresses (recipients)
934 as well as the number of messages.
936 SC/05 Eximstats V1.47
937 Added 'Message too big' to the list of mail rejection
938 reasons (thanks to Marco Gaiarin).
940 SC/06 Eximstats V1.48
941 Mainlog lines which have GMT offsets and are too short to
942 have a flag are now skipped.
944 SC/07 Eximstats V1.49 (Alain Williams)
945 Added the -emptyok flag.
947 SC/08 Eximstats V1.50
948 Fixes for obtaining the IP address from reject messages.
950 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
951 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
952 whitesspace changes from 4.64-PH/27
954 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
957 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
958 are found, allow negative numbers in numeric criteria)
960 JJ/06 exipick.20061117.2, added new $message_body_missing variable
962 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
963 to match changes made in 4.64-PH/43
965 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
967 PH/47 Put in an explicit test for a DNS lookup of an address record where the
968 "domain" is actually an IP address, and force a failure. This locks out
969 those revolvers/nameservers that support "A-for-A" lookups, in
970 contravention of the specifications.
972 PH/48 When a host name was looked up from an IP address, and the subsequent
973 forward lookup of the name timed out, the host name was left in
974 $sender_host_name, contrary to the specification.
976 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
977 restricted to single-key lookups, Exim was not diagnosing an error if
978 * or *@ was used with a query-style lookup.
980 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
982 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
983 long ago, but noone interested enough thought of it. Let's just say that
984 the "1.1" means that there are some new functions that weren't there at
985 some point in the past.
987 PH/51 Error processing for expansion failure of helo_data from an smtp
988 transport during callout processing was broken.
990 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
991 tested/used via the -bh/-bhc/-bs options.
993 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
994 bug, fixed in subsequent PCRE releases).
996 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
997 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
999 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1005 SC/01 Use a glob alias rather than an array ref in eximstats generated
1006 parser. This improves both readability and performance.
1008 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1009 Don't display local sender or destination tables in eximstats unless
1010 there is data to show.
1011 Added average volumes into the eximstats top table text output.
1013 SC/03 Collect data on the number of addresses (recipients) as well
1014 as the number of messages in eximstats.
1016 TF/01 Correct an error in the documentation for the redirect router. Exim
1017 does (usually) call initgroups() when daemonizing.
1019 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1020 with consistent privilege compared to when running as a daemon.
1022 TF/03 Note in the spec that $authenticated_id is not set for local
1023 submissions from trusted users.
1025 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1026 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1028 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1029 by adding some example configuration directives to the default
1030 configuration file. A little bit of work is required to uncomment the
1031 directives and define how usernames and passwords are checked, but
1032 there is now a framework to start from.
1034 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1035 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1036 without this. I don't know how relevant this is to other LDAP libraries.
1038 PH/02 Add the verb name to the "unknown ACL verb" error.
1040 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1042 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1044 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1045 directory not expanded when it should be if an expanded home directory
1046 was set for the address (which is overridden by the transport).
1048 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1051 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1052 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1053 because it is too late at that time, and has no effect.
1055 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1056 security issue with \' (bugzilla #107). I could not use the
1057 PQescapeStringConn() function, because it needs a PGconn value as one of
1060 PH/08 When testing addresses using -bt, indicate those final addresses that
1061 are duplicates that would not cause an additional delivery. At least one
1062 person was confused, thinking that -bt output corresponded to deliveries.
1063 (Suppressing duplicates isn't a good idea as you lose the information
1064 about possibly different redirections that led to the duplicates.)
1066 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1067 systems where poll() doesn't work, in particular OS X.
1069 PH/10 Added more information to debugging output for retry time not reached.
1071 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1072 operations in malware.c.
1074 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1077 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1078 syslog_duplication set false, log lines that would normally be written
1079 both the the main log and to the reject log were not written to syslog at
1082 PH/14 In the default configuration, change the use of "message" in ACL warn
1083 statements to "add_header".
1085 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1086 not followed by a command (e.g. "seen endif").
1088 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1089 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1092 PH/17 Added extra conditions to the default value of delay_warning_condition
1096 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1097 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1098 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1101 The Auto-Submitted: and various List- headers are standardised, whereas I
1102 don't think Precedence: ever was.
1104 PH/18 Refactored debugging code in route_finduser() to show more information,
1105 in particular, the error code if getpwnam() issues one.
1107 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1108 This is apparently needed in addition to the PH/07 change above to avoid
1109 any possible encoding problems.
1111 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1112 but not after initializing Perl.
1114 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1115 output them only if debugging. By default they are written stderr,
1116 apparently, which is not desirable.
1118 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1121 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1124 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1126 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1127 authenticated or an ident call has been made. Suppress the default
1128 values for $authenticated_id and $authenticated_sender (but permit -oMai
1129 and -oMas) when testing with -bh.
1131 PH/24 Re-jigged the order of the tests in the default configuration so that the
1132 tests for valid domains and recipients precede the DNS black list and CSA
1133 tests, on the grounds that those ones are more expensive.
1135 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1136 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1137 This bug exists in every version of Exim that I still have, right back to
1140 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1141 However, an attempt to turn on case-sensitivity in a regex key by
1142 including (?-i) didn't work because the subject string was already
1143 lowercased, and the effects were non-intuitive. It turns out that a
1144 one-line patch can be used to allow (?-i) to work as expected.
1150 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1151 other effects) broke the use of negated acl sub-conditions.
1153 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1156 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1157 "Deny" causes Exim to reject the incoming connection with a 554 error.
1158 Unfortunately, if there is a major crisis, such as a disk failure,
1159 tcp-wrappers gives "deny", whereas what one would like would be some
1160 kind of temporary error. A kludge has been added to help with this.
1161 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1162 554 error is used if errno is still zero or contains ENOENT (which occurs
1163 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1166 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1168 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1169 errors. Otherwise a message that provokes a temporary error (when other
1170 messages do not) can cause a whole host to time out.
1172 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1173 addresses were routed directly to files or pipes from a redirect router.
1174 File deliveries just didn't batch; pipe deliveries might have suffered
1177 PH/06 A failure to get a lock for a hints database would erroneously always say
1178 "Failed to get write lock", even when it was really a read lock.
1180 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1181 of 0600. This has been changed to use the value of the lockfile_mode
1182 option (which defaults to 0600).
1184 PH/08 Applied small patch from the Sieve maintainer.
1186 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1187 folder from quota calculations, a direct delivery into this folder messed
1188 up the contents of the maildirsize file. This was because the regex was
1189 used only to exclude .Trash (or whatever) when the size of the mailbox
1190 was calculated. There was no check that a delivery was happening into an
1191 excluded directory. This bug has been fixed by ignoring all quota
1192 processing for deliveries into excluded directories.
1194 PH/10 Added the maildirfolder_create_regex option to appendfile.
1200 PH/01 The code for finding all the local interface addresses on a FreeBSD
1201 system running IPv6 was broken. This may well have applied to all BSD
1202 systems, as well as to others that have similar system calls. The broken
1203 code found IPv4 interfaces correctly, but gave incorrect values for the
1204 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1205 that it would not match correctly against @[] and not recognize the IPv6
1208 PH/02 The ipliteral router was not recognizing addresses of the form user@
1209 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1211 PH/03 Added disable_ipv6.
1213 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1214 decoded form, because it is most often used to construct To: headers
1215 lines in autoreplies, and the decoded form may well be syntactically
1216 invalid. However, $reply_address has leading white space removed, and all
1217 newlines turned into spaces so that the autoreply transport does not
1220 PH/05 If group was specified without a user on a router, and no group or user
1221 was specified on a transport, the group from the router was ignored.
1223 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1224 for visible compile-time settings that can be used to change these
1225 numbers, for those that want even more. Backwards compatibility with old
1226 spool files has been maintained. However, going back to a previous Exim
1227 release will lost any variables that are in spool files.
1229 PH/07 Two small changes when running in the test harness: increase delay when
1230 passing a TCP/IP connection to a new process, in case the original
1231 process has to generate a bounce, and remove special handling of
1232 127.0.0.2 (sic), which is no longer necessary.
1234 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1235 be the same on different OS.
1237 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1240 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1241 whether --show-vars was specified or not
1243 JJ/02 exipick: Added support for new ACL variable spool format introduced
1246 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1247 syntactically invalid From: or Reply-to: line, and a filter used this to
1248 generate an autoreply, and therefore failed to obtain an address for the
1249 autoreply, Exim could try to deliver to a non-existent relative file
1250 name, causing unrelated and misleading errors. What now happens is that
1251 it logs this as a hard delivery error, but does not attempt to create a
1254 PH/11 The exinext utility has a -C option for testing purposes, but although
1255 the given file was scanned by exinext itself; it wasn't being passed on
1256 when Exim was called.
1258 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1259 an end-of-file indication when reading a command response.
1261 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1262 compiled. In many other places in Exim, IPv6 addresses are always
1263 recognized, so I have changed this. It also means that IPv4 domain
1264 literals of the form [IPV4:n.n.n.n] are now always recognized.
1266 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1267 used if the router is not running as root, for example, when verifying at
1268 ACL time, or when using -bh. The debugging output from this situation was
1269 non-existent - all you got was a failure to exec. I have made two
1272 (a) Failures to set uid/gid, the current directory, or a process leader
1273 in a subprocess such as that created by queryprogram now generate
1274 suitable debugging ouput when -d is set.
1276 (b) The queryprogram router detects when it is not running as root,
1277 outputs suitable debugging information if -d is set, and then runs
1278 the subprocess without attempting to change uid/gid.
1280 PH/15 Minor change to Makefile for building test_host (undocumented testing
1283 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1284 additional section of a DNS packet that returns MX or SRV records.
1285 Instead, it always explicitly searches for A/AAAA records. This avoids
1286 major problems that occur when a DNS server includes only records of one
1287 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1288 fixed another bug: if SRV records were looked up and the corresponding
1289 address records were *not* found in the additional section, the port
1290 values from the SRV records were lost.
1292 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1293 using the correct key (the original address) when searching the retry
1294 rules in order to find which one to use for generating the retry hint.
1296 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1297 adding the default one. Similarly, if it contains a Reply-To: header, the
1298 errors_reply_to option, if set, is not used.
1300 PH/19 When calculating a retry time, Exim used to measure the "time since
1301 failure" by looking at the "first failed" field in the retry record. Now
1302 it does not use this if it is later than than the arrival time of the
1303 message. Instead it uses the arrival time. This makes for better
1304 behaviour in cases where some deliveries succeed, thus re-setting the
1305 "first failed" field. An example is a quota failure for a huge message
1306 when small messages continue to be delivered. Without this change, the
1307 "time since failure" will always be short, possible causing more frequent
1308 delivery attempts for the huge message than are intended.
1309 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1311 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1312 $1, $2, $3) because the numerical variables can be reset during some
1313 expansion items (e.g. "match"), thereby losing the authentication data.
1315 PH/21 Make -bV show the size of off_t variables so that the test suite can
1316 decide whether to run tests for quotas > 2G.
1318 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1319 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1320 filecount value is greater than 2G or if a quota value is greater than 2G
1321 on a system where the size of off_t is not greater than 4, a panic error
1324 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1325 never match. The debug and -bh output now contains an explicit error
1326 message indicating a malformed IPv4 address or mask.
1328 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1329 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1330 PH/23 above applies.
1332 PH/25 Do not write to syslog when running in the test harness. The only
1333 occasion when this arises is a failure to open the main or panic logs
1334 (for which there is an explicit test).
1336 PH/26 Added the /no_tell option to "control=freeze".
1338 PH/27 If a host name lookup failed very early in a connection, for example, if
1339 the IP address matched host_lookup and the reverse lookup yielded a name
1340 that did not have a forward lookup, an error message of the form "no IP
1341 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1342 could be logged. Now it outputs the IP address instead of "NULL".
1344 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1345 allows the sender and the authenticated sender to be set when
1346 submitting a message from within Exim. Since child_open_exim() is
1347 documented for local_scan(), the new function should be too.
1349 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1350 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1351 results in an empty string is now treated as unset.
1353 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1355 PH/31 Added sender_verify_fail logging option.
1357 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1358 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1359 tidied the source and removed it altogether.
1361 PH/33 When a queue run was abandoned because the load average was too high, a
1362 log line was always written; now it is written only if the queue_run log
1363 selector is set. In addition, the log line for abandonment now contains
1364 information about the queue run such as the pid. This is always present
1365 in "start" and "stop" lines but was omitted from the "abandon" line.
1367 PH/34 Omit spaces between a header name and the colon in the error message that
1368 is given when verify = headers_syntax fails (if there are lots of them,
1369 the message gets confusing).
1371 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1372 names, as there are now some PTR records that contain slashes. This check
1373 is only to protect against broken name servers that fall over on strange
1374 characters, so the fact that it applies to all lookups doesn't matter.
1376 PH/36 Now that the new test suite is complete, we can remove some of the
1377 special code in Exim that was needed for the old test suite. For example,
1378 sorting DNS records because real resolvers return them in an arbitrary
1379 order. The new test suite's fake resolver always returns records in the
1382 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1383 bounces) except when queue_only is set, to avoid logging races between
1384 the different processes.
1386 PH/38 Panic-die if .include specifies a non-absolute path.
1388 PH/39 A tweak to the "H" retry rule from its user.
1390 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1391 a label. They prevented compilation on older perls.
1393 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1394 a warning to be raised on newish perls.
1396 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1397 on queue. Changes to match documented behaviour of showing count of
1398 messages matching specified criteria.
1400 PH/40 Changed the default ident timeout from 30s to 5s.
1402 PH/41 Added support for the use of login_cap features, on those BSD systems
1403 that have them, for controlling the resources used by pipe deliveries.
1405 PH/42 The content-scanning code uses fopen() to create files in which to put
1406 message data. Previously it was not paying any attention to the mode of
1407 the files. Exim runs with umask(0) because the rest of the code creates
1408 files with open(), and sets the required mode explicitly. Thus, these
1409 files were ending up world-writeable. This was not a big issue, because,
1410 being within the spool directory, they were not world-accessible. I have
1411 created a function called modefopen, which takes an additional mode
1412 argument. It sets umask(777), creates the file, chmods it to the required
1413 mode, then resets the umask. All the relevant calls to fopen() in the
1414 content scanning code have been changed to use this function.
1416 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1417 to 24 hours. This avoids potential overflow problems when processing G
1418 and H retry rules. I suspect nobody ever tinkers with this value.
1420 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1422 PH/45 When the plaintext authenticator is running as a client, the server's
1423 challenges are checked to ensure they are valid base64 strings. By
1424 default, the authentication attempt is cancelled if an invalid string is
1425 received. Setting client_ignore_invalid_base64 true ignores these errors.
1426 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1427 they are received. Thus, the responses can be made to depend on the
1428 challenges. If an invalid string is ignored, an empty string is placed in
1431 PH/46 Messages that are created by the autoreply transport now contains a
1432 References: header, in accordance with RFCs 2822 and 3834.
1434 PH/47 Added authenticated_sender_force to the smtp transport.
1436 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1438 PH/49 Installed latest patch from the Sieve maintainer.
1440 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1441 was also set, the appendfile transport was unnecessarily scanning a
1442 directory of message files (e.g. for maildir delivery) to find the count
1443 of files (along with the size), even though it did not need this
1444 information. It now does the scan only if it needs to find either the
1445 size of the count of files.
1447 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1449 PH/52 Two bugs concerned with error handling when the smtp transport is
1452 (i) Exim was not creating retry information for temporary errors given
1453 for individual recipients after the DATA command when the smtp transport
1454 was used in LMTP mode. This meant that they could be retried too
1455 frequently, and not timed out correctly.
1457 (ii) Exim was setting the flag that allows error details to be returned
1458 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1459 recipients that were returned after the DATA command.
1461 PH/53 This is related to PH/52, but is more general: for any failing address,
1462 when detailed error information was permitted to be returned to the
1463 sender, but the error was temporary, then after the final timeout, only
1464 "retry timeout exceeded" was returned. Now it returns the full error as
1465 well as "retry timeout exceeded".
1467 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1468 do this, and (what is worse) MTAs that accept it.
1470 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1471 will now be deprecated.
1473 PH/56 New os.c-cygwin from the Cygwin maintainer.
1475 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1476 formats (previously only available in exim formats via -bpr, -bpru,
1477 and -bpra. Now also available in native and exiqgrep formats)
1479 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1480 with very large, slow to parse queues
1482 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1484 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1486 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1487 responses to authentication challenges, though it was showing the
1488 challenges; (ii) I've removed the CR characters from the debug output for
1491 PH/58 Allow for the insertion of a newline as well as a space when a string
1492 is turned into more than one encoded-word during RFC 2047 encoding. The
1493 Sieve code now uses this.
1495 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1496 data_4xx, lost_connection, tls_required.
1498 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1499 message was being sent as an SMTP response.
1501 PH/61 Add -l and -k options to exicyclog.
1503 PH/62 When verifying, if an address was redirected to one new address, so that
1504 verification continued, and the new address failed or deferred after
1505 having set something in $address_data, the value of $address_data was not
1506 passed back to the ACL. This was different to the case when no
1507 redirection occurred. The value is now passed back in both cases.
1509 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1510 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1511 use login_cap.h, so on its own it isn't the distinguishing feature. The
1512 new name refers directly to the setclassresources() function.
1514 PH/65 Added configuration files for NetBSD3.
1516 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1518 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1519 is preferred over IPv4.
1521 PH/68 The bounce_return_message and bounce_return_body options were not being
1522 honoured for bounces generated during the reception of non-SMTP messages.
1523 In particular, this applied to messages rejected by the ACL. This bug has
1524 been fixed. However, if bounce_return_message is true and bounce_return_
1525 body is false, the headers that are returned for a non-SMTP message
1526 include only those that have been read before the error was detected.
1527 (In the case of an ACL rejection, they have all been read.)
1529 PH/69 The HTML version of the specification is now built in a directory called
1530 spec_html instead of spec.html, because the latter looks like a path with
1531 a MIME-type, and this confuses some software.
1533 PH/70 Catch two compiler warnings in sieve.c.
1535 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1536 function verify_get_ident() calls ip_connect() to connect a socket, but
1537 if the "connect()" function timed out, ip_connect() used to close the
1538 socket. However, verify_get_ident() also closes the socket later, and in
1539 between Exim writes to the log, which may get opened at this point. When
1540 the socket was closed in ip_connect(), the log could get the same file
1541 descriptor number as the socket. This naturally causes chaos. The fix is
1542 not to close the socket in ip_connect(); the socket should be closed by
1543 the function that creates it. There was only one place in the code where
1544 this was missing, in the iplookup router, which I don't think anybody now
1545 uses, but I've fixed it anyway.
1547 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1548 well as to direct DNS lookups. Otherwise the handling of names in host
1549 lists is inconsistent and therefore confusing.
1555 PH/01 Two changes to the default runtime configuration:
1557 (1) Move the checks for relay_from_hosts and authenticated clients from
1558 after to before the (commented out) DNS black list checks.
1560 (2) Add control=submission to the relay_from_hosts and authenticated
1561 clients checks, on the grounds that messages accepted by these
1562 statements are most likely to be submissions.
1564 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1566 (1) Generate an error if the third argument for the ${prvs expansion is
1569 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1572 (3) Reset the variables that are obtained from the first argument of
1573 ${prvscheck and used in the second argument before leaving the code,
1574 because their memory is reclaimed, so using them afterwards may do
1577 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1578 one (it's much easier than Tom thought :-).
1580 (5) Because of (4), we can now allow for the use of $prvscheck_result
1581 inside the third argument.
1583 PH/03 For some reason, the default setting of PATH when running a command from
1584 a pipe transport was just "/usr/bin". I have changed it to
1587 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1588 anything to be listed in the output from -bV.
1590 PH/05 When a filter generated an autoreply, the entire To: header line was
1591 quoted in the delivery log line, like this:
1593 => >A.N.Other <ano@some.domain> <original@ddress> ...
1595 This has been changed so that it extracts the operative address. There
1596 may be more than one such address. If so, they are comma-separated, like
1599 => >ano@some.domain,ona@other.domain <original@ddress> ...
1601 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1602 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1603 not being looked up in the rDNS to get a host name, Exim was showing the
1604 IP address twice in Received: lines, even though the IP addresses were
1605 identical. For example:
1607 Received: from [1.2.3.4] (helo=[1.2.3.4])
1609 However, if the real host name was known, it was omitting the HELO data
1610 if it matched the actual IP address. This has been tidied up so that it
1611 doesn't show the same IP address twice.
1613 PH/07 When both +timestamp and +memory debugging was on, the value given by
1614 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1615 called by the debug printing, thereby overwriting the timestamp buffer.
1616 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1618 PH/08 When the original message was included in an autoreply transport, it
1619 always said "this is a copy of the message, including all the headers",
1620 even if body_only or headers_only was set. It now gives an appropriate
1623 PH/09 Applied a patch from the Sieve maintainer which:
1625 o fixes some comments
1626 o adds the (disabled) notify extension core
1627 o adds some debug output for the result of if/elsif tests
1628 o points to the current vacation draft in the documentation
1629 and documents the missing references header update
1633 o fixes a bug in processing the envelope test (when testing
1634 multiple envelope elements, the last element determinted the
1637 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1638 Electronic Mail") by including:
1640 Auto-submitted: auto-generated
1642 in the messages that it generates (bounce messages and others, such as
1643 warnings). In the case of bounce messages for non-SMTP mesages, there was
1644 also a typo: it was using "Auto_submitted" (underscore instead of
1645 hyphen). Since every message generated by Exim is necessarily in response
1646 to another message, thes have all been changed to:
1648 Auto-Submitted: auto-replied
1650 in accordance with these statements in the RFC:
1652 The auto-replied keyword:
1654 - SHOULD be used on messages sent in direct response to another
1655 message by an automatic process,
1657 - MUST NOT be used on manually-generated messages,
1659 - MAY be used on Delivery Status Notifications (DSNs) and Message
1660 Disposition Notifications (MDNs),
1662 - MUST NOT be used on messages generated by automatic or periodic
1663 processes, except for messages which are automatic responses to
1666 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1667 to the default Received: header definition.
1669 PH/12 Added log selector acl_warn_skipped (default on).
1671 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1672 variables because (a) they are in the wrong storage pool and (b) even if
1673 they were copied, it wouldn't work properly because of the caching.
1675 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1676 checking when decoding. Apparently there are clients that generate
1677 overlong encoded strings. Why am I not surprised?
1679 PH/15 If the first argument of "${if match_address" was not empty, but did not
1680 contain an "@" character, Exim crashed. Now it writes a panic log message
1681 and treats the condition as false.
1683 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1685 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1686 extension "envelope-auth". The code is finished and in agreement with
1687 other implementations, but there is no documentation so far and in fact,
1688 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1689 not changing the active code.
1691 Print executed "if" and "elsif" statements when debugging is used. This
1692 helps a great deal to understand what a filter does.
1694 Document more things not specified clearly in RFC3028. I had all this
1695 sorted out, when out of a sudden new issues came to my mind. Oops."
1697 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1700 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1701 canonical form (as documented). However, after a host name lookup from
1702 the IP address, check_host() was doing a simple string comparison with
1703 addresses acquired from the DNS when checking that the found name did
1704 have the original IP as one of its addresses. Since any found IPv6
1705 addresses are likely to be in abbreviated form, the comparison could
1706 fail. Luckily, there already exists a function for doing the comparison
1707 by converting both addresses to binary, so now that is used instead of
1708 the text comparison.
1710 PH/20 There was another similar case to PH/19, when a complete host name was
1711 given in a host list; looking up its IP address could give an abbreviated
1712 form, whereas the current host's name might or might not be abbreviated.
1713 The same fix has been applied.
1719 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1720 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1723 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1724 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1726 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1728 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1729 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1730 Also, for "Auto-Submitted", treat anything other than "no" as
1731 non-personal, in accordance with RFC 3834. (Previously it treated
1732 anything starting "auto-" as non-personal.)
1734 TF/01 The control=submission/name=... option had a problem with syntax
1735 errors if the name included a slash character. The /name= option
1736 now slurps the rest of the string, so it can include any characters
1737 but it must come last in the list of options (after /sender_retain
1740 PH/05 Some modifications to the interface to the fake nameserver for the new
1748 TK/01 Added the "success_on_redirect" address verification option. See
1749 NewStuff for rationale and an example.
1751 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1753 PH/02 Patch to exigrep to allow it to work on syslog lines.
1755 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1756 fread() to skip over the body file's header line, because in Cygwin the
1757 header line is locked and is inaccessible.
1759 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1760 co-exist for some time) to make it clear that it is the Exim ID that is
1761 referenced, not the Message-ID: header line.
1763 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1764 string_format() function, because snprintf() does not exist on all
1767 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1768 ${stat: expansion item.
1770 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1771 protocol synchronization error", to keep the pedants happy.
1773 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1774 well as for IRIX systems, when gcc is being used. See the host.c source
1777 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1779 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1782 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1783 transport and to the smtp transport in LMTP mode.
1785 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1787 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1788 run a filter in a subprocess. This could lead to confusion in subsequent
1789 lookups in the parent process. There should also be a search_tidyup() at
1790 the end of the subprocess.
1792 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1793 only if the host matched helo_try_verify_hosts, which caused the
1794 verification to occur when the EHLO/HELO command was issued. The ACL just
1795 tested the remembered result. Now, if a previous verification attempt has
1796 not happened, "verify = helo" does it there and then.
1798 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1800 TK/03 Fix log output including CR from clamd.
1802 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1803 exist provoked a memory error which could cause a segfault.
1805 PH/15 Installed PCRE 6.2
1807 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1809 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1810 of the problem. Specifically, suggested +O2 rather than +O1 for the
1813 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1815 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1816 record, the port was not correctly logged when the outgoing_port log
1817 selector was set (it logged the transort's default port).
1819 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1820 fallback_hosts, and "hosts" in the smtp transport.
1822 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1823 host errors such as "Connection refused".
1825 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1826 authentication with radiusclient 0.4.9:
1828 - Error returned from rc_read_config was caught wrongly
1829 - Username/password not passed on to radius server due to wrong length.
1831 The presumption is that some radiusclient API changes for 4.51/PH/17
1832 were not taken care of correctly. The code is still untested by me (my
1833 Linux distribution still has 0.3.2 of radiusclient), but it was
1834 contributed by a Radius user.
1836 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1837 expanding the "port" option of the smtp transport.
1839 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1840 while reading a MIME header. Thanks to Tom Hughes for a patch.
1842 PH/24 Include config.h inside local_scan.h so that configuration settings are
1845 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1846 that in an ACL for RCPT (for example), you can examine exactly what was
1849 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1850 commands, but it was not correctly comparing the address with the actual
1851 client host address. Thus, it would show the EHLO address in Received:
1852 header lines when this was not necessary.
1854 PH/27 Added the % operator to ${eval:}.
1856 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1857 it should be ignoring failures (because with -C, for example, it has lost
1858 privilege). It wasn't ignoring creation failures other than "already
1861 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1864 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1865 because an input line was too long, either on its own, or by virtue of
1866 too many continuations, the temporary file was not being removed, and the
1867 return code was incorrect.
1869 PH/31 Missing "BOOL" in function definition in filtertest.c.
1871 PH/32 Applied Sieve patches from the maintainer.
1873 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1875 PH/33 Added "verify = not_blind".
1877 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1878 Local/Makefile (with some defaults set). These are used in built scripts
1879 such as exicyclog, but they have never been used in the exim_install
1880 script (though there are many overriding facilities there). I have
1881 arranged that the exim_install script now takes note of these two
1884 PH/35 Installed configuration files for Dragonfly.
1886 PH/36 When a locally submitted message by a trusted user did not contain a
1887 From: header, and the sender address was obtained from -f or from an SMTP
1888 MAIL command, and the trusted user did not use -F to supply a sender
1889 name, $originator_name was incorrectly used when constructing a From:
1890 header. Furthermore, $originator_name was used for submission mode
1891 messages from external hosts without From: headers in a similar way,
1892 which is clearly wrong.
1894 PH/37 Added control=suppress_local_fixups.
1896 PH/38 When log_selector = +received_sender was set, and the addition of the
1897 sender made the log line's construction buffer exactly full, or one byte
1898 less than full, an overflow happened when the terminating "\n" was
1901 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1902 when the result of a list match is failure because a DNS lookup failed.
1904 PH/40 RM_COMMAND is now used in the building process.
1906 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1907 the "build-* directories that it finds.
1909 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1910 address in a domain literal was a prefix of an interface address.
1912 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1913 when verifying a sender address, unless rewrite_headers is false.
1915 PH/44 Wrote a long comment about why errors_to addresses are verified as
1916 recipients, not senders.
1918 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1919 the ratelimit ACL was added.
1921 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1923 PH/46 Added extra information about PostgreSQL errors to the error string.
1925 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1926 avoiding the need to install special zones in a real server. This is
1927 backwards compatible; if it can't find the fake resolver, it drops back.
1928 Thus, both old and new test suites can be run.
1930 TF/02 Added util/ratelimit.pl
1932 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1933 clock is set back in time.
1935 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1936 Candler <B.Candler@pobox.com>.
1938 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1939 OK for addresses that are the result of redirections.
1941 PH/48 A number of further additions for the benefit of the new test suite,
1942 including a fake gethostbyname() that interfaces to the fake DNS resolver
1945 TF/06 The fix for widen_domains has also been applied to qualify_single and
1946 search_parents which are the other dnslookup options that can cause
1949 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1952 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1954 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1955 service" if it encountered an MX record with an empty target hostname.
1956 The message is now "an MX or SRV record indicated no SMTP service".
1958 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1959 if the DNS of the sending site is misconfigured. This is quite a
1960 common situation. This change restores the behaviour of treating a
1961 helo verification defer as a failure.
1963 PH/51 If self=fail was set on a router, the bounce message did not include the
1964 actual error message.
1970 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1972 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1973 command itself ended in error, the underlying message about the transport
1974 filter timeout was being overwritten with the pipe command error. Now the
1975 underlying error message should be appended to the second error message.
1977 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1979 PH/02 Reduce the amount of output that "make" produces by default. Full output
1980 can still be requested.
1982 PH/03 The warning log line about a condition test deferring for a "warn" verb
1983 was being output only once per connection, rather than after each
1984 occurrence (because it was using the same function as for successful
1985 "warn" verbs). This seems wrong, so I have changed it.
1987 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1988 it should not have, which might have caused a crash in the right
1989 circumstances, but probably never did.
1991 PH/04 Installed a modified version of Tony Finch's patch to make submission
1992 mode fix the return path as well as the Sender: header line, and to
1993 add a /name= option so that you can make the user's friendly name appear
1996 TF/03 Added the control = fakedefer ACL modifier.
1998 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1999 Mark Lowes for thorough testing.
2001 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2003 TK/03 Merged latest SRS patch from Miles Wilton.
2005 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2006 with the definition in sysexits.h (which is #included earlier).
2007 Fortunately, Exim does not actually use EX_OK. The code used to try to
2008 preserve the sysexits.h value, by assumimg that macro definitions were
2009 scanned for macro replacements. I have been disabused of this notion,
2010 so now the code just undefines EX_OK before #including unistd.h.
2012 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2013 in the smtp transport. When a block could not be written in a single
2014 write() function, the timeout was being re-applied to each part-write.
2015 This seems wrong - if the receiver was accepting one byte at a time it
2016 would take for ever. The timeout is now adjusted when this happens. It
2017 doesn't have to be particularly precise.
2019 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2020 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2022 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2023 without a domain if the check to <postmaster@domain> fails.
2025 SC/01 Eximstats: added -xls and the ability to specify output files
2026 (patch written by Frank Heydlauf).
2028 SC/02 Eximstats: use FileHandles for outputing results.
2030 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2032 SC/04 Eximstats: fixed display of large numbers with -nvr option
2034 SC/05 Eximstats: fixed merging of reports with empty tables.
2036 SC/06 Eximstats: added the -include_original_destination flag
2038 SC/07 Eximstats: removed tabs and trailing whitespace.
2040 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2042 TK/06 MBOX spool code: Add real "From " MBOX separator line
2043 so the .eml file is really in mbox format (even though
2044 most programs do not really care). Patch from Alex Miller.
2046 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2047 The latter is generated from $received_to and is only set if the
2048 message has one envelope recipient. SA can use these headers,
2049 obviously out-of-the-box. Patch from Alex Miller.
2051 PH/08 The ${def test on a variable was returning false if the variable's
2052 value was "0", contrary to what the specification has always said!
2053 The result should be true unless the variable is empty.
2055 PH/09 The syntax error of a character other than { following "${if
2056 def:variable_name" (after optional whitespace) was not being diagnosed.
2057 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2058 accidental colon was present, for example, could give incorrect results.
2060 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2061 result is used (not including appendfile, where other changes are about
2064 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2065 This involved changing a lot of size variables from int to off_t. It
2066 should work with maildirs and everything.
2068 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2069 spamd dying while we are connected to it.
2071 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2074 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2075 function for BDB 4.3.
2077 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2079 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2080 the sources for PCRE have been re-arranged and more files are now
2083 PH/15 The code I had for printing potentially long long variables in PH/11
2084 above was not the best (it lost precision). The length of off_t variables
2085 is now inspected at build time, and an appropriate printing format (%ld
2086 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2087 to be "long long int" or "long int". This is needed for the internal
2088 formatting function string_vformat().
2090 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2091 the configuration file to be ":syslog", then the script "guesses" where
2092 the logs files are, rather than using the compiled in default. In our
2093 case the guess is not the same as the compiled default, so the script
2094 suddenly stopped working when I started to use syslog. The patch checks
2095 to see if log_file_path is "". If so, it attempts to read it from exim
2096 with no configuration file to get the compiled in version, before it
2097 falls back to the previous guessing code."
2099 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2100 implementing BATV in an Exim configuration. See NewStuff for the gory
2103 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2104 Makefile that are specific to HP-UX.
2106 PH/18 If the "use_postmaster" option was set for a recipient callout together
2107 with the "random" option, the postmaster address was used as the MAIL
2108 FROM address for the random test, but not for the subsequent recipient
2109 test. It is now used for both.
2111 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2112 patch removes a few documentation additions to RFC 3028, because the
2113 latest draft now contains them. It adds the new en;ascii-case comparator
2114 and a new error check for 8bit text in MIME parts. Comparator and
2115 require names are now matched exactly. I enabled the subaddress
2116 extension, but it is not well tested yet (read: it works for me)."
2118 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2119 rework some of the code of TK/09 above to avoid the hardwired use of
2120 "%lld" and "long long". Replaced the call to snprintf() with a call to
2123 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2124 records point to non-existent hosts", "retry timeout exceeded", and
2125 "retry time not reached for any host after a long failure period".
2127 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2128 experimental DomainKeys support:
2130 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2131 (2) On an error such as an illegally used "control", the wrong name for
2132 the control was given.
2134 These problems did NOT occur unless DomainKeys support was compiled.
2136 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2138 PH/24 Added ${if match_ip condition.
2140 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2141 databases so that it will be absolutely obvious if a crash occurs in the
2142 DB library. This is a regular occurrence (often caused by mis-matched
2145 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2146 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2147 picked up on a user's system that detects such things. There doesn't seem
2148 to be a gcc warning option for this - only an attribute that has to be
2149 put on the function's prototype. It seems that in Fedora Core 4 they have
2150 set this on a number of new functions. No doubt there will be more in due
2153 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2154 specify a transport. However, if an address that was verified by such a
2155 router was the subject of a callout, Exim crashed because it tried to
2156 read the rcpt_include_affixes from the non-existent transport. Now it
2157 just assumes that the setting of that option is false. This bug was
2158 introduced by 4.51/PH/31.
2160 PH/28 Changed -d+all to exclude +memory, because that information is very
2161 rarely of interest, but it makes the output a lot bigger. People tend to
2162 do -d+all out of habit.
2164 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2165 code in os-type was giving problems when libc.so lives in lib64, like on
2168 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2169 aren't the modern standard, and it seems that some systems' include files
2170 don't always have them. Exim was already checking for some of the newer
2171 ones like T_AAAA, and defining it itself. I've added checks for all the
2172 record types that Exim uses.
2174 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2175 not automatically generating a new one, as it is supposed to. This
2176 prevented TLS from working. If the file did exist, but contained invalid
2177 data, a new version was generated, as expected. It was only the case of a
2178 non-existent file that was broken.
2180 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2181 with a change in libdomainkeys > 0.64.
2183 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2184 from DNS. If the selector record carries the flag, it now has
2185 precedence over the domain-wide flag.
2187 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2189 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2190 the use of a port name that isn't defined in /etc/services) occurred, the
2191 message was deferred as in a normal delivery, and thus remained on the
2192 spool, instead of being failed because of the mua_wrapper setting. This
2193 is now fixed, and I tidied up some of the mua_wrapper messages at the
2196 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2197 the messages in a hash of arrays rather than using individual hashes.
2198 This is a bit cleaner and results in dramatic memory savings, albeit
2199 at a slight CPU cost.
2201 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2202 as requested by Marc Sherman.
2204 SC/10 Eximstats: added histograms for user specified patterns as requested
2207 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2209 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2210 fopen() in the content-scanning modules that did not already have it.
2216 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2217 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2219 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2221 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2223 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2224 to test Sieve filters that use "vacation".
2226 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2227 that changes the way the GnuTLS parameters are stored in the cache file.
2228 The new format can be generated externally. For backward compatibility,
2229 if the data in the cache doesn't make sense, Exim assumes it has read an
2230 old-format file, and it generates new data and writes a new file. This
2231 means that you can't go back to an older release without removing the
2234 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2235 work if there are any delivery delays because "one_time" forces the
2236 parent to be marked "delivered", so its unseen clone is never tried
2237 again. For this reason, Exim now forbids the simultaneous setting of
2240 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2241 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2242 entry if you want to know the details. The fix, however, neglected to
2243 consider the case when local delivery batching is involved. The test for
2244 "previously delivered" was not happening when checking to see if an
2245 address could be batched with a previous (undelivered) one; under
2246 certain circumstances this could lead to multiple deliveries to the same
2249 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2250 in its include files, and this causes problems building Exim.
2252 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2253 header_syntax) but Exim was just ignoring anything given after a slash.
2254 In particular, this caused confusion with an attempt to use "verify =
2255 reverse_host_lookup/defer_ok". An error is now given when options are
2256 supplied for verify items that do not have them. (Maybe reverse_host_
2257 lookup should have a defer_ok option, but that's a different point.)
2259 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2260 defined by RFC 821) to 2048, because there were problems with some AUTH
2261 commands, and RFC 1869 says the size should be increased for extended
2262 SMTP commands that take arguments.
2264 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2267 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2268 "unknown" error; now it says that the functionality isn't in the binary.
2270 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2271 an address' error message when a string expansion fails (syntax or
2272 whatever). Otherwise the password may appear in the log. Following change
2273 PH/42 below, there is no longer a chance of it appearing in a bounce
2276 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2278 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2279 of its addresses was ever tried. (Bugzilla bug #2.)
2281 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2282 the result incorrectly in the debug output. (It correctly added a newline
2283 to what was transported.)
2285 TF/01 Added $received_time.
2287 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2288 commented out examples of how to interface to a virus scanner and to
2289 SpamAssassin. Also added commented examples of av_scanner and
2290 spamd_address settings.
2292 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2293 and controls are allowed in which ACLs. There were a couple of minor
2294 errors. Some of the entries in the conditions table (which is a table of
2295 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2296 negation of where the condition IS allowed.
2298 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2300 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2301 header file does not have a version number, so I've had to invent a new
2302 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2303 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2304 radiusclient), but it was contributed by a Radius user.
2306 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2307 files or directories, for OpenSSL.
2309 PH/19 When an Exim process that is running as root has to create an Exim log
2310 file, it does so in a subprocess that runs as exim:exim so as to get the
2311 ownership right at creation (otherwise, other Exim processes might see
2312 the file with the wrong ownership). There was no test for failure of this
2313 fork() call, which would lead to the process getting stuck as it waited
2314 for a non-existent subprocess. Forks do occasionally fail when resources
2315 run out. I reviewed all the other calls to fork(); they all seem to check
2318 PH/20 When checking for unexpected SMTP input at connect time (before writing
2319 the banner), Exim was not dealing correctly with a non-positive return
2320 from the read() function. If the client had disconnected by this time,
2321 the result was a log entry for a synchronization error with an empty
2322 string after "input=" when read() returned zero. If read() returned -1
2323 (an event I could not check), uninitialized data bytes were printed.
2324 There were reports of junk text (parts of files, etc) appearing after
2327 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2329 PH/22 Added support for macro redefinition, and (re)definition in between
2330 driver and ACL definitions.
2332 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2333 forgetting to use the resulting value; it was using the unexpanded value.
2335 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2336 hadn't been configured. The fix is from Juergen Kreileder, who
2337 understands it better than I do:
2339 "Here's what I see happening with three configured cyrus_sasl
2340 authenticators configured (plain, login, cram-md5):
2342 On startup auth_cyrus_sasl_init() gets called for each of these.
2343 This means three calls to sasl_listmech() without a specified mech_list.
2344 => SASL tests which mechs of all available mechs actually work
2345 => three warnings about OTP not working
2346 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2348 With the patch, sasl_listmech() also gets called three times. But now
2349 SASL's mech_list option is set to the server_mech specified in the the
2350 authenticator. Or in other words, the answer from sasl_listmech()
2351 gets limited to just the mech you're testing for (which is different
2353 => the return list contains just 'plain' or 'login', 'cram-md5' or
2354 nothing depending on the value of ob->server_mech.
2356 I've just tested the patch: Authentication still works fine,
2357 unavailable mechs specified in the exim configuration are still
2358 caught, and the auth.log warnings about OTP are gone."
2360 PH/25 When debugging is enabled, the contents of the command line are added
2361 to the debugging output, even when log_selector=+arguments is not
2364 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2365 answer is "GNU", and only if the return is "GNU/something" is the answer
2368 PH/27 $acl_verify_message is now set immediately after the failure of a
2369 verification in an ACL, and so is available in subsequent modifiers. In
2370 particular, the message can be preserved by coding like this:
2372 warn !verify = sender
2373 set acl_m0 = $acl_verify_message
2375 Previously, $acl_verify_message was set only while expanding "message"
2376 and "log_message" when a very denied access.
2378 PH/28 Modified OS/os.c-Linux with
2380 -#ifndef OS_LOAD_AVERAGE
2381 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2383 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2384 nomenclature these days.)
2386 PH/29 Installed patch from the Sieve maintainer that adds the options
2387 sieve_useraddress and sieve_subaddress to the redirect router.
2389 PH/30 In these circumstances:
2390 . Two addresses routed to the same list of hosts;
2391 . First host does not offer TLS;
2392 . First host accepts first address;
2393 . First host gives temporary error to second address;
2394 . Second host offers TLS and a TLS session is established;
2395 . Second host accepts second address.
2396 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2397 and peerdn, if requested) that were in fact used only for the second
2400 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2401 attention to any local part prefix or suffix that was matched by the
2402 router that accepted the address. It now behaves in the same way as it
2403 does for delivery: the affixes are removed from the local part unless
2404 rcpt_include_affixes is set on the transport.
2406 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2407 timeout during the DATA phase of an incoming message.
2409 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2410 applied a patch sanctioned by the Sieve maintainer.
2412 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2413 the uid or gid is negative. A case of a negative gid caused this to be
2414 noticed. The fix allows for either to be negative.
2416 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2417 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2420 PH/36 The change PH/12 above was broken. Fixed it.
2422 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2423 the grounds that routing the same address twice would always produce the
2424 same answer. This might have been true once, but it is certainly no
2425 longer true now. Routing a child address may depend on the previous
2426 routing that produced that child. Some complicated redirection strategies
2427 went wrong when messages had multiple recipients, and made Exim's
2428 behaviour dependent on the order in which the addresses were given.
2430 I have moved the duplicate checking until after the routing is complete.
2431 Exim scans the addresses that are assigned to local and remote
2432 transports, and removes any duplicates. This means that more work will be
2433 done, as duplicates will always all be routed, but duplicates are
2434 presumably rare, so I don't expect this is of any significance.
2436 For deliveries to pipes, files, and autoreplies, the duplicate checking
2437 still happens during the routing process, since they are not going to be
2440 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2441 It corrects a timeout issue with spamd. This is Ian's comment: "The
2442 background is that sometimes spamd either never reads data from a
2443 connection it has accepted, or it never writes response data. The exiscan
2444 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2445 blindly assumes that writes won't block so it may never time out."
2447 PH/39 Allow G after quota size as well as K and M.
2449 PH/40 The value set for $authenticated_id in an authenticator may not contain
2450 binary zeroes or newlines because the value is written to log lines and
2451 to spool files. There was no check on this. Now the value is run through
2452 the string_printing() function so that such characters are converted to
2453 printable escape sequences.
2455 PH/41 $message_linecount is a new variable that contains the total number of
2456 lines in the message. Compare $body_linecount, which is the count for the
2459 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2460 bounce and delay warning messages, except in certain special cases, which
2463 (a) An SMTP error message from a remote host;
2464 (b) A message specified in a :fail: redirection;
2465 (c) A message specified in a "fail" command in a system filter;
2466 (d) A message specified in a FAIL return from the queryprogram router;
2467 (e) A message specified by the cannot_route_message router option.
2469 In these cases only, Exim does include the error details in bounce and
2470 warning messages. There are also a few cases where bland messages such
2471 as "unrouteable address" or "local delivery error" are given.
2473 PH/43 $value is now also set for the "else" part of a ${run expansion.
2475 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2476 being worked on, but at least Exim now implements the latest version to
2479 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2480 process to complete was treated as a delivery failure, a timeout while
2481 writing the message to the pipe was logged, but erroneously treated as a
2482 successful delivery. Such timeouts include transport filter timeouts. For
2483 consistency with the overall process timeout, these timeouts are now
2484 treated as errors, giving rise to delivery failures by default. However,
2485 there is now a new Boolean option for the pipe transport called
2486 timeout_defer, which, if set TRUE, converts the failures into defers for
2487 both kinds of timeout. A transport filter timeout is now identified in
2490 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2491 systems where "make" and "gmake" are different, calling "gmake" at top
2492 level broke things. I've arranged for the value of $(MAKE) to be passed
2493 from the Makefile to this script so that it can call the same version of
2497 A note about Exim versions 4.44 and 4.50
2498 ----------------------------------------
2500 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2501 changes of various kinds. As a consequence, a big documentation update was
2502 needed. This delayed the release for rather longer than seemed good, especially
2503 in the light of a couple of (minor) security issues. Therefore, the changes
2504 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2505 release. So 4.44 and 4.50 are in effect two different branches that both start
2508 I have left the 4.50 change log unchanged; it contains all the changes since
2509 4.43. The change log for 4.44 is below; many of its items are identical to
2510 those for 4.50. This seems to be the most sensible way to preserve the
2511 historical information.
2517 1. Minor wording change to the doc/README.SIEVE file.
2519 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2520 computation of the current number of files was incorrect.
2522 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2523 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2526 4. Give more explanation in the error message when the command for a transport
2527 filter fails to execute.
2529 5. There are several places where Exim runs a non-Exim command in a
2530 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2531 was being done only for the command run by the queryprogram router. It is
2532 now done for all such subprocesses. The other cases are: ${run, transport
2533 filters, and the commands run by the lmtp and pipe transports.
2535 6. Added CONFIGURE_GROUP build-time option.
2537 7. Some older OS have a limit of 256 on the maximum number of file
2538 descriptors. Exim was using setrlimit() to set 1000 as a large value
2539 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2540 systems. I've change it so that if it can't get 1000, it tries for 256.
2542 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2543 was an oversight, and furthermore, ever since the addition of extra
2544 controls (e.g. 4.43/32), the checks on when to allow different forms of
2545 "control" were broken. There should now be diagnostics for all cases when a
2546 control that does not make sense is encountered.
2548 9. Added the /retain_sender option to "control=submission".
2550 10. $recipients is now available in the predata ACL (oversight).
2552 11. Tidy the search cache before the fork to do a delivery from a message
2553 received from the command line. Otherwise the child will trigger a lookup
2554 failure and thereby defer the delivery if it tries to use (for example) a
2555 cached ldap connection that the parent has called unbind on.
2557 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2558 of $address_data from the recipient verification was clobbered by the
2559 sender verification.
2561 13. The value of address_data from a sender verification is now available in
2562 $sender_address_data in subsequent conditions in the ACL statement.
2564 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2566 15. Added a new option "connect=<time>" to callout options, to set a different
2569 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2570 was its contents. (It was OK if the option was not defined at all.)
2572 17. A "Completed" log line is now written for messages that are removed from
2573 the spool by the -Mrm option.
2575 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2576 information about exactly what failed.
2578 19. Added -dd to debug only the daemon process.
2580 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2581 handles timeouts, both on the server side and network timeouts. Renamed the
2582 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2584 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2585 It is now set to "smtps".
2587 22. $host_address is now set to the target address during the checking of
2588 ignore_target_hosts.
2590 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2591 being passed; this would have caused $sender_host_name to have been used if
2592 matching the list had actually called for a host name (not very likely,
2593 since this list is usually IP addresses). A host name is now passed as
2596 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2597 code that specifies a non-restarting handler (typically sigaction() in
2598 modern systems) in an attempt to fix a rare and obscure crash bug.
2600 25. Narrowed the window for a race in the daemon that could cause it to ignore
2601 SIGCHLD signals. This is not a major problem, because they are used only to
2602 wake it up if nothing else does.
2604 26. A malformed maildirsize file could cause Exim to calculate negative values
2605 for the mailbox size or file count. Odd effects could occur as a result.
2606 The maildirsize information is now recalculated if the size or filecount
2609 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2610 support for a long time. Removed HAVE_SYS_VFS_H.
2612 28. Installed the latest version of exipick from John Jetmore.
2614 29. In an address list, if the pattern was not a regular expression, an empty
2615 subject address (from a bounce message) matched only if the pattern was an
2616 empty string. Non-empty patterns were not even tested. This was the wrong
2617 because it is perfectly reasonable to use an empty address as part of a
2618 database query. An empty address is now tested by patterns that are
2619 lookups. However, all the other forms of pattern expect the subject to
2620 contain a local part and a domain, and therefore, for them, an empty
2621 address still always fails if the pattern is not itself empty.
2623 30. Exim went into a mad DNS loop when attempting to do a callout where the
2624 host was specified on an smtp transport, and looking it up yielded more
2625 than one IP address.
2627 31. Re-factored the code for checking spool and log partition space into a
2628 function that finds that data and another that does the check. The former
2629 is then used to implement four new variables: $spool_space, $log_space,
2630 $spool_inodes, and $log_inodes.
2632 32. The RFC2047 encoding function was originally intended for short strings
2633 such as real names; it was not keeping to the 75-character limit for
2634 encoded words that the RFC imposes. It now respects the limit, and
2635 generates multiple encoded words if necessary. To be on the safe side, I
2636 have increased the buffer size for the ${rfc2047: expansion operator from
2639 33. It is now permitted to omit both strings after an "if" condition; if the
2640 condition is true, the result is "true". As before, when the second string
2641 is omitted, a false condition yields an empty string. This makes it less
2642 cumbersome to write custom ACL and router conditions.
2644 34. Failure to deliver a bounce message always caused it to be frozen, even if
2645 there was an errors_to setting on the router. The errors_to setting is now
2648 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2649 canonical form (fully expanded) before being placed in
2650 $sender_host_address.
2652 36. The table in the code that translates DNS record types into text (T_A to
2653 "A" for instance) was missing entries for NS and CNAME. It is just possible
2654 that this could have caused confusion if both these types were looked up
2655 for the same domain, because the text type is used as part of Exim's
2656 per-process caching. But the chance of anyone hitting this buglet seems
2659 37. The dnsdb lookup has been extended in a number of ways.
2661 (1) There is a new type, "zns", which walks up the domain tree until it
2662 finds some nameserver records. It should be used with care.
2664 (2) There is a new type, "mxh", which is like "mx" except that it returns
2665 just the host names, not the priorities.
2667 (3) It is now possible to give a list of domains (or IP addresses) to be
2668 looked up. The behaviour when one of the lookups defers can be
2669 controlled by a keyword.
2671 (4) It is now possible to specify the separator character for use when
2672 multiple records are returned.
2674 38. The dnslists ACL condition has been extended: it is now possible to supply
2675 a list of IP addresses and/or domains to be looked up in a particular DNS
2678 39. Added log_selector=+queue_time_overall.
2680 40. When running the queue in the test harness, wait just a tad after forking a
2681 delivery process, to get repeatability of debugging output.
2683 41. Include certificate and key file names in error message when GnuTLS fails
2684 to set them up, because the GnuTLS error message doesn't include the name
2685 of the failing file when there is a problem reading it.
2687 42. Allow both -bf and -bF in the same test run.
2689 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2691 44. The "Exiscan patch" is now merged into the mainline Exim source.
2693 45. Sometimes the final signoff response after QUIT could fail to get
2694 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2695 < 0 before doing a fflush(). This bug looks as though it goes back to the
2696 introduction of TLS in release 3.20, but "sometimes" must have been rare
2697 because the tests only now provoked it.
2699 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2700 (this can affect the format of dates).
2702 47. exim_tidydb, when checking for the continued existence of a message for
2703 which it has found a message-specific retry record, was not finding
2704 messages that were in split spool directories. Consequently, it was
2705 deleting retry records that should have stayed in existence.
2707 48. Steve fixed some bugs in eximstats.
2709 49. The SPA authentication driver was not abandoning authentication and moving
2710 on to the next authenticator when an expansion was forced to fail,
2711 contradicting the general specification for all authenticators. Instead it
2712 was generating a temporary error. It now behaves as specified.
2714 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2715 (the order specifies the preference for clients). The order is now AES256,
2716 AES128, 3DES, ARCFOUR128.
2718 51. Small patch to Sieve code - explicitly set From: when generating an
2721 52. Exim crashed if a remote delivery caused a very long error message to be
2722 recorded - for instance if somebody sent an entire SpamAssassin report back
2723 as a large number of 550 error lines. This bug was coincidentally fixed by
2724 increasing the size of one of Exim's internal buffers (big_buffer) that
2725 happened as part of the Exiscan merge. However, to be on the safe side, I
2726 have made the code more robust (and fixed the comments that describe what
2729 53. Now that there can be additional text after "Completed" in log lines (if
2730 the queue_time_overall log selector is set), a one-byte patch to exigrep
2731 was needed to allow it to recognize "Completed" as not the last thing in
2734 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2735 patch that reportedly fixes this has been added. I am not expert enough to
2736 create a test for it. This is what the patch creator wrote:
2738 "I found a little strange behaviour of ldap code when working with
2739 Windows 2003 AD Domain, where users was placed in more than one
2740 Organization Units. When I tried to give exim partial DN, the exit code
2741 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2742 But simultaneously result of request was absolutely normal ldap result,
2743 so I produce this patch..."
2745 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2746 REFERENCE, so I have modified the code to exclude the patch when that macro
2749 55. Some experimental protocols are using DNS PTR records for new purposes. The
2750 keys for these records are domain names, not reversed IP addresses. The
2751 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2752 leaves it alone. Component reversal etc. now happens only for IP addresses.
2755 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2757 57. Double the size of the debug message buffer (to 2048) so that more of very
2758 long debug lines gets shown.
2760 58. The exicyclog utility now does better if the number of log files to keep
2761 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2763 59. Two changes related to the smtp_active_hostname option:
2765 (1) $smtp_active_hostname is now available as a variable.
2766 (2) The default for smtp_banner uses $smtp_active_hostname instead
2767 of $primary_hostname.
2769 60. The host_aton() function is supposed to be passed a string that is known
2770 to be a valid IP address. However, in the case of IPv6 addresses, it was
2771 not checking this. This is a hostage to fortune. Exim now panics and dies
2772 if the condition is not met. A case was found where this could be provoked
2773 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2774 components; fortuitously, this particular loophole had already been fixed
2775 by change 4.50/55 above.
2777 If there are any other similar loopholes, the new check in host_aton()
2778 itself should stop them being exploited. The report I received stated that
2779 data on the command line could provoke the exploit when Exim was running as
2780 exim, but did not say which command line option was involved. All I could
2781 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2782 running as the user.
2785 61. There was a buffer overflow vulnerability in the SPA authentication code
2786 (which came originally from the Samba project). I have added a test to the
2787 spa_base64_to_bits() function which I hope fixes it.
2790 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2791 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2793 63. The daemon start-up calls getloadavg() while still root for those OS that
2794 need the first call to be done as root, but it missed one case: when
2795 deliver_queue_load_max is set with deliver_drop_privilege. This is
2796 necessary for the benefit of the queue runner, because there is no re-exec
2797 when deliver_drop_privilege is set.
2799 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2800 This has been fixed.
2802 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2803 was in use, was not putting the data itself into the right store pool;
2804 consequently, it could be overwritten for a subsequent message in the same
2805 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2808 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2810 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2811 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2812 the function were treating the return as a boolean value, which happened to
2813 work because 0=false and not-0=true, but is not correct code.
2815 68. The host_aton() function was not handling scoped IPv6 addresses (those
2816 with, for example, "%eth0" on the end) correctly.
2818 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2819 negated items (that is, ~something) in unsigned ints. Some compilers
2820 apparently mutter when there is no cast.
2822 70. If an address verification called from an ACL failed, and did not produce a
2823 user-specific message (i.e. there was only a "system" message), nothing was
2824 put in $acl_verify_message. In this situation, it now puts the system
2827 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2828 session; change 4.31/43 added the unwanted input to the log line - except
2829 that it did not do this in the start of session case. It now does.
2831 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2832 This is wrong and can cause the other end to generate a synchronization
2833 error if it is another Exim or anything else that does the synchronization
2834 check. A QUIT command is no longer sent after a timeout.
2836 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2837 during host lookups.
2839 74. The defer_ok option of callout verification was not working if it was used
2840 when verifying addresses in header lines, that is, for this case:
2842 verify = header_sender/callout=defer_ok
2844 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2845 those file descriptors could be used for SMTP connections. If anything
2846 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2847 could be sent to the SMTP client, causing chaos. The daemon now opens
2848 stdin, stdout, and stderr to /dev/null when it puts itself into the
2851 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2852 log by default. The user can override this with suitable Perl magic.
2854 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2855 the log message when discard triggers, was not working for the DATA ACL or
2856 for the non-SMTP ACL.
2858 78. Error message wording change in sieve.c.
2860 79. If smtp_accept_max_per_host was set, the number of connections could be
2861 restricted to fewer than expected, because the daemon was trying to set up
2862 a new connection before checking whether the processes handling previous
2863 connections had finished. The check for completed processes is now done
2864 earlier. On busy systems, this bug wouldn't be noticed because something
2865 else would have woken the daemon, and it would have reaped the completed
2868 80. If a message was submitted locally by a user whose login name contained one
2869 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2870 It caused a spool format error. I have fixed the spool reading code. A
2871 related problem was that the "from" clause in the Received: line became
2872 illegal because of the space(s). It is now covered by ${quote_local_part.
2874 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2877 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2879 83. Patch from Sieve maintainer to fix unterminated string problem in
2880 "vacation" handling.
2882 84. Some minor changes to the Linux configuration files to help with other
2883 OS variants using glibc.
2885 85. One more patch for Sieve to update vacation handling to latest spec.
2888 ----------------------------------------------------
2889 See the note above about the 4.44 and 4.50 releases.
2890 ----------------------------------------------------
2896 1. Change 4.43/35 introduced a bug that caused file counts to be
2897 incorrectly computed when quota_filecount was set in an appendfile
2900 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2901 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2904 3. Give more explanation in the error message when the command for a transport
2905 filter fails to execute.
2907 4. There are several places where Exim runs a non-Exim command in a
2908 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2909 was being done only for the command run by the queryprogram router. It is
2910 now done for all such subprocesses. The other cases are: ${run, transport
2911 filters, and the commands run by the lmtp and pipe transports.
2913 5. Some older OS have a limit of 256 on the maximum number of file
2914 descriptors. Exim was using setrlimit() to set 1000 as a large value
2915 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2916 systems. I've change it so that if it can't get 1000, it tries for 256.
2918 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2919 was an oversight, and furthermore, ever since the addition of extra
2920 controls (e.g. 4.43/32), the checks on when to allow different forms of
2921 "control" were broken. There should now be diagnostics for all cases when a
2922 control that does not make sense is encountered.
2924 7. $recipients is now available in the predata ACL (oversight).
2926 8. Tidy the search cache before the fork to do a delivery from a message
2927 received from the command line. Otherwise the child will trigger a lookup
2928 failure and thereby defer the delivery if it tries to use (for example) a
2929 cached ldap connection that the parent has called unbind on.
2931 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2932 of $address_data from the recipient verification was clobbered by the
2933 sender verification.
2935 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2936 was its contents. (It was OK if the option was not defined at all.)
2938 11. A "Completed" log line is now written for messages that are removed from
2939 the spool by the -Mrm option.
2941 12. $host_address is now set to the target address during the checking of
2942 ignore_target_hosts.
2944 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2945 being passed; this would have caused $sender_host_name to have been used if
2946 matching the list had actually called for a host name (not very likely,
2947 since this list is usually IP addresses). A host name is now passed as
2950 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2951 code that specifies a non-restarting handler (typically sigaction() in
2952 modern systems) in an attempt to fix a rare and obscure crash bug.
2954 15. Narrowed the window for a race in the daemon that could cause it to ignore
2955 SIGCHLD signals. This is not a major problem, because they are used only to
2956 wake it up if nothing else does.
2958 16. A malformed maildirsize file could cause Exim to calculate negative values
2959 for the mailbox size or file count. Odd effects could occur as a result.
2960 The maildirsize information is now recalculated if the size or filecount
2963 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2964 support for a long time. Removed HAVE_SYS_VFS_H.
2966 18. Updated exipick to current release from John Jetmore.
2968 19. Allow an empty sender to be matched against a lookup in an address list.
2969 Previously the only cases considered were a regular expression, or an
2972 20. Exim went into a mad DNS lookup loop when doing a callout where the
2973 host was specified on the transport, if the DNS lookup yielded more than
2976 21. The RFC2047 encoding function was originally intended for short strings
2977 such as real names; it was not keeping to the 75-character limit for
2978 encoded words that the RFC imposes. It now respects the limit, and
2979 generates multiple encoded words if necessary. To be on the safe side, I
2980 have increased the buffer size for the ${rfc2047: expansion operator from
2983 22. Failure to deliver a bounce message always caused it to be frozen, even if
2984 there was an errors_to setting on the router. The errors_to setting is now
2987 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2988 canonical form (fully expanded) before being placed in
2989 $sender_host_address.
2991 24. Updated eximstats to version 1.33
2993 25. Include certificate and key file names in error message when GnuTLS fails
2994 to set them up, because the GnuTLS error message doesn't include the name
2995 of the failing file when there is a problem reading it.
2997 26. Expand error message when OpenSSL has problems setting up cert/key files.
3000 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3001 (this can affect the format of dates).
3003 28. exim_tidydb, when checking for the continued existence of a message for
3004 which it has found a message-specific retry record, was not finding
3005 messages that were in split spool directories. Consequently, it was
3006 deleting retry records that should have stayed in existence.
3008 29. eximstats updated to version 1.35
3009 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3010 1.35 - bugfix such that pie charts by volume are generated correctly
3012 30. The SPA authentication driver was not abandoning authentication and moving
3013 on to the next authenticator when an expansion was forced to fail,
3014 contradicting the general specification for all authenticators. Instead it
3015 was generating a temporary error. It now behaves as specified.
3017 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3018 (the order specifies the preference for clients). The order is now AES256,
3019 AES128, 3DES, ARCFOUR128.
3021 31. Small patch to Sieve code - explicitly set From: when generating an
3024 32. Exim crashed if a remote delivery caused a very long error message to be
3025 recorded - for instance if somebody sent an entire SpamAssassin report back
3026 as a large number of 550 error lines. This bug was coincidentally fixed by
3027 increasing the size of one of Exim's internal buffers (big_buffer) that
3028 happened as part of the Exiscan merge. However, to be on the safe side, I
3029 have made the code more robust (and fixed the comments that describe what
3032 33. Some experimental protocols are using DNS PTR records for new purposes. The
3033 keys for these records are domain names, not reversed IP addresses. The
3034 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3035 leaves it alone. Component reversal etc. now happens only for IP addresses.
3038 34. The host_aton() function is supposed to be passed a string that is known
3039 to be a valid IP address. However, in the case of IPv6 addresses, it was
3040 not checking this. This is a hostage to fortune. Exim now panics and dies
3041 if the condition is not met. A case was found where this could be provoked
3042 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3043 components; fortuitously, this particular loophole had already been fixed
3044 by change 4.50/55 or 4.44/33 above.
3046 If there are any other similar loopholes, the new check in host_aton()
3047 itself should stop them being exploited. The report I received stated that
3048 data on the command line could provoke the exploit when Exim was running as
3049 exim, but did not say which command line option was involved. All I could
3050 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3051 running as the user.
3054 35. There was a buffer overflow vulnerability in the SPA authentication code
3055 (which came originally from the Samba project). I have added a test to the
3056 spa_base64_to_bits() function which I hope fixes it.
3059 36. The daemon start-up calls getloadavg() while still root for those OS that
3060 need the first call to be done as root, but it missed one case: when
3061 deliver_queue_load_max is set with deliver_drop_privilege. This is
3062 necessary for the benefit of the queue runner, because there is no re-exec
3063 when deliver_drop_privilege is set.
3065 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3066 was in use, was not putting the data itself into the right store pool;
3067 consequently, it could be overwritten for a subsequent message in the same
3068 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3071 38. Sometimes the final signoff response after QUIT could fail to get
3072 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3073 < 0 before doing a fflush(). This bug looks as though it goes back to the
3074 introduction of TLS in release 3.20, but "sometimes" must have been rare
3075 because the tests only now provoked it.
3081 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3082 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3083 Now it returns an integer. A number of calls were still expecting a T/F
3084 return. Fortuitously, in all cases, the tests worked in OK situations,
3085 which is the norm. However, things would have gone wrong on any write
3086 failures on the smtp file descriptor. This function is used when sending
3087 messages over SMTP and also when doing verify callouts.
3089 2. When Exim is called to do synchronous delivery of a locally submitted
3090 message (the -odf or -odi options), it no longer closes stderr before doing
3093 3. Implemented the mua_wrapper option.
3095 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3097 5. Implemented the functions header_remove(), header_testname(),
3098 header_add_at_position(), and receive_remove_recipient(), and exported them
3101 6. If an ACL "warn" statement specified the addition of headers, Exim already
3102 inserted X-ACL-Warn: at the start if there was no header name. However, it
3103 was not making this test for the second and subsequent header lines if
3104 there were newlines in the string. This meant that an invalid header could
3105 be inserted if Exim was badly configured.
3107 7. Allow an ACL "warn" statement to add header lines at the start or after all
3108 the Received: headers, as well as at the end.
3110 8. Added the rcpt_4xx retry error code.
3112 9. Added postmaster_mailfrom=xxx to callout verification option.
3114 10. Added mailfrom=xxxx to the callout verification option, for verify=
3117 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3118 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3120 12. Inserted some casts to stop certain compilers warning when using pointer
3121 differences as field lengths or precisions in printf-type calls (mostly
3122 affecting debugging statements).
3124 13. Added optional readline() support for -be (dynamically loaded).
3126 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3127 same clock tick as a message's arrival, so that its received time was the
3128 same as the "first fail" time on the retry record, and that message
3129 remained on the queue past the ultimate address timeout, every queue runner
3130 would try a delivery (because it was past the ultimate address timeout) but
3131 after another failure, the ultimate address timeout, which should have then
3132 bounced the address, did not kick in. This was a "< instead of <=" error;
3133 in most cases the first failure would have been in the next clock tick
3134 after the received time, and all would be well.
3136 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3137 being recognized when the domain list was tested by the match_domain
3138 condition in an expansion string.
3140 16. Added the ${str2b64: operator.
3142 17. Exim was always calling setrlimit() to set a large limit for the number of
3143 processes, without checking whether the existing limit was already
3144 adequate. (It did check for the limit on file descriptors.) Furthermore,
3145 errors from getrlimit() and setrlimit() were being ignored. Now they are
3146 logged to the main and panic logs, but Exim does carry on, to try to do its
3147 job under whatever limits there are.
3149 18. Imported PCRE 5.0.
3151 19. Trivial typo in log message " temporarily refused connection" (the leading
3154 20. If the log selector return_path_on_delivery was set and an address was
3155 redirected to /dev/null, the delivery process crashed because it assumed
3156 that a return path would always be set for a "successful" delivery. In this
3157 case, the whole delivery is bypassed as an optimization, and therefore no
3160 21. Internal re-arrangement: the function for sending a challenge and reading
3161 a response while authentication was assuming a zero-terminated challenge
3162 string. It's now changed to take a pointer and a length, to allow for
3163 binary data in such strings.
3165 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3167 23. Exim was not respecting finduser_retries when seeking the login of the
3168 uid under which it was called; it was always trying 10 times. (The default
3169 setting of finduser_retries is zero.) Also, it was sleeping after the final
3170 failure, which is pointless.
3172 24. Implemented tls_on_connect_ports.
3174 25. Implemented acl_smtp_predata.
3176 26. If the domain in control=submission is set empty, Exim assumes that the
3177 authenticated id is a complete email address when it generates From: or
3178 Sender: header lines.
3180 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3181 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3182 chown and chgrp in /bin and hostname in /usr/bin.
3184 28. Exim was keeping the "process log" file open after each use, just as it
3185 does for the main log. This opens the possibility of it remaining open for
3186 long periods when the USR1 signal hits a daemon. Occasional processlog
3187 errors were reported, that could have been caused by this. Anyway, it seems
3188 much more sensible not to leave this file open at all, so that is what now
3191 29. The long-running daemon process does not normally write to the log once it
3192 has entered its main loop, and it closes the log before doing so. This is
3193 so that log files can straightforwardly be renamed and moved. However,
3194 there are a couple of unusual error situations where the daemon does write
3195 log entries, and I had neglected to close the log afterwards.
3197 30. The text of an SMTP error response that was received during a remote
3198 delivery was being truncated at 512 bytes. This is too short for some of
3199 the long messages that one sometimes sees. I've increased the limit to
3202 31. It is now possible to make retry rules that apply only when a message has a
3203 specific sender, in particular, an empty sender.
3205 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3206 it possible to be selective about when SMTP synchronization is enforced.
3208 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3210 32. Implemented hosts_connection_nolog.
3212 33. Added an ACL for QUIT.
3214 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3217 35. Added mailbox_size and mailbox_filecount to appendfile.
3219 36. Added control = no_multiline_responses to ACLs.
3221 37. There was a bug in the logic of the code that waits for the clock to tick
3222 in the case where the clock went backwards by a substantial amount such
3223 that the microsecond fraction of "now" was more than the microsecond
3224 fraction of "then" (but the whole seconds number was less).
3226 38. Added support for the libradius Radius client library this is found on
3227 FreeBSD (previously only the radiusclient library was supported).
3233 1. When certain lookups returned multiple values in the form name=value, the
3234 quoting of the values was not always being done properly. Specifically:
3235 (a) If the value started with a double quote, but contained no whitespace,
3237 (b) If the value contained whitespace other than a space character (i.e.
3238 tabs or newlines or carriage returns) it was not quoted.
3239 This fix has been applied to the mysql and pgsql lookups by writing a
3240 separate quoting function and calling it from the lookup code. The fix
3241 should probably also be applied to nisplus, ibase and oracle lookups, but
3242 since I cannot test any of those, I have not disturbed their existing code.
3244 2. A hit in the callout cache for a specific address caused a log line with no
3245 reason for rejecting RCPT. Now it says "Previous (cached) callout
3246 verification failure".
3248 3. There was an off-by-one bug in the queryprogram router. An over-long
3249 return line was truncated at 256 instead of 255 characters, thereby
3250 overflowing its buffer with the terminating zero. As well as fixing this, I
3251 have increased the buffer size to 1024 (and made a note to document this).
3253 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3254 when Exim is waiting for an SMTP response from a remote server, Exim
3255 restarts its select() call on the socket, thereby resetting its timeout.
3256 This is not a problem when such interrupts are rare. Somebody set up a cron
3257 job to run exiwhat every 2 minutes, which is less than the normal select()
3258 timeout (5 or 10 minutes). This meant that the select() timeout never
3259 kicked in because it was always reset. I have fixed this by comparing the
3260 time when an interrupt arrives with the time at the start of the first call
3261 to select(). If more time than the timeout has elapsed, the interrupt is
3262 treated as a timeout.
3264 5. Some internal re-factoring in preparation for the addition of Sieve
3265 extensions (by MH). In particular, the "personal" test is moved to a
3266 separate function, and given an option for scanning Cc: and Bcc: (which is
3267 not set for Exim filters).
3269 6. When Exim created an email address using the login of the caller as the
3270 local part (e.g. when creating a From: or Sender: header line), it was not
3271 quoting the local part when it contained special characters such as @.
3273 7. Installed new OpenBSD configuration files.
3275 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3276 try to make them clearer.
3278 9. Callout options, other than the timeout value, were being ignored when
3279 verifying sender addresses in header lines. For example, when using
3281 verify = header_sender/callout=no_cache
3283 the cache was (incorrectly) being used.
3285 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3286 only the Cygwin environment.
3288 11. When return_path_on_delivery was set as a log selector, if different remote
3289 addresses in the same message used different return paths and parallel
3290 remote delivery occurred, the wrong values would sometimes be logged.
3291 (Whenever a remote delivery process finished, the return path value from
3292 the most recently started remote delivery process was logged.)
3294 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3295 lines when AUTH and/or TLS are in use. This is the "received protocol"
3296 field. Exim used to use "asmtp" for authenticated SMTP, without any
3297 indication (in the protocol name) for TLS use. Now it follows the RFC and
3298 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3299 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3300 names appear in log lines as well as in Received: header lines.
3302 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3303 extensions, and comparison tests, and to fix some bugs.
3305 14. Changes to the "personal" filter test:
3307 (1) The test was buggy in that it was just doing the equivalent of
3308 "contains" tests on header lines. For example, if a user's address was
3309 anne@some.where, the "personal" test would incorrectly be true for
3311 To: susanne@some.where
3313 This test is now done by extracting each address from the header in turn,
3314 and checking the entire address. Other tests that are part of "personal"
3315 are now done using regular expressions (for example, to check local parts
3316 of addresses in From: header lines).
3318 (2) The list of non-personal local parts in From: addresses has been
3319 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3320 taken from the Sieve specification recommendations.
3322 (3) If the message contains any header line starting with "List-" it is
3323 treated as non-personal.
3325 (4) The test for "circular" in the Subject: header line has been removed
3326 because it now seems ill-conceived.
3328 15. Minor typos in src/EDITME comments corrected.
3330 16. Installed latest exipick from John Jetmore.
3332 17. If headers_add on a router specified a text string that was too long for
3333 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3334 of string_sprintf() is now avoided.
3336 18. $message_body_size was not set (it was always zero) when running the DATA
3337 ACL and the local_scan() function.
3339 19. For the "mail" command in an Exim filter, no default was being set for
3340 the once_repeat time, causing a random time value to be used if "once" was
3341 specified. (If the value happened to be <= 0, no repeat happened.) The
3342 default is now 0s, meaning "never repeat". The "vacation" command was OK
3343 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3344 (I found it when inspecting the code).
3346 20. There is now an overall timeout for performing a callout verification. It
3347 defaults to 4 times the callout timeout, which applies to individual SMTP
3348 commands during the callout. The overall timeout applies when there is more
3349 than one host that can be tried. The timeout is checked before trying the
3350 next host. This prevents very long delays if there are a large number of
3351 hosts and all are timing out (e.g. when the network connections are timing
3352 out). The value of the overall timeout can be changed by specifying an
3353 additional sub-option for "callout", called "maxwait". For example:
3355 verify = sender/callout=5s,maxwait=20s
3357 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3358 to the end before writing, but this should make it even safer).
3360 22. Exim was forgetting that it had advertised PIPELINING for the second and
3361 subsequent messages on an SMTP connection. It was also not resetting its
3362 memory on STARTTLS and an internal HELO.
3364 23. When Exim logs an SMTP synchronization error within a session, it now
3365 records whether PIPELINING has been advertised or not.
3367 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3368 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3369 rather than long int.
3371 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3373 26. Added the never_mail option to autoreply.
3379 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3380 crash if the getsockname() call failed; this can happen if a connection is
3381 closed very soon after it is established. The problem was simply in the
3382 order in which certain operations were done, causing Exim to try to write
3383 to the SMTP stream before it had set up the file descriptor. The bug has
3384 been fixed by making things happen in the correct order.
3390 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3391 before the connection was closed, thus losing the rejection response.
3393 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3394 some early Solaris releases, but causes trouble in current releases where
3395 socklen_t is defined.
3397 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3400 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3403 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3404 that releasing the top of it at the end releases what was used for sub-
3405 expansions (unless the block got too big). However, discard this block if
3406 the first thing is a variable or header, so that we can use its block when
3407 it is dynamic (useful for very large $message_headers, for example).
3409 6. Lookups now cache *every* query, not just the most recent. A new, separate
3410 store pool is used for this. It can be recovered when all lookup caches are
3411 flushed. Lookups now release memory at the end of their result strings.
3412 This has involved some general refactoring of the lookup sources.
3414 7. Some code has been added to the store_xxx() functions to reduce the amount
3415 of flapping under certain conditions.
3417 8. log_incoming_interface used to affect only the <= reception log lines. Now
3418 it causes the local interface and port to be added to several more SMTP log
3419 lines, for example "SMTP connection from", and rejection lines.
3421 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3423 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3425 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3426 could be overwritten at the end of the current message (or the start of a
3427 new message if it was set in a HELO ACL). The value is now preserved for
3428 the duration of the SMTP connection.
3430 12. If a transport had a headers_rewrite setting, and a matching header line
3431 contained an unqualified address, that address was qualified, even if it
3432 did not match any rewriting rules. The underlying bug was that the values
3433 of the flags that permit the existence of unqualified sender and recipient
3434 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3435 non-local messages, and by -bnq for local messages) were not being
3436 preserved with the message after it was received.
3438 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3439 "next input=" as part of the text comprising the host identity instead of
3440 the correct text. The code was using the same buffer for two different
3441 strings. However, depending on which order the printing function evaluated
3442 its arguments, the bug did not always show up. Under Linux, for example, my
3443 test suite worked just fine.
3445 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3446 doesn't work with some older versions of Perl. It has been changed to "my",
3447 which in any case is probably the better facility to use.
3449 15. A really picky compiler found some instances of statements for creating
3450 error messages that either had too many or two few arguments for the format
3453 16. The size of the buffer for calls to the DNS resolver has been increased
3454 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3455 for addresses that have a lot of PTR records. This alleviates a problem; it
3456 does not fully solve it.
3458 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3459 buffer now truncates the list and logs the incident, which is the same
3460 action as happens when Exim is looking up a host name and its aliases.
3461 Previously in this situation something unpredictable would happen;
3462 sometimes it was "internal error: store_reset failed".
3464 18. If a server dropped the connection unexpectedly when an Exim client was
3465 using GnuTLS and trying to read a response, the client delivery process
3466 crashed while trying to generate an error log message.
3468 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3469 string, for example:
3471 warn message = H1: something\nH2: something
3473 the text was added as a single header line from Exim's point of view
3474 though it ended up OK in the delivered message. However, searching for the
3475 second and subsequent header lines using $h_h2: did not work. This has been
3476 fixed. Similarly, if a system filter added multiple headers in this way,
3477 the routers could not see them.
3479 20. Expanded the error message when iplsearch is called with an invalid key to
3480 suggest using net-iplsearch in a host list.
3482 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3483 ACLs are no longer actually imposed (and a message to that effect is
3486 22. If a "gecos" field in a passwd entry contained escaped characters, in
3487 particular, if it contained a \" sequence, Exim got it wrong when building
3488 a From: or a Sender: header from that name. A second bug also caused
3489 incorrect handling when an unquoted " was present following a character
3490 that needed quoting.
3492 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3493 was not being matched caselessly.
3495 24. Arranged for all hyphens in the exim.8 source to be escaped with
3498 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3499 properly. Recipient callout cache records were still being keyed to include
3500 the sender, even when use_sender was set false. This led to far more
3501 callouts that were necessary. The sender is no longer included in the key
3502 when use_sender is false.
3504 26. Added "control = submission" modifier to ACLs.
3506 27. Added the ${base62d: operator to decode base 62 numbers.
3508 28. dnsdb lookups can now access SRV records.
3510 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3511 the configuration file.
3513 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3514 (-v) mode. This makes the output for a verbose queue run more intelligible.
3516 31. Added a use_postmaster feature to recipient callouts.
3518 32. Added the $body_zerocount variable, containing the number of binary zero
3519 bytes in the message body.
3521 33. The time of last modification of the "new" subdirectory is now used as the
3522 "mailbox time last read" when there is a quota error for a maildir
3525 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3527 35. Added +ignore_unknown as a special item in host lists.
3529 36. Code for decoding IPv6 addresses in host lists is now included, even if
3530 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3531 address was recognized as an IP address, but was then not correctly decoded
3532 into binary, causing unexpected and incorrect effects when compared with
3539 1. Very minor rewording of debugging text in manualroute to say "list of
3540 hosts" instead of "hostlist".
3542 2. If verify=header_syntax was set, and a header line with an unqualified
3543 address (no domain) and a large number of spaces between the end of the
3544 name and the colon was received, the reception process suffered a buffer
3545 overflow, and (when I tested it) crashed. This was caused by some obsolete
3546 code that should have been removed. The fix is to remove it!
3548 3. When running in the test harness, delay a bit after writing a bounce
3549 message to get a bit more predictability in the log output.
3551 4. Added a call to search_tidyup() just before forking a reception process. In
3552 theory, someone could use a lookup in the expansion of smtp_accept_max_
3553 per_host which, without the tidyup, could leave open a database connection.
3555 5. Added the variables $recipient_data and $sender_data which get set from a
3556 lookup success in an ACL "recipients" or "senders" condition, or a router
3557 "senders" option, similar to $domain_data and $local_part_data.
3559 6. Moved the writing of debug_print from before to after the "senders" test
3562 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3563 problems for message scanning, either using a data ACL, or using
3564 local_scan() because the Received: header was not generated till after they
3565 were called (in order to set the time as the time of reception completion).
3566 I have revised the way this works. The header is now generated after the
3567 body is received, but before the ACL or local_scan() are called. After they
3568 are run, the timestamp in the header is updated.
3574 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3575 before starting a queue runner without re-exec. This happened only when
3576 deliver_drop_privilege was set or when the Exim user was set to root. The
3577 effect of the bug was that timeouts during subsequent deliveries caused
3578 crashes instead of being properly handled. The handler is now left at its
3579 default (and expected) setting.
3581 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3582 message, again when deliver_drop_privilege is set or Exim is run as root.
3583 The bug described in (1) was not present in this case, but the tidying up
3584 of the other signals was missing. I have made the two cases consistent.
3586 3. The ignore_target_hosts setting on a manualroute router was being ignored
3587 for hosts that were looked up using the /MX notation.
3589 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3592 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3593 operated on the sender address. After changing the $sender_address to <>
3594 for the sender address verify, Exim was re-instated it as the original
3595 (before rewriting) address, but remembering that it had rewritten it, so it
3596 wasn't rewriting it again. This bug also had the effect of breaking the
3597 sender address verification caching when the sender address was rewritten.
3599 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3600 This has been changed so that if the ip literal address matches
3601 ignore_target_hosts, the router declines.
3603 7. Added expansion conditions match_domain, match_address, and match_local_
3604 part (NOT match_host).
3606 8. The placeholder for the Received: header didn't have a length field set.
3608 9. Added code to Exim itself and to exim_lock to test for a specific race
3609 condition that could lead to file corruption when using MBX delivery. The
3610 issue is with the lockfile that is created in /tmp. If this file is removed
3611 after a process has opened it but before that process has acquired a lock,
3612 there is the potential for a second process to recreate the file and also
3613 acquire a lock. This could lead to two Exim processes writing to the file
3614 at the same time. The added code performs the same test as UW imapd; it
3615 checks after acquiring the lock that its file descriptor still refers to
3616 the same named file.
3618 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3619 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3622 11. Added the smtp_active_hostname option. If used, this will typically be made
3623 to depend on the incoming interface address. Because $interface_address is
3624 not set up until the daemon has forked a reception process, error responses
3625 that can happen earlier (such as "too many connections") no longer contain
3628 12. If an expansion in a condition on a "warn" statement fails because a lookup
3629 defers, the "warn" statement is abandoned, and the next ACL statement is
3630 processed. Previously this caused the whole ACL to be aborted.
3632 13. Added the iplsearch lookup type.
3634 14. Added ident_timeout as a log selector.
3636 15. Added tls_certificate_verified as a log selector.
3638 16. Added a global option tls_require_ciphers (compare the smtp transport
3639 option of the same name). This controls incoming TLS connections.
3641 17. I finally figured out how to make tls_require_ciphers do a similar thing
3642 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3643 before starting the TLS session.
3645 18. Tabs are now shown as \t in -bP output.
3647 19. If the log selector return_path_on_delivery was set, Exim crashed when
3648 bouncing a message because it had too many Received: header lines.
3650 20. If two routers both had headers_remove settings, and the first one included
3651 a superfluous trailing colon, the final name in the first list and the
3652 first name in the second list were incorrectly joined into one item (with a
3653 colon in the middle).
3659 1. Added -C and -D options to the exinext utility, mainly to make it easier
3660 to include in the automated testing, but these could be helpful when
3661 multiple configurations are in use.
3663 2. The exinext utility was not formatting the output nicely when there was
3664 an alternate port involved in the retry record key, nor when there was a
3665 message id as well (for retries that were specific to a specific message
3666 and a specific host). It was also confused by IPv6 addresses, because of
3667 the additional colons they contain. I have fixed the IPv4 problem, and
3668 patched it up to do a reasonable job for IPv6.
3670 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3671 delivery, the log line now contains "pipelined" if PIPELINING was used.
3673 4. An SMTP transport process used to panic and die if the bind() call to set
3674 an explicit outgoing interface failed. This has been changed; it is now
3675 treated in the same way as a connect() failure.
3677 5. A reference to $sender_host_name in the part of a conditional expansion
3678 that was being skipped was still causing a DNS lookup. This no longer
3681 6. The def: expansion condition was not recognizing references to header lines
3682 that used bh_ and bheader_.
3684 7. Added the _cache feature to named lists.
3686 8. The code for checking quota_filecount in the appendfile transport was
3687 allowing one more file than it should have been.
3689 9. For compatibility with Sendmail, the command line option
3697 and sets the incoming protocol and host name (for trusted callers). The
3698 host name and its colon can be omitted when only the protocol is to be set.
3699 Note the Exim already has two private options, -pd and -ps, that refer to
3700 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3701 "s", but I don't think that's a major issue.
3703 10. A number of refactoring changes to the code, none of which should affect
3706 (a) The number of logging options was getting close to filling up the
3707 32-bit word that was used as a bit map. I have split them into two classes:
3708 those that are passed in the argument to log_write(), and those that are
3709 only ever tested independently outside of that function. These are now in
3710 separate 32-bit words, so there is plenty of room for expansion again.
3711 There is no change in the user interface or the logging behaviour.
3713 (b) When building, for example, log lines, the code previously used a
3714 macro that called string_cat() twice, in order to add two strings. This is
3715 not really sufficiently general. Furthermore, there was one instance where
3716 it was actually wrong because one of the argument was used twice, and in
3717 one call a function was used. (As it happened, calling the function twice
3718 did not affect the overall behaviour.) The macro has been replaced by a
3719 function that can join an arbitrary number of extra strings onto a growing
3722 (c) The code for expansion conditions now uses a table and a binary chop
3723 instead of a serial search (which was left over from when there were very
3724 few conditions). Also, it now recognizes conditions like "pam" even when
3725 the relevant support is not compiled in: a suitably worded error message is
3726 given if an attempt is made to use such a condition.
3728 11. Added ${time_interval:xxxxx}.
3730 12. A bug was causing one of the ddress fields not to be passed back correctly
3731 from remote delivery subprocesses. The field in question was not being
3732 subsequently used, so this caused to problems in practice.
3734 13. Added new log selectors queue_time and deliver_time.
3736 14. Might have fixed a bug in maildirsizefile handling that threw up
3737 "unexpected character" debug warnings, and recalculated the data
3738 unnecessarily. In any case, I expanded the warning message to give more
3741 15. Added the message "Restricted characters in address" to the statements in
3742 the default ACL that block characters like @ and % in local parts.
3744 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3745 Three changes have been made:
3747 (a) There was a serious bug; a negative response to MAIL caused the whole
3748 recipient domain to be cached as invalid, thereby blocking all messages
3749 to all local parts at the same domain, from all senders. This bug has
3750 been fixed. The domain is no longer cached after a negative response to
3751 MAIL if the sender used is not empty.
3753 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3756 (c) A new callout option, "use_sender" has been added for people who want
3757 the modified behaviour.
3763 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3766 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3767 indeed breaks things for older releases.
3769 3. Added additional logging to the case where there is a problem reading data
3770 from a filter that is running in a subprocess using a pipe, in order to
3771 try to track down a specific problem.
3773 4. Testing facility fudge: when running in the test harness and attempting
3774 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3775 getting "No route to host". Convert this to a timeout.
3777 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3780 6. Some OS don't have socklen_t but use size_t instead. This affects the
3781 fifth argument of getsockopt() amongst other things. This is now
3782 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3783 can be set for individual OS. I have set it for SunOS5, OSF1, and
3784 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3785 some earlier ones do not.
3787 7. Change 4.30/15 was not doing the test caselessly.
3789 8. The standard form for an IPv6 address literal was being rejected by address
3790 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3791 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3792 this, as well as the form without the "IPv6" on the front (but only when
3793 address literals are enabled, of course).
3795 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3797 10. Exim crashed if a message with an empty sender address specified by -f
3798 encountered a router with an errors_to setting. This could be provoked only
3799 by a command such as
3803 where an empty string was supplied; "<>" did not hit this bug.
3805 11. Installed PCRE release 4.5.
3807 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3808 remained set. It is now erased.
3810 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3811 times from message ids (which are base 36 rather than the normal 62).
3813 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3814 were being counted as actual protocol errors, and logged if the log
3815 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3816 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3817 and DATA following a set of rejected RCPTs do not count as protocol errors.
3818 In other words, Exim assumes they were pipelined, though this may not
3819 actually be the case. Of course, in all cases the client gets an
3820 appropriate error code.
3822 15. If a lookup fails in an ACL condition, a message about the failure may
3823 be available; it is used if testing the ACL cannot continue, because most
3824 such messages specify what the cause of the deferral is. However, some
3825 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3826 that caused an old message to be retained and used if a later statement
3827 caused a defer, replacing the real cause of the deferral.
3829 16. If an IP address had so many PTR records that the DNS lookup buffer
3830 was not large enough to hold them, Exim could crash while trying to process
3831 the truncated data. It now detects and logs this case.
3833 17. Further to 4.21/58, another change has been made: if (and only if) the
3834 first line of a message (the first header line) ends with CRLF, a bare LF
3835 in a subsequent header line has a space inserted after it, so as not to
3836 terminate the header.
3838 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3839 unnecessarily, used atoi() instead of strtol(), and didn't check the
3840 termination when getting file sizes from file names by regex.
3842 19. Completely re-implemented the support for maildirsize files, in the light
3843 of a number of problems with the previous contributed implementation
3844 (4.30/29). In particular:
3846 . If the quota is zero, the maildirsize file is maintained, but no quota is
3849 . If the maildir directory does not exist, it is created before any attempt
3850 to write a maildirsize file.
3852 . The quota value in the file is just a cache; if the quota is changed in
3853 the transport, the new value overrides.
3855 . A regular expression is available for excluding directories from the
3858 20. The autoreply transport checks the characters in options that define the
3859 message's headers; it allows continued headers, but it was checking with
3860 isspace() after an embedded newline instead of explicitly looking for a
3863 21. If all the "regular" hosts to which an address was routed had passed their
3864 expiry times, and had not reached their retry times, the address was
3865 bounced, even if fallback hosts were defined. Now Exim should go on to try
3868 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3869 equivalent code in the SMTP transport. Some hosts send humungous responses
3870 to HELO/EHLO, more than 1024 it seems.
3872 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3873 gives compiler warnings in some environments. I've now done it "properly",
3876 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3877 (because of problems with the built-in one) was declared to return uschar *
3878 instead of char *, causing compiler failure.
3880 25. Fixed a file descriptor leak when processing alias/forward files.
3882 26. Fixed a minor format string issue in dbfn.c.
3884 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3886 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3887 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3890 29. When a sender address is verified, it is cached, to save repeating the test
3891 when there is more than one recipient in a message. However, when the
3892 verification involves a callout, it is possible for different callout
3893 options to be set for different recipients. It is too complicated to keep
3894 track of this in the cache, so now Exim always runs a verification when a
3895 callout is required, relying on the callout cache for the optimization.
3896 The overhead is duplication of the address routing, but this should not be
3899 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3900 to be verified with callout=postmaster, and the main callout worked but the
3901 postmaster check failed, the verification correctly failed. However, if a
3902 subsequent RCPT command asked for sender verification *without* the
3903 postmaster check, incorrect caching caused this verification also to fail,
3906 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3907 it was not caching the DNS options (qualify_single, search_parents) that
3908 were used when the lookup failed. A subsequent lookup with different
3909 options therefore always gave the same answer, though there were cases
3910 where it should not have. (Example: a "domains = !$mx_any" option on a
3911 dnslookup router: the "domains" option is always processed without any
3912 widening, but the router might have qualify_single set.) Now Exim uses the
3913 cached value only when the same options are set.
3915 32. Added John Jetmore's "exipick" utility to the distribution.
3917 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3918 than a timeout (e.g. a certificate is required, and is not provided), an
3919 Exim server now closes the connection immediately. Previously it waited for
3920 the client to close - but if the client is SSL, it seems that they each
3921 wait for each other, leading to a delay before one of them times out.
3923 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3924 maintained 0.8.x compatibility because I don't think many are using it, and
3925 it is clearly obsolete.
3927 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3930 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3931 client certificate was expired. A simple patch fixes this, though I don't
3932 understand the full logic of why the verify callback is called multiple
3935 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3936 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3937 which causes problems with some clients (such as the Certicom SSL Plus
3938 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3939 disables the coutermeasure allowing Eudora to connect."
3941 38. Exim was not checking that a write() to a log file succeeded. This could
3942 lead to Bad Things if a log got too big, in particular if it hit a file
3943 size limit. Exim now panics and dies if it cannot write to a log file, just
3944 as it does if it cannot open a log file.
3946 39. Modified OS/Makefile-Linux so that it now contains
3948 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3950 The two -D definitions ensure that Exim is compiled with large file
3951 support, which makes it possible to handle log files that are bigger than
3954 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3955 instance) a domain was checked against a named list that involved a lookup,
3956 causing $domain_data to be set, then another domain was checked against the
3957 same list, then the first domain was re-checked, the value of $domain_data
3958 after the final check could be wrong. In particular, if the second check
3959 failed, it could be set empty. This bug probably also applied to
3962 41. The strip_trailing_dot option was not being applied to the address given
3963 with the -f command-line option.
3965 42. The code for reading a message's header from the spool was incrementing
3966 $received_count, but never initializing it. This meant that the value was
3967 incorrect (doubled) while delivering a message in the same process in which
3968 it was received. In the most common configuration of Exim, this never
3969 happens - a fresh exec is done - but it can happen when
3970 deliver_drop_privilege is set.
3972 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3973 it now includes up to 150 characters of the unexpected data in the log
3976 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3977 and building data strings. The size of both of these buffers was 10 000
3978 bytes - far larger than anybody would *ever* want, thought I. Needless to
3979 say, somebody hit the limit. I have increased the maximum line length to
3980 20 000 and the maximum data length of concatenated lines to 100 000. I have
3981 also fixed two bugs, because there was no checking on these buffers. Tsk,
3982 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3983 buffer is too small.
3985 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3986 lsearch lookups. Now it does.
3988 46. When parsing a route_list item in a manualroute router, a fixed-length
3989 buffer was used for the list of hosts. I made this 1024 bytes long,
3990 thinking that nobody would ever have a list of hosts that long. Wrong.
3991 Somebody had a whole pile of complicated expansion conditions, and the
3992 string was silently truncated, leading to an expansion error. It turns out
3993 that it is easier to change to an unlimited length (owing to other changes
3994 that have happened since this code was originally written) than to build
3995 structure for giving a limitation error. The length of the item that
3996 expands into the list of hosts is now unlimited.
3998 47. The lsearch lookup could not handle data where the length of text line was
3999 more than 4095 characters. Such lines were truncated, leading to shortened
4000 data being returned. It should now handle lines of any length.
4002 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4003 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4006 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4007 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4008 no longer happens in comment lines. A list of replacements is now placed
4009 at the head of all of the source files, except those whose only change is
4010 to replace PERL_COMMAND in the very first #! line.
4012 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4013 messages on the queue, with a bottom-up merge sort, using code contributed
4014 by Michael Haardt. This should make operations like -bp somewhat faster on
4015 large queues. It won't affect queue runners, except when queue_run_in_order
4018 51. Installed eximstats 1.31 in the distribution.
4020 52. Added support for SRV lookups to the dnslookup router.
4022 53. If an ACL referred to $message_body or $message_body_end, the value was not
4023 reset for any messages that followed in the same SMTP session.
4025 54. The store-handling optimization for building very long strings was not
4026 differentiating between the different store pools. I don't think this
4027 actually made any difference in practice, but I've tidied it.
4029 55. While running the routers to verify a sender address, $sender_address
4030 was still set to the sender address. This is wrong, because when routing to
4031 send a bounce to the sender, it would be empty. Therefore, I have changed
4032 it so that, while verifying a sender address, $sender_address is set to <>.
4033 (There is no change to what happens when verifying a recipient address.)
4035 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4036 target A or AAAA records (if not already returned) without resetting the
4037 qualify_single or search_parents options of the DNS resolver. These are
4038 inappropriate in this case because the targets of MX and SRV records must
4039 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4040 target that, when qualified, matched something in the local domain. These
4041 two options are now turned off when doing these lookups.
4043 57. It seems that at least some releases of Reiserfs (which does not have the
4044 concept of a fixed number of inodes) returns zero and not -1 for the
4045 number of available inodes. This interacted badly with check_spool_inodes,
4046 which assumed that -1 was the "no such thing" setting. What I have done is
4047 to check that the total number of inodes is greater than zero before doing
4048 the test of how many are available.
4050 58. When a "warn" ACL statement has a log_message modifier, the message is
4051 remembered, and not repeated. This is to avoid a lot of repetition when a
4052 message has many recipients that cause the same warning to be written.
4053 Howewer, Exim was preserving the list of already written lines for an
4054 entire SMTP session, which doesn't seem right. The memory is now reset if a
4055 new message is started.
4057 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4058 debugging output unless log_rewrite was also set.
4060 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4061 of (int)(handle) when we know that handle contains (void *)(-1).
4063 61. The Exim daemon panic-logs an error return when it closes the incoming
4064 connection. However "connection reset by peer" seems to be common, and
4065 isn't really an error worthy of noting specially, so that particular error
4068 62. When Exim is trying to find all the local interfaces, it used to panic and
4069 die if the ioctl to get the interface flags failed. However, it seems that
4070 on at least one OS (Solaris 9) it is possible to have an interface that is
4071 included in the list of interfaces, but for which you get a failure error
4072 for this call. This happens when the interface is not "plumbed" into a
4073 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4074 failure of the "get flags" call assumes that the interface is down.
4076 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4077 makes life easier for people who are doing arithmetic on fields extracted
4078 from dates, where you often get leading zeros that should not be
4079 interpreted as octal.
4081 64. Added qualify_domain to the redirect router, to override the global
4084 65. If a pathologically long header line contained very many addresses (the
4085 report of this problem mentioned 10 000) and each of them was rewritten,
4086 Exim could use up a very large amount of memory. (It kept on making new
4087 copies of the header line as it rewrote, and never released the old ones.)
4088 At the expense of a bit more processing, the header rewriting function has
4089 been changed so that it no longer eats memory in this way.
4091 66. The generation of the Received: header has been moved from the time that a
4092 message starts to be received, to the time that it finishes. The timestamp
4093 in the Received: header should now be very close to that of the <= log
4094 line. There are two side-effects of this change:
4096 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4097 logged header lines no longer include the local Received: line, because
4098 it has not yet been created. The same applies to a copy of the message
4099 that is returned to a non-SMTP sender when a message is rejected.
4101 (b) When a filter file is tested using -bf, no additional Received: header
4102 is added to the test message. After some thought, I decided that this
4105 This change does not affect the value of $received_for. It is still set
4106 after address rewriting, but before local_scan() is called.
4108 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4110 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4111 gave an unhelpful panic error message, and a defer error. I have managed to
4112 change this behaviour so that it now rejects any supplied certificate,
4113 which seems right, as the list of acceptable certificates is empty.
4115 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4116 gave an unhelpful defer error. I have not managed to make this reject any
4117 supplied certificates, but the error message it gives is "no certificate
4118 supplied", which is not helpful.
4120 70. exigrep's output now also includes lines that are not associated with any
4121 message, but which match the given pattern. Implemented by a patch from
4122 Martin Sluka, which also tidied up the Perl a bit.
4124 71. Recipient callout verification, like sender verification, was using <> in
4125 the MAIL FROM command. This isn't really the right thing, since the actual
4126 sender may affect whether the remote host accepts the recipient or not. I
4127 have changed it to use the actual sender in the callout; this means that
4128 the cache record is now keyed on a recipient/sender pair, not just the
4129 recipient address. There doesn't seem to be a real danger of callout loops,
4130 since a callout by the remote host to check the sender would use <>.
4131 [SEE ABOVE: changed after hitting problems.]
4133 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4134 temporary errors. However, in the case of such a code being given after
4135 the end of a data transmission (i.e. after ".") Exim was failing to write
4136 a retry record for the message. (Yes, there was some broken host that was
4137 actually sending 8xx at this point.)
4139 73. An unknown lookup type in a host list could cause Exim to panic-die when
4140 the list was checked. (An example that provoked this was putting <; in the
4141 middle of a list instead of at the start.) If this happened during a DATA
4142 ACL check, a -D file could be left lying around. This kind of configuration
4143 error no longer causes Exim to die; instead it causes a defer errror. The
4144 incident is still logged to the main and panic logs.
4146 74. Buglet left over from Exim 3 conversion. The message "too many messages
4147 in one connection" was written to the rejectlog but not the mainlog, except
4148 when address rewriting (yes!) was being logged.
4150 75. Added write_rejectlog option.
4152 76. When a system filter was run not as root (that is, when system_filter_user
4153 was set), the values of the $n variables were not being returned to the
4154 main process; thus, they were not subsequently available in the $sn
4157 77. Added +return_path_on_delivery log selector.
4159 78. A connection timeout was being treated differently from recipients deferred
4160 when testing hosts_max_try with a message that was older than the host's
4161 retry timeout. (The host should not be counted, thus allowing all hosts to
4162 be tried at least once before bouncing.) This may have been the cause of an
4163 occasionally reported bug whereby a message would remain on the queue
4164 longer than the retry timeout, but would be bounced if a delivery was
4165 forced. I say "may" because I never totally pinned down the problem;
4166 setting up timeout/retry tests is difficult. See also the next item.
4168 79. The ultimate address timeout was not being applied to errors that involved
4169 a combination of host plus message (for example, a timeout on a MAIL
4170 command). When an address resolved to a number of possible hosts, and they
4171 were not all tried for each delivery (e.g. because of hosts_max_try), a
4172 message could remain on the queue longer than the retry timeout.
4174 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4177 81. Fixed an obscure SMTP outgoing bug which required at least the following
4178 conditions: (a) there was another message waiting for the same server;
4179 (b) the server returned 5xx to all RCPT commands in the first message so
4180 that the message was not completed; (c) the server dropped the connection
4181 or gave a negative response to the RSET that Exim sends to abort the
4182 transaction. The observed case was a dropped connection after DATA that had
4183 been sent in pipelining mode. That is, the server had advertised PIPELINING
4184 but was not implementing it correctly. The effect of the bug was incorrect
4185 behaviour, such as trying another host, and this could lead to a crash.
4191 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4192 and daemon.c were passed as pointers to ints; they should have been
4193 pointers to socklen_t variables (which are typically unsigned ints).
4195 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4198 3. Fixed a really odd bug that affected only the testing scheme; patching a
4199 certain fixed string in the binary changed the value of another string that
4200 happened to be identical to the end of the original first string.
4202 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4203 name", it returns that address as the IP address. On some operating
4204 systems (e.g. Solaris), it also passes back the IP address string as the
4205 "host name". However, on others (e.g. Linux), it passes back an empty
4206 string. Exim wasn't checking for this, and was changing the host name to an
4207 empty string, assuming it had been canonicized.
4209 5. Although rare, it is permitted to have more than one PTR record for a given
4210 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4211 all the names associated with an address, because they do in Solaris.
4212 However, it seems that they do not in Linux for data that comes from the
4213 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4214 I found this out when I moved to a new Linux workstation and tried to run
4215 the Exim test suite.
4217 To get round this problem I have changed the code so that it now does its
4218 own call to the DNS to look up PTR records when searching for a host name.
4219 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4220 addresses that are only in /etc/hosts are still found.
4222 This behaviour is, however, controlled by an option called host_lookup_
4223 order, which defaults to "bydns:byaddr". If people want to use the other
4224 order, or indeed, just use one or the other means of lookup, they can
4225 specify it in this variable.
4227 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4228 some operating systems, this comes back from gethostbyaddr() as an empty
4229 string, and this is what Exim used to test for. However, it seems that in
4230 other systems, "." is yielded. Exim now tests for this case too.
4232 7. The values of check_spool_space and check_log_space are now held internally
4233 as a number of kilobytes instead of an absolute number of bytes. If a
4234 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4235 kilobyte. This means that much larger values can be stored.
4237 8. Exim monitor: an attempt to get the action menu when not actually pointing
4238 at a message produces an empty menu entitled "No message selected". This
4239 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4240 no entries in it ("Shell widget menu has zero width and/or height"). So I
4241 have added a single, blank menu entry in this case.
4243 9. Added ${quote_local_part.
4245 10. MIME decoding is now applied to the contents of Subject: header lines when
4248 11. Now that a reference to $sender_host_address automatically causes a reverse
4249 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4250 host lookup before query-style lookups in lists that might use this
4251 variable. This has therefore been abolished, and the "net-" prefix is no
4252 longer necessary for query-style lookups.
4254 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4255 have been a typo for LFLAGS, so it has been changed.
4257 13. The install script calls Exim with "-C /dev/null" in order to find the
4258 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4259 to be output. Howeve, since Exim outputs its version number before the
4260 error, it didn't break the script. It just looked ugly. I fixed this by
4261 always allowing "-C /dev/null" if the caller is root.
4263 14. Ignore overlarge ACL variable number when reading spool file - insurance
4264 against a later release with more variables having written the file.
4266 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4267 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4268 this, as well as the form without the "IPv6" on the front.
4270 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4271 OS/Makefile-Darwin file.
4273 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4274 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4276 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4277 variable. However, some OS use ints for inodes. I've added cast to long int
4278 to get rid of the compiler warning.
4280 19. I had forgotten to lock out "/../" in configuration file names when
4281 ALT_CONFIG_PREFIX was set.
4283 20. Routers used for verification do not need to specify transports. However,
4284 if such a router generated a host list, and callout was configured, Exim
4285 crashed, because it could not find a port number from the (non-existent)
4286 transport. It now assumes port 25 in this circumstance.
4288 21. Added the -t option to exigrep.
4290 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4291 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4292 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4293 them is set, LOOKUP_LSEARCH is forced.
4295 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4297 24. Added sender and host information to the "rejected by local_scan()" log
4298 line; previously there was no indication of these.
4300 25. Added .include_if_exists.
4302 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4303 Linux. It turns out that not all file systems support this. Apparently some
4304 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4305 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4306 not supported on the file descriptor, is now ignored when Exim is trying to
4307 sync a directory. This applies only to Linux.
4309 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4311 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4312 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4315 29. Added support for maildirsize files from supplied patch (modified a bit).
4317 30. The use of :fail: followed by an empty string could lead Exim to respond to
4318 sender verification failures with (e.g.):
4320 550 Verification failed for <xxx>
4321 550 Sender verify failed
4323 where the first response line was missing the '-' that indicates it is not
4324 the final line of the response.
4326 31. The loop for finding the name of the user that called Exim had a hardwired
4327 limit of 10; it now uses the value of finduser_retries, which is used for
4328 all other user lookups.
4330 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4333 33. Exim was neglecting to zero errno before one call of strtol() when
4334 expanding a string and expecting an integer value. On some systems this
4335 resulted in spurious "integer overflow" errors. Also, it was casting the
4336 result into an int without checking.
4338 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4339 did not work. The code looks as if it has *never* worked, though it appears
4340 to have been documented since at least releast 1.62. I have made it work.
4342 35. The "timeout_DNS" error in retry rules, also documented since at least
4343 1.62, also never worked. As it isn't clear exactly what this means, and
4344 clearly it isn't a major issue, I have abolished the feature by treating it
4345 as "timeout", and writing a warning to the main and panic logs.
4347 36. The display of retry rules for -brt wasn't always showing the error code
4350 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4351 timeout_connect_A, timeout_connect_MX.
4353 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4354 to the empty sender.
4356 39. The daemon was not analysing the content of -oX till after it had closed
4357 stderr and disconnected from the controlling terminal. This meant that any
4358 syntax errors were only noted on the panic log, and the return code from
4359 the command was 0. By re-arranging the code a little, I've made the
4360 decoding happen first, so such errors now appear on stderr, and the return
4361 code is 1. However, the actual setting up of the sockets still happens in
4362 the disconnected process, so errors there are still only recorded on the
4365 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4366 connections (as happens on some IP stacks) was logged at start up time as
4367 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4368 it from "IPv6 and IPv4", which means that two separate sockets are being
4371 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4372 says whether AF_INET or AF_INET6 was passed as an argument.
4374 42. Exiwhat output was messed up when time zones were included in log
4377 43. Exiwhat now gives more information about the daemon's listening ports,
4378 and whether -tls-on-connect was used.
4380 44. The "port" option of the smtp transport is now expanded.
4382 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4383 silently ignored. Now an error message is written to the main and panic
4386 46. There's a new ACL modifier called "logwrite" which writes to a log file
4387 as soon as it is encountered.
4389 47. Added $local_user_uid and $local_user_gid at routing time.
4391 48. Exim crashed when trying to verify a sender address that was being
4394 49. Exim was recognizing only a space character after ".include". It now also
4395 recognizes a tab character.
4397 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4398 extracting the relevant information from the specification. The man page no
4399 longer contains scrambled data for the -d option, and I've added a section
4400 at the front about calling Exim under different names.
4402 51. Added "extra_headers" argument to the "mail" command in filter files.
4404 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4407 53. Installed eximstats 1.29.
4409 54. Added transport_filter_timeout as a generic transport option.
4411 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4412 Cc: header lines. This was required by RFC 822, but it not required by RFC
4415 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4416 incoming messages that did not have them. Now it does so only if the
4417 message originates locally, that is, if there is no associated remote host
4418 address. When Resent- header lines are present, this applies to the Resent-
4419 lines rather than the non-Resent- lines.
4421 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4422 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4424 58. Messages for configuration errors now include the name of the main
4425 configuration file - useful now that there may be more than one file in a
4426 list (.included file names were always shown).
4428 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4429 for those rare installations that do not start the daemon as root or run it
4430 setuid root. I've cut out the call to initgroups() if the daemon is not
4433 60. The Exim user and group can now be bound into the binary as text strings
4434 that are looked up at the start of Exim's processing.
4436 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4438 62. Added $mailstore_basename variable.
4440 63. Installed patch to sieve.c from Michael Haardt.
4442 64. When Exim failed to open the panic log after failing to open the main log,
4443 the original message it was trying to log was written to stderr and debug
4444 output, but if they were not available (the usual case in production), it
4445 was lost. Now it is written to syslog before the two lines that record the
4446 failures to open the logs.
4448 65. Users' Exim filters run in subprocesses under the user's uid. It is
4449 possible for a "deliver" command or an alias in a "personal" command to
4450 provoke an address rewrite. If logging of address rewriting is configured,
4451 this fails because the process is not running as root or exim. There may be
4452 a better way of dealing with this, but for the moment (because 4.30 needs
4453 to be released), I have disabled address rewrite logging when running a
4454 filter in a non-root, non-exim process.
4460 1. The buildconfig auxiliary program wasn't quoting the value set for
4461 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4462 not defined. This bug was masked in 4.22 by the effect that was fixed in
4465 2. Some messages that were rejected after a message id was allocated were
4466 shown as "incomplete" by exigrep. It no longer does this for messages that
4467 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4469 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4470 have allow_domain_literals set, the ID did not get logged in the <= line.
4471 Domain literals are now always recognized in Message-ID: header lines.
4473 4. The first argument for a ${extract expansion item is the key name or field
4474 number. Leading and trailing spaces in this item were not being ignored,
4475 causing some misleading effects.
4477 5. When deliver_drop_privilege was set, single queue runner processes started
4478 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4479 same command in the process it spins off) were not dropping privilege.
4481 6. When the daemon running as "exim" started a queue runner, it always
4482 re-executed Exim in the spun-off process. This is a waste of effort when
4483 deliver_drop_privilege is set. The new process now just calls the
4484 queue-runner function directly.
4490 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4493 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4494 ignored. Though the use of -oP was forcing the writing of a pid file, it
4495 was always written to the default place.
4497 3. If the message "no IP address found for host xxxx" is generated during
4498 incoming verification, it is now followed by identification of the incoming
4499 connection (so you can more easily find what provoked it).
4501 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4503 5. Added some features to "harden" Exim a bit more against certain attacks:
4505 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4506 be put in Local/Makefile. This is like the never_users runtime option,
4507 but it cannot be overridden. The default setting is "root".
4509 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4510 prefix string with which any file named in a -C command line option
4513 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4514 is retained for -C and -D only if the caller of Exim is root. Without
4515 it, the exim user may also use -C and -D and retain privilege.
4517 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4518 command line option is disabled.
4520 6. Macro names set by the -D option must start with an upper case letter, just
4521 like macro names defined in the configuration file.
4523 7. Added "dereference=" facility to LDAP.
4525 8. Two instances of the typo "uknown" in the source files are fixed.
4527 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4528 the Configure-Makefile script screwed up while processing it.
4530 10. Incorporated PCRE 4.4.
4532 11. The SMTP synchronization check was not operating right at the start of an
4533 SMTP session. For example, it could not catch a HELO sent before the client
4534 waited for the greeting. There is now a check for outstanding input at the
4535 point when the greeting is written. Because of the duplex, asynchronous
4536 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4537 way, but not yet received, when the check is performed.
4539 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4540 on TCP/IP sockets, because this apparently causes some broken clients to
4543 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4544 unchanged) from the Cygwin maintainer.
4546 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4547 was supported instead of testing for mbx. Effectively a typo.
4549 15. The spa authenticator server code was not checking that the input it
4550 received was valid base64.
4552 16. The debug output line for the "set" modifier in ACLs was not showing the
4553 name of the variable that was being set.
4555 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4557 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4558 lookup on its own. Something else was needed to trigger the lookup. For
4559 example, a match in host_lookup or the need for a host name in a host list.
4560 Now, if $sender_host_name is referenced and the host name has not yet been
4561 looked up, a lookup is performed. If the lookup fails, the variable remains
4562 empty, and $host_lookup_failed is set to "1".
4564 19. Added "eqi" as a case-independent comparison operator.
4566 20. The saslauthd authentication condition could segfault if neither service
4567 nor realm was specified.
4569 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4570 error message that was logged was misleading, and incoming SMTP
4571 connections were dropped. The message is now more accurate, and temporary
4572 errors are given to SMTP connections.
4574 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4575 (except RSET) with a 421 error, until QUIT is received. However, it was
4576 failing to send a response to QUIT.
4578 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4579 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4580 ACL is now run after the helo_try_verify_hosts code.
4582 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4583 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4584 case-independent, so other case variants are also recognized). Apparently
4585 some systems use these upper case variants.
4587 25. If more than two messages were waiting for the same host, and a transport
4588 filter was specified for the transport, Exim sent two messages over the
4589 same TCP/IP connection, and then failed with "socket operation on non-
4590 socket" when it tried to send the third.
4592 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4594 27. The extern definition of crypt16() in expand.c was not being excluded when
4595 the OS had its own crypt16() function.
4597 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4598 as a preferred synonym for return_size_limit, both as an option and as an
4601 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4603 30. Changed the default configuration ACL to relax the local part checking rule
4604 for addresses that are not in any local domains. For these addresses,
4605 slashes and pipe symbols are allowed within local parts, but the sequence
4606 /../ is explicitly forbidden.
4608 31. SPA server authentication was not clearing the challenge buffer before
4611 32. log_message in a "warn" ACL statement was writing to the reject log as
4612 well as to the main log, which contradicts the documentation and doesn't
4613 seem right (because no rejection is happening). So I have stopped it.
4615 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4616 However, I am unable to do any testing of this.
4618 34. Fixed an infelicity in the appendfile transport. When checking directories
4619 for a mailbox, to see if any needed to be created, it was accidentally
4620 using path names with one or more superfluous leading slashes; tracing
4621 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4623 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4624 contents are added to the log line that is written for every discarded
4625 recipient. (Previously a log_message setting was ignored.)
4627 36. The ${quote: operator now quotes the string if it is empty.
4629 37. The install script runs exim in order to find its version number. If for
4630 some reason other than non-existence or emptiness, which it checks, it
4631 could not run './exim', it was installing it with an empty version number,
4632 i.e. as "exim-". This error state is now caught, and the installation is
4635 38. An argument was missing from the function that creates an error message
4636 when Exim fails to connect to the socket for saslauthd authentication.
4637 This could cause Exim to crash, or give a corrupted message.
4639 39. Added isip, isip4, and isip6 to ${if conditions.
4641 40. The ACL variables $acl_xx are now saved with the message, and can be
4642 accessed later in routers, transports, and filters.
4644 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4645 strings in the file are not string-expanded.
4647 42. If a MAIL command specified a SIZE value that was too large to fit into an
4648 int variable, the check against message_size_limit failed. Such values are
4649 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4650 day this will have to be increased, but I don't think I want to be around
4651 when emails are that large.
4658 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4659 iconv() is not standard in FreeBSD.
4661 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4662 IPv6 enabled. The observed symptom was a segmentation fault on return from
4663 the function os_common_find_running_interfaces() in src/os.c.
4665 3. In the check_special_case() function in daemon.c I had used "errno" as an
4666 argument name, which causes warnings on some systems. This was basically a
4667 typo, since it was named "eno" in the comments!
4669 4. The code that waits for the clock to tick (at a resolution of some fraction
4670 of a second) so as to ensure message-id uniqueness was always waiting for
4671 at least one whole tick, when it could have waited for less. [This is
4672 almost certainly not relevant at current processor speeds, where it is
4673 unlikely to ever wait at all. But we try to future-proof.]
4675 5. The function that sleeps for a time interval that includes fractions of a
4676 second contained a race. It did not block SIGALRM between setting the
4677 timer, and suspending (a couple of lines later). If the interval was short
4678 and the sigsuspend() was delayed until after it had expired, the suspension
4679 never ended. On busy systems this could lead to processes getting stuck for
4682 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4683 process, before it forks any delivery processes. The open lookup caching
4684 mechanism meant that the open file or database connection was passed into
4685 the delivery process. The problem was that delivery processes always tidy
4686 up cached lookup data. This could cause a problem for the next delivery
4687 process started by the queue runner, because the external queue runner
4688 process does not know about the closure. So the next delivery process
4689 still has data in the lookup cache. In the case of a file lookup, there was
4690 no problem because closing a file descriptor in a subprocess doesn't affect
4691 the parent. However, if the lookup was caching a connection to a database,
4692 the connection was closed, and the second delivery process was likely to
4693 see errors such as "PGSQL: query failed: server closed the connection
4694 unexpectedly". The problem has been fixed by closing all cached lookups
4695 in a queue runner before running a delivery process.
4697 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4698 seem to have the "const" qualifier which it has on other OS. I've
4701 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4702 *of the same type* (client or server) with the same public name that an
4703 error should be diagnosed.
4705 9. When Exim looked up a host name for an IP address, but failed to find the
4706 original IP address when looking up the host name (a safety check), it
4707 output the message "<ip address> does not match any IP for NULL", which was
4708 confusing, to say the least. The bug was that the host name should have
4709 appeared instead of "NULL".
4711 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4712 user that is built into the binary, and the -C or -D options is used, root
4713 privilege is dropped before the configuration file is read. In addition,
4714 logging is switched to stderr instead of the normal log files. If the
4715 configuration then re-defines the Exim user, the unprivileged environment
4716 is probably not what is expected, so Exim logs a panic warning message (but
4719 However, if deliver_drop_privilege is set, the unprivileged state may well
4720 be exactly what is intended, so the warning has been cut out in that case,
4721 and Exim is allowed to try to write to its normal log files.
4727 1. smtp_return_error_details was not giving details for temporary sender
4728 or receiver verification errors.
4730 2. Diagnose a configuration error if two authenticators have the same public
4733 3. Exim used not to create the message log file for a message until the first
4734 delivery attempt. This could be confusing when incoming messages were held
4735 for policy or load reasons. The message log file is now created at the time
4736 the message is received, and an initial "Received" line is written to it.
4738 4. The automatically generated man page for command line options had a minor
4739 bug that caused no ill effects; however, a more serious problem was that
4740 the procedure for building the man page automatically didn't always
4741 operate. Consequently, release 4.20 contains an out-of-date version. This
4742 shouldn't happen again.
4744 5. When building Exim with embedded Perl support, the script that builds the
4745 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4746 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4748 6. The freeze_tell option was not being used for messages that were frozen on
4749 arrival, either by an ACL or by local_scan().
4751 7. Added the smtp_incomplete_transaction log selector.
4753 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4754 was accepting AUTH without a new EHLO.
4756 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4757 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4760 10. Logging of TCP/IP connections (when configured) now happens in the main
4761 daemon process instead of the child process, so that the TCP/IP connection
4762 count is more accurate (but it can never be perfect).
4764 11. The use of "drop" in a nested ACL was not being handled correctly in the
4765 outer ACL. Now, if condition failure induced by the nested "drop" causes
4766 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4767 or "require"), the connection is dropped.
4769 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4770 that yield "discard" can now be used with an "accept" or a "discard" verb,
4771 but an error is generated for any others (because I can't see a useful way
4772 to define what should happen).
4774 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4775 are now processed in the same way as lines in the Exim configuration file.
4776 In particular, continuation lines are supported.
4778 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4780 15. Added -ti meaning -t -i.
4782 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4783 domains, and warn by logging if others are found.
4785 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4786 returned from GIFCONF when Exim is trying to find the list of interfaces on
4787 a host. The code in os.c has been modified to copy each ifreq to an aligned
4788 structure in all cases.
4790 Also, in some cases, the returned ifreq's were being copied to a 'struct
4791 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4792 means the last couple of bytes of an IPv6 address could be chopped if the
4793 ifreq contained only a normal sockaddr (14 bytes storage).
4795 18. Named domain lists were not supported in the hosts_treat_as_local option.
4796 An entry such as +xxxx was not recognized, and was treated as a literal
4799 19. Ensure that header lines added by a DATA ACL are included in the reject log
4800 if the ACL subsequently rejects the message.
4802 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4803 MD5 (which is deprecated).
4805 21. When testing a filter file using -bf, Exim was writing a message when it
4806 took the sender from a "From " line in the message, but it was not doing so
4807 when it took $return_path from a Return-Path: header line. It now does.
4809 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4810 with a valid header line field name (a series of printing characters
4811 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4813 23. Changed "disc" in the source to "disk" to conform to the documentation and
4814 the book and for uniformity.
4816 24. Ignore Sendmail's -Ooption=value command line item.
4818 25. When execve() failed while trying to run a command in a pipe transport,
4819 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4820 could be confused with a return value of 69 from the command itself. This
4821 has been changed to 127, the value the shell returns if it is asked to run
4822 a non-existent command. The wording for the related log line suggests a
4823 non-existent command as the problem.
4825 26. If received_header_text expands to an empty string, do not add a Received:
4826 header line to the message. (Well, it adds a token one on the spool, but
4827 marks it "old" so that it doesn't get used or transmitted.)
4829 27. Installed eximstats 1.28 (addition of -nt option).
4831 28. There was no check for failure on the call to getsockname() in the daemon
4832 code. This can fail if there is a shortage of resources on the system, with
4833 ENOMEM, for example. A temporary error is now given on failure.
4835 29. Contrary to the C standard, it seems that in some environments, the
4836 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4837 program. Exim now does this explicitly; it affects the formatting of
4838 timestamps using strftime().
4840 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4841 complaints. I've now initialized all the variables, to avoid this.
4843 32. Header lines added by a system filter were not being "seen" during
4844 transport-time rewrites.
4846 33. The info_callback() function passed to OpenSSL is set up with type void
4847 (*)(SSL *, int, int), as described somewhere. However, when calling the
4848 function (actually a macro) that sets it up, the type void(*)() is
4849 expected. I've put in a cast to prevent warnings from picky compilers.
4851 34. If a DNS black list lookup found a CNAME record, but there were no A
4852 records associated with the domain it pointed at, Exim crashed.
4854 35. If a DNS black list lookup returned more than one A record, Exim ignored
4855 all but the first. It now scans all returned addresses if a particular IP
4856 value is being sought. In this situation, the contents of the
4857 $dnslist_value variable are a list of all the addresses, separated by a
4860 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4861 to accept a host name and all its aliases if the forward lookup for any of
4862 them yielded the IP address of the incoming connection. Now it accepts only
4863 those names whose forward lookup yields the correct IP address. Any other
4864 names are discarded. This closes a loophole whereby a rogue DNS
4865 administrator could create reverse DNS records to break through a
4866 wildcarded host restriction in an ACL.
4868 37. If a user filter or a system filter that ran in a subprocess used any of
4869 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4870 the wrong values were passed to the pipe command ($thisaddress had the
4871 value of $0, $0 had the value of $1, etc). This bug was introduced by
4872 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4874 38. Improved the line breaking for long SMTP error messages from ACLs.
4875 Previously, if there was no break point between 40 and 75 characters, Exim
4876 left the rest of the message alone. Two changes have been made: (a) I've
4877 reduced the minimum length to 35 characters; (b) if it can't find a break
4878 point between 35 and 75 characters, it looks ahead and uses the first one
4879 that it finds. This may give the occasional overlong line, but at least the
4880 remaining text gets split now.
4882 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4883 file descriptors might be low, and that setting 1000 would always raise it.
4884 It turns out that in some environments, the limit is already over 1000 and
4885 that lowering it causes trouble. So now Exim takes care not to decrease it.
4887 40. When delivering a message, the value of $return_path is set to $sender_
4888 address at the start of routing (routers may change the value). By an
4889 oversight, this default was not being set up when an address was tested by
4890 -bt or -bv, which affected the outcome if any router or filter referred to
4893 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4894 log only when "message" or "log_message" are set. However, if one of the
4895 conditions was an address verification, or a call to a nested ACL, the
4896 messages generated by the underlying test were being passed through. This
4897 no longer happens. The underlying message is available in $acl_verify_
4898 message for both "message" and "log_message" expansions, so it can be
4899 passed through if needed.
4901 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4902 new expansion $bh_ to give the encoded byte string without charset
4903 translation. Translation happens only if iconv() is available; HAVE_ICONV
4904 indicates this at build time. HEADERS_CHARSET gives the charset to
4905 translate to; headers_charset can change it in the configuration, and
4906 "headers charset" can change it in an individual filter file.
4908 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4909 that creates RFC 2047 encoded "words" labels them as that charset instead
4910 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4911 expansion operator; (ii) when Exim creates a From: line for a local
4912 message; (iii) when a header line is rewritten to include a "phrase" part.
4914 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4915 buggy, causing it to skip the first lines of messages whose message ID
4916 ended in 'D'. This would not have bitten before Exim release 4.14, because
4917 message IDs were unlikely to end in 'D' before then. The effect was to have
4918 incorrect size information for certain domains.
4920 45. #include "config.h" was missing at the start of the crypt16.c module. This
4921 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4924 46. If there was a timeout during a "random" callout check, Exim treated it as
4925 a failure of the random address, and carried on sending RSET and the real
4926 address. If the delay was just some slowness somewhere, the response to the
4927 original RCPT would be taken as a response to RSET and so on, causing
4928 mayhem of various kinds.
4930 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4931 when I implemented it. It didn't allow for the fact that some option values
4932 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4933 the right test for positive values.
4935 48. Domain names in DNS records are case-independent. Exim always looks them up
4936 in lower case. Some resolvers return domain names in exactly the case they
4937 appear in the zone file, that is, they may contain uppercase letters. Not
4938 all resolvers do this - some return always lower case. Exim was treating a
4939 change of case by a resolver as a change of domain, similar to a widening
4940 of a domain abbreviation. This triggered its re-routing code and so it was
4941 trying to route what was effectively the same domain again. This normally
4942 caused routing to fail (because the router wouldn't handle the domain
4943 twice). Now Exim checks for this case specially, and just changes the
4944 casing of the domain that it ultimately uses when it transmits the message
4947 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4950 50. If a filter generated a file delivery with a non-absolute name (possible if
4951 no home directory exists for the router), the forbid_file option was not
4954 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4955 the existing equality matching.
4957 52. Exim was using ints instead of ino_t variables in some places where it was
4958 dealing with inode numbers.
4960 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4961 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4962 called TMPDIR, and if it finds it is different, it changes its value.
4964 54. The smtp_printf() function is now made available to local_scan() so
4965 additional output lines can be written before returning. There is also an
4966 smtp_fflush() function to enable the detection of a dropped connection.
4967 The variables smtp_input and smtp_batched_input are exported to
4970 55. Changed the default runtime configuration: the message "Unknown user"
4971 has been removed from the ACL, and instead placed on the localuser router,
4972 using the cannot_route_message feature. This means that any verification
4973 failures that generate their own messages won't get overridden. Similarly,
4974 the "Unrouteable address" message that was in the ACL for unverifiable
4975 relay addresses has also been removed.
4977 56. Added hosts_avoid_esmtp to the smtp transport.
4979 57. The exicyclog script was not checking for the esoteric option
4980 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4981 will work only if exicyclog is run under the appropriate euid.
4983 58. Following a discussion on the list, the rules by which Exim recognises line
4984 endings on incoming messages have been changed. The -dropcr and drop_cr
4985 options are now no-ops, retained only for backwards compatibility. The
4986 following line terminators are recognized: LF CRLF CR. However, special
4987 processing applies to CR:
4989 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4990 nor a local message in the state where . is a terminator.
4992 (ii) If a bare CR is encountered in a header line, an extra space is added
4993 after the line terminator so as not to end the header. The reasoning
4994 behind this is that bare CRs in header lines are most likely either
4995 to be mistakes, or people trying to play silly games.
4997 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4998 was being incorrectly given as 18 bytes larger than it should have been.
4999 This is a VOB (very old bug).
5001 60. This may never have affected anything current, but just in case it has:
5002 When the local host is found other than at the start of a list of hosts,
5003 the local host, those with the same MX, and any that follow, are discarded.
5004 When the list in question was part of a longer list of hosts, the following
5005 hosts (not currently being processed) were also being discarded. This no
5006 longer happens. I'm not sure if this situation could ever has previously
5009 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5012 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5013 header. This is something that is recommended in a new Internet Draft, and
5014 is something that is documented as being done by Sendmail. There are two
5015 possible values. For messages generated by the autoreply transport, Exim
5018 Auto-Submitted: auto-replied
5020 whereas for all other generated messages (e.g. bounces) it adds
5022 Auto-Submitted: auto-generated
5024 63. The "personal" condition in filters now includes a test for the
5025 Auto-Submitted: header. If it contains the string "auto-" the message it
5026 not considered personal.
5028 64. Added rcpt_include_affixes as a generic transport option.
5030 65. Added queue_only_override (default true).
5032 66. Added the syslog_duplication option.
5034 67. If what should have been the first header line of a message consisted of
5035 a space followed by a colon, Exim was mis-interpreting it as a header line.
5036 It isn't of course - it is syntactically invalid and should therefore be
5037 treated as the start of the message body. The misbehaviour could have
5038 caused a number of strange effects, including loss of data in subsequent
5039 header lines, and spool format errors.
5041 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5042 client host had authenticated. This control can now be exercised by an ACL
5043 for more flexibility.
5045 69. By default, callouts do not happen when testing with -bh. There is now a
5046 variant, -bhc, which does actually run the callout code, including
5047 consulting and updating the callout cache.
5049 70. Added support for saslauthd authentication, courtesy of Alexander
5052 71. If statvfs() failed on the spool or log directories while checking their
5053 size for availability, Exim confusingly gave the error "space shortage".
5054 Furthermore, in debugging mode it crashed with a floating point exception.
5055 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5056 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5057 serious problem, Exim now writes to the main and panic logs when this
5058 happens, with details of the failure. It then refuses to accept the
5059 incoming message, giving the message "spool directory problem" or "log
5060 directory problem" with a 421 code for SMTP messages.
5062 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5063 0, 1, and 2 exist, because some OS complain for execs without them (see
5064 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5065 descriptors. However, the code omitted to check that the open succeeded,
5066 causing mysterious errors if for some reason the permissions on /dev/null
5067 got screwed. Now Exim writes a message to the main and panic logs, and
5068 bombs out if it can't open /dev/null.
5070 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5071 interact so that it is all more flexible. It is supposed to remain
5072 backwards compatible. Also added extra_local_interfaces.
5074 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5075 to bomb out with an assertion failure - to the client this appears as a
5076 connection drop. This problem occurs in the part of the code that was taken
5077 from the Samba project. Fortunately, the assertion is in a very simple
5078 function, so I have fixed this by reproducing the function inline in the
5079 one place where it is called, and arranging for authentication to fail
5080 instead of killing the process with assert().
5082 75. The SPA client code was not working when the server requested OEM rather
5083 than Unicode encoding.
5085 76. Added code to make require_files with a specific uid setting more usable in
5086 the case where statting the file as root fails - usually a non-root-mounted
5087 NFS file system. When this happens and the failure is EACCES, Exim now
5088 forks a subprocess and does the per-uid checking as the relevant uid.
5090 77. Added process_log_path.
5092 78. If log_file_path was not explicitly set, a setting of check_log_space or
5093 check_log_inodes was ignored.
5095 79. If a space check for the spool or log partitions fails, the incident is now
5096 logged. Of course, in the latter case the data may get lost...
5098 80. Added the %p formatting code to string_format() so that it can be used to
5099 print addresses in debug_print(). Adjusted all the address printing in the
5100 debugging in store.c to use %p rather than %d.
5102 81. There was a concern that a line of code in smtp_in.c could overflow a
5103 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5104 initially expressed, the concern was not well-founded, because trailing
5105 spaces are removed early. However, if the trailing spaces were followed by
5106 a NULL, they did not get removed, so the overflow was possible. Two fixes
5109 (a) I re-wrote the offending code in a cleaner fashion.
5110 (b) If an incoming SMTP command contains a NULL character, it is rejected
5113 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5114 runs initgroups(), so that if the Exim user is in any additional groups,
5115 they will be used during message reception.
5121 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob