1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PP/01 Date: & Message-Id: revert to normally being appended to a message,
10 only prepend for the Resent-* case. Fixes regression introduced in
11 Exim 4.70 by NM/22 for Bugzilla 607.
13 PP/02 Include check_rfc2047_length in configure.default because we're seeing
14 increasing numbers of administrators be bitten by this.
16 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
18 PP/03 Bugzilla 994: added openssl_options main configuration option.
20 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
22 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
24 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
26 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
27 without a peer certificate, leading to a segfault because of an
28 assumption that peers always have certificates. Be a little more
29 paranoid. Problem reported by Martin Tscholak.
31 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
32 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
33 NB: ClamAV planning to remove STREAM in "middle of 2010".
34 CL also introduces -bmalware, various -d+acl logging additions and
35 more caution in buffer sizes.
37 PP/09 Implemented reverse_ip expansion operator.
39 PP/10 Bugzilla 937: provide a "debug" ACL control.
41 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
43 PP/12 Bugzilla 973: Implement --version.
45 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
47 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
49 PP/15 Bugzilla 816: support multiple condition rules on Routers.
51 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
52 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
53 ignore trailing whitespace.
55 JJ/02 prevent non-panic DKIM error from being sent to paniclog
57 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
60 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
61 Notification from Dr Andrew Aitchison.
63 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
64 ExtendedDetectionInfo response format.
65 Notification from John Horne.
67 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
70 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
71 XSL and documented dependency on system catalogs, with examples of how
78 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
79 $data_path, and $header_path variables; fixed documentation bugs and
82 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
83 exipick to access non-standard spools, including the "frozen" queue
86 NM/01 Bugzilla 965: Support mysql stored procedures.
87 Patch from Alain Williams
89 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
91 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
92 Patch from Andreas Metzler
94 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
95 Patch from Kirill Miazine
97 NM/05 Bugzilla 671: Added umask to procmail example.
99 JJ/03 installed exipick 20100323.0, fixing doc bug
101 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
102 directory. Notification and patch from Dan Rosenberg.
104 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
106 TK/02 Improve log output when DKIM signing operation fails.
108 MH/01 Treat the transport option dkim_domain as a colon separated
109 list, not as a single string, and sign the message with each element,
110 omitting multiple occurences of the same signer.
112 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
113 Bugzilla 985, 986. Patch by Simon Arlott
115 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
116 Patch by Simon Arlott
118 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
119 MBX locking. Notification from Dan Rosenberg.
125 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
127 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
129 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
131 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
133 NM/04 Bugzilla 915: Flag broken perl installation during build.
139 TK/01 Added patch by Johannes Berg that expands the main option
140 "spamd_address" if it starts with a dollar sign.
142 TK/02 Write list of recipients to X-Envelope-Sender header when building
143 the mbox-format spool file for content scanning (suggested by Jakob
146 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
147 (http://www.dcc-servers.net/) support via dccifd. Activated by
148 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
150 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
151 by Mark Daniel Reidel <mr@df.eu>.
153 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
154 When building exim an external PCRE library is now needed -
155 PCRE is a system library on the majority of modern systems.
156 See entry on PCRE_LIBS in EDITME file.
158 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
159 conversation. Added nologin parameter to request.
160 Patch contributed by Kirill Miazine.
162 TF/01 Do not log submission mode rewrites if they do not change the address.
164 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
166 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
167 log files in place. Contributed by Roberto Lima.
169 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
171 TF/03 Bugzilla 615: When checking the local_parts router precondition
172 after a local_part_suffix or local_part_prefix option, Exim now
173 does not use the address's named list lookup cache, since this
174 contains cached lookups for the whole local part.
176 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
177 Robert Millan. Documentation is in experimental-spec.txt.
179 TF/04 Bugzilla 668: Fix parallel build (make -j).
181 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
183 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
184 Patch provided by Jan Srzednicki.
186 TF/05 Leading white space used to be stripped from $spam_report which
187 wrecked the formatting. Now it is preserved.
189 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
190 that they are available at delivery time.
192 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
194 TF/08 TLS error reporting now respects the incoming_interface and
195 incoming_port log selectors.
197 TF/09 Produce a more useful error message if an SMTP transport's hosts
198 setting expands to an empty string.
200 NM/06 Bugzilla 744: EXPN did not work under TLS.
201 Patch provided by Phil Pennock.
203 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
204 Patch provided by Richard Godbee.
206 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
207 acl_smtp_notquit, added index entry.
209 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
210 Patch provided by Eugene Bujak.
212 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
213 accept(). Patch provided by Maxim Dounin.
215 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
216 Patch provided by Phil Pennock.
218 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
220 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
221 Patch provided by Brad "anomie" Jorsch.
223 NM/14 Bugzilla 622: Added timeout setting to transport filter.
224 Patch provided by Dean Brooks.
226 TK/05 Add native DKIM support (does not depend on external libraries).
228 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
229 Patch provided by Graeme Fowler.
231 NM/16 Bugzilla 851: Documentation example syntax fix.
233 NM/17 Changed NOTICE file to remove references to embedded PCRE.
235 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
238 NM/19 Bugzilla 745: TLS version reporting.
239 Patch provided by Phil Pennock.
241 NM/20 Bugzilla 167: bool: condition support.
242 Patch provided by Phil Pennock.
244 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
245 clients. Patch provided by Phil Pennock.
247 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
248 Patch provided by Brad "anomie" Jorsch.
250 NM/23 Bugzilla 687: Fix misparses in eximstats.
251 Patch provided by Heiko Schlittermann.
253 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
254 Patch provided by Heiko Schlittermann.
256 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
257 plus update to original patch.
259 NM/26 Bugzilla 799: Documentation correction for ratelimit.
261 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
262 Patch provided by David Brownlee.
264 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
266 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
268 NM/30 Bugzilla 888: TLS documentation bugfixes.
270 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
272 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
273 Unlike the original bugzilla I have changed all shell scripts in src tree.
275 NM/33 Bugzilla 898: Transport filter timeout fix.
276 Patch by Todd Rinaldo.
278 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
279 Patch by Serge Demonchaux.
281 NM/35 Bugzilla 39: Base64 decode bug fixes.
282 Patch by Jakob Hirsch.
284 NM/36 Bugzilla 909: Correct connect() call in dcc code.
286 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
288 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
290 NM/39 Bugzilla 911: Fixed MakeLinks build script.
296 TK/01 Add preliminary DKIM support. Currently requires a forked version of
297 ALT-N's libdkim that I have put here:
298 http://duncanthrax.net/exim-experimental/
300 Note to Michael Haardt: I had to rename some vars in sieve.c. They
301 were called 'true' and it seems that C99 defines that as a reserved
302 keyword to be used with 'bool' variable types. That means you could
303 not include C99-style headers which use bools without triggering
304 build errors in sieve.c.
306 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
307 as mailq or other aliases. Changed the --help handling significantly
308 to do whats expected. exim_usage() emits usage/help information.
310 SC/01 Added the -bylocaldomain option to eximstats.
312 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
314 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
316 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
322 PH/01 Another patch from the Sieve maintainer.
324 PH/02 When an IPv6 address is converted to a string for single-key lookup
325 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
326 dots are used instead of colons so that keys in lsearch files need not
327 contain colons. This was done some time before quoting was made available
328 in lsearch files. However, iplsearch files do require colons in IPv6 keys
329 (notated using the quote facility) so as to distinguish them from IPv4
330 keys. This meant that lookups for IP addresses in host lists did not work
331 for iplsearch lookups.
333 This has been fixed by arranging for IPv6 addresses to be expressed with
334 colons if the lookup type is iplsearch. This is not incompatible, because
335 previously such lookups could never work.
337 The situation is now rather anomolous, since one *can* have colons in
338 ordinary lsearch keys. However, making the change in all cases is
339 incompatible and would probably break a number of configurations.
341 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
344 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
345 conversion specification without a maximum field width, thereby enabling
346 a rogue spamd server to cause a buffer overflow. While nobody in their
347 right mind would setup Exim to query an untrusted spamd server, an
348 attacker that gains access to a server running spamd could potentially
349 exploit this vulnerability to run arbitrary code as the Exim user.
351 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
352 $primary_hostname instead of what libspf2 thinks the hosts name is.
354 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
355 a directory entry by the name of the lookup key. Previously, if a
356 symlink pointed to a non-existing file or a file in a directory that
357 Exim lacked permissions to read, a lookup for a key matching that
358 symlink would fail. Now it is enough that a matching directory entry
359 exists, symlink or not. (Bugzilla 503.)
361 PH/03 The body_linecount and body_zerocount variables are now exported in the
364 PH/04 Added the $dnslist_matched variable.
366 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
367 This means they are set thereafter only if the connection becomes
370 PH/06 Added the client_condition to authenticators so that some can be skipped
371 by clients under certain conditions.
373 PH/07 The error message for a badly-placed control=no_multiline_responses left
374 "_responses" off the end of the name.
376 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
378 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
379 (without spaces) instead of just copying the configuration text.
381 PH/10 Added the /noupdate option to the ratelimit ACL condition.
383 PH/11 Added $max_received_linelength.
385 PH/12 Added +ignore_defer and +include_defer to host lists.
387 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
388 way in which PCRE > 7.0 is built.
390 PH/14 Implemented queue_only_load_latch.
392 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
393 MAIL command. The effect was to mangle the value on 64-bit systems.
395 PH/16 Another patch from the Sieve maintainer.
397 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
399 PH/18 If a system quota error occurred while trying to create the file for
400 a maildir delivery, the message "Mailbox is full" was not appended to the
401 bounce if the delivery eventually timed out. Change 4.67/27 below applied
402 only to a quota excession during the actual writing of the file.
404 PH/19 It seems that peer DN values may contain newlines (and other non-printing
405 characters?) which causes problems in log lines. The DN values are now
406 passed through string_printing() before being added to log lines.
408 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
409 and InterBase are left for another time.)
411 PH/21 Added message_body_newlines option.
413 PH/22 Guard against possible overflow in moan_check_errorcopy().
415 PH/23 POSIX allows open() to be a macro; guard against that.
417 PH/24 If the recipient of an error message contained an @ in the local part
418 (suitably quoted, of course), incorrect values were put in $domain and
419 $local_part during the evaluation of errors_copy.
425 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
426 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
429 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
430 issue a MAIL command.
432 PH/02 In an ACL statement such as
434 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
436 if a client was not listed at all, or was listed with a value other than
437 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
438 the condition was not true (as it should be), so access was not denied.
439 The bug was that the ! inversion was incorrectly passed on to the second
440 item. This has been fixed.
442 PH/03 Added additional dnslists conditions == and =& which are different from
443 = and & when the dns lookup returns more than one IP address.
445 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
446 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
448 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
449 FSYNC, which compiles an option called disable_fsync that allows for
450 bypassing fsync(). The documentation is heavily laced with warnings.
452 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
454 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
455 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
456 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
457 including adding "make clean"; (3) Added -fPIC when compiling the test
458 dynamically loaded module, to get rid of a warning.
460 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
461 message fails, move_frozen_messages = true and ignore_bounce_errors_after
462 = 0s. The bug is otherwise harmless.
464 PH/07 There was a bug in the dovecot authenticator such that the value of
465 $auth1 could be overwritten, and so not correctly preserved, after a
466 successful authentication. This usually meant that the value preserved by
467 the server_setid option was incorrect.
469 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
471 PH/09 Installed PCRE release 7.0.
473 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
474 run for batched SMTP input. It is now run at the start of every message
475 in the batch. While fixing this I discovered that the process information
476 (output by running exiwhat) was not always getting set for -bs and -bS
477 input. This is fixed, and it now also says "batched" for BSMTP.
479 PH/11 Added control=no_pipelining.
481 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
482 patch, slightly modified), and move the expansion of helo_data till after
483 the connection is made in the smtp transport (so it can use these
486 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
488 PH/14 Added log_selector = +pid.
490 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
492 PH/16 Add ${if forany and ${if forall.
494 PH/17 Added dsn_from option to vary the From: line in DSNs.
496 PH/18 Flush SMTP output before performing a callout, unless control =
497 no_callout_flush is set.
499 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
500 was true (the default) a successful delivery failed to delete the retry
501 item, thus causing premature timeout of the address. The bug is now
504 PH/20 Added hosts_avoid_pipelining to the smtp transport.
506 PH/21 Long custom messages for fakedefer and fakereject are now split up
507 into multiline reponses in the same way that messages for "deny" and
508 other ACL rejections are.
510 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
511 with slight modification.
513 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
514 draft, changing the syntax and factoring some duplicate code".
516 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
517 for deliveries of the second and subsequent messages over the same SMTP
520 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
521 ${reduce, with only minor "tidies".
523 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
525 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
526 expansion side effects.
528 PH/27 When a message times out after an over-quota error from an Exim-imposed
529 quota, the bounce message says "mailbox is full". This message was not
530 being given when it was a system quota that was exceeded. It now should
533 MH/03 Made $recipients available in local_scan(). local_scan() already has
534 better access to the recipient list through recipients_list[], but
535 $recipients can be useful in postmaster-provided expansion strings.
537 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
538 in the case of a MAIL command with additional options following the
539 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
540 were accidentally chopped off.
542 PH/29 SMTP synchronization checks are implemented when a command is read -
543 there is a check that no more input is waiting when there shouldn't be
544 any. However, for some commands, a delay in an ACL can mean that it is
545 some time before the response is written. In this time, more input might
546 arrive, invalidly. So now there are extra checks after an ACL has run for
547 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
548 pipelining has not been advertised.
550 PH/30 MH's patch to allow iscntrl() characters to be list separators.
552 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
553 returned in the SMTP response when smtp_return_error_details was false.
556 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
557 instead of the C I/O that was originally supplied, because problems were
560 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
561 Exim which did not show up earlier: it was assuming that a call to
562 SSL_CTX_set_info_callback() might give an error value. In fact, there is
563 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
564 was a macro that became an assignment, so it seemed to work. This has
565 changed to a proper function call with a void return, hence the compile
566 error. Exim's code has been fixed.
568 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
571 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
573 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
575 PH/37 If a message is not accepted after it has had an id assigned (e.g.
576 because it turns out to be too big or there is a timeout) there is no
577 "Completed" line in the log. When some messages of this type were
578 selected by exigrep, they were listed as "not completed". Others were
579 picked up by some special patterns. I have improved the selection
580 criteria to be more general.
582 PH/38 The host_find_failed option in the manualroute router can now be set
583 to "ignore", to completely ignore a host whose IP address cannot be
584 found. If all hosts are ignored, the behaviour is controlled by the new
585 host_all_ignored option.
587 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
588 homing or because of multiple MX records with /mx) generated more than
589 one IP address, and the following item turned out to be the local host,
590 all the secondary addresses of the first item were incorrectly removed
591 from the list, along with the local host and any following hosts (which
592 is what is supposed to happen).
594 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
595 whoever called Exim into the -H file. In the case of the daemon it was
596 behaving confusingly. When first started, it used values for whoever
597 started the daemon, but after a SIGHUP it used the Exim user (because it
598 calls itself on a restart). I have changed the code so that it now always
601 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
602 message are rejected with the same error (e.g. no authentication or bad
603 sender address), and a DATA command is nevertheless sent (as can happen
604 with PIPELINING or a stupid MUA), the error message that was given to the
605 RCPT commands is included in the rejection of the DATA command. This is
606 intended to be helpful for MUAs that show only the final error to their
609 PH/42 Another patch from the Sieve maintainer.
611 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
612 Eximstats - Fixed some broken HTML links and added missing column headers
614 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
615 columns for Rejects, Temp Rejects, Ham, and Spam rows.
617 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
619 PH/43 Yet another patch from the Sieve maintainer.
621 PH/44 I found a way to check for a TCP/IP connection going away before sending
622 the response to the final '.' that terminates a message, but only in the
623 case where the client has not sent further data following the '.'
624 (unfortunately, this is allowed). However, in many cases there won't be
625 any further data because there won't be any more messages to send. A call
626 to select() can be used: if it shows that the input is "ready", there is
627 either input waiting, or the socket has been closed. An attempt to read
628 the next input character can distinguish the two cases. Previously, Exim
629 would have sent an OK response which the client would never have see.
630 This could lead to message repetition. This fix should cure that, at
631 least in a lot of common cases.
633 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
634 advertised in response to EHLO.
640 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
641 fixed by 4.65/MH/01 (is this a record?) are fixed:
643 (i) An empty string was always treated as zero by the numeric comparison
644 operators. This behaviour has been restored.
646 (ii) It is documented that the numeric comparison operators always treat
647 their arguments as decimal numbers. This was broken in that numbers
648 starting with 0 were being interpreted as octal.
650 While fixing these problems I realized that there was another issue that
651 hadn't been noticed. Values of message_size_limit (both the global option
652 and the transport option) were treated as octal if they started with 0.
653 The documentation was vague. These values are now always treated as
654 decimal, and I will make that clear in the documentation.
660 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
661 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
664 MH/01 Don't check that the operands of numeric comparison operators are
665 integers when their expansion is in "skipping" mode (fixes bug
666 introduced by 4.64-PH/07).
668 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
669 child addresses, Exim now panics and dies. Previously, because the count
670 is held in a short int, deliveries were likely to be lost. As such a
671 large number of recipients for a single message is ridiculous
672 (performance will be very, very poor), I have chosen to impose a limit
673 rather than extend the field.
679 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
680 leftover -K file (the existence of which was triggered by #402).
681 While we were at it, introduced process PID as part of the -K
682 filename. This should rule out race conditions when creating
685 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
686 processing considerably. Previous code took too long for large mails,
687 triggering a timeout which in turn triggers #401.
689 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
690 in the DK code in transports.c. sendfile() is not really portable,
691 hence the _LINUX specificness.
693 TF/01 In the add_headers option to the mail command in an Exim filter,
694 there was a bug that Exim would claim a syntax error in any
695 header after the first one which had an odd number of characters
698 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
699 callout verification, Exim cached a "reject" for the entire domain. This
700 is correct for most verifications, but it is not correct for a recipient
701 verification with use_sender or use_postmaster set, because in that case
702 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
703 case of MAIL FROM:<> rejection from other early rejections (e.g.
704 rejection of HELO). When verifying a recipient using a non-null MAIL
705 address, the cache is ignored if it shows MAIL FROM:<> rejection.
706 Whatever the result of the callout, the value of the domain cache is
707 left unchanged (for any other kind of callout, getting as far as trying
708 RCPT means that the domain itself is ok).
710 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
713 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
714 manifest itself as EPIPE rather than ECONNECT. When tidying away a
715 session, the daemon ignores ECONNECT errors and logs others; it now
716 ignores EPIPE as well.
718 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
719 (quoted-printable decoding).
721 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
722 later the small subsequent patch to fix an introduced bug.
724 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
726 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
728 PH/08 An error is now given if message_size_limit is specified negative.
730 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
731 to be given (somewhat) arbitrary names.
733 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
736 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
737 miscellaneous code fixes
739 PH/10 Added the log_reject_target ACL modifier to specify where to log
742 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
743 hostname. This is wrong, because it relates to the incoming message (and
744 probably the interface on which it is arriving) and not to the outgoing
745 callout (which could be using a different interface). This has been
746 changed to use the value of the helo_data option from the smtp transport
747 instead - this is what is used when a message is actually being sent. If
748 there is no remote transport (possible with a router that sets up host
749 addresses), $smtp_active_hostname is used.
751 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
752 tweaks were necessary in order to get it to work (see also 21 below):
753 (a) The code assumed that strncpy() returns a negative number on buffer
754 overflow, which isn't the case. Replaced with Exim's string_format()
756 (b) There were several signed/unsigned issues. I just did the minimum
757 hacking in of casts. There is scope for a larger refactoring.
758 (c) The code used strcasecmp() which is not a standard C function.
759 Replaced with Exim's strcmpic() function.
760 (d) The code set only $1; it now sets $auth1 as well.
761 (e) A simple test gave the error "authentication client didn't specify
762 service in request". It would seem that Dovecot has changed its
763 interface. Fortunately there's a specification; I followed it and
764 changed what the client sends and it appears to be working now.
766 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
769 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
770 address A is aliased to B and C, where B exists and C does not. Without
771 -v the output is "A verified" because verification stops after a
772 successful redirection if more than one address is generated. However,
773 with -v the child addresses are also verified. Exim was outputting "A
774 failed to verify" and then showing the successful verification for C,
775 with its parentage. It now outputs "B failed to verify", showing B's
776 parentage before showing the successful verification of C.
778 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
779 look up a TXT record in a specific list after matching in a combined
782 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
783 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
784 they consult the DNS. I had assumed they would set it the way they
785 wanted; and indeed my experiments on Linux seem to show that in some
786 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
787 To be on the safe side, however, I have now made the interface to
788 host_find_byname() similar to host_find_bydns(), with an argument
789 containing the DNS resolver options. The host_find_byname() function now
790 sets these options at its start, just as host_find_bydns() does. The smtp
791 transport options dns_qualify_single and dns_search_parents are passed to
792 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
793 of host_find_byname() use the default settings of RES_DEFNAMES
794 (qualify_single) but not RES_DNSRCH (search_parents).
796 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
797 spool_read_header() do less string testing, by means of a preliminary
798 switch on the second character of optional "-foo" lines. (This is
799 overdue, caused by the large number of possibilities that now exist.
800 Originally there were few.) While I was there, I also converted the
801 str(n)cmp tests so they don't re-test the leading "-" and the first
802 character, in the hope this might squeeze out yet more improvement.
804 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
805 flag allowing group syntax was set by the header_syntax check but not
806 turned off, possible causing trouble later; (2) The flag was not being
807 set at all for the header_verify test, causing "group"-style headers to
808 be rejected. I have now set it in this case, and also caused header_
809 verify to ignore an empty address taken from a group. While doing this, I
810 came across some other cases where the code for allowing group syntax
811 while scanning a header line wasn't quite right (mostly, not resetting
812 the flag correctly in the right place). These bugs could have caused
813 trouble for malformed header lines. I hope it is now all correct.
815 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
816 with the "reply" argument non-NULL. The code, however (which originally
817 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
818 but it didn't always do it. This confused somebody who was copying the
819 code for some other use. I have removed all the tests.
821 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
822 feature that was used to support insecure browsers during the U.S. crypto
823 embargo. It requires special client support, and Exim is probably the
824 only MTA that supported it -- and would never use it because real RSA is
825 always available. This code has been removed, because it had the bad
826 effect of slowing Exim down by computing (never used) parameters for the
827 RSA_EXPORT functionality.
829 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
830 authenticator to fail if there's a tab character in the incoming data
831 (there should never be unless someone is messing about, as it's supposed
832 to be base64-encoded). Also added, on Timo's advice, the "secured" option
833 if the connection is using TLS or if the remote IP is the same as the
834 local IP, and the "valid-client-cert option" if a client certificate has
837 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
838 authenticators. This can be used for authorization after authentication
839 succeeds. (In the case of plaintext, it servers for both authentication
842 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
843 if any retry times were supplied.
845 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
846 connection, where there is no client IP address to check. In this
847 situation, the verify now always succeeds.
849 PH/25 Applied John Jetmore's -Mset patch.
851 PH/26 Added -bem to be like -Mset, but loading a message from a file.
853 PH/27 In a string expansion for a processed (not raw) header when multiple
854 headers of the same name were present, leading whitespace was being
855 removed from all of them, but trailing whitespace was being removed only
856 from the last one. Now trailing whitespace is removed from each header
857 before concatenation. Completely empty headers in a concatenation (as
860 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
861 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
863 PH/29 [Removed. This was a change that I later backed out, and forgot to
864 correct the ChangeLog entry (that I had efficiently created) before
865 committing the later change.]
867 PH/30 Exim was sometimes attempting to deliver messages that had suffered
868 address errors (4xx response to RCPT) over the same connection as other
869 messages routed to the same hosts. Such deliveries are always "forced",
870 so retry times are not inspected. This resulted in far too many retries
871 for the affected addresses. The effect occurred only when there were more
872 hosts than the hosts_max_try setting in the smtp transport when it had
873 the 4xx errors. Those hosts that it had tried were not added to the list
874 of hosts for which the message was waiting, so if all were tried, there
875 was no problem. Two fixes have been applied:
877 (i) If there are any address or message errors in an SMTP delivery, none
878 of the hosts (tried or untried) are now added to the list of hosts
879 for which the message is waiting, so the message should not be a
880 candidate for sending over the same connection that was used for a
881 successful delivery of some other message. This seems entirely
882 reasonable: after all the message is NOT "waiting for some host".
883 This is so "obvious" that I'm not sure why it wasn't done
884 previously. Hope I haven't missed anything, but it can't do any
885 harm, as the worst effect is to miss an optimization.
887 (ii) If, despite (i), such a delivery is accidentally attempted, the
888 routing retry time is respected, so at least it doesn't keep
889 hammering the server.
891 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
892 in ${readsocket because some servers need this prod.
894 PH/32 Added some extra debug output when updating a wait-xxx database.
896 PH/33 The hint "could be header name not terminated by colon", which has been
897 given for certain expansion errors for a long time, was not being given
898 for the ${if def:h_colon_omitted{... case.
900 PH/34 The spec says: "With one important exception, whenever a domain list is
901 being scanned, $domain contains the subject domain." There was at least
902 one case where this was not true.
904 PH/35 The error "getsockname() failed: connection reset by peer" was being
905 written to the panic log as well as the main log, but it isn't really
906 panic-worthy as it just means the connection died rather early on. I have
907 removed the panic log writing for the ECONNRESET error when getsockname()
910 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
911 runs only) independently of the message's sender address. This meant
912 that, if the 4xx error was in fact related to the sender, a different
913 message to the same recipient with a different sender could confuse
914 things. In particualar, this can happen when sending to a greylisting
915 server, but other circumstances could also provoke similar problems.
916 I have changed the default so that the retry time for these errors is now
917 based a combination of the sender and recipient addresses. This change
918 can be overridden by setting address_retry_include_sender=false in the
921 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
922 remote server are returned as part of bounce messages. This was not
923 happening for LMTP over a pipe (the lmtp transport), but now it is the
924 same for both kinds of LMTP.
926 PH/38 Despite being documented as not happening, Exim was rewriting addresses
927 in header lines that were in fact CNAMEs. This is no longer the case.
929 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
930 and queue runs started by the daemon processed all messages. This has
931 been fixed so that -R and -S can now usefully be given with -q<time>.
933 PH/40 Import PCRE release 6.7 (fixes some bugs).
935 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
937 PH/42 Give an error if -q is specified more than once.
939 PH/43 Renamed the variables $interface_address and $interface_port as
940 $received_ip_address and $received_port, to make it clear that these
941 values apply to message reception, and not to the outgoing interface when
942 a message is delivered. (The old names remain recognized, of course.)
944 PH/44 There was no timeout on the connect() call when using a Unix domain
945 socket in the ${readsocket expansion. There now is.
947 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
948 be meaningful with "accept".
950 SC/01 Eximstats V1.43
951 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
953 SC/02 Eximstats V1.44
954 Use a glob alias rather than an array ref in the generated
955 parser. This improves both readability and performance.
957 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
958 Collect SpamAssassin and rejection statistics.
959 Don't display local sender or destination tables unless
960 there is data to show.
961 Added average volumes into the top table text output.
963 SC/04 Eximstats V1.46
964 Collect data on the number of addresses (recipients)
965 as well as the number of messages.
967 SC/05 Eximstats V1.47
968 Added 'Message too big' to the list of mail rejection
969 reasons (thanks to Marco Gaiarin).
971 SC/06 Eximstats V1.48
972 Mainlog lines which have GMT offsets and are too short to
973 have a flag are now skipped.
975 SC/07 Eximstats V1.49 (Alain Williams)
976 Added the -emptyok flag.
978 SC/08 Eximstats V1.50
979 Fixes for obtaining the IP address from reject messages.
981 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
982 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
983 whitesspace changes from 4.64-PH/27
985 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
988 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
989 are found, allow negative numbers in numeric criteria)
991 JJ/06 exipick.20061117.2, added new $message_body_missing variable
993 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
994 to match changes made in 4.64-PH/43
996 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
998 PH/47 Put in an explicit test for a DNS lookup of an address record where the
999 "domain" is actually an IP address, and force a failure. This locks out
1000 those revolvers/nameservers that support "A-for-A" lookups, in
1001 contravention of the specifications.
1003 PH/48 When a host name was looked up from an IP address, and the subsequent
1004 forward lookup of the name timed out, the host name was left in
1005 $sender_host_name, contrary to the specification.
1007 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1008 restricted to single-key lookups, Exim was not diagnosing an error if
1009 * or *@ was used with a query-style lookup.
1011 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1013 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1014 long ago, but noone interested enough thought of it. Let's just say that
1015 the "1.1" means that there are some new functions that weren't there at
1016 some point in the past.
1018 PH/51 Error processing for expansion failure of helo_data from an smtp
1019 transport during callout processing was broken.
1021 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1022 tested/used via the -bh/-bhc/-bs options.
1024 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1025 bug, fixed in subsequent PCRE releases).
1027 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1028 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1030 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1036 SC/01 Use a glob alias rather than an array ref in eximstats generated
1037 parser. This improves both readability and performance.
1039 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1040 Don't display local sender or destination tables in eximstats unless
1041 there is data to show.
1042 Added average volumes into the eximstats top table text output.
1044 SC/03 Collect data on the number of addresses (recipients) as well
1045 as the number of messages in eximstats.
1047 TF/01 Correct an error in the documentation for the redirect router. Exim
1048 does (usually) call initgroups() when daemonizing.
1050 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1051 with consistent privilege compared to when running as a daemon.
1053 TF/03 Note in the spec that $authenticated_id is not set for local
1054 submissions from trusted users.
1056 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1057 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1059 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1060 by adding some example configuration directives to the default
1061 configuration file. A little bit of work is required to uncomment the
1062 directives and define how usernames and passwords are checked, but
1063 there is now a framework to start from.
1065 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1066 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1067 without this. I don't know how relevant this is to other LDAP libraries.
1069 PH/02 Add the verb name to the "unknown ACL verb" error.
1071 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1073 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1075 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1076 directory not expanded when it should be if an expanded home directory
1077 was set for the address (which is overridden by the transport).
1079 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1082 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1083 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1084 because it is too late at that time, and has no effect.
1086 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1087 security issue with \' (bugzilla #107). I could not use the
1088 PQescapeStringConn() function, because it needs a PGconn value as one of
1091 PH/08 When testing addresses using -bt, indicate those final addresses that
1092 are duplicates that would not cause an additional delivery. At least one
1093 person was confused, thinking that -bt output corresponded to deliveries.
1094 (Suppressing duplicates isn't a good idea as you lose the information
1095 about possibly different redirections that led to the duplicates.)
1097 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1098 systems where poll() doesn't work, in particular OS X.
1100 PH/10 Added more information to debugging output for retry time not reached.
1102 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1103 operations in malware.c.
1105 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1108 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1109 syslog_duplication set false, log lines that would normally be written
1110 both the the main log and to the reject log were not written to syslog at
1113 PH/14 In the default configuration, change the use of "message" in ACL warn
1114 statements to "add_header".
1116 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1117 not followed by a command (e.g. "seen endif").
1119 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1120 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1123 PH/17 Added extra conditions to the default value of delay_warning_condition
1127 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1128 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1129 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1132 The Auto-Submitted: and various List- headers are standardised, whereas I
1133 don't think Precedence: ever was.
1135 PH/18 Refactored debugging code in route_finduser() to show more information,
1136 in particular, the error code if getpwnam() issues one.
1138 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1139 This is apparently needed in addition to the PH/07 change above to avoid
1140 any possible encoding problems.
1142 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1143 but not after initializing Perl.
1145 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1146 output them only if debugging. By default they are written stderr,
1147 apparently, which is not desirable.
1149 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1152 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1155 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1157 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1158 authenticated or an ident call has been made. Suppress the default
1159 values for $authenticated_id and $authenticated_sender (but permit -oMai
1160 and -oMas) when testing with -bh.
1162 PH/24 Re-jigged the order of the tests in the default configuration so that the
1163 tests for valid domains and recipients precede the DNS black list and CSA
1164 tests, on the grounds that those ones are more expensive.
1166 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1167 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1168 This bug exists in every version of Exim that I still have, right back to
1171 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1172 However, an attempt to turn on case-sensitivity in a regex key by
1173 including (?-i) didn't work because the subject string was already
1174 lowercased, and the effects were non-intuitive. It turns out that a
1175 one-line patch can be used to allow (?-i) to work as expected.
1181 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1182 other effects) broke the use of negated acl sub-conditions.
1184 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1187 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1188 "Deny" causes Exim to reject the incoming connection with a 554 error.
1189 Unfortunately, if there is a major crisis, such as a disk failure,
1190 tcp-wrappers gives "deny", whereas what one would like would be some
1191 kind of temporary error. A kludge has been added to help with this.
1192 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1193 554 error is used if errno is still zero or contains ENOENT (which occurs
1194 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1197 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1199 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1200 errors. Otherwise a message that provokes a temporary error (when other
1201 messages do not) can cause a whole host to time out.
1203 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1204 addresses were routed directly to files or pipes from a redirect router.
1205 File deliveries just didn't batch; pipe deliveries might have suffered
1208 PH/06 A failure to get a lock for a hints database would erroneously always say
1209 "Failed to get write lock", even when it was really a read lock.
1211 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1212 of 0600. This has been changed to use the value of the lockfile_mode
1213 option (which defaults to 0600).
1215 PH/08 Applied small patch from the Sieve maintainer.
1217 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1218 folder from quota calculations, a direct delivery into this folder messed
1219 up the contents of the maildirsize file. This was because the regex was
1220 used only to exclude .Trash (or whatever) when the size of the mailbox
1221 was calculated. There was no check that a delivery was happening into an
1222 excluded directory. This bug has been fixed by ignoring all quota
1223 processing for deliveries into excluded directories.
1225 PH/10 Added the maildirfolder_create_regex option to appendfile.
1231 PH/01 The code for finding all the local interface addresses on a FreeBSD
1232 system running IPv6 was broken. This may well have applied to all BSD
1233 systems, as well as to others that have similar system calls. The broken
1234 code found IPv4 interfaces correctly, but gave incorrect values for the
1235 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1236 that it would not match correctly against @[] and not recognize the IPv6
1239 PH/02 The ipliteral router was not recognizing addresses of the form user@
1240 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1242 PH/03 Added disable_ipv6.
1244 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1245 decoded form, because it is most often used to construct To: headers
1246 lines in autoreplies, and the decoded form may well be syntactically
1247 invalid. However, $reply_address has leading white space removed, and all
1248 newlines turned into spaces so that the autoreply transport does not
1251 PH/05 If group was specified without a user on a router, and no group or user
1252 was specified on a transport, the group from the router was ignored.
1254 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1255 for visible compile-time settings that can be used to change these
1256 numbers, for those that want even more. Backwards compatibility with old
1257 spool files has been maintained. However, going back to a previous Exim
1258 release will lost any variables that are in spool files.
1260 PH/07 Two small changes when running in the test harness: increase delay when
1261 passing a TCP/IP connection to a new process, in case the original
1262 process has to generate a bounce, and remove special handling of
1263 127.0.0.2 (sic), which is no longer necessary.
1265 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1266 be the same on different OS.
1268 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1271 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1272 whether --show-vars was specified or not
1274 JJ/02 exipick: Added support for new ACL variable spool format introduced
1277 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1278 syntactically invalid From: or Reply-to: line, and a filter used this to
1279 generate an autoreply, and therefore failed to obtain an address for the
1280 autoreply, Exim could try to deliver to a non-existent relative file
1281 name, causing unrelated and misleading errors. What now happens is that
1282 it logs this as a hard delivery error, but does not attempt to create a
1285 PH/11 The exinext utility has a -C option for testing purposes, but although
1286 the given file was scanned by exinext itself; it wasn't being passed on
1287 when Exim was called.
1289 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1290 an end-of-file indication when reading a command response.
1292 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1293 compiled. In many other places in Exim, IPv6 addresses are always
1294 recognized, so I have changed this. It also means that IPv4 domain
1295 literals of the form [IPV4:n.n.n.n] are now always recognized.
1297 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1298 used if the router is not running as root, for example, when verifying at
1299 ACL time, or when using -bh. The debugging output from this situation was
1300 non-existent - all you got was a failure to exec. I have made two
1303 (a) Failures to set uid/gid, the current directory, or a process leader
1304 in a subprocess such as that created by queryprogram now generate
1305 suitable debugging ouput when -d is set.
1307 (b) The queryprogram router detects when it is not running as root,
1308 outputs suitable debugging information if -d is set, and then runs
1309 the subprocess without attempting to change uid/gid.
1311 PH/15 Minor change to Makefile for building test_host (undocumented testing
1314 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1315 additional section of a DNS packet that returns MX or SRV records.
1316 Instead, it always explicitly searches for A/AAAA records. This avoids
1317 major problems that occur when a DNS server includes only records of one
1318 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1319 fixed another bug: if SRV records were looked up and the corresponding
1320 address records were *not* found in the additional section, the port
1321 values from the SRV records were lost.
1323 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1324 using the correct key (the original address) when searching the retry
1325 rules in order to find which one to use for generating the retry hint.
1327 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1328 adding the default one. Similarly, if it contains a Reply-To: header, the
1329 errors_reply_to option, if set, is not used.
1331 PH/19 When calculating a retry time, Exim used to measure the "time since
1332 failure" by looking at the "first failed" field in the retry record. Now
1333 it does not use this if it is later than than the arrival time of the
1334 message. Instead it uses the arrival time. This makes for better
1335 behaviour in cases where some deliveries succeed, thus re-setting the
1336 "first failed" field. An example is a quota failure for a huge message
1337 when small messages continue to be delivered. Without this change, the
1338 "time since failure" will always be short, possible causing more frequent
1339 delivery attempts for the huge message than are intended.
1340 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1342 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1343 $1, $2, $3) because the numerical variables can be reset during some
1344 expansion items (e.g. "match"), thereby losing the authentication data.
1346 PH/21 Make -bV show the size of off_t variables so that the test suite can
1347 decide whether to run tests for quotas > 2G.
1349 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1350 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1351 filecount value is greater than 2G or if a quota value is greater than 2G
1352 on a system where the size of off_t is not greater than 4, a panic error
1355 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1356 never match. The debug and -bh output now contains an explicit error
1357 message indicating a malformed IPv4 address or mask.
1359 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1360 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1361 PH/23 above applies.
1363 PH/25 Do not write to syslog when running in the test harness. The only
1364 occasion when this arises is a failure to open the main or panic logs
1365 (for which there is an explicit test).
1367 PH/26 Added the /no_tell option to "control=freeze".
1369 PH/27 If a host name lookup failed very early in a connection, for example, if
1370 the IP address matched host_lookup and the reverse lookup yielded a name
1371 that did not have a forward lookup, an error message of the form "no IP
1372 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1373 could be logged. Now it outputs the IP address instead of "NULL".
1375 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1376 allows the sender and the authenticated sender to be set when
1377 submitting a message from within Exim. Since child_open_exim() is
1378 documented for local_scan(), the new function should be too.
1380 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1381 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1382 results in an empty string is now treated as unset.
1384 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1386 PH/31 Added sender_verify_fail logging option.
1388 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1389 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1390 tidied the source and removed it altogether.
1392 PH/33 When a queue run was abandoned because the load average was too high, a
1393 log line was always written; now it is written only if the queue_run log
1394 selector is set. In addition, the log line for abandonment now contains
1395 information about the queue run such as the pid. This is always present
1396 in "start" and "stop" lines but was omitted from the "abandon" line.
1398 PH/34 Omit spaces between a header name and the colon in the error message that
1399 is given when verify = headers_syntax fails (if there are lots of them,
1400 the message gets confusing).
1402 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1403 names, as there are now some PTR records that contain slashes. This check
1404 is only to protect against broken name servers that fall over on strange
1405 characters, so the fact that it applies to all lookups doesn't matter.
1407 PH/36 Now that the new test suite is complete, we can remove some of the
1408 special code in Exim that was needed for the old test suite. For example,
1409 sorting DNS records because real resolvers return them in an arbitrary
1410 order. The new test suite's fake resolver always returns records in the
1413 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1414 bounces) except when queue_only is set, to avoid logging races between
1415 the different processes.
1417 PH/38 Panic-die if .include specifies a non-absolute path.
1419 PH/39 A tweak to the "H" retry rule from its user.
1421 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1422 a label. They prevented compilation on older perls.
1424 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1425 a warning to be raised on newish perls.
1427 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1428 on queue. Changes to match documented behaviour of showing count of
1429 messages matching specified criteria.
1431 PH/40 Changed the default ident timeout from 30s to 5s.
1433 PH/41 Added support for the use of login_cap features, on those BSD systems
1434 that have them, for controlling the resources used by pipe deliveries.
1436 PH/42 The content-scanning code uses fopen() to create files in which to put
1437 message data. Previously it was not paying any attention to the mode of
1438 the files. Exim runs with umask(0) because the rest of the code creates
1439 files with open(), and sets the required mode explicitly. Thus, these
1440 files were ending up world-writeable. This was not a big issue, because,
1441 being within the spool directory, they were not world-accessible. I have
1442 created a function called modefopen, which takes an additional mode
1443 argument. It sets umask(777), creates the file, chmods it to the required
1444 mode, then resets the umask. All the relevant calls to fopen() in the
1445 content scanning code have been changed to use this function.
1447 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1448 to 24 hours. This avoids potential overflow problems when processing G
1449 and H retry rules. I suspect nobody ever tinkers with this value.
1451 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1453 PH/45 When the plaintext authenticator is running as a client, the server's
1454 challenges are checked to ensure they are valid base64 strings. By
1455 default, the authentication attempt is cancelled if an invalid string is
1456 received. Setting client_ignore_invalid_base64 true ignores these errors.
1457 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1458 they are received. Thus, the responses can be made to depend on the
1459 challenges. If an invalid string is ignored, an empty string is placed in
1462 PH/46 Messages that are created by the autoreply transport now contains a
1463 References: header, in accordance with RFCs 2822 and 3834.
1465 PH/47 Added authenticated_sender_force to the smtp transport.
1467 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1469 PH/49 Installed latest patch from the Sieve maintainer.
1471 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1472 was also set, the appendfile transport was unnecessarily scanning a
1473 directory of message files (e.g. for maildir delivery) to find the count
1474 of files (along with the size), even though it did not need this
1475 information. It now does the scan only if it needs to find either the
1476 size of the count of files.
1478 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1480 PH/52 Two bugs concerned with error handling when the smtp transport is
1483 (i) Exim was not creating retry information for temporary errors given
1484 for individual recipients after the DATA command when the smtp transport
1485 was used in LMTP mode. This meant that they could be retried too
1486 frequently, and not timed out correctly.
1488 (ii) Exim was setting the flag that allows error details to be returned
1489 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1490 recipients that were returned after the DATA command.
1492 PH/53 This is related to PH/52, but is more general: for any failing address,
1493 when detailed error information was permitted to be returned to the
1494 sender, but the error was temporary, then after the final timeout, only
1495 "retry timeout exceeded" was returned. Now it returns the full error as
1496 well as "retry timeout exceeded".
1498 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1499 do this, and (what is worse) MTAs that accept it.
1501 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1502 will now be deprecated.
1504 PH/56 New os.c-cygwin from the Cygwin maintainer.
1506 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1507 formats (previously only available in exim formats via -bpr, -bpru,
1508 and -bpra. Now also available in native and exiqgrep formats)
1510 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1511 with very large, slow to parse queues
1513 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1515 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1517 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1518 responses to authentication challenges, though it was showing the
1519 challenges; (ii) I've removed the CR characters from the debug output for
1522 PH/58 Allow for the insertion of a newline as well as a space when a string
1523 is turned into more than one encoded-word during RFC 2047 encoding. The
1524 Sieve code now uses this.
1526 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1527 data_4xx, lost_connection, tls_required.
1529 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1530 message was being sent as an SMTP response.
1532 PH/61 Add -l and -k options to exicyclog.
1534 PH/62 When verifying, if an address was redirected to one new address, so that
1535 verification continued, and the new address failed or deferred after
1536 having set something in $address_data, the value of $address_data was not
1537 passed back to the ACL. This was different to the case when no
1538 redirection occurred. The value is now passed back in both cases.
1540 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1541 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1542 use login_cap.h, so on its own it isn't the distinguishing feature. The
1543 new name refers directly to the setclassresources() function.
1545 PH/65 Added configuration files for NetBSD3.
1547 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1549 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1550 is preferred over IPv4.
1552 PH/68 The bounce_return_message and bounce_return_body options were not being
1553 honoured for bounces generated during the reception of non-SMTP messages.
1554 In particular, this applied to messages rejected by the ACL. This bug has
1555 been fixed. However, if bounce_return_message is true and bounce_return_
1556 body is false, the headers that are returned for a non-SMTP message
1557 include only those that have been read before the error was detected.
1558 (In the case of an ACL rejection, they have all been read.)
1560 PH/69 The HTML version of the specification is now built in a directory called
1561 spec_html instead of spec.html, because the latter looks like a path with
1562 a MIME-type, and this confuses some software.
1564 PH/70 Catch two compiler warnings in sieve.c.
1566 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1567 function verify_get_ident() calls ip_connect() to connect a socket, but
1568 if the "connect()" function timed out, ip_connect() used to close the
1569 socket. However, verify_get_ident() also closes the socket later, and in
1570 between Exim writes to the log, which may get opened at this point. When
1571 the socket was closed in ip_connect(), the log could get the same file
1572 descriptor number as the socket. This naturally causes chaos. The fix is
1573 not to close the socket in ip_connect(); the socket should be closed by
1574 the function that creates it. There was only one place in the code where
1575 this was missing, in the iplookup router, which I don't think anybody now
1576 uses, but I've fixed it anyway.
1578 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1579 well as to direct DNS lookups. Otherwise the handling of names in host
1580 lists is inconsistent and therefore confusing.
1586 PH/01 Two changes to the default runtime configuration:
1588 (1) Move the checks for relay_from_hosts and authenticated clients from
1589 after to before the (commented out) DNS black list checks.
1591 (2) Add control=submission to the relay_from_hosts and authenticated
1592 clients checks, on the grounds that messages accepted by these
1593 statements are most likely to be submissions.
1595 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1597 (1) Generate an error if the third argument for the ${prvs expansion is
1600 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1603 (3) Reset the variables that are obtained from the first argument of
1604 ${prvscheck and used in the second argument before leaving the code,
1605 because their memory is reclaimed, so using them afterwards may do
1608 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1609 one (it's much easier than Tom thought :-).
1611 (5) Because of (4), we can now allow for the use of $prvscheck_result
1612 inside the third argument.
1614 PH/03 For some reason, the default setting of PATH when running a command from
1615 a pipe transport was just "/usr/bin". I have changed it to
1618 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1619 anything to be listed in the output from -bV.
1621 PH/05 When a filter generated an autoreply, the entire To: header line was
1622 quoted in the delivery log line, like this:
1624 => >A.N.Other <ano@some.domain> <original@ddress> ...
1626 This has been changed so that it extracts the operative address. There
1627 may be more than one such address. If so, they are comma-separated, like
1630 => >ano@some.domain,ona@other.domain <original@ddress> ...
1632 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1633 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1634 not being looked up in the rDNS to get a host name, Exim was showing the
1635 IP address twice in Received: lines, even though the IP addresses were
1636 identical. For example:
1638 Received: from [1.2.3.4] (helo=[1.2.3.4])
1640 However, if the real host name was known, it was omitting the HELO data
1641 if it matched the actual IP address. This has been tidied up so that it
1642 doesn't show the same IP address twice.
1644 PH/07 When both +timestamp and +memory debugging was on, the value given by
1645 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1646 called by the debug printing, thereby overwriting the timestamp buffer.
1647 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1649 PH/08 When the original message was included in an autoreply transport, it
1650 always said "this is a copy of the message, including all the headers",
1651 even if body_only or headers_only was set. It now gives an appropriate
1654 PH/09 Applied a patch from the Sieve maintainer which:
1656 o fixes some comments
1657 o adds the (disabled) notify extension core
1658 o adds some debug output for the result of if/elsif tests
1659 o points to the current vacation draft in the documentation
1660 and documents the missing references header update
1664 o fixes a bug in processing the envelope test (when testing
1665 multiple envelope elements, the last element determinted the
1668 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1669 Electronic Mail") by including:
1671 Auto-submitted: auto-generated
1673 in the messages that it generates (bounce messages and others, such as
1674 warnings). In the case of bounce messages for non-SMTP mesages, there was
1675 also a typo: it was using "Auto_submitted" (underscore instead of
1676 hyphen). Since every message generated by Exim is necessarily in response
1677 to another message, thes have all been changed to:
1679 Auto-Submitted: auto-replied
1681 in accordance with these statements in the RFC:
1683 The auto-replied keyword:
1685 - SHOULD be used on messages sent in direct response to another
1686 message by an automatic process,
1688 - MUST NOT be used on manually-generated messages,
1690 - MAY be used on Delivery Status Notifications (DSNs) and Message
1691 Disposition Notifications (MDNs),
1693 - MUST NOT be used on messages generated by automatic or periodic
1694 processes, except for messages which are automatic responses to
1697 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1698 to the default Received: header definition.
1700 PH/12 Added log selector acl_warn_skipped (default on).
1702 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1703 variables because (a) they are in the wrong storage pool and (b) even if
1704 they were copied, it wouldn't work properly because of the caching.
1706 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1707 checking when decoding. Apparently there are clients that generate
1708 overlong encoded strings. Why am I not surprised?
1710 PH/15 If the first argument of "${if match_address" was not empty, but did not
1711 contain an "@" character, Exim crashed. Now it writes a panic log message
1712 and treats the condition as false.
1714 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1716 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1717 extension "envelope-auth". The code is finished and in agreement with
1718 other implementations, but there is no documentation so far and in fact,
1719 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1720 not changing the active code.
1722 Print executed "if" and "elsif" statements when debugging is used. This
1723 helps a great deal to understand what a filter does.
1725 Document more things not specified clearly in RFC3028. I had all this
1726 sorted out, when out of a sudden new issues came to my mind. Oops."
1728 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1731 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1732 canonical form (as documented). However, after a host name lookup from
1733 the IP address, check_host() was doing a simple string comparison with
1734 addresses acquired from the DNS when checking that the found name did
1735 have the original IP as one of its addresses. Since any found IPv6
1736 addresses are likely to be in abbreviated form, the comparison could
1737 fail. Luckily, there already exists a function for doing the comparison
1738 by converting both addresses to binary, so now that is used instead of
1739 the text comparison.
1741 PH/20 There was another similar case to PH/19, when a complete host name was
1742 given in a host list; looking up its IP address could give an abbreviated
1743 form, whereas the current host's name might or might not be abbreviated.
1744 The same fix has been applied.
1750 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1751 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1754 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1755 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1757 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1759 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1760 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1761 Also, for "Auto-Submitted", treat anything other than "no" as
1762 non-personal, in accordance with RFC 3834. (Previously it treated
1763 anything starting "auto-" as non-personal.)
1765 TF/01 The control=submission/name=... option had a problem with syntax
1766 errors if the name included a slash character. The /name= option
1767 now slurps the rest of the string, so it can include any characters
1768 but it must come last in the list of options (after /sender_retain
1771 PH/05 Some modifications to the interface to the fake nameserver for the new
1779 TK/01 Added the "success_on_redirect" address verification option. See
1780 NewStuff for rationale and an example.
1782 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1784 PH/02 Patch to exigrep to allow it to work on syslog lines.
1786 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1787 fread() to skip over the body file's header line, because in Cygwin the
1788 header line is locked and is inaccessible.
1790 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1791 co-exist for some time) to make it clear that it is the Exim ID that is
1792 referenced, not the Message-ID: header line.
1794 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1795 string_format() function, because snprintf() does not exist on all
1798 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1799 ${stat: expansion item.
1801 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1802 protocol synchronization error", to keep the pedants happy.
1804 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1805 well as for IRIX systems, when gcc is being used. See the host.c source
1808 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1810 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1813 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1814 transport and to the smtp transport in LMTP mode.
1816 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1818 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1819 run a filter in a subprocess. This could lead to confusion in subsequent
1820 lookups in the parent process. There should also be a search_tidyup() at
1821 the end of the subprocess.
1823 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1824 only if the host matched helo_try_verify_hosts, which caused the
1825 verification to occur when the EHLO/HELO command was issued. The ACL just
1826 tested the remembered result. Now, if a previous verification attempt has
1827 not happened, "verify = helo" does it there and then.
1829 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1831 TK/03 Fix log output including CR from clamd.
1833 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1834 exist provoked a memory error which could cause a segfault.
1836 PH/15 Installed PCRE 6.2
1838 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1840 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1841 of the problem. Specifically, suggested +O2 rather than +O1 for the
1844 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1846 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1847 record, the port was not correctly logged when the outgoing_port log
1848 selector was set (it logged the transort's default port).
1850 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1851 fallback_hosts, and "hosts" in the smtp transport.
1853 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1854 host errors such as "Connection refused".
1856 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1857 authentication with radiusclient 0.4.9:
1859 - Error returned from rc_read_config was caught wrongly
1860 - Username/password not passed on to radius server due to wrong length.
1862 The presumption is that some radiusclient API changes for 4.51/PH/17
1863 were not taken care of correctly. The code is still untested by me (my
1864 Linux distribution still has 0.3.2 of radiusclient), but it was
1865 contributed by a Radius user.
1867 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1868 expanding the "port" option of the smtp transport.
1870 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1871 while reading a MIME header. Thanks to Tom Hughes for a patch.
1873 PH/24 Include config.h inside local_scan.h so that configuration settings are
1876 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1877 that in an ACL for RCPT (for example), you can examine exactly what was
1880 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1881 commands, but it was not correctly comparing the address with the actual
1882 client host address. Thus, it would show the EHLO address in Received:
1883 header lines when this was not necessary.
1885 PH/27 Added the % operator to ${eval:}.
1887 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1888 it should be ignoring failures (because with -C, for example, it has lost
1889 privilege). It wasn't ignoring creation failures other than "already
1892 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1895 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1896 because an input line was too long, either on its own, or by virtue of
1897 too many continuations, the temporary file was not being removed, and the
1898 return code was incorrect.
1900 PH/31 Missing "BOOL" in function definition in filtertest.c.
1902 PH/32 Applied Sieve patches from the maintainer.
1904 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1906 PH/33 Added "verify = not_blind".
1908 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1909 Local/Makefile (with some defaults set). These are used in built scripts
1910 such as exicyclog, but they have never been used in the exim_install
1911 script (though there are many overriding facilities there). I have
1912 arranged that the exim_install script now takes note of these two
1915 PH/35 Installed configuration files for Dragonfly.
1917 PH/36 When a locally submitted message by a trusted user did not contain a
1918 From: header, and the sender address was obtained from -f or from an SMTP
1919 MAIL command, and the trusted user did not use -F to supply a sender
1920 name, $originator_name was incorrectly used when constructing a From:
1921 header. Furthermore, $originator_name was used for submission mode
1922 messages from external hosts without From: headers in a similar way,
1923 which is clearly wrong.
1925 PH/37 Added control=suppress_local_fixups.
1927 PH/38 When log_selector = +received_sender was set, and the addition of the
1928 sender made the log line's construction buffer exactly full, or one byte
1929 less than full, an overflow happened when the terminating "\n" was
1932 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1933 when the result of a list match is failure because a DNS lookup failed.
1935 PH/40 RM_COMMAND is now used in the building process.
1937 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1938 the "build-* directories that it finds.
1940 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1941 address in a domain literal was a prefix of an interface address.
1943 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1944 when verifying a sender address, unless rewrite_headers is false.
1946 PH/44 Wrote a long comment about why errors_to addresses are verified as
1947 recipients, not senders.
1949 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1950 the ratelimit ACL was added.
1952 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1954 PH/46 Added extra information about PostgreSQL errors to the error string.
1956 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1957 avoiding the need to install special zones in a real server. This is
1958 backwards compatible; if it can't find the fake resolver, it drops back.
1959 Thus, both old and new test suites can be run.
1961 TF/02 Added util/ratelimit.pl
1963 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1964 clock is set back in time.
1966 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1967 Candler <B.Candler@pobox.com>.
1969 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1970 OK for addresses that are the result of redirections.
1972 PH/48 A number of further additions for the benefit of the new test suite,
1973 including a fake gethostbyname() that interfaces to the fake DNS resolver
1976 TF/06 The fix for widen_domains has also been applied to qualify_single and
1977 search_parents which are the other dnslookup options that can cause
1980 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1983 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1985 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1986 service" if it encountered an MX record with an empty target hostname.
1987 The message is now "an MX or SRV record indicated no SMTP service".
1989 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1990 if the DNS of the sending site is misconfigured. This is quite a
1991 common situation. This change restores the behaviour of treating a
1992 helo verification defer as a failure.
1994 PH/51 If self=fail was set on a router, the bounce message did not include the
1995 actual error message.
2001 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2003 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2004 command itself ended in error, the underlying message about the transport
2005 filter timeout was being overwritten with the pipe command error. Now the
2006 underlying error message should be appended to the second error message.
2008 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2010 PH/02 Reduce the amount of output that "make" produces by default. Full output
2011 can still be requested.
2013 PH/03 The warning log line about a condition test deferring for a "warn" verb
2014 was being output only once per connection, rather than after each
2015 occurrence (because it was using the same function as for successful
2016 "warn" verbs). This seems wrong, so I have changed it.
2018 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2019 it should not have, which might have caused a crash in the right
2020 circumstances, but probably never did.
2022 PH/04 Installed a modified version of Tony Finch's patch to make submission
2023 mode fix the return path as well as the Sender: header line, and to
2024 add a /name= option so that you can make the user's friendly name appear
2027 TF/03 Added the control = fakedefer ACL modifier.
2029 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2030 Mark Lowes for thorough testing.
2032 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2034 TK/03 Merged latest SRS patch from Miles Wilton.
2036 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2037 with the definition in sysexits.h (which is #included earlier).
2038 Fortunately, Exim does not actually use EX_OK. The code used to try to
2039 preserve the sysexits.h value, by assumimg that macro definitions were
2040 scanned for macro replacements. I have been disabused of this notion,
2041 so now the code just undefines EX_OK before #including unistd.h.
2043 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2044 in the smtp transport. When a block could not be written in a single
2045 write() function, the timeout was being re-applied to each part-write.
2046 This seems wrong - if the receiver was accepting one byte at a time it
2047 would take for ever. The timeout is now adjusted when this happens. It
2048 doesn't have to be particularly precise.
2050 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2051 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2053 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2054 without a domain if the check to <postmaster@domain> fails.
2056 SC/01 Eximstats: added -xls and the ability to specify output files
2057 (patch written by Frank Heydlauf).
2059 SC/02 Eximstats: use FileHandles for outputing results.
2061 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2063 SC/04 Eximstats: fixed display of large numbers with -nvr option
2065 SC/05 Eximstats: fixed merging of reports with empty tables.
2067 SC/06 Eximstats: added the -include_original_destination flag
2069 SC/07 Eximstats: removed tabs and trailing whitespace.
2071 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2073 TK/06 MBOX spool code: Add real "From " MBOX separator line
2074 so the .eml file is really in mbox format (even though
2075 most programs do not really care). Patch from Alex Miller.
2077 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2078 The latter is generated from $received_to and is only set if the
2079 message has one envelope recipient. SA can use these headers,
2080 obviously out-of-the-box. Patch from Alex Miller.
2082 PH/08 The ${def test on a variable was returning false if the variable's
2083 value was "0", contrary to what the specification has always said!
2084 The result should be true unless the variable is empty.
2086 PH/09 The syntax error of a character other than { following "${if
2087 def:variable_name" (after optional whitespace) was not being diagnosed.
2088 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2089 accidental colon was present, for example, could give incorrect results.
2091 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2092 result is used (not including appendfile, where other changes are about
2095 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2096 This involved changing a lot of size variables from int to off_t. It
2097 should work with maildirs and everything.
2099 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2100 spamd dying while we are connected to it.
2102 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2105 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2106 function for BDB 4.3.
2108 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2110 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2111 the sources for PCRE have been re-arranged and more files are now
2114 PH/15 The code I had for printing potentially long long variables in PH/11
2115 above was not the best (it lost precision). The length of off_t variables
2116 is now inspected at build time, and an appropriate printing format (%ld
2117 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2118 to be "long long int" or "long int". This is needed for the internal
2119 formatting function string_vformat().
2121 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2122 the configuration file to be ":syslog", then the script "guesses" where
2123 the logs files are, rather than using the compiled in default. In our
2124 case the guess is not the same as the compiled default, so the script
2125 suddenly stopped working when I started to use syslog. The patch checks
2126 to see if log_file_path is "". If so, it attempts to read it from exim
2127 with no configuration file to get the compiled in version, before it
2128 falls back to the previous guessing code."
2130 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2131 implementing BATV in an Exim configuration. See NewStuff for the gory
2134 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2135 Makefile that are specific to HP-UX.
2137 PH/18 If the "use_postmaster" option was set for a recipient callout together
2138 with the "random" option, the postmaster address was used as the MAIL
2139 FROM address for the random test, but not for the subsequent recipient
2140 test. It is now used for both.
2142 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2143 patch removes a few documentation additions to RFC 3028, because the
2144 latest draft now contains them. It adds the new en;ascii-case comparator
2145 and a new error check for 8bit text in MIME parts. Comparator and
2146 require names are now matched exactly. I enabled the subaddress
2147 extension, but it is not well tested yet (read: it works for me)."
2149 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2150 rework some of the code of TK/09 above to avoid the hardwired use of
2151 "%lld" and "long long". Replaced the call to snprintf() with a call to
2154 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2155 records point to non-existent hosts", "retry timeout exceeded", and
2156 "retry time not reached for any host after a long failure period".
2158 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2159 experimental DomainKeys support:
2161 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2162 (2) On an error such as an illegally used "control", the wrong name for
2163 the control was given.
2165 These problems did NOT occur unless DomainKeys support was compiled.
2167 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2169 PH/24 Added ${if match_ip condition.
2171 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2172 databases so that it will be absolutely obvious if a crash occurs in the
2173 DB library. This is a regular occurrence (often caused by mis-matched
2176 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2177 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2178 picked up on a user's system that detects such things. There doesn't seem
2179 to be a gcc warning option for this - only an attribute that has to be
2180 put on the function's prototype. It seems that in Fedora Core 4 they have
2181 set this on a number of new functions. No doubt there will be more in due
2184 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2185 specify a transport. However, if an address that was verified by such a
2186 router was the subject of a callout, Exim crashed because it tried to
2187 read the rcpt_include_affixes from the non-existent transport. Now it
2188 just assumes that the setting of that option is false. This bug was
2189 introduced by 4.51/PH/31.
2191 PH/28 Changed -d+all to exclude +memory, because that information is very
2192 rarely of interest, but it makes the output a lot bigger. People tend to
2193 do -d+all out of habit.
2195 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2196 code in os-type was giving problems when libc.so lives in lib64, like on
2199 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2200 aren't the modern standard, and it seems that some systems' include files
2201 don't always have them. Exim was already checking for some of the newer
2202 ones like T_AAAA, and defining it itself. I've added checks for all the
2203 record types that Exim uses.
2205 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2206 not automatically generating a new one, as it is supposed to. This
2207 prevented TLS from working. If the file did exist, but contained invalid
2208 data, a new version was generated, as expected. It was only the case of a
2209 non-existent file that was broken.
2211 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2212 with a change in libdomainkeys > 0.64.
2214 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2215 from DNS. If the selector record carries the flag, it now has
2216 precedence over the domain-wide flag.
2218 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2220 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2221 the use of a port name that isn't defined in /etc/services) occurred, the
2222 message was deferred as in a normal delivery, and thus remained on the
2223 spool, instead of being failed because of the mua_wrapper setting. This
2224 is now fixed, and I tidied up some of the mua_wrapper messages at the
2227 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2228 the messages in a hash of arrays rather than using individual hashes.
2229 This is a bit cleaner and results in dramatic memory savings, albeit
2230 at a slight CPU cost.
2232 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2233 as requested by Marc Sherman.
2235 SC/10 Eximstats: added histograms for user specified patterns as requested
2238 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2240 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2241 fopen() in the content-scanning modules that did not already have it.
2247 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2248 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2250 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2252 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2254 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2255 to test Sieve filters that use "vacation".
2257 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2258 that changes the way the GnuTLS parameters are stored in the cache file.
2259 The new format can be generated externally. For backward compatibility,
2260 if the data in the cache doesn't make sense, Exim assumes it has read an
2261 old-format file, and it generates new data and writes a new file. This
2262 means that you can't go back to an older release without removing the
2265 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2266 work if there are any delivery delays because "one_time" forces the
2267 parent to be marked "delivered", so its unseen clone is never tried
2268 again. For this reason, Exim now forbids the simultaneous setting of
2271 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2272 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2273 entry if you want to know the details. The fix, however, neglected to
2274 consider the case when local delivery batching is involved. The test for
2275 "previously delivered" was not happening when checking to see if an
2276 address could be batched with a previous (undelivered) one; under
2277 certain circumstances this could lead to multiple deliveries to the same
2280 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2281 in its include files, and this causes problems building Exim.
2283 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2284 header_syntax) but Exim was just ignoring anything given after a slash.
2285 In particular, this caused confusion with an attempt to use "verify =
2286 reverse_host_lookup/defer_ok". An error is now given when options are
2287 supplied for verify items that do not have them. (Maybe reverse_host_
2288 lookup should have a defer_ok option, but that's a different point.)
2290 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2291 defined by RFC 821) to 2048, because there were problems with some AUTH
2292 commands, and RFC 1869 says the size should be increased for extended
2293 SMTP commands that take arguments.
2295 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2298 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2299 "unknown" error; now it says that the functionality isn't in the binary.
2301 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2302 an address' error message when a string expansion fails (syntax or
2303 whatever). Otherwise the password may appear in the log. Following change
2304 PH/42 below, there is no longer a chance of it appearing in a bounce
2307 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2309 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2310 of its addresses was ever tried. (Bugzilla bug #2.)
2312 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2313 the result incorrectly in the debug output. (It correctly added a newline
2314 to what was transported.)
2316 TF/01 Added $received_time.
2318 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2319 commented out examples of how to interface to a virus scanner and to
2320 SpamAssassin. Also added commented examples of av_scanner and
2321 spamd_address settings.
2323 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2324 and controls are allowed in which ACLs. There were a couple of minor
2325 errors. Some of the entries in the conditions table (which is a table of
2326 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2327 negation of where the condition IS allowed.
2329 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2331 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2332 header file does not have a version number, so I've had to invent a new
2333 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2334 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2335 radiusclient), but it was contributed by a Radius user.
2337 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2338 files or directories, for OpenSSL.
2340 PH/19 When an Exim process that is running as root has to create an Exim log
2341 file, it does so in a subprocess that runs as exim:exim so as to get the
2342 ownership right at creation (otherwise, other Exim processes might see
2343 the file with the wrong ownership). There was no test for failure of this
2344 fork() call, which would lead to the process getting stuck as it waited
2345 for a non-existent subprocess. Forks do occasionally fail when resources
2346 run out. I reviewed all the other calls to fork(); they all seem to check
2349 PH/20 When checking for unexpected SMTP input at connect time (before writing
2350 the banner), Exim was not dealing correctly with a non-positive return
2351 from the read() function. If the client had disconnected by this time,
2352 the result was a log entry for a synchronization error with an empty
2353 string after "input=" when read() returned zero. If read() returned -1
2354 (an event I could not check), uninitialized data bytes were printed.
2355 There were reports of junk text (parts of files, etc) appearing after
2358 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2360 PH/22 Added support for macro redefinition, and (re)definition in between
2361 driver and ACL definitions.
2363 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2364 forgetting to use the resulting value; it was using the unexpanded value.
2366 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2367 hadn't been configured. The fix is from Juergen Kreileder, who
2368 understands it better than I do:
2370 "Here's what I see happening with three configured cyrus_sasl
2371 authenticators configured (plain, login, cram-md5):
2373 On startup auth_cyrus_sasl_init() gets called for each of these.
2374 This means three calls to sasl_listmech() without a specified mech_list.
2375 => SASL tests which mechs of all available mechs actually work
2376 => three warnings about OTP not working
2377 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2379 With the patch, sasl_listmech() also gets called three times. But now
2380 SASL's mech_list option is set to the server_mech specified in the the
2381 authenticator. Or in other words, the answer from sasl_listmech()
2382 gets limited to just the mech you're testing for (which is different
2384 => the return list contains just 'plain' or 'login', 'cram-md5' or
2385 nothing depending on the value of ob->server_mech.
2387 I've just tested the patch: Authentication still works fine,
2388 unavailable mechs specified in the exim configuration are still
2389 caught, and the auth.log warnings about OTP are gone."
2391 PH/25 When debugging is enabled, the contents of the command line are added
2392 to the debugging output, even when log_selector=+arguments is not
2395 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2396 answer is "GNU", and only if the return is "GNU/something" is the answer
2399 PH/27 $acl_verify_message is now set immediately after the failure of a
2400 verification in an ACL, and so is available in subsequent modifiers. In
2401 particular, the message can be preserved by coding like this:
2403 warn !verify = sender
2404 set acl_m0 = $acl_verify_message
2406 Previously, $acl_verify_message was set only while expanding "message"
2407 and "log_message" when a very denied access.
2409 PH/28 Modified OS/os.c-Linux with
2411 -#ifndef OS_LOAD_AVERAGE
2412 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2414 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2415 nomenclature these days.)
2417 PH/29 Installed patch from the Sieve maintainer that adds the options
2418 sieve_useraddress and sieve_subaddress to the redirect router.
2420 PH/30 In these circumstances:
2421 . Two addresses routed to the same list of hosts;
2422 . First host does not offer TLS;
2423 . First host accepts first address;
2424 . First host gives temporary error to second address;
2425 . Second host offers TLS and a TLS session is established;
2426 . Second host accepts second address.
2427 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2428 and peerdn, if requested) that were in fact used only for the second
2431 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2432 attention to any local part prefix or suffix that was matched by the
2433 router that accepted the address. It now behaves in the same way as it
2434 does for delivery: the affixes are removed from the local part unless
2435 rcpt_include_affixes is set on the transport.
2437 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2438 timeout during the DATA phase of an incoming message.
2440 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2441 applied a patch sanctioned by the Sieve maintainer.
2443 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2444 the uid or gid is negative. A case of a negative gid caused this to be
2445 noticed. The fix allows for either to be negative.
2447 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2448 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2451 PH/36 The change PH/12 above was broken. Fixed it.
2453 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2454 the grounds that routing the same address twice would always produce the
2455 same answer. This might have been true once, but it is certainly no
2456 longer true now. Routing a child address may depend on the previous
2457 routing that produced that child. Some complicated redirection strategies
2458 went wrong when messages had multiple recipients, and made Exim's
2459 behaviour dependent on the order in which the addresses were given.
2461 I have moved the duplicate checking until after the routing is complete.
2462 Exim scans the addresses that are assigned to local and remote
2463 transports, and removes any duplicates. This means that more work will be
2464 done, as duplicates will always all be routed, but duplicates are
2465 presumably rare, so I don't expect this is of any significance.
2467 For deliveries to pipes, files, and autoreplies, the duplicate checking
2468 still happens during the routing process, since they are not going to be
2471 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2472 It corrects a timeout issue with spamd. This is Ian's comment: "The
2473 background is that sometimes spamd either never reads data from a
2474 connection it has accepted, or it never writes response data. The exiscan
2475 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2476 blindly assumes that writes won't block so it may never time out."
2478 PH/39 Allow G after quota size as well as K and M.
2480 PH/40 The value set for $authenticated_id in an authenticator may not contain
2481 binary zeroes or newlines because the value is written to log lines and
2482 to spool files. There was no check on this. Now the value is run through
2483 the string_printing() function so that such characters are converted to
2484 printable escape sequences.
2486 PH/41 $message_linecount is a new variable that contains the total number of
2487 lines in the message. Compare $body_linecount, which is the count for the
2490 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2491 bounce and delay warning messages, except in certain special cases, which
2494 (a) An SMTP error message from a remote host;
2495 (b) A message specified in a :fail: redirection;
2496 (c) A message specified in a "fail" command in a system filter;
2497 (d) A message specified in a FAIL return from the queryprogram router;
2498 (e) A message specified by the cannot_route_message router option.
2500 In these cases only, Exim does include the error details in bounce and
2501 warning messages. There are also a few cases where bland messages such
2502 as "unrouteable address" or "local delivery error" are given.
2504 PH/43 $value is now also set for the "else" part of a ${run expansion.
2506 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2507 being worked on, but at least Exim now implements the latest version to
2510 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2511 process to complete was treated as a delivery failure, a timeout while
2512 writing the message to the pipe was logged, but erroneously treated as a
2513 successful delivery. Such timeouts include transport filter timeouts. For
2514 consistency with the overall process timeout, these timeouts are now
2515 treated as errors, giving rise to delivery failures by default. However,
2516 there is now a new Boolean option for the pipe transport called
2517 timeout_defer, which, if set TRUE, converts the failures into defers for
2518 both kinds of timeout. A transport filter timeout is now identified in
2521 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2522 systems where "make" and "gmake" are different, calling "gmake" at top
2523 level broke things. I've arranged for the value of $(MAKE) to be passed
2524 from the Makefile to this script so that it can call the same version of
2528 A note about Exim versions 4.44 and 4.50
2529 ----------------------------------------
2531 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2532 changes of various kinds. As a consequence, a big documentation update was
2533 needed. This delayed the release for rather longer than seemed good, especially
2534 in the light of a couple of (minor) security issues. Therefore, the changes
2535 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2536 release. So 4.44 and 4.50 are in effect two different branches that both start
2539 I have left the 4.50 change log unchanged; it contains all the changes since
2540 4.43. The change log for 4.44 is below; many of its items are identical to
2541 those for 4.50. This seems to be the most sensible way to preserve the
2542 historical information.
2548 1. Minor wording change to the doc/README.SIEVE file.
2550 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2551 computation of the current number of files was incorrect.
2553 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2554 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2557 4. Give more explanation in the error message when the command for a transport
2558 filter fails to execute.
2560 5. There are several places where Exim runs a non-Exim command in a
2561 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2562 was being done only for the command run by the queryprogram router. It is
2563 now done for all such subprocesses. The other cases are: ${run, transport
2564 filters, and the commands run by the lmtp and pipe transports.
2566 6. Added CONFIGURE_GROUP build-time option.
2568 7. Some older OS have a limit of 256 on the maximum number of file
2569 descriptors. Exim was using setrlimit() to set 1000 as a large value
2570 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2571 systems. I've change it so that if it can't get 1000, it tries for 256.
2573 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2574 was an oversight, and furthermore, ever since the addition of extra
2575 controls (e.g. 4.43/32), the checks on when to allow different forms of
2576 "control" were broken. There should now be diagnostics for all cases when a
2577 control that does not make sense is encountered.
2579 9. Added the /retain_sender option to "control=submission".
2581 10. $recipients is now available in the predata ACL (oversight).
2583 11. Tidy the search cache before the fork to do a delivery from a message
2584 received from the command line. Otherwise the child will trigger a lookup
2585 failure and thereby defer the delivery if it tries to use (for example) a
2586 cached ldap connection that the parent has called unbind on.
2588 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2589 of $address_data from the recipient verification was clobbered by the
2590 sender verification.
2592 13. The value of address_data from a sender verification is now available in
2593 $sender_address_data in subsequent conditions in the ACL statement.
2595 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2597 15. Added a new option "connect=<time>" to callout options, to set a different
2600 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2601 was its contents. (It was OK if the option was not defined at all.)
2603 17. A "Completed" log line is now written for messages that are removed from
2604 the spool by the -Mrm option.
2606 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2607 information about exactly what failed.
2609 19. Added -dd to debug only the daemon process.
2611 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2612 handles timeouts, both on the server side and network timeouts. Renamed the
2613 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2615 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2616 It is now set to "smtps".
2618 22. $host_address is now set to the target address during the checking of
2619 ignore_target_hosts.
2621 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2622 being passed; this would have caused $sender_host_name to have been used if
2623 matching the list had actually called for a host name (not very likely,
2624 since this list is usually IP addresses). A host name is now passed as
2627 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2628 code that specifies a non-restarting handler (typically sigaction() in
2629 modern systems) in an attempt to fix a rare and obscure crash bug.
2631 25. Narrowed the window for a race in the daemon that could cause it to ignore
2632 SIGCHLD signals. This is not a major problem, because they are used only to
2633 wake it up if nothing else does.
2635 26. A malformed maildirsize file could cause Exim to calculate negative values
2636 for the mailbox size or file count. Odd effects could occur as a result.
2637 The maildirsize information is now recalculated if the size or filecount
2640 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2641 support for a long time. Removed HAVE_SYS_VFS_H.
2643 28. Installed the latest version of exipick from John Jetmore.
2645 29. In an address list, if the pattern was not a regular expression, an empty
2646 subject address (from a bounce message) matched only if the pattern was an
2647 empty string. Non-empty patterns were not even tested. This was the wrong
2648 because it is perfectly reasonable to use an empty address as part of a
2649 database query. An empty address is now tested by patterns that are
2650 lookups. However, all the other forms of pattern expect the subject to
2651 contain a local part and a domain, and therefore, for them, an empty
2652 address still always fails if the pattern is not itself empty.
2654 30. Exim went into a mad DNS loop when attempting to do a callout where the
2655 host was specified on an smtp transport, and looking it up yielded more
2656 than one IP address.
2658 31. Re-factored the code for checking spool and log partition space into a
2659 function that finds that data and another that does the check. The former
2660 is then used to implement four new variables: $spool_space, $log_space,
2661 $spool_inodes, and $log_inodes.
2663 32. The RFC2047 encoding function was originally intended for short strings
2664 such as real names; it was not keeping to the 75-character limit for
2665 encoded words that the RFC imposes. It now respects the limit, and
2666 generates multiple encoded words if necessary. To be on the safe side, I
2667 have increased the buffer size for the ${rfc2047: expansion operator from
2670 33. It is now permitted to omit both strings after an "if" condition; if the
2671 condition is true, the result is "true". As before, when the second string
2672 is omitted, a false condition yields an empty string. This makes it less
2673 cumbersome to write custom ACL and router conditions.
2675 34. Failure to deliver a bounce message always caused it to be frozen, even if
2676 there was an errors_to setting on the router. The errors_to setting is now
2679 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2680 canonical form (fully expanded) before being placed in
2681 $sender_host_address.
2683 36. The table in the code that translates DNS record types into text (T_A to
2684 "A" for instance) was missing entries for NS and CNAME. It is just possible
2685 that this could have caused confusion if both these types were looked up
2686 for the same domain, because the text type is used as part of Exim's
2687 per-process caching. But the chance of anyone hitting this buglet seems
2690 37. The dnsdb lookup has been extended in a number of ways.
2692 (1) There is a new type, "zns", which walks up the domain tree until it
2693 finds some nameserver records. It should be used with care.
2695 (2) There is a new type, "mxh", which is like "mx" except that it returns
2696 just the host names, not the priorities.
2698 (3) It is now possible to give a list of domains (or IP addresses) to be
2699 looked up. The behaviour when one of the lookups defers can be
2700 controlled by a keyword.
2702 (4) It is now possible to specify the separator character for use when
2703 multiple records are returned.
2705 38. The dnslists ACL condition has been extended: it is now possible to supply
2706 a list of IP addresses and/or domains to be looked up in a particular DNS
2709 39. Added log_selector=+queue_time_overall.
2711 40. When running the queue in the test harness, wait just a tad after forking a
2712 delivery process, to get repeatability of debugging output.
2714 41. Include certificate and key file names in error message when GnuTLS fails
2715 to set them up, because the GnuTLS error message doesn't include the name
2716 of the failing file when there is a problem reading it.
2718 42. Allow both -bf and -bF in the same test run.
2720 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2722 44. The "Exiscan patch" is now merged into the mainline Exim source.
2724 45. Sometimes the final signoff response after QUIT could fail to get
2725 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2726 < 0 before doing a fflush(). This bug looks as though it goes back to the
2727 introduction of TLS in release 3.20, but "sometimes" must have been rare
2728 because the tests only now provoked it.
2730 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2731 (this can affect the format of dates).
2733 47. exim_tidydb, when checking for the continued existence of a message for
2734 which it has found a message-specific retry record, was not finding
2735 messages that were in split spool directories. Consequently, it was
2736 deleting retry records that should have stayed in existence.
2738 48. Steve fixed some bugs in eximstats.
2740 49. The SPA authentication driver was not abandoning authentication and moving
2741 on to the next authenticator when an expansion was forced to fail,
2742 contradicting the general specification for all authenticators. Instead it
2743 was generating a temporary error. It now behaves as specified.
2745 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2746 (the order specifies the preference for clients). The order is now AES256,
2747 AES128, 3DES, ARCFOUR128.
2749 51. Small patch to Sieve code - explicitly set From: when generating an
2752 52. Exim crashed if a remote delivery caused a very long error message to be
2753 recorded - for instance if somebody sent an entire SpamAssassin report back
2754 as a large number of 550 error lines. This bug was coincidentally fixed by
2755 increasing the size of one of Exim's internal buffers (big_buffer) that
2756 happened as part of the Exiscan merge. However, to be on the safe side, I
2757 have made the code more robust (and fixed the comments that describe what
2760 53. Now that there can be additional text after "Completed" in log lines (if
2761 the queue_time_overall log selector is set), a one-byte patch to exigrep
2762 was needed to allow it to recognize "Completed" as not the last thing in
2765 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2766 patch that reportedly fixes this has been added. I am not expert enough to
2767 create a test for it. This is what the patch creator wrote:
2769 "I found a little strange behaviour of ldap code when working with
2770 Windows 2003 AD Domain, where users was placed in more than one
2771 Organization Units. When I tried to give exim partial DN, the exit code
2772 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2773 But simultaneously result of request was absolutely normal ldap result,
2774 so I produce this patch..."
2776 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2777 REFERENCE, so I have modified the code to exclude the patch when that macro
2780 55. Some experimental protocols are using DNS PTR records for new purposes. The
2781 keys for these records are domain names, not reversed IP addresses. The
2782 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2783 leaves it alone. Component reversal etc. now happens only for IP addresses.
2786 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2788 57. Double the size of the debug message buffer (to 2048) so that more of very
2789 long debug lines gets shown.
2791 58. The exicyclog utility now does better if the number of log files to keep
2792 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2794 59. Two changes related to the smtp_active_hostname option:
2796 (1) $smtp_active_hostname is now available as a variable.
2797 (2) The default for smtp_banner uses $smtp_active_hostname instead
2798 of $primary_hostname.
2800 60. The host_aton() function is supposed to be passed a string that is known
2801 to be a valid IP address. However, in the case of IPv6 addresses, it was
2802 not checking this. This is a hostage to fortune. Exim now panics and dies
2803 if the condition is not met. A case was found where this could be provoked
2804 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2805 components; fortuitously, this particular loophole had already been fixed
2806 by change 4.50/55 above.
2808 If there are any other similar loopholes, the new check in host_aton()
2809 itself should stop them being exploited. The report I received stated that
2810 data on the command line could provoke the exploit when Exim was running as
2811 exim, but did not say which command line option was involved. All I could
2812 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2813 running as the user.
2816 61. There was a buffer overflow vulnerability in the SPA authentication code
2817 (which came originally from the Samba project). I have added a test to the
2818 spa_base64_to_bits() function which I hope fixes it.
2821 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2822 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2824 63. The daemon start-up calls getloadavg() while still root for those OS that
2825 need the first call to be done as root, but it missed one case: when
2826 deliver_queue_load_max is set with deliver_drop_privilege. This is
2827 necessary for the benefit of the queue runner, because there is no re-exec
2828 when deliver_drop_privilege is set.
2830 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2831 This has been fixed.
2833 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2834 was in use, was not putting the data itself into the right store pool;
2835 consequently, it could be overwritten for a subsequent message in the same
2836 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2839 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2841 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2842 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2843 the function were treating the return as a boolean value, which happened to
2844 work because 0=false and not-0=true, but is not correct code.
2846 68. The host_aton() function was not handling scoped IPv6 addresses (those
2847 with, for example, "%eth0" on the end) correctly.
2849 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2850 negated items (that is, ~something) in unsigned ints. Some compilers
2851 apparently mutter when there is no cast.
2853 70. If an address verification called from an ACL failed, and did not produce a
2854 user-specific message (i.e. there was only a "system" message), nothing was
2855 put in $acl_verify_message. In this situation, it now puts the system
2858 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2859 session; change 4.31/43 added the unwanted input to the log line - except
2860 that it did not do this in the start of session case. It now does.
2862 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2863 This is wrong and can cause the other end to generate a synchronization
2864 error if it is another Exim or anything else that does the synchronization
2865 check. A QUIT command is no longer sent after a timeout.
2867 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2868 during host lookups.
2870 74. The defer_ok option of callout verification was not working if it was used
2871 when verifying addresses in header lines, that is, for this case:
2873 verify = header_sender/callout=defer_ok
2875 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2876 those file descriptors could be used for SMTP connections. If anything
2877 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2878 could be sent to the SMTP client, causing chaos. The daemon now opens
2879 stdin, stdout, and stderr to /dev/null when it puts itself into the
2882 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2883 log by default. The user can override this with suitable Perl magic.
2885 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2886 the log message when discard triggers, was not working for the DATA ACL or
2887 for the non-SMTP ACL.
2889 78. Error message wording change in sieve.c.
2891 79. If smtp_accept_max_per_host was set, the number of connections could be
2892 restricted to fewer than expected, because the daemon was trying to set up
2893 a new connection before checking whether the processes handling previous
2894 connections had finished. The check for completed processes is now done
2895 earlier. On busy systems, this bug wouldn't be noticed because something
2896 else would have woken the daemon, and it would have reaped the completed
2899 80. If a message was submitted locally by a user whose login name contained one
2900 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2901 It caused a spool format error. I have fixed the spool reading code. A
2902 related problem was that the "from" clause in the Received: line became
2903 illegal because of the space(s). It is now covered by ${quote_local_part.
2905 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2908 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2910 83. Patch from Sieve maintainer to fix unterminated string problem in
2911 "vacation" handling.
2913 84. Some minor changes to the Linux configuration files to help with other
2914 OS variants using glibc.
2916 85. One more patch for Sieve to update vacation handling to latest spec.
2919 ----------------------------------------------------
2920 See the note above about the 4.44 and 4.50 releases.
2921 ----------------------------------------------------
2927 1. Change 4.43/35 introduced a bug that caused file counts to be
2928 incorrectly computed when quota_filecount was set in an appendfile
2931 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2932 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2935 3. Give more explanation in the error message when the command for a transport
2936 filter fails to execute.
2938 4. There are several places where Exim runs a non-Exim command in a
2939 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2940 was being done only for the command run by the queryprogram router. It is
2941 now done for all such subprocesses. The other cases are: ${run, transport
2942 filters, and the commands run by the lmtp and pipe transports.
2944 5. Some older OS have a limit of 256 on the maximum number of file
2945 descriptors. Exim was using setrlimit() to set 1000 as a large value
2946 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2947 systems. I've change it so that if it can't get 1000, it tries for 256.
2949 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2950 was an oversight, and furthermore, ever since the addition of extra
2951 controls (e.g. 4.43/32), the checks on when to allow different forms of
2952 "control" were broken. There should now be diagnostics for all cases when a
2953 control that does not make sense is encountered.
2955 7. $recipients is now available in the predata ACL (oversight).
2957 8. Tidy the search cache before the fork to do a delivery from a message
2958 received from the command line. Otherwise the child will trigger a lookup
2959 failure and thereby defer the delivery if it tries to use (for example) a
2960 cached ldap connection that the parent has called unbind on.
2962 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2963 of $address_data from the recipient verification was clobbered by the
2964 sender verification.
2966 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2967 was its contents. (It was OK if the option was not defined at all.)
2969 11. A "Completed" log line is now written for messages that are removed from
2970 the spool by the -Mrm option.
2972 12. $host_address is now set to the target address during the checking of
2973 ignore_target_hosts.
2975 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2976 being passed; this would have caused $sender_host_name to have been used if
2977 matching the list had actually called for a host name (not very likely,
2978 since this list is usually IP addresses). A host name is now passed as
2981 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2982 code that specifies a non-restarting handler (typically sigaction() in
2983 modern systems) in an attempt to fix a rare and obscure crash bug.
2985 15. Narrowed the window for a race in the daemon that could cause it to ignore
2986 SIGCHLD signals. This is not a major problem, because they are used only to
2987 wake it up if nothing else does.
2989 16. A malformed maildirsize file could cause Exim to calculate negative values
2990 for the mailbox size or file count. Odd effects could occur as a result.
2991 The maildirsize information is now recalculated if the size or filecount
2994 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2995 support for a long time. Removed HAVE_SYS_VFS_H.
2997 18. Updated exipick to current release from John Jetmore.
2999 19. Allow an empty sender to be matched against a lookup in an address list.
3000 Previously the only cases considered were a regular expression, or an
3003 20. Exim went into a mad DNS lookup loop when doing a callout where the
3004 host was specified on the transport, if the DNS lookup yielded more than
3007 21. The RFC2047 encoding function was originally intended for short strings
3008 such as real names; it was not keeping to the 75-character limit for
3009 encoded words that the RFC imposes. It now respects the limit, and
3010 generates multiple encoded words if necessary. To be on the safe side, I
3011 have increased the buffer size for the ${rfc2047: expansion operator from
3014 22. Failure to deliver a bounce message always caused it to be frozen, even if
3015 there was an errors_to setting on the router. The errors_to setting is now
3018 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3019 canonical form (fully expanded) before being placed in
3020 $sender_host_address.
3022 24. Updated eximstats to version 1.33
3024 25. Include certificate and key file names in error message when GnuTLS fails
3025 to set them up, because the GnuTLS error message doesn't include the name
3026 of the failing file when there is a problem reading it.
3028 26. Expand error message when OpenSSL has problems setting up cert/key files.
3031 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3032 (this can affect the format of dates).
3034 28. exim_tidydb, when checking for the continued existence of a message for
3035 which it has found a message-specific retry record, was not finding
3036 messages that were in split spool directories. Consequently, it was
3037 deleting retry records that should have stayed in existence.
3039 29. eximstats updated to version 1.35
3040 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3041 1.35 - bugfix such that pie charts by volume are generated correctly
3043 30. The SPA authentication driver was not abandoning authentication and moving
3044 on to the next authenticator when an expansion was forced to fail,
3045 contradicting the general specification for all authenticators. Instead it
3046 was generating a temporary error. It now behaves as specified.
3048 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3049 (the order specifies the preference for clients). The order is now AES256,
3050 AES128, 3DES, ARCFOUR128.
3052 31. Small patch to Sieve code - explicitly set From: when generating an
3055 32. Exim crashed if a remote delivery caused a very long error message to be
3056 recorded - for instance if somebody sent an entire SpamAssassin report back
3057 as a large number of 550 error lines. This bug was coincidentally fixed by
3058 increasing the size of one of Exim's internal buffers (big_buffer) that
3059 happened as part of the Exiscan merge. However, to be on the safe side, I
3060 have made the code more robust (and fixed the comments that describe what
3063 33. Some experimental protocols are using DNS PTR records for new purposes. The
3064 keys for these records are domain names, not reversed IP addresses. The
3065 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3066 leaves it alone. Component reversal etc. now happens only for IP addresses.
3069 34. The host_aton() function is supposed to be passed a string that is known
3070 to be a valid IP address. However, in the case of IPv6 addresses, it was
3071 not checking this. This is a hostage to fortune. Exim now panics and dies
3072 if the condition is not met. A case was found where this could be provoked
3073 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3074 components; fortuitously, this particular loophole had already been fixed
3075 by change 4.50/55 or 4.44/33 above.
3077 If there are any other similar loopholes, the new check in host_aton()
3078 itself should stop them being exploited. The report I received stated that
3079 data on the command line could provoke the exploit when Exim was running as
3080 exim, but did not say which command line option was involved. All I could
3081 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3082 running as the user.
3085 35. There was a buffer overflow vulnerability in the SPA authentication code
3086 (which came originally from the Samba project). I have added a test to the
3087 spa_base64_to_bits() function which I hope fixes it.
3090 36. The daemon start-up calls getloadavg() while still root for those OS that
3091 need the first call to be done as root, but it missed one case: when
3092 deliver_queue_load_max is set with deliver_drop_privilege. This is
3093 necessary for the benefit of the queue runner, because there is no re-exec
3094 when deliver_drop_privilege is set.
3096 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3097 was in use, was not putting the data itself into the right store pool;
3098 consequently, it could be overwritten for a subsequent message in the same
3099 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3102 38. Sometimes the final signoff response after QUIT could fail to get
3103 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3104 < 0 before doing a fflush(). This bug looks as though it goes back to the
3105 introduction of TLS in release 3.20, but "sometimes" must have been rare
3106 because the tests only now provoked it.
3112 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3113 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3114 Now it returns an integer. A number of calls were still expecting a T/F
3115 return. Fortuitously, in all cases, the tests worked in OK situations,
3116 which is the norm. However, things would have gone wrong on any write
3117 failures on the smtp file descriptor. This function is used when sending
3118 messages over SMTP and also when doing verify callouts.
3120 2. When Exim is called to do synchronous delivery of a locally submitted
3121 message (the -odf or -odi options), it no longer closes stderr before doing
3124 3. Implemented the mua_wrapper option.
3126 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3128 5. Implemented the functions header_remove(), header_testname(),
3129 header_add_at_position(), and receive_remove_recipient(), and exported them
3132 6. If an ACL "warn" statement specified the addition of headers, Exim already
3133 inserted X-ACL-Warn: at the start if there was no header name. However, it
3134 was not making this test for the second and subsequent header lines if
3135 there were newlines in the string. This meant that an invalid header could
3136 be inserted if Exim was badly configured.
3138 7. Allow an ACL "warn" statement to add header lines at the start or after all
3139 the Received: headers, as well as at the end.
3141 8. Added the rcpt_4xx retry error code.
3143 9. Added postmaster_mailfrom=xxx to callout verification option.
3145 10. Added mailfrom=xxxx to the callout verification option, for verify=
3148 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3149 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3151 12. Inserted some casts to stop certain compilers warning when using pointer
3152 differences as field lengths or precisions in printf-type calls (mostly
3153 affecting debugging statements).
3155 13. Added optional readline() support for -be (dynamically loaded).
3157 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3158 same clock tick as a message's arrival, so that its received time was the
3159 same as the "first fail" time on the retry record, and that message
3160 remained on the queue past the ultimate address timeout, every queue runner
3161 would try a delivery (because it was past the ultimate address timeout) but
3162 after another failure, the ultimate address timeout, which should have then
3163 bounced the address, did not kick in. This was a "< instead of <=" error;
3164 in most cases the first failure would have been in the next clock tick
3165 after the received time, and all would be well.
3167 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3168 being recognized when the domain list was tested by the match_domain
3169 condition in an expansion string.
3171 16. Added the ${str2b64: operator.
3173 17. Exim was always calling setrlimit() to set a large limit for the number of
3174 processes, without checking whether the existing limit was already
3175 adequate. (It did check for the limit on file descriptors.) Furthermore,
3176 errors from getrlimit() and setrlimit() were being ignored. Now they are
3177 logged to the main and panic logs, but Exim does carry on, to try to do its
3178 job under whatever limits there are.
3180 18. Imported PCRE 5.0.
3182 19. Trivial typo in log message " temporarily refused connection" (the leading
3185 20. If the log selector return_path_on_delivery was set and an address was
3186 redirected to /dev/null, the delivery process crashed because it assumed
3187 that a return path would always be set for a "successful" delivery. In this
3188 case, the whole delivery is bypassed as an optimization, and therefore no
3191 21. Internal re-arrangement: the function for sending a challenge and reading
3192 a response while authentication was assuming a zero-terminated challenge
3193 string. It's now changed to take a pointer and a length, to allow for
3194 binary data in such strings.
3196 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3198 23. Exim was not respecting finduser_retries when seeking the login of the
3199 uid under which it was called; it was always trying 10 times. (The default
3200 setting of finduser_retries is zero.) Also, it was sleeping after the final
3201 failure, which is pointless.
3203 24. Implemented tls_on_connect_ports.
3205 25. Implemented acl_smtp_predata.
3207 26. If the domain in control=submission is set empty, Exim assumes that the
3208 authenticated id is a complete email address when it generates From: or
3209 Sender: header lines.
3211 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3212 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3213 chown and chgrp in /bin and hostname in /usr/bin.
3215 28. Exim was keeping the "process log" file open after each use, just as it
3216 does for the main log. This opens the possibility of it remaining open for
3217 long periods when the USR1 signal hits a daemon. Occasional processlog
3218 errors were reported, that could have been caused by this. Anyway, it seems
3219 much more sensible not to leave this file open at all, so that is what now
3222 29. The long-running daemon process does not normally write to the log once it
3223 has entered its main loop, and it closes the log before doing so. This is
3224 so that log files can straightforwardly be renamed and moved. However,
3225 there are a couple of unusual error situations where the daemon does write
3226 log entries, and I had neglected to close the log afterwards.
3228 30. The text of an SMTP error response that was received during a remote
3229 delivery was being truncated at 512 bytes. This is too short for some of
3230 the long messages that one sometimes sees. I've increased the limit to
3233 31. It is now possible to make retry rules that apply only when a message has a
3234 specific sender, in particular, an empty sender.
3236 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3237 it possible to be selective about when SMTP synchronization is enforced.
3239 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3241 32. Implemented hosts_connection_nolog.
3243 33. Added an ACL for QUIT.
3245 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3248 35. Added mailbox_size and mailbox_filecount to appendfile.
3250 36. Added control = no_multiline_responses to ACLs.
3252 37. There was a bug in the logic of the code that waits for the clock to tick
3253 in the case where the clock went backwards by a substantial amount such
3254 that the microsecond fraction of "now" was more than the microsecond
3255 fraction of "then" (but the whole seconds number was less).
3257 38. Added support for the libradius Radius client library this is found on
3258 FreeBSD (previously only the radiusclient library was supported).
3264 1. When certain lookups returned multiple values in the form name=value, the
3265 quoting of the values was not always being done properly. Specifically:
3266 (a) If the value started with a double quote, but contained no whitespace,
3268 (b) If the value contained whitespace other than a space character (i.e.
3269 tabs or newlines or carriage returns) it was not quoted.
3270 This fix has been applied to the mysql and pgsql lookups by writing a
3271 separate quoting function and calling it from the lookup code. The fix
3272 should probably also be applied to nisplus, ibase and oracle lookups, but
3273 since I cannot test any of those, I have not disturbed their existing code.
3275 2. A hit in the callout cache for a specific address caused a log line with no
3276 reason for rejecting RCPT. Now it says "Previous (cached) callout
3277 verification failure".
3279 3. There was an off-by-one bug in the queryprogram router. An over-long
3280 return line was truncated at 256 instead of 255 characters, thereby
3281 overflowing its buffer with the terminating zero. As well as fixing this, I
3282 have increased the buffer size to 1024 (and made a note to document this).
3284 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3285 when Exim is waiting for an SMTP response from a remote server, Exim
3286 restarts its select() call on the socket, thereby resetting its timeout.
3287 This is not a problem when such interrupts are rare. Somebody set up a cron
3288 job to run exiwhat every 2 minutes, which is less than the normal select()
3289 timeout (5 or 10 minutes). This meant that the select() timeout never
3290 kicked in because it was always reset. I have fixed this by comparing the
3291 time when an interrupt arrives with the time at the start of the first call
3292 to select(). If more time than the timeout has elapsed, the interrupt is
3293 treated as a timeout.
3295 5. Some internal re-factoring in preparation for the addition of Sieve
3296 extensions (by MH). In particular, the "personal" test is moved to a
3297 separate function, and given an option for scanning Cc: and Bcc: (which is
3298 not set for Exim filters).
3300 6. When Exim created an email address using the login of the caller as the
3301 local part (e.g. when creating a From: or Sender: header line), it was not
3302 quoting the local part when it contained special characters such as @.
3304 7. Installed new OpenBSD configuration files.
3306 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3307 try to make them clearer.
3309 9. Callout options, other than the timeout value, were being ignored when
3310 verifying sender addresses in header lines. For example, when using
3312 verify = header_sender/callout=no_cache
3314 the cache was (incorrectly) being used.
3316 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3317 only the Cygwin environment.
3319 11. When return_path_on_delivery was set as a log selector, if different remote
3320 addresses in the same message used different return paths and parallel
3321 remote delivery occurred, the wrong values would sometimes be logged.
3322 (Whenever a remote delivery process finished, the return path value from
3323 the most recently started remote delivery process was logged.)
3325 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3326 lines when AUTH and/or TLS are in use. This is the "received protocol"
3327 field. Exim used to use "asmtp" for authenticated SMTP, without any
3328 indication (in the protocol name) for TLS use. Now it follows the RFC and
3329 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3330 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3331 names appear in log lines as well as in Received: header lines.
3333 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3334 extensions, and comparison tests, and to fix some bugs.
3336 14. Changes to the "personal" filter test:
3338 (1) The test was buggy in that it was just doing the equivalent of
3339 "contains" tests on header lines. For example, if a user's address was
3340 anne@some.where, the "personal" test would incorrectly be true for
3342 To: susanne@some.where
3344 This test is now done by extracting each address from the header in turn,
3345 and checking the entire address. Other tests that are part of "personal"
3346 are now done using regular expressions (for example, to check local parts
3347 of addresses in From: header lines).
3349 (2) The list of non-personal local parts in From: addresses has been
3350 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3351 taken from the Sieve specification recommendations.
3353 (3) If the message contains any header line starting with "List-" it is
3354 treated as non-personal.
3356 (4) The test for "circular" in the Subject: header line has been removed
3357 because it now seems ill-conceived.
3359 15. Minor typos in src/EDITME comments corrected.
3361 16. Installed latest exipick from John Jetmore.
3363 17. If headers_add on a router specified a text string that was too long for
3364 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3365 of string_sprintf() is now avoided.
3367 18. $message_body_size was not set (it was always zero) when running the DATA
3368 ACL and the local_scan() function.
3370 19. For the "mail" command in an Exim filter, no default was being set for
3371 the once_repeat time, causing a random time value to be used if "once" was
3372 specified. (If the value happened to be <= 0, no repeat happened.) The
3373 default is now 0s, meaning "never repeat". The "vacation" command was OK
3374 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3375 (I found it when inspecting the code).
3377 20. There is now an overall timeout for performing a callout verification. It
3378 defaults to 4 times the callout timeout, which applies to individual SMTP
3379 commands during the callout. The overall timeout applies when there is more
3380 than one host that can be tried. The timeout is checked before trying the
3381 next host. This prevents very long delays if there are a large number of
3382 hosts and all are timing out (e.g. when the network connections are timing
3383 out). The value of the overall timeout can be changed by specifying an
3384 additional sub-option for "callout", called "maxwait". For example:
3386 verify = sender/callout=5s,maxwait=20s
3388 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3389 to the end before writing, but this should make it even safer).
3391 22. Exim was forgetting that it had advertised PIPELINING for the second and
3392 subsequent messages on an SMTP connection. It was also not resetting its
3393 memory on STARTTLS and an internal HELO.
3395 23. When Exim logs an SMTP synchronization error within a session, it now
3396 records whether PIPELINING has been advertised or not.
3398 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3399 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3400 rather than long int.
3402 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3404 26. Added the never_mail option to autoreply.
3410 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3411 crash if the getsockname() call failed; this can happen if a connection is
3412 closed very soon after it is established. The problem was simply in the
3413 order in which certain operations were done, causing Exim to try to write
3414 to the SMTP stream before it had set up the file descriptor. The bug has
3415 been fixed by making things happen in the correct order.
3421 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3422 before the connection was closed, thus losing the rejection response.
3424 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3425 some early Solaris releases, but causes trouble in current releases where
3426 socklen_t is defined.
3428 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3431 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3434 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3435 that releasing the top of it at the end releases what was used for sub-
3436 expansions (unless the block got too big). However, discard this block if
3437 the first thing is a variable or header, so that we can use its block when
3438 it is dynamic (useful for very large $message_headers, for example).
3440 6. Lookups now cache *every* query, not just the most recent. A new, separate
3441 store pool is used for this. It can be recovered when all lookup caches are
3442 flushed. Lookups now release memory at the end of their result strings.
3443 This has involved some general refactoring of the lookup sources.
3445 7. Some code has been added to the store_xxx() functions to reduce the amount
3446 of flapping under certain conditions.
3448 8. log_incoming_interface used to affect only the <= reception log lines. Now
3449 it causes the local interface and port to be added to several more SMTP log
3450 lines, for example "SMTP connection from", and rejection lines.
3452 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3454 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3456 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3457 could be overwritten at the end of the current message (or the start of a
3458 new message if it was set in a HELO ACL). The value is now preserved for
3459 the duration of the SMTP connection.
3461 12. If a transport had a headers_rewrite setting, and a matching header line
3462 contained an unqualified address, that address was qualified, even if it
3463 did not match any rewriting rules. The underlying bug was that the values
3464 of the flags that permit the existence of unqualified sender and recipient
3465 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3466 non-local messages, and by -bnq for local messages) were not being
3467 preserved with the message after it was received.
3469 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3470 "next input=" as part of the text comprising the host identity instead of
3471 the correct text. The code was using the same buffer for two different
3472 strings. However, depending on which order the printing function evaluated
3473 its arguments, the bug did not always show up. Under Linux, for example, my
3474 test suite worked just fine.
3476 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3477 doesn't work with some older versions of Perl. It has been changed to "my",
3478 which in any case is probably the better facility to use.
3480 15. A really picky compiler found some instances of statements for creating
3481 error messages that either had too many or two few arguments for the format
3484 16. The size of the buffer for calls to the DNS resolver has been increased
3485 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3486 for addresses that have a lot of PTR records. This alleviates a problem; it
3487 does not fully solve it.
3489 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3490 buffer now truncates the list and logs the incident, which is the same
3491 action as happens when Exim is looking up a host name and its aliases.
3492 Previously in this situation something unpredictable would happen;
3493 sometimes it was "internal error: store_reset failed".
3495 18. If a server dropped the connection unexpectedly when an Exim client was
3496 using GnuTLS and trying to read a response, the client delivery process
3497 crashed while trying to generate an error log message.
3499 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3500 string, for example:
3502 warn message = H1: something\nH2: something
3504 the text was added as a single header line from Exim's point of view
3505 though it ended up OK in the delivered message. However, searching for the
3506 second and subsequent header lines using $h_h2: did not work. This has been
3507 fixed. Similarly, if a system filter added multiple headers in this way,
3508 the routers could not see them.
3510 20. Expanded the error message when iplsearch is called with an invalid key to
3511 suggest using net-iplsearch in a host list.
3513 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3514 ACLs are no longer actually imposed (and a message to that effect is
3517 22. If a "gecos" field in a passwd entry contained escaped characters, in
3518 particular, if it contained a \" sequence, Exim got it wrong when building
3519 a From: or a Sender: header from that name. A second bug also caused
3520 incorrect handling when an unquoted " was present following a character
3521 that needed quoting.
3523 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3524 was not being matched caselessly.
3526 24. Arranged for all hyphens in the exim.8 source to be escaped with
3529 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3530 properly. Recipient callout cache records were still being keyed to include
3531 the sender, even when use_sender was set false. This led to far more
3532 callouts that were necessary. The sender is no longer included in the key
3533 when use_sender is false.
3535 26. Added "control = submission" modifier to ACLs.
3537 27. Added the ${base62d: operator to decode base 62 numbers.
3539 28. dnsdb lookups can now access SRV records.
3541 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3542 the configuration file.
3544 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3545 (-v) mode. This makes the output for a verbose queue run more intelligible.
3547 31. Added a use_postmaster feature to recipient callouts.
3549 32. Added the $body_zerocount variable, containing the number of binary zero
3550 bytes in the message body.
3552 33. The time of last modification of the "new" subdirectory is now used as the
3553 "mailbox time last read" when there is a quota error for a maildir
3556 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3558 35. Added +ignore_unknown as a special item in host lists.
3560 36. Code for decoding IPv6 addresses in host lists is now included, even if
3561 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3562 address was recognized as an IP address, but was then not correctly decoded
3563 into binary, causing unexpected and incorrect effects when compared with
3570 1. Very minor rewording of debugging text in manualroute to say "list of
3571 hosts" instead of "hostlist".
3573 2. If verify=header_syntax was set, and a header line with an unqualified
3574 address (no domain) and a large number of spaces between the end of the
3575 name and the colon was received, the reception process suffered a buffer
3576 overflow, and (when I tested it) crashed. This was caused by some obsolete
3577 code that should have been removed. The fix is to remove it!
3579 3. When running in the test harness, delay a bit after writing a bounce
3580 message to get a bit more predictability in the log output.
3582 4. Added a call to search_tidyup() just before forking a reception process. In
3583 theory, someone could use a lookup in the expansion of smtp_accept_max_
3584 per_host which, without the tidyup, could leave open a database connection.
3586 5. Added the variables $recipient_data and $sender_data which get set from a
3587 lookup success in an ACL "recipients" or "senders" condition, or a router
3588 "senders" option, similar to $domain_data and $local_part_data.
3590 6. Moved the writing of debug_print from before to after the "senders" test
3593 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3594 problems for message scanning, either using a data ACL, or using
3595 local_scan() because the Received: header was not generated till after they
3596 were called (in order to set the time as the time of reception completion).
3597 I have revised the way this works. The header is now generated after the
3598 body is received, but before the ACL or local_scan() are called. After they
3599 are run, the timestamp in the header is updated.
3605 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3606 before starting a queue runner without re-exec. This happened only when
3607 deliver_drop_privilege was set or when the Exim user was set to root. The
3608 effect of the bug was that timeouts during subsequent deliveries caused
3609 crashes instead of being properly handled. The handler is now left at its
3610 default (and expected) setting.
3612 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3613 message, again when deliver_drop_privilege is set or Exim is run as root.
3614 The bug described in (1) was not present in this case, but the tidying up
3615 of the other signals was missing. I have made the two cases consistent.
3617 3. The ignore_target_hosts setting on a manualroute router was being ignored
3618 for hosts that were looked up using the /MX notation.
3620 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3623 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3624 operated on the sender address. After changing the $sender_address to <>
3625 for the sender address verify, Exim was re-instated it as the original
3626 (before rewriting) address, but remembering that it had rewritten it, so it
3627 wasn't rewriting it again. This bug also had the effect of breaking the
3628 sender address verification caching when the sender address was rewritten.
3630 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3631 This has been changed so that if the ip literal address matches
3632 ignore_target_hosts, the router declines.
3634 7. Added expansion conditions match_domain, match_address, and match_local_
3635 part (NOT match_host).
3637 8. The placeholder for the Received: header didn't have a length field set.
3639 9. Added code to Exim itself and to exim_lock to test for a specific race
3640 condition that could lead to file corruption when using MBX delivery. The
3641 issue is with the lockfile that is created in /tmp. If this file is removed
3642 after a process has opened it but before that process has acquired a lock,
3643 there is the potential for a second process to recreate the file and also
3644 acquire a lock. This could lead to two Exim processes writing to the file
3645 at the same time. The added code performs the same test as UW imapd; it
3646 checks after acquiring the lock that its file descriptor still refers to
3647 the same named file.
3649 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3650 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3653 11. Added the smtp_active_hostname option. If used, this will typically be made
3654 to depend on the incoming interface address. Because $interface_address is
3655 not set up until the daemon has forked a reception process, error responses
3656 that can happen earlier (such as "too many connections") no longer contain
3659 12. If an expansion in a condition on a "warn" statement fails because a lookup
3660 defers, the "warn" statement is abandoned, and the next ACL statement is
3661 processed. Previously this caused the whole ACL to be aborted.
3663 13. Added the iplsearch lookup type.
3665 14. Added ident_timeout as a log selector.
3667 15. Added tls_certificate_verified as a log selector.
3669 16. Added a global option tls_require_ciphers (compare the smtp transport
3670 option of the same name). This controls incoming TLS connections.
3672 17. I finally figured out how to make tls_require_ciphers do a similar thing
3673 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3674 before starting the TLS session.
3676 18. Tabs are now shown as \t in -bP output.
3678 19. If the log selector return_path_on_delivery was set, Exim crashed when
3679 bouncing a message because it had too many Received: header lines.
3681 20. If two routers both had headers_remove settings, and the first one included
3682 a superfluous trailing colon, the final name in the first list and the
3683 first name in the second list were incorrectly joined into one item (with a
3684 colon in the middle).
3690 1. Added -C and -D options to the exinext utility, mainly to make it easier
3691 to include in the automated testing, but these could be helpful when
3692 multiple configurations are in use.
3694 2. The exinext utility was not formatting the output nicely when there was
3695 an alternate port involved in the retry record key, nor when there was a
3696 message id as well (for retries that were specific to a specific message
3697 and a specific host). It was also confused by IPv6 addresses, because of
3698 the additional colons they contain. I have fixed the IPv4 problem, and
3699 patched it up to do a reasonable job for IPv6.
3701 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3702 delivery, the log line now contains "pipelined" if PIPELINING was used.
3704 4. An SMTP transport process used to panic and die if the bind() call to set
3705 an explicit outgoing interface failed. This has been changed; it is now
3706 treated in the same way as a connect() failure.
3708 5. A reference to $sender_host_name in the part of a conditional expansion
3709 that was being skipped was still causing a DNS lookup. This no longer
3712 6. The def: expansion condition was not recognizing references to header lines
3713 that used bh_ and bheader_.
3715 7. Added the _cache feature to named lists.
3717 8. The code for checking quota_filecount in the appendfile transport was
3718 allowing one more file than it should have been.
3720 9. For compatibility with Sendmail, the command line option
3728 and sets the incoming protocol and host name (for trusted callers). The
3729 host name and its colon can be omitted when only the protocol is to be set.
3730 Note the Exim already has two private options, -pd and -ps, that refer to
3731 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3732 "s", but I don't think that's a major issue.
3734 10. A number of refactoring changes to the code, none of which should affect
3737 (a) The number of logging options was getting close to filling up the
3738 32-bit word that was used as a bit map. I have split them into two classes:
3739 those that are passed in the argument to log_write(), and those that are
3740 only ever tested independently outside of that function. These are now in
3741 separate 32-bit words, so there is plenty of room for expansion again.
3742 There is no change in the user interface or the logging behaviour.
3744 (b) When building, for example, log lines, the code previously used a
3745 macro that called string_cat() twice, in order to add two strings. This is
3746 not really sufficiently general. Furthermore, there was one instance where
3747 it was actually wrong because one of the argument was used twice, and in
3748 one call a function was used. (As it happened, calling the function twice
3749 did not affect the overall behaviour.) The macro has been replaced by a
3750 function that can join an arbitrary number of extra strings onto a growing
3753 (c) The code for expansion conditions now uses a table and a binary chop
3754 instead of a serial search (which was left over from when there were very
3755 few conditions). Also, it now recognizes conditions like "pam" even when
3756 the relevant support is not compiled in: a suitably worded error message is
3757 given if an attempt is made to use such a condition.
3759 11. Added ${time_interval:xxxxx}.
3761 12. A bug was causing one of the ddress fields not to be passed back correctly
3762 from remote delivery subprocesses. The field in question was not being
3763 subsequently used, so this caused to problems in practice.
3765 13. Added new log selectors queue_time and deliver_time.
3767 14. Might have fixed a bug in maildirsizefile handling that threw up
3768 "unexpected character" debug warnings, and recalculated the data
3769 unnecessarily. In any case, I expanded the warning message to give more
3772 15. Added the message "Restricted characters in address" to the statements in
3773 the default ACL that block characters like @ and % in local parts.
3775 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3776 Three changes have been made:
3778 (a) There was a serious bug; a negative response to MAIL caused the whole
3779 recipient domain to be cached as invalid, thereby blocking all messages
3780 to all local parts at the same domain, from all senders. This bug has
3781 been fixed. The domain is no longer cached after a negative response to
3782 MAIL if the sender used is not empty.
3784 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3787 (c) A new callout option, "use_sender" has been added for people who want
3788 the modified behaviour.
3794 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3797 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3798 indeed breaks things for older releases.
3800 3. Added additional logging to the case where there is a problem reading data
3801 from a filter that is running in a subprocess using a pipe, in order to
3802 try to track down a specific problem.
3804 4. Testing facility fudge: when running in the test harness and attempting
3805 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3806 getting "No route to host". Convert this to a timeout.
3808 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3811 6. Some OS don't have socklen_t but use size_t instead. This affects the
3812 fifth argument of getsockopt() amongst other things. This is now
3813 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3814 can be set for individual OS. I have set it for SunOS5, OSF1, and
3815 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3816 some earlier ones do not.
3818 7. Change 4.30/15 was not doing the test caselessly.
3820 8. The standard form for an IPv6 address literal was being rejected by address
3821 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3822 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3823 this, as well as the form without the "IPv6" on the front (but only when
3824 address literals are enabled, of course).
3826 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3828 10. Exim crashed if a message with an empty sender address specified by -f
3829 encountered a router with an errors_to setting. This could be provoked only
3830 by a command such as
3834 where an empty string was supplied; "<>" did not hit this bug.
3836 11. Installed PCRE release 4.5.
3838 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3839 remained set. It is now erased.
3841 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3842 times from message ids (which are base 36 rather than the normal 62).
3844 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3845 were being counted as actual protocol errors, and logged if the log
3846 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3847 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3848 and DATA following a set of rejected RCPTs do not count as protocol errors.
3849 In other words, Exim assumes they were pipelined, though this may not
3850 actually be the case. Of course, in all cases the client gets an
3851 appropriate error code.
3853 15. If a lookup fails in an ACL condition, a message about the failure may
3854 be available; it is used if testing the ACL cannot continue, because most
3855 such messages specify what the cause of the deferral is. However, some
3856 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3857 that caused an old message to be retained and used if a later statement
3858 caused a defer, replacing the real cause of the deferral.
3860 16. If an IP address had so many PTR records that the DNS lookup buffer
3861 was not large enough to hold them, Exim could crash while trying to process
3862 the truncated data. It now detects and logs this case.
3864 17. Further to 4.21/58, another change has been made: if (and only if) the
3865 first line of a message (the first header line) ends with CRLF, a bare LF
3866 in a subsequent header line has a space inserted after it, so as not to
3867 terminate the header.
3869 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3870 unnecessarily, used atoi() instead of strtol(), and didn't check the
3871 termination when getting file sizes from file names by regex.
3873 19. Completely re-implemented the support for maildirsize files, in the light
3874 of a number of problems with the previous contributed implementation
3875 (4.30/29). In particular:
3877 . If the quota is zero, the maildirsize file is maintained, but no quota is
3880 . If the maildir directory does not exist, it is created before any attempt
3881 to write a maildirsize file.
3883 . The quota value in the file is just a cache; if the quota is changed in
3884 the transport, the new value overrides.
3886 . A regular expression is available for excluding directories from the
3889 20. The autoreply transport checks the characters in options that define the
3890 message's headers; it allows continued headers, but it was checking with
3891 isspace() after an embedded newline instead of explicitly looking for a
3894 21. If all the "regular" hosts to which an address was routed had passed their
3895 expiry times, and had not reached their retry times, the address was
3896 bounced, even if fallback hosts were defined. Now Exim should go on to try
3899 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3900 equivalent code in the SMTP transport. Some hosts send humungous responses
3901 to HELO/EHLO, more than 1024 it seems.
3903 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3904 gives compiler warnings in some environments. I've now done it "properly",
3907 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3908 (because of problems with the built-in one) was declared to return uschar *
3909 instead of char *, causing compiler failure.
3911 25. Fixed a file descriptor leak when processing alias/forward files.
3913 26. Fixed a minor format string issue in dbfn.c.
3915 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3917 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3918 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3921 29. When a sender address is verified, it is cached, to save repeating the test
3922 when there is more than one recipient in a message. However, when the
3923 verification involves a callout, it is possible for different callout
3924 options to be set for different recipients. It is too complicated to keep
3925 track of this in the cache, so now Exim always runs a verification when a
3926 callout is required, relying on the callout cache for the optimization.
3927 The overhead is duplication of the address routing, but this should not be
3930 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3931 to be verified with callout=postmaster, and the main callout worked but the
3932 postmaster check failed, the verification correctly failed. However, if a
3933 subsequent RCPT command asked for sender verification *without* the
3934 postmaster check, incorrect caching caused this verification also to fail,
3937 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3938 it was not caching the DNS options (qualify_single, search_parents) that
3939 were used when the lookup failed. A subsequent lookup with different
3940 options therefore always gave the same answer, though there were cases
3941 where it should not have. (Example: a "domains = !$mx_any" option on a
3942 dnslookup router: the "domains" option is always processed without any
3943 widening, but the router might have qualify_single set.) Now Exim uses the
3944 cached value only when the same options are set.
3946 32. Added John Jetmore's "exipick" utility to the distribution.
3948 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3949 than a timeout (e.g. a certificate is required, and is not provided), an
3950 Exim server now closes the connection immediately. Previously it waited for
3951 the client to close - but if the client is SSL, it seems that they each
3952 wait for each other, leading to a delay before one of them times out.
3954 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3955 maintained 0.8.x compatibility because I don't think many are using it, and
3956 it is clearly obsolete.
3958 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3961 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3962 client certificate was expired. A simple patch fixes this, though I don't
3963 understand the full logic of why the verify callback is called multiple
3966 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3967 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3968 which causes problems with some clients (such as the Certicom SSL Plus
3969 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3970 disables the coutermeasure allowing Eudora to connect."
3972 38. Exim was not checking that a write() to a log file succeeded. This could
3973 lead to Bad Things if a log got too big, in particular if it hit a file
3974 size limit. Exim now panics and dies if it cannot write to a log file, just
3975 as it does if it cannot open a log file.
3977 39. Modified OS/Makefile-Linux so that it now contains
3979 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3981 The two -D definitions ensure that Exim is compiled with large file
3982 support, which makes it possible to handle log files that are bigger than
3985 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3986 instance) a domain was checked against a named list that involved a lookup,
3987 causing $domain_data to be set, then another domain was checked against the
3988 same list, then the first domain was re-checked, the value of $domain_data
3989 after the final check could be wrong. In particular, if the second check
3990 failed, it could be set empty. This bug probably also applied to
3993 41. The strip_trailing_dot option was not being applied to the address given
3994 with the -f command-line option.
3996 42. The code for reading a message's header from the spool was incrementing
3997 $received_count, but never initializing it. This meant that the value was
3998 incorrect (doubled) while delivering a message in the same process in which
3999 it was received. In the most common configuration of Exim, this never
4000 happens - a fresh exec is done - but it can happen when
4001 deliver_drop_privilege is set.
4003 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4004 it now includes up to 150 characters of the unexpected data in the log
4007 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4008 and building data strings. The size of both of these buffers was 10 000
4009 bytes - far larger than anybody would *ever* want, thought I. Needless to
4010 say, somebody hit the limit. I have increased the maximum line length to
4011 20 000 and the maximum data length of concatenated lines to 100 000. I have
4012 also fixed two bugs, because there was no checking on these buffers. Tsk,
4013 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4014 buffer is too small.
4016 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4017 lsearch lookups. Now it does.
4019 46. When parsing a route_list item in a manualroute router, a fixed-length
4020 buffer was used for the list of hosts. I made this 1024 bytes long,
4021 thinking that nobody would ever have a list of hosts that long. Wrong.
4022 Somebody had a whole pile of complicated expansion conditions, and the
4023 string was silently truncated, leading to an expansion error. It turns out
4024 that it is easier to change to an unlimited length (owing to other changes
4025 that have happened since this code was originally written) than to build
4026 structure for giving a limitation error. The length of the item that
4027 expands into the list of hosts is now unlimited.
4029 47. The lsearch lookup could not handle data where the length of text line was
4030 more than 4095 characters. Such lines were truncated, leading to shortened
4031 data being returned. It should now handle lines of any length.
4033 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4034 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4037 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4038 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4039 no longer happens in comment lines. A list of replacements is now placed
4040 at the head of all of the source files, except those whose only change is
4041 to replace PERL_COMMAND in the very first #! line.
4043 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4044 messages on the queue, with a bottom-up merge sort, using code contributed
4045 by Michael Haardt. This should make operations like -bp somewhat faster on
4046 large queues. It won't affect queue runners, except when queue_run_in_order
4049 51. Installed eximstats 1.31 in the distribution.
4051 52. Added support for SRV lookups to the dnslookup router.
4053 53. If an ACL referred to $message_body or $message_body_end, the value was not
4054 reset for any messages that followed in the same SMTP session.
4056 54. The store-handling optimization for building very long strings was not
4057 differentiating between the different store pools. I don't think this
4058 actually made any difference in practice, but I've tidied it.
4060 55. While running the routers to verify a sender address, $sender_address
4061 was still set to the sender address. This is wrong, because when routing to
4062 send a bounce to the sender, it would be empty. Therefore, I have changed
4063 it so that, while verifying a sender address, $sender_address is set to <>.
4064 (There is no change to what happens when verifying a recipient address.)
4066 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4067 target A or AAAA records (if not already returned) without resetting the
4068 qualify_single or search_parents options of the DNS resolver. These are
4069 inappropriate in this case because the targets of MX and SRV records must
4070 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4071 target that, when qualified, matched something in the local domain. These
4072 two options are now turned off when doing these lookups.
4074 57. It seems that at least some releases of Reiserfs (which does not have the
4075 concept of a fixed number of inodes) returns zero and not -1 for the
4076 number of available inodes. This interacted badly with check_spool_inodes,
4077 which assumed that -1 was the "no such thing" setting. What I have done is
4078 to check that the total number of inodes is greater than zero before doing
4079 the test of how many are available.
4081 58. When a "warn" ACL statement has a log_message modifier, the message is
4082 remembered, and not repeated. This is to avoid a lot of repetition when a
4083 message has many recipients that cause the same warning to be written.
4084 Howewer, Exim was preserving the list of already written lines for an
4085 entire SMTP session, which doesn't seem right. The memory is now reset if a
4086 new message is started.
4088 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4089 debugging output unless log_rewrite was also set.
4091 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4092 of (int)(handle) when we know that handle contains (void *)(-1).
4094 61. The Exim daemon panic-logs an error return when it closes the incoming
4095 connection. However "connection reset by peer" seems to be common, and
4096 isn't really an error worthy of noting specially, so that particular error
4099 62. When Exim is trying to find all the local interfaces, it used to panic and
4100 die if the ioctl to get the interface flags failed. However, it seems that
4101 on at least one OS (Solaris 9) it is possible to have an interface that is
4102 included in the list of interfaces, but for which you get a failure error
4103 for this call. This happens when the interface is not "plumbed" into a
4104 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4105 failure of the "get flags" call assumes that the interface is down.
4107 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4108 makes life easier for people who are doing arithmetic on fields extracted
4109 from dates, where you often get leading zeros that should not be
4110 interpreted as octal.
4112 64. Added qualify_domain to the redirect router, to override the global
4115 65. If a pathologically long header line contained very many addresses (the
4116 report of this problem mentioned 10 000) and each of them was rewritten,
4117 Exim could use up a very large amount of memory. (It kept on making new
4118 copies of the header line as it rewrote, and never released the old ones.)
4119 At the expense of a bit more processing, the header rewriting function has
4120 been changed so that it no longer eats memory in this way.
4122 66. The generation of the Received: header has been moved from the time that a
4123 message starts to be received, to the time that it finishes. The timestamp
4124 in the Received: header should now be very close to that of the <= log
4125 line. There are two side-effects of this change:
4127 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4128 logged header lines no longer include the local Received: line, because
4129 it has not yet been created. The same applies to a copy of the message
4130 that is returned to a non-SMTP sender when a message is rejected.
4132 (b) When a filter file is tested using -bf, no additional Received: header
4133 is added to the test message. After some thought, I decided that this
4136 This change does not affect the value of $received_for. It is still set
4137 after address rewriting, but before local_scan() is called.
4139 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4141 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4142 gave an unhelpful panic error message, and a defer error. I have managed to
4143 change this behaviour so that it now rejects any supplied certificate,
4144 which seems right, as the list of acceptable certificates is empty.
4146 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4147 gave an unhelpful defer error. I have not managed to make this reject any
4148 supplied certificates, but the error message it gives is "no certificate
4149 supplied", which is not helpful.
4151 70. exigrep's output now also includes lines that are not associated with any
4152 message, but which match the given pattern. Implemented by a patch from
4153 Martin Sluka, which also tidied up the Perl a bit.
4155 71. Recipient callout verification, like sender verification, was using <> in
4156 the MAIL FROM command. This isn't really the right thing, since the actual
4157 sender may affect whether the remote host accepts the recipient or not. I
4158 have changed it to use the actual sender in the callout; this means that
4159 the cache record is now keyed on a recipient/sender pair, not just the
4160 recipient address. There doesn't seem to be a real danger of callout loops,
4161 since a callout by the remote host to check the sender would use <>.
4162 [SEE ABOVE: changed after hitting problems.]
4164 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4165 temporary errors. However, in the case of such a code being given after
4166 the end of a data transmission (i.e. after ".") Exim was failing to write
4167 a retry record for the message. (Yes, there was some broken host that was
4168 actually sending 8xx at this point.)
4170 73. An unknown lookup type in a host list could cause Exim to panic-die when
4171 the list was checked. (An example that provoked this was putting <; in the
4172 middle of a list instead of at the start.) If this happened during a DATA
4173 ACL check, a -D file could be left lying around. This kind of configuration
4174 error no longer causes Exim to die; instead it causes a defer errror. The
4175 incident is still logged to the main and panic logs.
4177 74. Buglet left over from Exim 3 conversion. The message "too many messages
4178 in one connection" was written to the rejectlog but not the mainlog, except
4179 when address rewriting (yes!) was being logged.
4181 75. Added write_rejectlog option.
4183 76. When a system filter was run not as root (that is, when system_filter_user
4184 was set), the values of the $n variables were not being returned to the
4185 main process; thus, they were not subsequently available in the $sn
4188 77. Added +return_path_on_delivery log selector.
4190 78. A connection timeout was being treated differently from recipients deferred
4191 when testing hosts_max_try with a message that was older than the host's
4192 retry timeout. (The host should not be counted, thus allowing all hosts to
4193 be tried at least once before bouncing.) This may have been the cause of an
4194 occasionally reported bug whereby a message would remain on the queue
4195 longer than the retry timeout, but would be bounced if a delivery was
4196 forced. I say "may" because I never totally pinned down the problem;
4197 setting up timeout/retry tests is difficult. See also the next item.
4199 79. The ultimate address timeout was not being applied to errors that involved
4200 a combination of host plus message (for example, a timeout on a MAIL
4201 command). When an address resolved to a number of possible hosts, and they
4202 were not all tried for each delivery (e.g. because of hosts_max_try), a
4203 message could remain on the queue longer than the retry timeout.
4205 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4208 81. Fixed an obscure SMTP outgoing bug which required at least the following
4209 conditions: (a) there was another message waiting for the same server;
4210 (b) the server returned 5xx to all RCPT commands in the first message so
4211 that the message was not completed; (c) the server dropped the connection
4212 or gave a negative response to the RSET that Exim sends to abort the
4213 transaction. The observed case was a dropped connection after DATA that had
4214 been sent in pipelining mode. That is, the server had advertised PIPELINING
4215 but was not implementing it correctly. The effect of the bug was incorrect
4216 behaviour, such as trying another host, and this could lead to a crash.
4222 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4223 and daemon.c were passed as pointers to ints; they should have been
4224 pointers to socklen_t variables (which are typically unsigned ints).
4226 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4229 3. Fixed a really odd bug that affected only the testing scheme; patching a
4230 certain fixed string in the binary changed the value of another string that
4231 happened to be identical to the end of the original first string.
4233 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4234 name", it returns that address as the IP address. On some operating
4235 systems (e.g. Solaris), it also passes back the IP address string as the
4236 "host name". However, on others (e.g. Linux), it passes back an empty
4237 string. Exim wasn't checking for this, and was changing the host name to an
4238 empty string, assuming it had been canonicized.
4240 5. Although rare, it is permitted to have more than one PTR record for a given
4241 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4242 all the names associated with an address, because they do in Solaris.
4243 However, it seems that they do not in Linux for data that comes from the
4244 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4245 I found this out when I moved to a new Linux workstation and tried to run
4246 the Exim test suite.
4248 To get round this problem I have changed the code so that it now does its
4249 own call to the DNS to look up PTR records when searching for a host name.
4250 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4251 addresses that are only in /etc/hosts are still found.
4253 This behaviour is, however, controlled by an option called host_lookup_
4254 order, which defaults to "bydns:byaddr". If people want to use the other
4255 order, or indeed, just use one or the other means of lookup, they can
4256 specify it in this variable.
4258 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4259 some operating systems, this comes back from gethostbyaddr() as an empty
4260 string, and this is what Exim used to test for. However, it seems that in
4261 other systems, "." is yielded. Exim now tests for this case too.
4263 7. The values of check_spool_space and check_log_space are now held internally
4264 as a number of kilobytes instead of an absolute number of bytes. If a
4265 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4266 kilobyte. This means that much larger values can be stored.
4268 8. Exim monitor: an attempt to get the action menu when not actually pointing
4269 at a message produces an empty menu entitled "No message selected". This
4270 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4271 no entries in it ("Shell widget menu has zero width and/or height"). So I
4272 have added a single, blank menu entry in this case.
4274 9. Added ${quote_local_part.
4276 10. MIME decoding is now applied to the contents of Subject: header lines when
4279 11. Now that a reference to $sender_host_address automatically causes a reverse
4280 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4281 host lookup before query-style lookups in lists that might use this
4282 variable. This has therefore been abolished, and the "net-" prefix is no
4283 longer necessary for query-style lookups.
4285 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4286 have been a typo for LFLAGS, so it has been changed.
4288 13. The install script calls Exim with "-C /dev/null" in order to find the
4289 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4290 to be output. Howeve, since Exim outputs its version number before the
4291 error, it didn't break the script. It just looked ugly. I fixed this by
4292 always allowing "-C /dev/null" if the caller is root.
4294 14. Ignore overlarge ACL variable number when reading spool file - insurance
4295 against a later release with more variables having written the file.
4297 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4298 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4299 this, as well as the form without the "IPv6" on the front.
4301 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4302 OS/Makefile-Darwin file.
4304 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4305 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4307 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4308 variable. However, some OS use ints for inodes. I've added cast to long int
4309 to get rid of the compiler warning.
4311 19. I had forgotten to lock out "/../" in configuration file names when
4312 ALT_CONFIG_PREFIX was set.
4314 20. Routers used for verification do not need to specify transports. However,
4315 if such a router generated a host list, and callout was configured, Exim
4316 crashed, because it could not find a port number from the (non-existent)
4317 transport. It now assumes port 25 in this circumstance.
4319 21. Added the -t option to exigrep.
4321 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4322 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4323 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4324 them is set, LOOKUP_LSEARCH is forced.
4326 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4328 24. Added sender and host information to the "rejected by local_scan()" log
4329 line; previously there was no indication of these.
4331 25. Added .include_if_exists.
4333 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4334 Linux. It turns out that not all file systems support this. Apparently some
4335 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4336 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4337 not supported on the file descriptor, is now ignored when Exim is trying to
4338 sync a directory. This applies only to Linux.
4340 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4342 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4343 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4346 29. Added support for maildirsize files from supplied patch (modified a bit).
4348 30. The use of :fail: followed by an empty string could lead Exim to respond to
4349 sender verification failures with (e.g.):
4351 550 Verification failed for <xxx>
4352 550 Sender verify failed
4354 where the first response line was missing the '-' that indicates it is not
4355 the final line of the response.
4357 31. The loop for finding the name of the user that called Exim had a hardwired
4358 limit of 10; it now uses the value of finduser_retries, which is used for
4359 all other user lookups.
4361 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4364 33. Exim was neglecting to zero errno before one call of strtol() when
4365 expanding a string and expecting an integer value. On some systems this
4366 resulted in spurious "integer overflow" errors. Also, it was casting the
4367 result into an int without checking.
4369 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4370 did not work. The code looks as if it has *never* worked, though it appears
4371 to have been documented since at least releast 1.62. I have made it work.
4373 35. The "timeout_DNS" error in retry rules, also documented since at least
4374 1.62, also never worked. As it isn't clear exactly what this means, and
4375 clearly it isn't a major issue, I have abolished the feature by treating it
4376 as "timeout", and writing a warning to the main and panic logs.
4378 36. The display of retry rules for -brt wasn't always showing the error code
4381 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4382 timeout_connect_A, timeout_connect_MX.
4384 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4385 to the empty sender.
4387 39. The daemon was not analysing the content of -oX till after it had closed
4388 stderr and disconnected from the controlling terminal. This meant that any
4389 syntax errors were only noted on the panic log, and the return code from
4390 the command was 0. By re-arranging the code a little, I've made the
4391 decoding happen first, so such errors now appear on stderr, and the return
4392 code is 1. However, the actual setting up of the sockets still happens in
4393 the disconnected process, so errors there are still only recorded on the
4396 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4397 connections (as happens on some IP stacks) was logged at start up time as
4398 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4399 it from "IPv6 and IPv4", which means that two separate sockets are being
4402 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4403 says whether AF_INET or AF_INET6 was passed as an argument.
4405 42. Exiwhat output was messed up when time zones were included in log
4408 43. Exiwhat now gives more information about the daemon's listening ports,
4409 and whether -tls-on-connect was used.
4411 44. The "port" option of the smtp transport is now expanded.
4413 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4414 silently ignored. Now an error message is written to the main and panic
4417 46. There's a new ACL modifier called "logwrite" which writes to a log file
4418 as soon as it is encountered.
4420 47. Added $local_user_uid and $local_user_gid at routing time.
4422 48. Exim crashed when trying to verify a sender address that was being
4425 49. Exim was recognizing only a space character after ".include". It now also
4426 recognizes a tab character.
4428 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4429 extracting the relevant information from the specification. The man page no
4430 longer contains scrambled data for the -d option, and I've added a section
4431 at the front about calling Exim under different names.
4433 51. Added "extra_headers" argument to the "mail" command in filter files.
4435 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4438 53. Installed eximstats 1.29.
4440 54. Added transport_filter_timeout as a generic transport option.
4442 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4443 Cc: header lines. This was required by RFC 822, but it not required by RFC
4446 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4447 incoming messages that did not have them. Now it does so only if the
4448 message originates locally, that is, if there is no associated remote host
4449 address. When Resent- header lines are present, this applies to the Resent-
4450 lines rather than the non-Resent- lines.
4452 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4453 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4455 58. Messages for configuration errors now include the name of the main
4456 configuration file - useful now that there may be more than one file in a
4457 list (.included file names were always shown).
4459 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4460 for those rare installations that do not start the daemon as root or run it
4461 setuid root. I've cut out the call to initgroups() if the daemon is not
4464 60. The Exim user and group can now be bound into the binary as text strings
4465 that are looked up at the start of Exim's processing.
4467 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4469 62. Added $mailstore_basename variable.
4471 63. Installed patch to sieve.c from Michael Haardt.
4473 64. When Exim failed to open the panic log after failing to open the main log,
4474 the original message it was trying to log was written to stderr and debug
4475 output, but if they were not available (the usual case in production), it
4476 was lost. Now it is written to syslog before the two lines that record the
4477 failures to open the logs.
4479 65. Users' Exim filters run in subprocesses under the user's uid. It is
4480 possible for a "deliver" command or an alias in a "personal" command to
4481 provoke an address rewrite. If logging of address rewriting is configured,
4482 this fails because the process is not running as root or exim. There may be
4483 a better way of dealing with this, but for the moment (because 4.30 needs
4484 to be released), I have disabled address rewrite logging when running a
4485 filter in a non-root, non-exim process.
4491 1. The buildconfig auxiliary program wasn't quoting the value set for
4492 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4493 not defined. This bug was masked in 4.22 by the effect that was fixed in
4496 2. Some messages that were rejected after a message id was allocated were
4497 shown as "incomplete" by exigrep. It no longer does this for messages that
4498 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4500 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4501 have allow_domain_literals set, the ID did not get logged in the <= line.
4502 Domain literals are now always recognized in Message-ID: header lines.
4504 4. The first argument for a ${extract expansion item is the key name or field
4505 number. Leading and trailing spaces in this item were not being ignored,
4506 causing some misleading effects.
4508 5. When deliver_drop_privilege was set, single queue runner processes started
4509 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4510 same command in the process it spins off) were not dropping privilege.
4512 6. When the daemon running as "exim" started a queue runner, it always
4513 re-executed Exim in the spun-off process. This is a waste of effort when
4514 deliver_drop_privilege is set. The new process now just calls the
4515 queue-runner function directly.
4521 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4524 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4525 ignored. Though the use of -oP was forcing the writing of a pid file, it
4526 was always written to the default place.
4528 3. If the message "no IP address found for host xxxx" is generated during
4529 incoming verification, it is now followed by identification of the incoming
4530 connection (so you can more easily find what provoked it).
4532 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4534 5. Added some features to "harden" Exim a bit more against certain attacks:
4536 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4537 be put in Local/Makefile. This is like the never_users runtime option,
4538 but it cannot be overridden. The default setting is "root".
4540 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4541 prefix string with which any file named in a -C command line option
4544 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4545 is retained for -C and -D only if the caller of Exim is root. Without
4546 it, the exim user may also use -C and -D and retain privilege.
4548 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4549 command line option is disabled.
4551 6. Macro names set by the -D option must start with an upper case letter, just
4552 like macro names defined in the configuration file.
4554 7. Added "dereference=" facility to LDAP.
4556 8. Two instances of the typo "uknown" in the source files are fixed.
4558 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4559 the Configure-Makefile script screwed up while processing it.
4561 10. Incorporated PCRE 4.4.
4563 11. The SMTP synchronization check was not operating right at the start of an
4564 SMTP session. For example, it could not catch a HELO sent before the client
4565 waited for the greeting. There is now a check for outstanding input at the
4566 point when the greeting is written. Because of the duplex, asynchronous
4567 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4568 way, but not yet received, when the check is performed.
4570 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4571 on TCP/IP sockets, because this apparently causes some broken clients to
4574 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4575 unchanged) from the Cygwin maintainer.
4577 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4578 was supported instead of testing for mbx. Effectively a typo.
4580 15. The spa authenticator server code was not checking that the input it
4581 received was valid base64.
4583 16. The debug output line for the "set" modifier in ACLs was not showing the
4584 name of the variable that was being set.
4586 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4588 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4589 lookup on its own. Something else was needed to trigger the lookup. For
4590 example, a match in host_lookup or the need for a host name in a host list.
4591 Now, if $sender_host_name is referenced and the host name has not yet been
4592 looked up, a lookup is performed. If the lookup fails, the variable remains
4593 empty, and $host_lookup_failed is set to "1".
4595 19. Added "eqi" as a case-independent comparison operator.
4597 20. The saslauthd authentication condition could segfault if neither service
4598 nor realm was specified.
4600 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4601 error message that was logged was misleading, and incoming SMTP
4602 connections were dropped. The message is now more accurate, and temporary
4603 errors are given to SMTP connections.
4605 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4606 (except RSET) with a 421 error, until QUIT is received. However, it was
4607 failing to send a response to QUIT.
4609 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4610 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4611 ACL is now run after the helo_try_verify_hosts code.
4613 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4614 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4615 case-independent, so other case variants are also recognized). Apparently
4616 some systems use these upper case variants.
4618 25. If more than two messages were waiting for the same host, and a transport
4619 filter was specified for the transport, Exim sent two messages over the
4620 same TCP/IP connection, and then failed with "socket operation on non-
4621 socket" when it tried to send the third.
4623 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4625 27. The extern definition of crypt16() in expand.c was not being excluded when
4626 the OS had its own crypt16() function.
4628 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4629 as a preferred synonym for return_size_limit, both as an option and as an
4632 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4634 30. Changed the default configuration ACL to relax the local part checking rule
4635 for addresses that are not in any local domains. For these addresses,
4636 slashes and pipe symbols are allowed within local parts, but the sequence
4637 /../ is explicitly forbidden.
4639 31. SPA server authentication was not clearing the challenge buffer before
4642 32. log_message in a "warn" ACL statement was writing to the reject log as
4643 well as to the main log, which contradicts the documentation and doesn't
4644 seem right (because no rejection is happening). So I have stopped it.
4646 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4647 However, I am unable to do any testing of this.
4649 34. Fixed an infelicity in the appendfile transport. When checking directories
4650 for a mailbox, to see if any needed to be created, it was accidentally
4651 using path names with one or more superfluous leading slashes; tracing
4652 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4654 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4655 contents are added to the log line that is written for every discarded
4656 recipient. (Previously a log_message setting was ignored.)
4658 36. The ${quote: operator now quotes the string if it is empty.
4660 37. The install script runs exim in order to find its version number. If for
4661 some reason other than non-existence or emptiness, which it checks, it
4662 could not run './exim', it was installing it with an empty version number,
4663 i.e. as "exim-". This error state is now caught, and the installation is
4666 38. An argument was missing from the function that creates an error message
4667 when Exim fails to connect to the socket for saslauthd authentication.
4668 This could cause Exim to crash, or give a corrupted message.
4670 39. Added isip, isip4, and isip6 to ${if conditions.
4672 40. The ACL variables $acl_xx are now saved with the message, and can be
4673 accessed later in routers, transports, and filters.
4675 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4676 strings in the file are not string-expanded.
4678 42. If a MAIL command specified a SIZE value that was too large to fit into an
4679 int variable, the check against message_size_limit failed. Such values are
4680 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4681 day this will have to be increased, but I don't think I want to be around
4682 when emails are that large.
4689 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4690 iconv() is not standard in FreeBSD.
4692 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4693 IPv6 enabled. The observed symptom was a segmentation fault on return from
4694 the function os_common_find_running_interfaces() in src/os.c.
4696 3. In the check_special_case() function in daemon.c I had used "errno" as an
4697 argument name, which causes warnings on some systems. This was basically a
4698 typo, since it was named "eno" in the comments!
4700 4. The code that waits for the clock to tick (at a resolution of some fraction
4701 of a second) so as to ensure message-id uniqueness was always waiting for
4702 at least one whole tick, when it could have waited for less. [This is
4703 almost certainly not relevant at current processor speeds, where it is
4704 unlikely to ever wait at all. But we try to future-proof.]
4706 5. The function that sleeps for a time interval that includes fractions of a
4707 second contained a race. It did not block SIGALRM between setting the
4708 timer, and suspending (a couple of lines later). If the interval was short
4709 and the sigsuspend() was delayed until after it had expired, the suspension
4710 never ended. On busy systems this could lead to processes getting stuck for
4713 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4714 process, before it forks any delivery processes. The open lookup caching
4715 mechanism meant that the open file or database connection was passed into
4716 the delivery process. The problem was that delivery processes always tidy
4717 up cached lookup data. This could cause a problem for the next delivery
4718 process started by the queue runner, because the external queue runner
4719 process does not know about the closure. So the next delivery process
4720 still has data in the lookup cache. In the case of a file lookup, there was
4721 no problem because closing a file descriptor in a subprocess doesn't affect
4722 the parent. However, if the lookup was caching a connection to a database,
4723 the connection was closed, and the second delivery process was likely to
4724 see errors such as "PGSQL: query failed: server closed the connection
4725 unexpectedly". The problem has been fixed by closing all cached lookups
4726 in a queue runner before running a delivery process.
4728 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4729 seem to have the "const" qualifier which it has on other OS. I've
4732 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4733 *of the same type* (client or server) with the same public name that an
4734 error should be diagnosed.
4736 9. When Exim looked up a host name for an IP address, but failed to find the
4737 original IP address when looking up the host name (a safety check), it
4738 output the message "<ip address> does not match any IP for NULL", which was
4739 confusing, to say the least. The bug was that the host name should have
4740 appeared instead of "NULL".
4742 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4743 user that is built into the binary, and the -C or -D options is used, root
4744 privilege is dropped before the configuration file is read. In addition,
4745 logging is switched to stderr instead of the normal log files. If the
4746 configuration then re-defines the Exim user, the unprivileged environment
4747 is probably not what is expected, so Exim logs a panic warning message (but
4750 However, if deliver_drop_privilege is set, the unprivileged state may well
4751 be exactly what is intended, so the warning has been cut out in that case,
4752 and Exim is allowed to try to write to its normal log files.
4758 1. smtp_return_error_details was not giving details for temporary sender
4759 or receiver verification errors.
4761 2. Diagnose a configuration error if two authenticators have the same public
4764 3. Exim used not to create the message log file for a message until the first
4765 delivery attempt. This could be confusing when incoming messages were held
4766 for policy or load reasons. The message log file is now created at the time
4767 the message is received, and an initial "Received" line is written to it.
4769 4. The automatically generated man page for command line options had a minor
4770 bug that caused no ill effects; however, a more serious problem was that
4771 the procedure for building the man page automatically didn't always
4772 operate. Consequently, release 4.20 contains an out-of-date version. This
4773 shouldn't happen again.
4775 5. When building Exim with embedded Perl support, the script that builds the
4776 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4777 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4779 6. The freeze_tell option was not being used for messages that were frozen on
4780 arrival, either by an ACL or by local_scan().
4782 7. Added the smtp_incomplete_transaction log selector.
4784 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4785 was accepting AUTH without a new EHLO.
4787 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4788 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4791 10. Logging of TCP/IP connections (when configured) now happens in the main
4792 daemon process instead of the child process, so that the TCP/IP connection
4793 count is more accurate (but it can never be perfect).
4795 11. The use of "drop" in a nested ACL was not being handled correctly in the
4796 outer ACL. Now, if condition failure induced by the nested "drop" causes
4797 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4798 or "require"), the connection is dropped.
4800 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4801 that yield "discard" can now be used with an "accept" or a "discard" verb,
4802 but an error is generated for any others (because I can't see a useful way
4803 to define what should happen).
4805 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4806 are now processed in the same way as lines in the Exim configuration file.
4807 In particular, continuation lines are supported.
4809 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4811 15. Added -ti meaning -t -i.
4813 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4814 domains, and warn by logging if others are found.
4816 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4817 returned from GIFCONF when Exim is trying to find the list of interfaces on
4818 a host. The code in os.c has been modified to copy each ifreq to an aligned
4819 structure in all cases.
4821 Also, in some cases, the returned ifreq's were being copied to a 'struct
4822 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4823 means the last couple of bytes of an IPv6 address could be chopped if the
4824 ifreq contained only a normal sockaddr (14 bytes storage).
4826 18. Named domain lists were not supported in the hosts_treat_as_local option.
4827 An entry such as +xxxx was not recognized, and was treated as a literal
4830 19. Ensure that header lines added by a DATA ACL are included in the reject log
4831 if the ACL subsequently rejects the message.
4833 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4834 MD5 (which is deprecated).
4836 21. When testing a filter file using -bf, Exim was writing a message when it
4837 took the sender from a "From " line in the message, but it was not doing so
4838 when it took $return_path from a Return-Path: header line. It now does.
4840 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4841 with a valid header line field name (a series of printing characters
4842 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4844 23. Changed "disc" in the source to "disk" to conform to the documentation and
4845 the book and for uniformity.
4847 24. Ignore Sendmail's -Ooption=value command line item.
4849 25. When execve() failed while trying to run a command in a pipe transport,
4850 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4851 could be confused with a return value of 69 from the command itself. This
4852 has been changed to 127, the value the shell returns if it is asked to run
4853 a non-existent command. The wording for the related log line suggests a
4854 non-existent command as the problem.
4856 26. If received_header_text expands to an empty string, do not add a Received:
4857 header line to the message. (Well, it adds a token one on the spool, but
4858 marks it "old" so that it doesn't get used or transmitted.)
4860 27. Installed eximstats 1.28 (addition of -nt option).
4862 28. There was no check for failure on the call to getsockname() in the daemon
4863 code. This can fail if there is a shortage of resources on the system, with
4864 ENOMEM, for example. A temporary error is now given on failure.
4866 29. Contrary to the C standard, it seems that in some environments, the
4867 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4868 program. Exim now does this explicitly; it affects the formatting of
4869 timestamps using strftime().
4871 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4872 complaints. I've now initialized all the variables, to avoid this.
4874 32. Header lines added by a system filter were not being "seen" during
4875 transport-time rewrites.
4877 33. The info_callback() function passed to OpenSSL is set up with type void
4878 (*)(SSL *, int, int), as described somewhere. However, when calling the
4879 function (actually a macro) that sets it up, the type void(*)() is
4880 expected. I've put in a cast to prevent warnings from picky compilers.
4882 34. If a DNS black list lookup found a CNAME record, but there were no A
4883 records associated with the domain it pointed at, Exim crashed.
4885 35. If a DNS black list lookup returned more than one A record, Exim ignored
4886 all but the first. It now scans all returned addresses if a particular IP
4887 value is being sought. In this situation, the contents of the
4888 $dnslist_value variable are a list of all the addresses, separated by a
4891 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4892 to accept a host name and all its aliases if the forward lookup for any of
4893 them yielded the IP address of the incoming connection. Now it accepts only
4894 those names whose forward lookup yields the correct IP address. Any other
4895 names are discarded. This closes a loophole whereby a rogue DNS
4896 administrator could create reverse DNS records to break through a
4897 wildcarded host restriction in an ACL.
4899 37. If a user filter or a system filter that ran in a subprocess used any of
4900 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4901 the wrong values were passed to the pipe command ($thisaddress had the
4902 value of $0, $0 had the value of $1, etc). This bug was introduced by
4903 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4905 38. Improved the line breaking for long SMTP error messages from ACLs.
4906 Previously, if there was no break point between 40 and 75 characters, Exim
4907 left the rest of the message alone. Two changes have been made: (a) I've
4908 reduced the minimum length to 35 characters; (b) if it can't find a break
4909 point between 35 and 75 characters, it looks ahead and uses the first one
4910 that it finds. This may give the occasional overlong line, but at least the
4911 remaining text gets split now.
4913 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4914 file descriptors might be low, and that setting 1000 would always raise it.
4915 It turns out that in some environments, the limit is already over 1000 and
4916 that lowering it causes trouble. So now Exim takes care not to decrease it.
4918 40. When delivering a message, the value of $return_path is set to $sender_
4919 address at the start of routing (routers may change the value). By an
4920 oversight, this default was not being set up when an address was tested by
4921 -bt or -bv, which affected the outcome if any router or filter referred to
4924 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4925 log only when "message" or "log_message" are set. However, if one of the
4926 conditions was an address verification, or a call to a nested ACL, the
4927 messages generated by the underlying test were being passed through. This
4928 no longer happens. The underlying message is available in $acl_verify_
4929 message for both "message" and "log_message" expansions, so it can be
4930 passed through if needed.
4932 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4933 new expansion $bh_ to give the encoded byte string without charset
4934 translation. Translation happens only if iconv() is available; HAVE_ICONV
4935 indicates this at build time. HEADERS_CHARSET gives the charset to
4936 translate to; headers_charset can change it in the configuration, and
4937 "headers charset" can change it in an individual filter file.
4939 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4940 that creates RFC 2047 encoded "words" labels them as that charset instead
4941 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4942 expansion operator; (ii) when Exim creates a From: line for a local
4943 message; (iii) when a header line is rewritten to include a "phrase" part.
4945 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4946 buggy, causing it to skip the first lines of messages whose message ID
4947 ended in 'D'. This would not have bitten before Exim release 4.14, because
4948 message IDs were unlikely to end in 'D' before then. The effect was to have
4949 incorrect size information for certain domains.
4951 45. #include "config.h" was missing at the start of the crypt16.c module. This
4952 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4955 46. If there was a timeout during a "random" callout check, Exim treated it as
4956 a failure of the random address, and carried on sending RSET and the real
4957 address. If the delay was just some slowness somewhere, the response to the
4958 original RCPT would be taken as a response to RSET and so on, causing
4959 mayhem of various kinds.
4961 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4962 when I implemented it. It didn't allow for the fact that some option values
4963 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4964 the right test for positive values.
4966 48. Domain names in DNS records are case-independent. Exim always looks them up
4967 in lower case. Some resolvers return domain names in exactly the case they
4968 appear in the zone file, that is, they may contain uppercase letters. Not
4969 all resolvers do this - some return always lower case. Exim was treating a
4970 change of case by a resolver as a change of domain, similar to a widening
4971 of a domain abbreviation. This triggered its re-routing code and so it was
4972 trying to route what was effectively the same domain again. This normally
4973 caused routing to fail (because the router wouldn't handle the domain
4974 twice). Now Exim checks for this case specially, and just changes the
4975 casing of the domain that it ultimately uses when it transmits the message
4978 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4981 50. If a filter generated a file delivery with a non-absolute name (possible if
4982 no home directory exists for the router), the forbid_file option was not
4985 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4986 the existing equality matching.
4988 52. Exim was using ints instead of ino_t variables in some places where it was
4989 dealing with inode numbers.
4991 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4992 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4993 called TMPDIR, and if it finds it is different, it changes its value.
4995 54. The smtp_printf() function is now made available to local_scan() so
4996 additional output lines can be written before returning. There is also an
4997 smtp_fflush() function to enable the detection of a dropped connection.
4998 The variables smtp_input and smtp_batched_input are exported to
5001 55. Changed the default runtime configuration: the message "Unknown user"
5002 has been removed from the ACL, and instead placed on the localuser router,
5003 using the cannot_route_message feature. This means that any verification
5004 failures that generate their own messages won't get overridden. Similarly,
5005 the "Unrouteable address" message that was in the ACL for unverifiable
5006 relay addresses has also been removed.
5008 56. Added hosts_avoid_esmtp to the smtp transport.
5010 57. The exicyclog script was not checking for the esoteric option
5011 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5012 will work only if exicyclog is run under the appropriate euid.
5014 58. Following a discussion on the list, the rules by which Exim recognises line
5015 endings on incoming messages have been changed. The -dropcr and drop_cr
5016 options are now no-ops, retained only for backwards compatibility. The
5017 following line terminators are recognized: LF CRLF CR. However, special
5018 processing applies to CR:
5020 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5021 nor a local message in the state where . is a terminator.
5023 (ii) If a bare CR is encountered in a header line, an extra space is added
5024 after the line terminator so as not to end the header. The reasoning
5025 behind this is that bare CRs in header lines are most likely either
5026 to be mistakes, or people trying to play silly games.
5028 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5029 was being incorrectly given as 18 bytes larger than it should have been.
5030 This is a VOB (very old bug).
5032 60. This may never have affected anything current, but just in case it has:
5033 When the local host is found other than at the start of a list of hosts,
5034 the local host, those with the same MX, and any that follow, are discarded.
5035 When the list in question was part of a longer list of hosts, the following
5036 hosts (not currently being processed) were also being discarded. This no
5037 longer happens. I'm not sure if this situation could ever has previously
5040 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5043 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5044 header. This is something that is recommended in a new Internet Draft, and
5045 is something that is documented as being done by Sendmail. There are two
5046 possible values. For messages generated by the autoreply transport, Exim
5049 Auto-Submitted: auto-replied
5051 whereas for all other generated messages (e.g. bounces) it adds
5053 Auto-Submitted: auto-generated
5055 63. The "personal" condition in filters now includes a test for the
5056 Auto-Submitted: header. If it contains the string "auto-" the message it
5057 not considered personal.
5059 64. Added rcpt_include_affixes as a generic transport option.
5061 65. Added queue_only_override (default true).
5063 66. Added the syslog_duplication option.
5065 67. If what should have been the first header line of a message consisted of
5066 a space followed by a colon, Exim was mis-interpreting it as a header line.
5067 It isn't of course - it is syntactically invalid and should therefore be
5068 treated as the start of the message body. The misbehaviour could have
5069 caused a number of strange effects, including loss of data in subsequent
5070 header lines, and spool format errors.
5072 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5073 client host had authenticated. This control can now be exercised by an ACL
5074 for more flexibility.
5076 69. By default, callouts do not happen when testing with -bh. There is now a
5077 variant, -bhc, which does actually run the callout code, including
5078 consulting and updating the callout cache.
5080 70. Added support for saslauthd authentication, courtesy of Alexander
5083 71. If statvfs() failed on the spool or log directories while checking their
5084 size for availability, Exim confusingly gave the error "space shortage".
5085 Furthermore, in debugging mode it crashed with a floating point exception.
5086 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5087 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5088 serious problem, Exim now writes to the main and panic logs when this
5089 happens, with details of the failure. It then refuses to accept the
5090 incoming message, giving the message "spool directory problem" or "log
5091 directory problem" with a 421 code for SMTP messages.
5093 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5094 0, 1, and 2 exist, because some OS complain for execs without them (see
5095 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5096 descriptors. However, the code omitted to check that the open succeeded,
5097 causing mysterious errors if for some reason the permissions on /dev/null
5098 got screwed. Now Exim writes a message to the main and panic logs, and
5099 bombs out if it can't open /dev/null.
5101 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5102 interact so that it is all more flexible. It is supposed to remain
5103 backwards compatible. Also added extra_local_interfaces.
5105 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5106 to bomb out with an assertion failure - to the client this appears as a
5107 connection drop. This problem occurs in the part of the code that was taken
5108 from the Samba project. Fortunately, the assertion is in a very simple
5109 function, so I have fixed this by reproducing the function inline in the
5110 one place where it is called, and arranging for authentication to fail
5111 instead of killing the process with assert().
5113 75. The SPA client code was not working when the server requested OEM rather
5114 than Unicode encoding.
5116 76. Added code to make require_files with a specific uid setting more usable in
5117 the case where statting the file as root fails - usually a non-root-mounted
5118 NFS file system. When this happens and the failure is EACCES, Exim now
5119 forks a subprocess and does the per-uid checking as the relevant uid.
5121 77. Added process_log_path.
5123 78. If log_file_path was not explicitly set, a setting of check_log_space or
5124 check_log_inodes was ignored.
5126 79. If a space check for the spool or log partitions fails, the incident is now
5127 logged. Of course, in the latter case the data may get lost...
5129 80. Added the %p formatting code to string_format() so that it can be used to
5130 print addresses in debug_print(). Adjusted all the address printing in the
5131 debugging in store.c to use %p rather than %d.
5133 81. There was a concern that a line of code in smtp_in.c could overflow a
5134 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5135 initially expressed, the concern was not well-founded, because trailing
5136 spaces are removed early. However, if the trailing spaces were followed by
5137 a NULL, they did not get removed, so the overflow was possible. Two fixes
5140 (a) I re-wrote the offending code in a cleaner fashion.
5141 (b) If an incoming SMTP command contains a NULL character, it is rejected
5144 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5145 runs initgroups(), so that if the Exim user is in any additional groups,
5146 they will be used during message reception.
5152 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob