1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.632 2010/06/12 15:21:25 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PP/01 Date: & Message-Id: revert to normally being appended to a message,
10 only prepend for the Resent-* case. Fixes regression introduced in
11 Exim 4.70 by NM/22 for Bugzilla 607.
13 PP/02 Include check_rfc2047_length in configure.default because we're seeing
14 increasing numbers of administrators be bitten by this.
16 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
18 PP/03 Bugzilla 994: added openssl_options main configuration option.
20 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
22 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
24 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
26 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
27 without a peer certificate, leading to a segfault because of an
28 assumption that peers always have certificates. Be a little more
29 paranoid. Problem reported by Martin Tscholak.
31 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
32 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
33 NB: ClamAV planning to remove STREAM in "middle of 2010".
34 CL also introduces -bmalware, various -d+acl logging additions and
35 more caution in buffer sizes.
37 PP/09 Implemented reverse_ip expansion operator.
39 PP/10 Bugzilla 937: provide a "debug" ACL control.
41 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
43 PP/12 Bugzilla 973: Implement --version.
45 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
47 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
49 PP/15 Bugzilla 816: support multiple condition rules on Routers.
51 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
52 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
53 ignore trailing whitespace.
55 JJ/02 prevent non-panic DKIM error from being sent to paniclog
57 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
60 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
61 Notification from Dr Andrew Aitchison.
63 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
64 ExtendedDetectionInfo response format.
65 Notification from John Horne.
67 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
70 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
71 XSL and documented dependency on system catalogs, with examples of how
74 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
77 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
78 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
79 configuration file which is writeable by the Exim user or group.
85 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
86 $data_path, and $header_path variables; fixed documentation bugs and
89 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
90 exipick to access non-standard spools, including the "frozen" queue
93 NM/01 Bugzilla 965: Support mysql stored procedures.
94 Patch from Alain Williams
96 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
98 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
99 Patch from Andreas Metzler
101 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
102 Patch from Kirill Miazine
104 NM/05 Bugzilla 671: Added umask to procmail example.
106 JJ/03 installed exipick 20100323.0, fixing doc bug
108 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
109 directory. Notification and patch from Dan Rosenberg.
111 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
113 TK/02 Improve log output when DKIM signing operation fails.
115 MH/01 Treat the transport option dkim_domain as a colon separated
116 list, not as a single string, and sign the message with each element,
117 omitting multiple occurences of the same signer.
119 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
120 Bugzilla 985, 986. Patch by Simon Arlott
122 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
123 Patch by Simon Arlott
125 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
126 MBX locking. Notification from Dan Rosenberg.
132 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
134 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
136 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
138 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
140 NM/04 Bugzilla 915: Flag broken perl installation during build.
146 TK/01 Added patch by Johannes Berg that expands the main option
147 "spamd_address" if it starts with a dollar sign.
149 TK/02 Write list of recipients to X-Envelope-Sender header when building
150 the mbox-format spool file for content scanning (suggested by Jakob
153 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
154 (http://www.dcc-servers.net/) support via dccifd. Activated by
155 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
157 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
158 by Mark Daniel Reidel <mr@df.eu>.
160 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
161 When building exim an external PCRE library is now needed -
162 PCRE is a system library on the majority of modern systems.
163 See entry on PCRE_LIBS in EDITME file.
165 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
166 conversation. Added nologin parameter to request.
167 Patch contributed by Kirill Miazine.
169 TF/01 Do not log submission mode rewrites if they do not change the address.
171 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
173 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
174 log files in place. Contributed by Roberto Lima.
176 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
178 TF/03 Bugzilla 615: When checking the local_parts router precondition
179 after a local_part_suffix or local_part_prefix option, Exim now
180 does not use the address's named list lookup cache, since this
181 contains cached lookups for the whole local part.
183 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
184 Robert Millan. Documentation is in experimental-spec.txt.
186 TF/04 Bugzilla 668: Fix parallel build (make -j).
188 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
190 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
191 Patch provided by Jan Srzednicki.
193 TF/05 Leading white space used to be stripped from $spam_report which
194 wrecked the formatting. Now it is preserved.
196 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
197 that they are available at delivery time.
199 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
201 TF/08 TLS error reporting now respects the incoming_interface and
202 incoming_port log selectors.
204 TF/09 Produce a more useful error message if an SMTP transport's hosts
205 setting expands to an empty string.
207 NM/06 Bugzilla 744: EXPN did not work under TLS.
208 Patch provided by Phil Pennock.
210 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
211 Patch provided by Richard Godbee.
213 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
214 acl_smtp_notquit, added index entry.
216 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
217 Patch provided by Eugene Bujak.
219 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
220 accept(). Patch provided by Maxim Dounin.
222 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
223 Patch provided by Phil Pennock.
225 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
227 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
228 Patch provided by Brad "anomie" Jorsch.
230 NM/14 Bugzilla 622: Added timeout setting to transport filter.
231 Patch provided by Dean Brooks.
233 TK/05 Add native DKIM support (does not depend on external libraries).
235 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
236 Patch provided by Graeme Fowler.
238 NM/16 Bugzilla 851: Documentation example syntax fix.
240 NM/17 Changed NOTICE file to remove references to embedded PCRE.
242 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
245 NM/19 Bugzilla 745: TLS version reporting.
246 Patch provided by Phil Pennock.
248 NM/20 Bugzilla 167: bool: condition support.
249 Patch provided by Phil Pennock.
251 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
252 clients. Patch provided by Phil Pennock.
254 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
255 Patch provided by Brad "anomie" Jorsch.
257 NM/23 Bugzilla 687: Fix misparses in eximstats.
258 Patch provided by Heiko Schlittermann.
260 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
261 Patch provided by Heiko Schlittermann.
263 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
264 plus update to original patch.
266 NM/26 Bugzilla 799: Documentation correction for ratelimit.
268 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
269 Patch provided by David Brownlee.
271 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
273 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
275 NM/30 Bugzilla 888: TLS documentation bugfixes.
277 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
279 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
280 Unlike the original bugzilla I have changed all shell scripts in src tree.
282 NM/33 Bugzilla 898: Transport filter timeout fix.
283 Patch by Todd Rinaldo.
285 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
286 Patch by Serge Demonchaux.
288 NM/35 Bugzilla 39: Base64 decode bug fixes.
289 Patch by Jakob Hirsch.
291 NM/36 Bugzilla 909: Correct connect() call in dcc code.
293 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
295 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
297 NM/39 Bugzilla 911: Fixed MakeLinks build script.
303 TK/01 Add preliminary DKIM support. Currently requires a forked version of
304 ALT-N's libdkim that I have put here:
305 http://duncanthrax.net/exim-experimental/
307 Note to Michael Haardt: I had to rename some vars in sieve.c. They
308 were called 'true' and it seems that C99 defines that as a reserved
309 keyword to be used with 'bool' variable types. That means you could
310 not include C99-style headers which use bools without triggering
311 build errors in sieve.c.
313 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
314 as mailq or other aliases. Changed the --help handling significantly
315 to do whats expected. exim_usage() emits usage/help information.
317 SC/01 Added the -bylocaldomain option to eximstats.
319 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
321 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
323 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
329 PH/01 Another patch from the Sieve maintainer.
331 PH/02 When an IPv6 address is converted to a string for single-key lookup
332 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
333 dots are used instead of colons so that keys in lsearch files need not
334 contain colons. This was done some time before quoting was made available
335 in lsearch files. However, iplsearch files do require colons in IPv6 keys
336 (notated using the quote facility) so as to distinguish them from IPv4
337 keys. This meant that lookups for IP addresses in host lists did not work
338 for iplsearch lookups.
340 This has been fixed by arranging for IPv6 addresses to be expressed with
341 colons if the lookup type is iplsearch. This is not incompatible, because
342 previously such lookups could never work.
344 The situation is now rather anomolous, since one *can* have colons in
345 ordinary lsearch keys. However, making the change in all cases is
346 incompatible and would probably break a number of configurations.
348 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
351 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
352 conversion specification without a maximum field width, thereby enabling
353 a rogue spamd server to cause a buffer overflow. While nobody in their
354 right mind would setup Exim to query an untrusted spamd server, an
355 attacker that gains access to a server running spamd could potentially
356 exploit this vulnerability to run arbitrary code as the Exim user.
358 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
359 $primary_hostname instead of what libspf2 thinks the hosts name is.
361 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
362 a directory entry by the name of the lookup key. Previously, if a
363 symlink pointed to a non-existing file or a file in a directory that
364 Exim lacked permissions to read, a lookup for a key matching that
365 symlink would fail. Now it is enough that a matching directory entry
366 exists, symlink or not. (Bugzilla 503.)
368 PH/03 The body_linecount and body_zerocount variables are now exported in the
371 PH/04 Added the $dnslist_matched variable.
373 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
374 This means they are set thereafter only if the connection becomes
377 PH/06 Added the client_condition to authenticators so that some can be skipped
378 by clients under certain conditions.
380 PH/07 The error message for a badly-placed control=no_multiline_responses left
381 "_responses" off the end of the name.
383 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
385 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
386 (without spaces) instead of just copying the configuration text.
388 PH/10 Added the /noupdate option to the ratelimit ACL condition.
390 PH/11 Added $max_received_linelength.
392 PH/12 Added +ignore_defer and +include_defer to host lists.
394 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
395 way in which PCRE > 7.0 is built.
397 PH/14 Implemented queue_only_load_latch.
399 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
400 MAIL command. The effect was to mangle the value on 64-bit systems.
402 PH/16 Another patch from the Sieve maintainer.
404 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
406 PH/18 If a system quota error occurred while trying to create the file for
407 a maildir delivery, the message "Mailbox is full" was not appended to the
408 bounce if the delivery eventually timed out. Change 4.67/27 below applied
409 only to a quota excession during the actual writing of the file.
411 PH/19 It seems that peer DN values may contain newlines (and other non-printing
412 characters?) which causes problems in log lines. The DN values are now
413 passed through string_printing() before being added to log lines.
415 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
416 and InterBase are left for another time.)
418 PH/21 Added message_body_newlines option.
420 PH/22 Guard against possible overflow in moan_check_errorcopy().
422 PH/23 POSIX allows open() to be a macro; guard against that.
424 PH/24 If the recipient of an error message contained an @ in the local part
425 (suitably quoted, of course), incorrect values were put in $domain and
426 $local_part during the evaluation of errors_copy.
432 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
433 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
436 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
437 issue a MAIL command.
439 PH/02 In an ACL statement such as
441 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
443 if a client was not listed at all, or was listed with a value other than
444 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
445 the condition was not true (as it should be), so access was not denied.
446 The bug was that the ! inversion was incorrectly passed on to the second
447 item. This has been fixed.
449 PH/03 Added additional dnslists conditions == and =& which are different from
450 = and & when the dns lookup returns more than one IP address.
452 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
453 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
455 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
456 FSYNC, which compiles an option called disable_fsync that allows for
457 bypassing fsync(). The documentation is heavily laced with warnings.
459 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
461 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
462 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
463 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
464 including adding "make clean"; (3) Added -fPIC when compiling the test
465 dynamically loaded module, to get rid of a warning.
467 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
468 message fails, move_frozen_messages = true and ignore_bounce_errors_after
469 = 0s. The bug is otherwise harmless.
471 PH/07 There was a bug in the dovecot authenticator such that the value of
472 $auth1 could be overwritten, and so not correctly preserved, after a
473 successful authentication. This usually meant that the value preserved by
474 the server_setid option was incorrect.
476 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
478 PH/09 Installed PCRE release 7.0.
480 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
481 run for batched SMTP input. It is now run at the start of every message
482 in the batch. While fixing this I discovered that the process information
483 (output by running exiwhat) was not always getting set for -bs and -bS
484 input. This is fixed, and it now also says "batched" for BSMTP.
486 PH/11 Added control=no_pipelining.
488 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
489 patch, slightly modified), and move the expansion of helo_data till after
490 the connection is made in the smtp transport (so it can use these
493 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
495 PH/14 Added log_selector = +pid.
497 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
499 PH/16 Add ${if forany and ${if forall.
501 PH/17 Added dsn_from option to vary the From: line in DSNs.
503 PH/18 Flush SMTP output before performing a callout, unless control =
504 no_callout_flush is set.
506 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
507 was true (the default) a successful delivery failed to delete the retry
508 item, thus causing premature timeout of the address. The bug is now
511 PH/20 Added hosts_avoid_pipelining to the smtp transport.
513 PH/21 Long custom messages for fakedefer and fakereject are now split up
514 into multiline reponses in the same way that messages for "deny" and
515 other ACL rejections are.
517 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
518 with slight modification.
520 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
521 draft, changing the syntax and factoring some duplicate code".
523 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
524 for deliveries of the second and subsequent messages over the same SMTP
527 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
528 ${reduce, with only minor "tidies".
530 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
532 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
533 expansion side effects.
535 PH/27 When a message times out after an over-quota error from an Exim-imposed
536 quota, the bounce message says "mailbox is full". This message was not
537 being given when it was a system quota that was exceeded. It now should
540 MH/03 Made $recipients available in local_scan(). local_scan() already has
541 better access to the recipient list through recipients_list[], but
542 $recipients can be useful in postmaster-provided expansion strings.
544 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
545 in the case of a MAIL command with additional options following the
546 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
547 were accidentally chopped off.
549 PH/29 SMTP synchronization checks are implemented when a command is read -
550 there is a check that no more input is waiting when there shouldn't be
551 any. However, for some commands, a delay in an ACL can mean that it is
552 some time before the response is written. In this time, more input might
553 arrive, invalidly. So now there are extra checks after an ACL has run for
554 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
555 pipelining has not been advertised.
557 PH/30 MH's patch to allow iscntrl() characters to be list separators.
559 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
560 returned in the SMTP response when smtp_return_error_details was false.
563 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
564 instead of the C I/O that was originally supplied, because problems were
567 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
568 Exim which did not show up earlier: it was assuming that a call to
569 SSL_CTX_set_info_callback() might give an error value. In fact, there is
570 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
571 was a macro that became an assignment, so it seemed to work. This has
572 changed to a proper function call with a void return, hence the compile
573 error. Exim's code has been fixed.
575 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
578 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
580 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
582 PH/37 If a message is not accepted after it has had an id assigned (e.g.
583 because it turns out to be too big or there is a timeout) there is no
584 "Completed" line in the log. When some messages of this type were
585 selected by exigrep, they were listed as "not completed". Others were
586 picked up by some special patterns. I have improved the selection
587 criteria to be more general.
589 PH/38 The host_find_failed option in the manualroute router can now be set
590 to "ignore", to completely ignore a host whose IP address cannot be
591 found. If all hosts are ignored, the behaviour is controlled by the new
592 host_all_ignored option.
594 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
595 homing or because of multiple MX records with /mx) generated more than
596 one IP address, and the following item turned out to be the local host,
597 all the secondary addresses of the first item were incorrectly removed
598 from the list, along with the local host and any following hosts (which
599 is what is supposed to happen).
601 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
602 whoever called Exim into the -H file. In the case of the daemon it was
603 behaving confusingly. When first started, it used values for whoever
604 started the daemon, but after a SIGHUP it used the Exim user (because it
605 calls itself on a restart). I have changed the code so that it now always
608 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
609 message are rejected with the same error (e.g. no authentication or bad
610 sender address), and a DATA command is nevertheless sent (as can happen
611 with PIPELINING or a stupid MUA), the error message that was given to the
612 RCPT commands is included in the rejection of the DATA command. This is
613 intended to be helpful for MUAs that show only the final error to their
616 PH/42 Another patch from the Sieve maintainer.
618 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
619 Eximstats - Fixed some broken HTML links and added missing column headers
621 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
622 columns for Rejects, Temp Rejects, Ham, and Spam rows.
624 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
626 PH/43 Yet another patch from the Sieve maintainer.
628 PH/44 I found a way to check for a TCP/IP connection going away before sending
629 the response to the final '.' that terminates a message, but only in the
630 case where the client has not sent further data following the '.'
631 (unfortunately, this is allowed). However, in many cases there won't be
632 any further data because there won't be any more messages to send. A call
633 to select() can be used: if it shows that the input is "ready", there is
634 either input waiting, or the socket has been closed. An attempt to read
635 the next input character can distinguish the two cases. Previously, Exim
636 would have sent an OK response which the client would never have see.
637 This could lead to message repetition. This fix should cure that, at
638 least in a lot of common cases.
640 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
641 advertised in response to EHLO.
647 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
648 fixed by 4.65/MH/01 (is this a record?) are fixed:
650 (i) An empty string was always treated as zero by the numeric comparison
651 operators. This behaviour has been restored.
653 (ii) It is documented that the numeric comparison operators always treat
654 their arguments as decimal numbers. This was broken in that numbers
655 starting with 0 were being interpreted as octal.
657 While fixing these problems I realized that there was another issue that
658 hadn't been noticed. Values of message_size_limit (both the global option
659 and the transport option) were treated as octal if they started with 0.
660 The documentation was vague. These values are now always treated as
661 decimal, and I will make that clear in the documentation.
667 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
668 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
671 MH/01 Don't check that the operands of numeric comparison operators are
672 integers when their expansion is in "skipping" mode (fixes bug
673 introduced by 4.64-PH/07).
675 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
676 child addresses, Exim now panics and dies. Previously, because the count
677 is held in a short int, deliveries were likely to be lost. As such a
678 large number of recipients for a single message is ridiculous
679 (performance will be very, very poor), I have chosen to impose a limit
680 rather than extend the field.
686 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
687 leftover -K file (the existence of which was triggered by #402).
688 While we were at it, introduced process PID as part of the -K
689 filename. This should rule out race conditions when creating
692 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
693 processing considerably. Previous code took too long for large mails,
694 triggering a timeout which in turn triggers #401.
696 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
697 in the DK code in transports.c. sendfile() is not really portable,
698 hence the _LINUX specificness.
700 TF/01 In the add_headers option to the mail command in an Exim filter,
701 there was a bug that Exim would claim a syntax error in any
702 header after the first one which had an odd number of characters
705 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
706 callout verification, Exim cached a "reject" for the entire domain. This
707 is correct for most verifications, but it is not correct for a recipient
708 verification with use_sender or use_postmaster set, because in that case
709 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
710 case of MAIL FROM:<> rejection from other early rejections (e.g.
711 rejection of HELO). When verifying a recipient using a non-null MAIL
712 address, the cache is ignored if it shows MAIL FROM:<> rejection.
713 Whatever the result of the callout, the value of the domain cache is
714 left unchanged (for any other kind of callout, getting as far as trying
715 RCPT means that the domain itself is ok).
717 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
720 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
721 manifest itself as EPIPE rather than ECONNECT. When tidying away a
722 session, the daemon ignores ECONNECT errors and logs others; it now
723 ignores EPIPE as well.
725 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
726 (quoted-printable decoding).
728 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
729 later the small subsequent patch to fix an introduced bug.
731 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
733 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
735 PH/08 An error is now given if message_size_limit is specified negative.
737 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
738 to be given (somewhat) arbitrary names.
740 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
743 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
744 miscellaneous code fixes
746 PH/10 Added the log_reject_target ACL modifier to specify where to log
749 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
750 hostname. This is wrong, because it relates to the incoming message (and
751 probably the interface on which it is arriving) and not to the outgoing
752 callout (which could be using a different interface). This has been
753 changed to use the value of the helo_data option from the smtp transport
754 instead - this is what is used when a message is actually being sent. If
755 there is no remote transport (possible with a router that sets up host
756 addresses), $smtp_active_hostname is used.
758 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
759 tweaks were necessary in order to get it to work (see also 21 below):
760 (a) The code assumed that strncpy() returns a negative number on buffer
761 overflow, which isn't the case. Replaced with Exim's string_format()
763 (b) There were several signed/unsigned issues. I just did the minimum
764 hacking in of casts. There is scope for a larger refactoring.
765 (c) The code used strcasecmp() which is not a standard C function.
766 Replaced with Exim's strcmpic() function.
767 (d) The code set only $1; it now sets $auth1 as well.
768 (e) A simple test gave the error "authentication client didn't specify
769 service in request". It would seem that Dovecot has changed its
770 interface. Fortunately there's a specification; I followed it and
771 changed what the client sends and it appears to be working now.
773 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
776 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
777 address A is aliased to B and C, where B exists and C does not. Without
778 -v the output is "A verified" because verification stops after a
779 successful redirection if more than one address is generated. However,
780 with -v the child addresses are also verified. Exim was outputting "A
781 failed to verify" and then showing the successful verification for C,
782 with its parentage. It now outputs "B failed to verify", showing B's
783 parentage before showing the successful verification of C.
785 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
786 look up a TXT record in a specific list after matching in a combined
789 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
790 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
791 they consult the DNS. I had assumed they would set it the way they
792 wanted; and indeed my experiments on Linux seem to show that in some
793 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
794 To be on the safe side, however, I have now made the interface to
795 host_find_byname() similar to host_find_bydns(), with an argument
796 containing the DNS resolver options. The host_find_byname() function now
797 sets these options at its start, just as host_find_bydns() does. The smtp
798 transport options dns_qualify_single and dns_search_parents are passed to
799 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
800 of host_find_byname() use the default settings of RES_DEFNAMES
801 (qualify_single) but not RES_DNSRCH (search_parents).
803 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
804 spool_read_header() do less string testing, by means of a preliminary
805 switch on the second character of optional "-foo" lines. (This is
806 overdue, caused by the large number of possibilities that now exist.
807 Originally there were few.) While I was there, I also converted the
808 str(n)cmp tests so they don't re-test the leading "-" and the first
809 character, in the hope this might squeeze out yet more improvement.
811 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
812 flag allowing group syntax was set by the header_syntax check but not
813 turned off, possible causing trouble later; (2) The flag was not being
814 set at all for the header_verify test, causing "group"-style headers to
815 be rejected. I have now set it in this case, and also caused header_
816 verify to ignore an empty address taken from a group. While doing this, I
817 came across some other cases where the code for allowing group syntax
818 while scanning a header line wasn't quite right (mostly, not resetting
819 the flag correctly in the right place). These bugs could have caused
820 trouble for malformed header lines. I hope it is now all correct.
822 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
823 with the "reply" argument non-NULL. The code, however (which originally
824 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
825 but it didn't always do it. This confused somebody who was copying the
826 code for some other use. I have removed all the tests.
828 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
829 feature that was used to support insecure browsers during the U.S. crypto
830 embargo. It requires special client support, and Exim is probably the
831 only MTA that supported it -- and would never use it because real RSA is
832 always available. This code has been removed, because it had the bad
833 effect of slowing Exim down by computing (never used) parameters for the
834 RSA_EXPORT functionality.
836 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
837 authenticator to fail if there's a tab character in the incoming data
838 (there should never be unless someone is messing about, as it's supposed
839 to be base64-encoded). Also added, on Timo's advice, the "secured" option
840 if the connection is using TLS or if the remote IP is the same as the
841 local IP, and the "valid-client-cert option" if a client certificate has
844 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
845 authenticators. This can be used for authorization after authentication
846 succeeds. (In the case of plaintext, it servers for both authentication
849 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
850 if any retry times were supplied.
852 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
853 connection, where there is no client IP address to check. In this
854 situation, the verify now always succeeds.
856 PH/25 Applied John Jetmore's -Mset patch.
858 PH/26 Added -bem to be like -Mset, but loading a message from a file.
860 PH/27 In a string expansion for a processed (not raw) header when multiple
861 headers of the same name were present, leading whitespace was being
862 removed from all of them, but trailing whitespace was being removed only
863 from the last one. Now trailing whitespace is removed from each header
864 before concatenation. Completely empty headers in a concatenation (as
867 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
868 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
870 PH/29 [Removed. This was a change that I later backed out, and forgot to
871 correct the ChangeLog entry (that I had efficiently created) before
872 committing the later change.]
874 PH/30 Exim was sometimes attempting to deliver messages that had suffered
875 address errors (4xx response to RCPT) over the same connection as other
876 messages routed to the same hosts. Such deliveries are always "forced",
877 so retry times are not inspected. This resulted in far too many retries
878 for the affected addresses. The effect occurred only when there were more
879 hosts than the hosts_max_try setting in the smtp transport when it had
880 the 4xx errors. Those hosts that it had tried were not added to the list
881 of hosts for which the message was waiting, so if all were tried, there
882 was no problem. Two fixes have been applied:
884 (i) If there are any address or message errors in an SMTP delivery, none
885 of the hosts (tried or untried) are now added to the list of hosts
886 for which the message is waiting, so the message should not be a
887 candidate for sending over the same connection that was used for a
888 successful delivery of some other message. This seems entirely
889 reasonable: after all the message is NOT "waiting for some host".
890 This is so "obvious" that I'm not sure why it wasn't done
891 previously. Hope I haven't missed anything, but it can't do any
892 harm, as the worst effect is to miss an optimization.
894 (ii) If, despite (i), such a delivery is accidentally attempted, the
895 routing retry time is respected, so at least it doesn't keep
896 hammering the server.
898 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
899 in ${readsocket because some servers need this prod.
901 PH/32 Added some extra debug output when updating a wait-xxx database.
903 PH/33 The hint "could be header name not terminated by colon", which has been
904 given for certain expansion errors for a long time, was not being given
905 for the ${if def:h_colon_omitted{... case.
907 PH/34 The spec says: "With one important exception, whenever a domain list is
908 being scanned, $domain contains the subject domain." There was at least
909 one case where this was not true.
911 PH/35 The error "getsockname() failed: connection reset by peer" was being
912 written to the panic log as well as the main log, but it isn't really
913 panic-worthy as it just means the connection died rather early on. I have
914 removed the panic log writing for the ECONNRESET error when getsockname()
917 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
918 runs only) independently of the message's sender address. This meant
919 that, if the 4xx error was in fact related to the sender, a different
920 message to the same recipient with a different sender could confuse
921 things. In particualar, this can happen when sending to a greylisting
922 server, but other circumstances could also provoke similar problems.
923 I have changed the default so that the retry time for these errors is now
924 based a combination of the sender and recipient addresses. This change
925 can be overridden by setting address_retry_include_sender=false in the
928 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
929 remote server are returned as part of bounce messages. This was not
930 happening for LMTP over a pipe (the lmtp transport), but now it is the
931 same for both kinds of LMTP.
933 PH/38 Despite being documented as not happening, Exim was rewriting addresses
934 in header lines that were in fact CNAMEs. This is no longer the case.
936 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
937 and queue runs started by the daemon processed all messages. This has
938 been fixed so that -R and -S can now usefully be given with -q<time>.
940 PH/40 Import PCRE release 6.7 (fixes some bugs).
942 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
944 PH/42 Give an error if -q is specified more than once.
946 PH/43 Renamed the variables $interface_address and $interface_port as
947 $received_ip_address and $received_port, to make it clear that these
948 values apply to message reception, and not to the outgoing interface when
949 a message is delivered. (The old names remain recognized, of course.)
951 PH/44 There was no timeout on the connect() call when using a Unix domain
952 socket in the ${readsocket expansion. There now is.
954 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
955 be meaningful with "accept".
957 SC/01 Eximstats V1.43
958 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
960 SC/02 Eximstats V1.44
961 Use a glob alias rather than an array ref in the generated
962 parser. This improves both readability and performance.
964 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
965 Collect SpamAssassin and rejection statistics.
966 Don't display local sender or destination tables unless
967 there is data to show.
968 Added average volumes into the top table text output.
970 SC/04 Eximstats V1.46
971 Collect data on the number of addresses (recipients)
972 as well as the number of messages.
974 SC/05 Eximstats V1.47
975 Added 'Message too big' to the list of mail rejection
976 reasons (thanks to Marco Gaiarin).
978 SC/06 Eximstats V1.48
979 Mainlog lines which have GMT offsets and are too short to
980 have a flag are now skipped.
982 SC/07 Eximstats V1.49 (Alain Williams)
983 Added the -emptyok flag.
985 SC/08 Eximstats V1.50
986 Fixes for obtaining the IP address from reject messages.
988 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
989 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
990 whitesspace changes from 4.64-PH/27
992 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
995 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
996 are found, allow negative numbers in numeric criteria)
998 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1000 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1001 to match changes made in 4.64-PH/43
1003 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1005 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1006 "domain" is actually an IP address, and force a failure. This locks out
1007 those revolvers/nameservers that support "A-for-A" lookups, in
1008 contravention of the specifications.
1010 PH/48 When a host name was looked up from an IP address, and the subsequent
1011 forward lookup of the name timed out, the host name was left in
1012 $sender_host_name, contrary to the specification.
1014 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1015 restricted to single-key lookups, Exim was not diagnosing an error if
1016 * or *@ was used with a query-style lookup.
1018 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1020 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1021 long ago, but noone interested enough thought of it. Let's just say that
1022 the "1.1" means that there are some new functions that weren't there at
1023 some point in the past.
1025 PH/51 Error processing for expansion failure of helo_data from an smtp
1026 transport during callout processing was broken.
1028 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1029 tested/used via the -bh/-bhc/-bs options.
1031 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1032 bug, fixed in subsequent PCRE releases).
1034 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1035 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1037 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1043 SC/01 Use a glob alias rather than an array ref in eximstats generated
1044 parser. This improves both readability and performance.
1046 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1047 Don't display local sender or destination tables in eximstats unless
1048 there is data to show.
1049 Added average volumes into the eximstats top table text output.
1051 SC/03 Collect data on the number of addresses (recipients) as well
1052 as the number of messages in eximstats.
1054 TF/01 Correct an error in the documentation for the redirect router. Exim
1055 does (usually) call initgroups() when daemonizing.
1057 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1058 with consistent privilege compared to when running as a daemon.
1060 TF/03 Note in the spec that $authenticated_id is not set for local
1061 submissions from trusted users.
1063 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1064 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1066 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1067 by adding some example configuration directives to the default
1068 configuration file. A little bit of work is required to uncomment the
1069 directives and define how usernames and passwords are checked, but
1070 there is now a framework to start from.
1072 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1073 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1074 without this. I don't know how relevant this is to other LDAP libraries.
1076 PH/02 Add the verb name to the "unknown ACL verb" error.
1078 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1080 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1082 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1083 directory not expanded when it should be if an expanded home directory
1084 was set for the address (which is overridden by the transport).
1086 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1089 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1090 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1091 because it is too late at that time, and has no effect.
1093 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1094 security issue with \' (bugzilla #107). I could not use the
1095 PQescapeStringConn() function, because it needs a PGconn value as one of
1098 PH/08 When testing addresses using -bt, indicate those final addresses that
1099 are duplicates that would not cause an additional delivery. At least one
1100 person was confused, thinking that -bt output corresponded to deliveries.
1101 (Suppressing duplicates isn't a good idea as you lose the information
1102 about possibly different redirections that led to the duplicates.)
1104 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1105 systems where poll() doesn't work, in particular OS X.
1107 PH/10 Added more information to debugging output for retry time not reached.
1109 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1110 operations in malware.c.
1112 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1115 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1116 syslog_duplication set false, log lines that would normally be written
1117 both the the main log and to the reject log were not written to syslog at
1120 PH/14 In the default configuration, change the use of "message" in ACL warn
1121 statements to "add_header".
1123 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1124 not followed by a command (e.g. "seen endif").
1126 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1127 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1130 PH/17 Added extra conditions to the default value of delay_warning_condition
1134 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1135 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1136 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1139 The Auto-Submitted: and various List- headers are standardised, whereas I
1140 don't think Precedence: ever was.
1142 PH/18 Refactored debugging code in route_finduser() to show more information,
1143 in particular, the error code if getpwnam() issues one.
1145 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1146 This is apparently needed in addition to the PH/07 change above to avoid
1147 any possible encoding problems.
1149 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1150 but not after initializing Perl.
1152 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1153 output them only if debugging. By default they are written stderr,
1154 apparently, which is not desirable.
1156 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1159 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1162 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1164 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1165 authenticated or an ident call has been made. Suppress the default
1166 values for $authenticated_id and $authenticated_sender (but permit -oMai
1167 and -oMas) when testing with -bh.
1169 PH/24 Re-jigged the order of the tests in the default configuration so that the
1170 tests for valid domains and recipients precede the DNS black list and CSA
1171 tests, on the grounds that those ones are more expensive.
1173 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1174 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1175 This bug exists in every version of Exim that I still have, right back to
1178 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1179 However, an attempt to turn on case-sensitivity in a regex key by
1180 including (?-i) didn't work because the subject string was already
1181 lowercased, and the effects were non-intuitive. It turns out that a
1182 one-line patch can be used to allow (?-i) to work as expected.
1188 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1189 other effects) broke the use of negated acl sub-conditions.
1191 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1194 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1195 "Deny" causes Exim to reject the incoming connection with a 554 error.
1196 Unfortunately, if there is a major crisis, such as a disk failure,
1197 tcp-wrappers gives "deny", whereas what one would like would be some
1198 kind of temporary error. A kludge has been added to help with this.
1199 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1200 554 error is used if errno is still zero or contains ENOENT (which occurs
1201 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1204 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1206 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1207 errors. Otherwise a message that provokes a temporary error (when other
1208 messages do not) can cause a whole host to time out.
1210 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1211 addresses were routed directly to files or pipes from a redirect router.
1212 File deliveries just didn't batch; pipe deliveries might have suffered
1215 PH/06 A failure to get a lock for a hints database would erroneously always say
1216 "Failed to get write lock", even when it was really a read lock.
1218 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1219 of 0600. This has been changed to use the value of the lockfile_mode
1220 option (which defaults to 0600).
1222 PH/08 Applied small patch from the Sieve maintainer.
1224 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1225 folder from quota calculations, a direct delivery into this folder messed
1226 up the contents of the maildirsize file. This was because the regex was
1227 used only to exclude .Trash (or whatever) when the size of the mailbox
1228 was calculated. There was no check that a delivery was happening into an
1229 excluded directory. This bug has been fixed by ignoring all quota
1230 processing for deliveries into excluded directories.
1232 PH/10 Added the maildirfolder_create_regex option to appendfile.
1238 PH/01 The code for finding all the local interface addresses on a FreeBSD
1239 system running IPv6 was broken. This may well have applied to all BSD
1240 systems, as well as to others that have similar system calls. The broken
1241 code found IPv4 interfaces correctly, but gave incorrect values for the
1242 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1243 that it would not match correctly against @[] and not recognize the IPv6
1246 PH/02 The ipliteral router was not recognizing addresses of the form user@
1247 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1249 PH/03 Added disable_ipv6.
1251 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1252 decoded form, because it is most often used to construct To: headers
1253 lines in autoreplies, and the decoded form may well be syntactically
1254 invalid. However, $reply_address has leading white space removed, and all
1255 newlines turned into spaces so that the autoreply transport does not
1258 PH/05 If group was specified without a user on a router, and no group or user
1259 was specified on a transport, the group from the router was ignored.
1261 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1262 for visible compile-time settings that can be used to change these
1263 numbers, for those that want even more. Backwards compatibility with old
1264 spool files has been maintained. However, going back to a previous Exim
1265 release will lost any variables that are in spool files.
1267 PH/07 Two small changes when running in the test harness: increase delay when
1268 passing a TCP/IP connection to a new process, in case the original
1269 process has to generate a bounce, and remove special handling of
1270 127.0.0.2 (sic), which is no longer necessary.
1272 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1273 be the same on different OS.
1275 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1278 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1279 whether --show-vars was specified or not
1281 JJ/02 exipick: Added support for new ACL variable spool format introduced
1284 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1285 syntactically invalid From: or Reply-to: line, and a filter used this to
1286 generate an autoreply, and therefore failed to obtain an address for the
1287 autoreply, Exim could try to deliver to a non-existent relative file
1288 name, causing unrelated and misleading errors. What now happens is that
1289 it logs this as a hard delivery error, but does not attempt to create a
1292 PH/11 The exinext utility has a -C option for testing purposes, but although
1293 the given file was scanned by exinext itself; it wasn't being passed on
1294 when Exim was called.
1296 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1297 an end-of-file indication when reading a command response.
1299 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1300 compiled. In many other places in Exim, IPv6 addresses are always
1301 recognized, so I have changed this. It also means that IPv4 domain
1302 literals of the form [IPV4:n.n.n.n] are now always recognized.
1304 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1305 used if the router is not running as root, for example, when verifying at
1306 ACL time, or when using -bh. The debugging output from this situation was
1307 non-existent - all you got was a failure to exec. I have made two
1310 (a) Failures to set uid/gid, the current directory, or a process leader
1311 in a subprocess such as that created by queryprogram now generate
1312 suitable debugging ouput when -d is set.
1314 (b) The queryprogram router detects when it is not running as root,
1315 outputs suitable debugging information if -d is set, and then runs
1316 the subprocess without attempting to change uid/gid.
1318 PH/15 Minor change to Makefile for building test_host (undocumented testing
1321 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1322 additional section of a DNS packet that returns MX or SRV records.
1323 Instead, it always explicitly searches for A/AAAA records. This avoids
1324 major problems that occur when a DNS server includes only records of one
1325 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1326 fixed another bug: if SRV records were looked up and the corresponding
1327 address records were *not* found in the additional section, the port
1328 values from the SRV records were lost.
1330 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1331 using the correct key (the original address) when searching the retry
1332 rules in order to find which one to use for generating the retry hint.
1334 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1335 adding the default one. Similarly, if it contains a Reply-To: header, the
1336 errors_reply_to option, if set, is not used.
1338 PH/19 When calculating a retry time, Exim used to measure the "time since
1339 failure" by looking at the "first failed" field in the retry record. Now
1340 it does not use this if it is later than than the arrival time of the
1341 message. Instead it uses the arrival time. This makes for better
1342 behaviour in cases where some deliveries succeed, thus re-setting the
1343 "first failed" field. An example is a quota failure for a huge message
1344 when small messages continue to be delivered. Without this change, the
1345 "time since failure" will always be short, possible causing more frequent
1346 delivery attempts for the huge message than are intended.
1347 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1349 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1350 $1, $2, $3) because the numerical variables can be reset during some
1351 expansion items (e.g. "match"), thereby losing the authentication data.
1353 PH/21 Make -bV show the size of off_t variables so that the test suite can
1354 decide whether to run tests for quotas > 2G.
1356 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1357 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1358 filecount value is greater than 2G or if a quota value is greater than 2G
1359 on a system where the size of off_t is not greater than 4, a panic error
1362 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1363 never match. The debug and -bh output now contains an explicit error
1364 message indicating a malformed IPv4 address or mask.
1366 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1367 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1368 PH/23 above applies.
1370 PH/25 Do not write to syslog when running in the test harness. The only
1371 occasion when this arises is a failure to open the main or panic logs
1372 (for which there is an explicit test).
1374 PH/26 Added the /no_tell option to "control=freeze".
1376 PH/27 If a host name lookup failed very early in a connection, for example, if
1377 the IP address matched host_lookup and the reverse lookup yielded a name
1378 that did not have a forward lookup, an error message of the form "no IP
1379 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1380 could be logged. Now it outputs the IP address instead of "NULL".
1382 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1383 allows the sender and the authenticated sender to be set when
1384 submitting a message from within Exim. Since child_open_exim() is
1385 documented for local_scan(), the new function should be too.
1387 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1388 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1389 results in an empty string is now treated as unset.
1391 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1393 PH/31 Added sender_verify_fail logging option.
1395 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1396 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1397 tidied the source and removed it altogether.
1399 PH/33 When a queue run was abandoned because the load average was too high, a
1400 log line was always written; now it is written only if the queue_run log
1401 selector is set. In addition, the log line for abandonment now contains
1402 information about the queue run such as the pid. This is always present
1403 in "start" and "stop" lines but was omitted from the "abandon" line.
1405 PH/34 Omit spaces between a header name and the colon in the error message that
1406 is given when verify = headers_syntax fails (if there are lots of them,
1407 the message gets confusing).
1409 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1410 names, as there are now some PTR records that contain slashes. This check
1411 is only to protect against broken name servers that fall over on strange
1412 characters, so the fact that it applies to all lookups doesn't matter.
1414 PH/36 Now that the new test suite is complete, we can remove some of the
1415 special code in Exim that was needed for the old test suite. For example,
1416 sorting DNS records because real resolvers return them in an arbitrary
1417 order. The new test suite's fake resolver always returns records in the
1420 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1421 bounces) except when queue_only is set, to avoid logging races between
1422 the different processes.
1424 PH/38 Panic-die if .include specifies a non-absolute path.
1426 PH/39 A tweak to the "H" retry rule from its user.
1428 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1429 a label. They prevented compilation on older perls.
1431 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1432 a warning to be raised on newish perls.
1434 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1435 on queue. Changes to match documented behaviour of showing count of
1436 messages matching specified criteria.
1438 PH/40 Changed the default ident timeout from 30s to 5s.
1440 PH/41 Added support for the use of login_cap features, on those BSD systems
1441 that have them, for controlling the resources used by pipe deliveries.
1443 PH/42 The content-scanning code uses fopen() to create files in which to put
1444 message data. Previously it was not paying any attention to the mode of
1445 the files. Exim runs with umask(0) because the rest of the code creates
1446 files with open(), and sets the required mode explicitly. Thus, these
1447 files were ending up world-writeable. This was not a big issue, because,
1448 being within the spool directory, they were not world-accessible. I have
1449 created a function called modefopen, which takes an additional mode
1450 argument. It sets umask(777), creates the file, chmods it to the required
1451 mode, then resets the umask. All the relevant calls to fopen() in the
1452 content scanning code have been changed to use this function.
1454 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1455 to 24 hours. This avoids potential overflow problems when processing G
1456 and H retry rules. I suspect nobody ever tinkers with this value.
1458 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1460 PH/45 When the plaintext authenticator is running as a client, the server's
1461 challenges are checked to ensure they are valid base64 strings. By
1462 default, the authentication attempt is cancelled if an invalid string is
1463 received. Setting client_ignore_invalid_base64 true ignores these errors.
1464 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1465 they are received. Thus, the responses can be made to depend on the
1466 challenges. If an invalid string is ignored, an empty string is placed in
1469 PH/46 Messages that are created by the autoreply transport now contains a
1470 References: header, in accordance with RFCs 2822 and 3834.
1472 PH/47 Added authenticated_sender_force to the smtp transport.
1474 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1476 PH/49 Installed latest patch from the Sieve maintainer.
1478 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1479 was also set, the appendfile transport was unnecessarily scanning a
1480 directory of message files (e.g. for maildir delivery) to find the count
1481 of files (along with the size), even though it did not need this
1482 information. It now does the scan only if it needs to find either the
1483 size of the count of files.
1485 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1487 PH/52 Two bugs concerned with error handling when the smtp transport is
1490 (i) Exim was not creating retry information for temporary errors given
1491 for individual recipients after the DATA command when the smtp transport
1492 was used in LMTP mode. This meant that they could be retried too
1493 frequently, and not timed out correctly.
1495 (ii) Exim was setting the flag that allows error details to be returned
1496 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1497 recipients that were returned after the DATA command.
1499 PH/53 This is related to PH/52, but is more general: for any failing address,
1500 when detailed error information was permitted to be returned to the
1501 sender, but the error was temporary, then after the final timeout, only
1502 "retry timeout exceeded" was returned. Now it returns the full error as
1503 well as "retry timeout exceeded".
1505 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1506 do this, and (what is worse) MTAs that accept it.
1508 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1509 will now be deprecated.
1511 PH/56 New os.c-cygwin from the Cygwin maintainer.
1513 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1514 formats (previously only available in exim formats via -bpr, -bpru,
1515 and -bpra. Now also available in native and exiqgrep formats)
1517 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1518 with very large, slow to parse queues
1520 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1522 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1524 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1525 responses to authentication challenges, though it was showing the
1526 challenges; (ii) I've removed the CR characters from the debug output for
1529 PH/58 Allow for the insertion of a newline as well as a space when a string
1530 is turned into more than one encoded-word during RFC 2047 encoding. The
1531 Sieve code now uses this.
1533 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1534 data_4xx, lost_connection, tls_required.
1536 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1537 message was being sent as an SMTP response.
1539 PH/61 Add -l and -k options to exicyclog.
1541 PH/62 When verifying, if an address was redirected to one new address, so that
1542 verification continued, and the new address failed or deferred after
1543 having set something in $address_data, the value of $address_data was not
1544 passed back to the ACL. This was different to the case when no
1545 redirection occurred. The value is now passed back in both cases.
1547 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1548 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1549 use login_cap.h, so on its own it isn't the distinguishing feature. The
1550 new name refers directly to the setclassresources() function.
1552 PH/65 Added configuration files for NetBSD3.
1554 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1556 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1557 is preferred over IPv4.
1559 PH/68 The bounce_return_message and bounce_return_body options were not being
1560 honoured for bounces generated during the reception of non-SMTP messages.
1561 In particular, this applied to messages rejected by the ACL. This bug has
1562 been fixed. However, if bounce_return_message is true and bounce_return_
1563 body is false, the headers that are returned for a non-SMTP message
1564 include only those that have been read before the error was detected.
1565 (In the case of an ACL rejection, they have all been read.)
1567 PH/69 The HTML version of the specification is now built in a directory called
1568 spec_html instead of spec.html, because the latter looks like a path with
1569 a MIME-type, and this confuses some software.
1571 PH/70 Catch two compiler warnings in sieve.c.
1573 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1574 function verify_get_ident() calls ip_connect() to connect a socket, but
1575 if the "connect()" function timed out, ip_connect() used to close the
1576 socket. However, verify_get_ident() also closes the socket later, and in
1577 between Exim writes to the log, which may get opened at this point. When
1578 the socket was closed in ip_connect(), the log could get the same file
1579 descriptor number as the socket. This naturally causes chaos. The fix is
1580 not to close the socket in ip_connect(); the socket should be closed by
1581 the function that creates it. There was only one place in the code where
1582 this was missing, in the iplookup router, which I don't think anybody now
1583 uses, but I've fixed it anyway.
1585 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1586 well as to direct DNS lookups. Otherwise the handling of names in host
1587 lists is inconsistent and therefore confusing.
1593 PH/01 Two changes to the default runtime configuration:
1595 (1) Move the checks for relay_from_hosts and authenticated clients from
1596 after to before the (commented out) DNS black list checks.
1598 (2) Add control=submission to the relay_from_hosts and authenticated
1599 clients checks, on the grounds that messages accepted by these
1600 statements are most likely to be submissions.
1602 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1604 (1) Generate an error if the third argument for the ${prvs expansion is
1607 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1610 (3) Reset the variables that are obtained from the first argument of
1611 ${prvscheck and used in the second argument before leaving the code,
1612 because their memory is reclaimed, so using them afterwards may do
1615 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1616 one (it's much easier than Tom thought :-).
1618 (5) Because of (4), we can now allow for the use of $prvscheck_result
1619 inside the third argument.
1621 PH/03 For some reason, the default setting of PATH when running a command from
1622 a pipe transport was just "/usr/bin". I have changed it to
1625 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1626 anything to be listed in the output from -bV.
1628 PH/05 When a filter generated an autoreply, the entire To: header line was
1629 quoted in the delivery log line, like this:
1631 => >A.N.Other <ano@some.domain> <original@ddress> ...
1633 This has been changed so that it extracts the operative address. There
1634 may be more than one such address. If so, they are comma-separated, like
1637 => >ano@some.domain,ona@other.domain <original@ddress> ...
1639 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1640 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1641 not being looked up in the rDNS to get a host name, Exim was showing the
1642 IP address twice in Received: lines, even though the IP addresses were
1643 identical. For example:
1645 Received: from [1.2.3.4] (helo=[1.2.3.4])
1647 However, if the real host name was known, it was omitting the HELO data
1648 if it matched the actual IP address. This has been tidied up so that it
1649 doesn't show the same IP address twice.
1651 PH/07 When both +timestamp and +memory debugging was on, the value given by
1652 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1653 called by the debug printing, thereby overwriting the timestamp buffer.
1654 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1656 PH/08 When the original message was included in an autoreply transport, it
1657 always said "this is a copy of the message, including all the headers",
1658 even if body_only or headers_only was set. It now gives an appropriate
1661 PH/09 Applied a patch from the Sieve maintainer which:
1663 o fixes some comments
1664 o adds the (disabled) notify extension core
1665 o adds some debug output for the result of if/elsif tests
1666 o points to the current vacation draft in the documentation
1667 and documents the missing references header update
1671 o fixes a bug in processing the envelope test (when testing
1672 multiple envelope elements, the last element determinted the
1675 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1676 Electronic Mail") by including:
1678 Auto-submitted: auto-generated
1680 in the messages that it generates (bounce messages and others, such as
1681 warnings). In the case of bounce messages for non-SMTP mesages, there was
1682 also a typo: it was using "Auto_submitted" (underscore instead of
1683 hyphen). Since every message generated by Exim is necessarily in response
1684 to another message, thes have all been changed to:
1686 Auto-Submitted: auto-replied
1688 in accordance with these statements in the RFC:
1690 The auto-replied keyword:
1692 - SHOULD be used on messages sent in direct response to another
1693 message by an automatic process,
1695 - MUST NOT be used on manually-generated messages,
1697 - MAY be used on Delivery Status Notifications (DSNs) and Message
1698 Disposition Notifications (MDNs),
1700 - MUST NOT be used on messages generated by automatic or periodic
1701 processes, except for messages which are automatic responses to
1704 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1705 to the default Received: header definition.
1707 PH/12 Added log selector acl_warn_skipped (default on).
1709 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1710 variables because (a) they are in the wrong storage pool and (b) even if
1711 they were copied, it wouldn't work properly because of the caching.
1713 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1714 checking when decoding. Apparently there are clients that generate
1715 overlong encoded strings. Why am I not surprised?
1717 PH/15 If the first argument of "${if match_address" was not empty, but did not
1718 contain an "@" character, Exim crashed. Now it writes a panic log message
1719 and treats the condition as false.
1721 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1723 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1724 extension "envelope-auth". The code is finished and in agreement with
1725 other implementations, but there is no documentation so far and in fact,
1726 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1727 not changing the active code.
1729 Print executed "if" and "elsif" statements when debugging is used. This
1730 helps a great deal to understand what a filter does.
1732 Document more things not specified clearly in RFC3028. I had all this
1733 sorted out, when out of a sudden new issues came to my mind. Oops."
1735 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1738 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1739 canonical form (as documented). However, after a host name lookup from
1740 the IP address, check_host() was doing a simple string comparison with
1741 addresses acquired from the DNS when checking that the found name did
1742 have the original IP as one of its addresses. Since any found IPv6
1743 addresses are likely to be in abbreviated form, the comparison could
1744 fail. Luckily, there already exists a function for doing the comparison
1745 by converting both addresses to binary, so now that is used instead of
1746 the text comparison.
1748 PH/20 There was another similar case to PH/19, when a complete host name was
1749 given in a host list; looking up its IP address could give an abbreviated
1750 form, whereas the current host's name might or might not be abbreviated.
1751 The same fix has been applied.
1757 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1758 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1761 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1762 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1764 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1766 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1767 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1768 Also, for "Auto-Submitted", treat anything other than "no" as
1769 non-personal, in accordance with RFC 3834. (Previously it treated
1770 anything starting "auto-" as non-personal.)
1772 TF/01 The control=submission/name=... option had a problem with syntax
1773 errors if the name included a slash character. The /name= option
1774 now slurps the rest of the string, so it can include any characters
1775 but it must come last in the list of options (after /sender_retain
1778 PH/05 Some modifications to the interface to the fake nameserver for the new
1786 TK/01 Added the "success_on_redirect" address verification option. See
1787 NewStuff for rationale and an example.
1789 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1791 PH/02 Patch to exigrep to allow it to work on syslog lines.
1793 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1794 fread() to skip over the body file's header line, because in Cygwin the
1795 header line is locked and is inaccessible.
1797 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1798 co-exist for some time) to make it clear that it is the Exim ID that is
1799 referenced, not the Message-ID: header line.
1801 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1802 string_format() function, because snprintf() does not exist on all
1805 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1806 ${stat: expansion item.
1808 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1809 protocol synchronization error", to keep the pedants happy.
1811 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1812 well as for IRIX systems, when gcc is being used. See the host.c source
1815 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1817 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1820 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1821 transport and to the smtp transport in LMTP mode.
1823 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1825 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1826 run a filter in a subprocess. This could lead to confusion in subsequent
1827 lookups in the parent process. There should also be a search_tidyup() at
1828 the end of the subprocess.
1830 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1831 only if the host matched helo_try_verify_hosts, which caused the
1832 verification to occur when the EHLO/HELO command was issued. The ACL just
1833 tested the remembered result. Now, if a previous verification attempt has
1834 not happened, "verify = helo" does it there and then.
1836 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1838 TK/03 Fix log output including CR from clamd.
1840 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1841 exist provoked a memory error which could cause a segfault.
1843 PH/15 Installed PCRE 6.2
1845 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1847 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1848 of the problem. Specifically, suggested +O2 rather than +O1 for the
1851 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1853 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1854 record, the port was not correctly logged when the outgoing_port log
1855 selector was set (it logged the transort's default port).
1857 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1858 fallback_hosts, and "hosts" in the smtp transport.
1860 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1861 host errors such as "Connection refused".
1863 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1864 authentication with radiusclient 0.4.9:
1866 - Error returned from rc_read_config was caught wrongly
1867 - Username/password not passed on to radius server due to wrong length.
1869 The presumption is that some radiusclient API changes for 4.51/PH/17
1870 were not taken care of correctly. The code is still untested by me (my
1871 Linux distribution still has 0.3.2 of radiusclient), but it was
1872 contributed by a Radius user.
1874 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1875 expanding the "port" option of the smtp transport.
1877 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1878 while reading a MIME header. Thanks to Tom Hughes for a patch.
1880 PH/24 Include config.h inside local_scan.h so that configuration settings are
1883 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1884 that in an ACL for RCPT (for example), you can examine exactly what was
1887 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1888 commands, but it was not correctly comparing the address with the actual
1889 client host address. Thus, it would show the EHLO address in Received:
1890 header lines when this was not necessary.
1892 PH/27 Added the % operator to ${eval:}.
1894 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1895 it should be ignoring failures (because with -C, for example, it has lost
1896 privilege). It wasn't ignoring creation failures other than "already
1899 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1902 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1903 because an input line was too long, either on its own, or by virtue of
1904 too many continuations, the temporary file was not being removed, and the
1905 return code was incorrect.
1907 PH/31 Missing "BOOL" in function definition in filtertest.c.
1909 PH/32 Applied Sieve patches from the maintainer.
1911 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1913 PH/33 Added "verify = not_blind".
1915 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1916 Local/Makefile (with some defaults set). These are used in built scripts
1917 such as exicyclog, but they have never been used in the exim_install
1918 script (though there are many overriding facilities there). I have
1919 arranged that the exim_install script now takes note of these two
1922 PH/35 Installed configuration files for Dragonfly.
1924 PH/36 When a locally submitted message by a trusted user did not contain a
1925 From: header, and the sender address was obtained from -f or from an SMTP
1926 MAIL command, and the trusted user did not use -F to supply a sender
1927 name, $originator_name was incorrectly used when constructing a From:
1928 header. Furthermore, $originator_name was used for submission mode
1929 messages from external hosts without From: headers in a similar way,
1930 which is clearly wrong.
1932 PH/37 Added control=suppress_local_fixups.
1934 PH/38 When log_selector = +received_sender was set, and the addition of the
1935 sender made the log line's construction buffer exactly full, or one byte
1936 less than full, an overflow happened when the terminating "\n" was
1939 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1940 when the result of a list match is failure because a DNS lookup failed.
1942 PH/40 RM_COMMAND is now used in the building process.
1944 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1945 the "build-* directories that it finds.
1947 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1948 address in a domain literal was a prefix of an interface address.
1950 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1951 when verifying a sender address, unless rewrite_headers is false.
1953 PH/44 Wrote a long comment about why errors_to addresses are verified as
1954 recipients, not senders.
1956 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1957 the ratelimit ACL was added.
1959 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1961 PH/46 Added extra information about PostgreSQL errors to the error string.
1963 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1964 avoiding the need to install special zones in a real server. This is
1965 backwards compatible; if it can't find the fake resolver, it drops back.
1966 Thus, both old and new test suites can be run.
1968 TF/02 Added util/ratelimit.pl
1970 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1971 clock is set back in time.
1973 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1974 Candler <B.Candler@pobox.com>.
1976 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1977 OK for addresses that are the result of redirections.
1979 PH/48 A number of further additions for the benefit of the new test suite,
1980 including a fake gethostbyname() that interfaces to the fake DNS resolver
1983 TF/06 The fix for widen_domains has also been applied to qualify_single and
1984 search_parents which are the other dnslookup options that can cause
1987 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1990 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1992 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1993 service" if it encountered an MX record with an empty target hostname.
1994 The message is now "an MX or SRV record indicated no SMTP service".
1996 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1997 if the DNS of the sending site is misconfigured. This is quite a
1998 common situation. This change restores the behaviour of treating a
1999 helo verification defer as a failure.
2001 PH/51 If self=fail was set on a router, the bounce message did not include the
2002 actual error message.
2008 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2010 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2011 command itself ended in error, the underlying message about the transport
2012 filter timeout was being overwritten with the pipe command error. Now the
2013 underlying error message should be appended to the second error message.
2015 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2017 PH/02 Reduce the amount of output that "make" produces by default. Full output
2018 can still be requested.
2020 PH/03 The warning log line about a condition test deferring for a "warn" verb
2021 was being output only once per connection, rather than after each
2022 occurrence (because it was using the same function as for successful
2023 "warn" verbs). This seems wrong, so I have changed it.
2025 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2026 it should not have, which might have caused a crash in the right
2027 circumstances, but probably never did.
2029 PH/04 Installed a modified version of Tony Finch's patch to make submission
2030 mode fix the return path as well as the Sender: header line, and to
2031 add a /name= option so that you can make the user's friendly name appear
2034 TF/03 Added the control = fakedefer ACL modifier.
2036 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2037 Mark Lowes for thorough testing.
2039 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2041 TK/03 Merged latest SRS patch from Miles Wilton.
2043 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2044 with the definition in sysexits.h (which is #included earlier).
2045 Fortunately, Exim does not actually use EX_OK. The code used to try to
2046 preserve the sysexits.h value, by assumimg that macro definitions were
2047 scanned for macro replacements. I have been disabused of this notion,
2048 so now the code just undefines EX_OK before #including unistd.h.
2050 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2051 in the smtp transport. When a block could not be written in a single
2052 write() function, the timeout was being re-applied to each part-write.
2053 This seems wrong - if the receiver was accepting one byte at a time it
2054 would take for ever. The timeout is now adjusted when this happens. It
2055 doesn't have to be particularly precise.
2057 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2058 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2060 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2061 without a domain if the check to <postmaster@domain> fails.
2063 SC/01 Eximstats: added -xls and the ability to specify output files
2064 (patch written by Frank Heydlauf).
2066 SC/02 Eximstats: use FileHandles for outputing results.
2068 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2070 SC/04 Eximstats: fixed display of large numbers with -nvr option
2072 SC/05 Eximstats: fixed merging of reports with empty tables.
2074 SC/06 Eximstats: added the -include_original_destination flag
2076 SC/07 Eximstats: removed tabs and trailing whitespace.
2078 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2080 TK/06 MBOX spool code: Add real "From " MBOX separator line
2081 so the .eml file is really in mbox format (even though
2082 most programs do not really care). Patch from Alex Miller.
2084 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2085 The latter is generated from $received_to and is only set if the
2086 message has one envelope recipient. SA can use these headers,
2087 obviously out-of-the-box. Patch from Alex Miller.
2089 PH/08 The ${def test on a variable was returning false if the variable's
2090 value was "0", contrary to what the specification has always said!
2091 The result should be true unless the variable is empty.
2093 PH/09 The syntax error of a character other than { following "${if
2094 def:variable_name" (after optional whitespace) was not being diagnosed.
2095 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2096 accidental colon was present, for example, could give incorrect results.
2098 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2099 result is used (not including appendfile, where other changes are about
2102 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2103 This involved changing a lot of size variables from int to off_t. It
2104 should work with maildirs and everything.
2106 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2107 spamd dying while we are connected to it.
2109 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2112 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2113 function for BDB 4.3.
2115 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2117 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2118 the sources for PCRE have been re-arranged and more files are now
2121 PH/15 The code I had for printing potentially long long variables in PH/11
2122 above was not the best (it lost precision). The length of off_t variables
2123 is now inspected at build time, and an appropriate printing format (%ld
2124 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2125 to be "long long int" or "long int". This is needed for the internal
2126 formatting function string_vformat().
2128 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2129 the configuration file to be ":syslog", then the script "guesses" where
2130 the logs files are, rather than using the compiled in default. In our
2131 case the guess is not the same as the compiled default, so the script
2132 suddenly stopped working when I started to use syslog. The patch checks
2133 to see if log_file_path is "". If so, it attempts to read it from exim
2134 with no configuration file to get the compiled in version, before it
2135 falls back to the previous guessing code."
2137 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2138 implementing BATV in an Exim configuration. See NewStuff for the gory
2141 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2142 Makefile that are specific to HP-UX.
2144 PH/18 If the "use_postmaster" option was set for a recipient callout together
2145 with the "random" option, the postmaster address was used as the MAIL
2146 FROM address for the random test, but not for the subsequent recipient
2147 test. It is now used for both.
2149 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2150 patch removes a few documentation additions to RFC 3028, because the
2151 latest draft now contains them. It adds the new en;ascii-case comparator
2152 and a new error check for 8bit text in MIME parts. Comparator and
2153 require names are now matched exactly. I enabled the subaddress
2154 extension, but it is not well tested yet (read: it works for me)."
2156 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2157 rework some of the code of TK/09 above to avoid the hardwired use of
2158 "%lld" and "long long". Replaced the call to snprintf() with a call to
2161 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2162 records point to non-existent hosts", "retry timeout exceeded", and
2163 "retry time not reached for any host after a long failure period".
2165 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2166 experimental DomainKeys support:
2168 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2169 (2) On an error such as an illegally used "control", the wrong name for
2170 the control was given.
2172 These problems did NOT occur unless DomainKeys support was compiled.
2174 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2176 PH/24 Added ${if match_ip condition.
2178 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2179 databases so that it will be absolutely obvious if a crash occurs in the
2180 DB library. This is a regular occurrence (often caused by mis-matched
2183 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2184 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2185 picked up on a user's system that detects such things. There doesn't seem
2186 to be a gcc warning option for this - only an attribute that has to be
2187 put on the function's prototype. It seems that in Fedora Core 4 they have
2188 set this on a number of new functions. No doubt there will be more in due
2191 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2192 specify a transport. However, if an address that was verified by such a
2193 router was the subject of a callout, Exim crashed because it tried to
2194 read the rcpt_include_affixes from the non-existent transport. Now it
2195 just assumes that the setting of that option is false. This bug was
2196 introduced by 4.51/PH/31.
2198 PH/28 Changed -d+all to exclude +memory, because that information is very
2199 rarely of interest, but it makes the output a lot bigger. People tend to
2200 do -d+all out of habit.
2202 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2203 code in os-type was giving problems when libc.so lives in lib64, like on
2206 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2207 aren't the modern standard, and it seems that some systems' include files
2208 don't always have them. Exim was already checking for some of the newer
2209 ones like T_AAAA, and defining it itself. I've added checks for all the
2210 record types that Exim uses.
2212 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2213 not automatically generating a new one, as it is supposed to. This
2214 prevented TLS from working. If the file did exist, but contained invalid
2215 data, a new version was generated, as expected. It was only the case of a
2216 non-existent file that was broken.
2218 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2219 with a change in libdomainkeys > 0.64.
2221 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2222 from DNS. If the selector record carries the flag, it now has
2223 precedence over the domain-wide flag.
2225 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2227 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2228 the use of a port name that isn't defined in /etc/services) occurred, the
2229 message was deferred as in a normal delivery, and thus remained on the
2230 spool, instead of being failed because of the mua_wrapper setting. This
2231 is now fixed, and I tidied up some of the mua_wrapper messages at the
2234 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2235 the messages in a hash of arrays rather than using individual hashes.
2236 This is a bit cleaner and results in dramatic memory savings, albeit
2237 at a slight CPU cost.
2239 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2240 as requested by Marc Sherman.
2242 SC/10 Eximstats: added histograms for user specified patterns as requested
2245 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2247 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2248 fopen() in the content-scanning modules that did not already have it.
2254 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2255 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2257 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2259 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2261 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2262 to test Sieve filters that use "vacation".
2264 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2265 that changes the way the GnuTLS parameters are stored in the cache file.
2266 The new format can be generated externally. For backward compatibility,
2267 if the data in the cache doesn't make sense, Exim assumes it has read an
2268 old-format file, and it generates new data and writes a new file. This
2269 means that you can't go back to an older release without removing the
2272 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2273 work if there are any delivery delays because "one_time" forces the
2274 parent to be marked "delivered", so its unseen clone is never tried
2275 again. For this reason, Exim now forbids the simultaneous setting of
2278 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2279 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2280 entry if you want to know the details. The fix, however, neglected to
2281 consider the case when local delivery batching is involved. The test for
2282 "previously delivered" was not happening when checking to see if an
2283 address could be batched with a previous (undelivered) one; under
2284 certain circumstances this could lead to multiple deliveries to the same
2287 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2288 in its include files, and this causes problems building Exim.
2290 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2291 header_syntax) but Exim was just ignoring anything given after a slash.
2292 In particular, this caused confusion with an attempt to use "verify =
2293 reverse_host_lookup/defer_ok". An error is now given when options are
2294 supplied for verify items that do not have them. (Maybe reverse_host_
2295 lookup should have a defer_ok option, but that's a different point.)
2297 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2298 defined by RFC 821) to 2048, because there were problems with some AUTH
2299 commands, and RFC 1869 says the size should be increased for extended
2300 SMTP commands that take arguments.
2302 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2305 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2306 "unknown" error; now it says that the functionality isn't in the binary.
2308 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2309 an address' error message when a string expansion fails (syntax or
2310 whatever). Otherwise the password may appear in the log. Following change
2311 PH/42 below, there is no longer a chance of it appearing in a bounce
2314 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2316 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2317 of its addresses was ever tried. (Bugzilla bug #2.)
2319 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2320 the result incorrectly in the debug output. (It correctly added a newline
2321 to what was transported.)
2323 TF/01 Added $received_time.
2325 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2326 commented out examples of how to interface to a virus scanner and to
2327 SpamAssassin. Also added commented examples of av_scanner and
2328 spamd_address settings.
2330 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2331 and controls are allowed in which ACLs. There were a couple of minor
2332 errors. Some of the entries in the conditions table (which is a table of
2333 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2334 negation of where the condition IS allowed.
2336 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2338 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2339 header file does not have a version number, so I've had to invent a new
2340 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2341 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2342 radiusclient), but it was contributed by a Radius user.
2344 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2345 files or directories, for OpenSSL.
2347 PH/19 When an Exim process that is running as root has to create an Exim log
2348 file, it does so in a subprocess that runs as exim:exim so as to get the
2349 ownership right at creation (otherwise, other Exim processes might see
2350 the file with the wrong ownership). There was no test for failure of this
2351 fork() call, which would lead to the process getting stuck as it waited
2352 for a non-existent subprocess. Forks do occasionally fail when resources
2353 run out. I reviewed all the other calls to fork(); they all seem to check
2356 PH/20 When checking for unexpected SMTP input at connect time (before writing
2357 the banner), Exim was not dealing correctly with a non-positive return
2358 from the read() function. If the client had disconnected by this time,
2359 the result was a log entry for a synchronization error with an empty
2360 string after "input=" when read() returned zero. If read() returned -1
2361 (an event I could not check), uninitialized data bytes were printed.
2362 There were reports of junk text (parts of files, etc) appearing after
2365 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2367 PH/22 Added support for macro redefinition, and (re)definition in between
2368 driver and ACL definitions.
2370 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2371 forgetting to use the resulting value; it was using the unexpanded value.
2373 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2374 hadn't been configured. The fix is from Juergen Kreileder, who
2375 understands it better than I do:
2377 "Here's what I see happening with three configured cyrus_sasl
2378 authenticators configured (plain, login, cram-md5):
2380 On startup auth_cyrus_sasl_init() gets called for each of these.
2381 This means three calls to sasl_listmech() without a specified mech_list.
2382 => SASL tests which mechs of all available mechs actually work
2383 => three warnings about OTP not working
2384 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2386 With the patch, sasl_listmech() also gets called three times. But now
2387 SASL's mech_list option is set to the server_mech specified in the the
2388 authenticator. Or in other words, the answer from sasl_listmech()
2389 gets limited to just the mech you're testing for (which is different
2391 => the return list contains just 'plain' or 'login', 'cram-md5' or
2392 nothing depending on the value of ob->server_mech.
2394 I've just tested the patch: Authentication still works fine,
2395 unavailable mechs specified in the exim configuration are still
2396 caught, and the auth.log warnings about OTP are gone."
2398 PH/25 When debugging is enabled, the contents of the command line are added
2399 to the debugging output, even when log_selector=+arguments is not
2402 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2403 answer is "GNU", and only if the return is "GNU/something" is the answer
2406 PH/27 $acl_verify_message is now set immediately after the failure of a
2407 verification in an ACL, and so is available in subsequent modifiers. In
2408 particular, the message can be preserved by coding like this:
2410 warn !verify = sender
2411 set acl_m0 = $acl_verify_message
2413 Previously, $acl_verify_message was set only while expanding "message"
2414 and "log_message" when a very denied access.
2416 PH/28 Modified OS/os.c-Linux with
2418 -#ifndef OS_LOAD_AVERAGE
2419 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2421 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2422 nomenclature these days.)
2424 PH/29 Installed patch from the Sieve maintainer that adds the options
2425 sieve_useraddress and sieve_subaddress to the redirect router.
2427 PH/30 In these circumstances:
2428 . Two addresses routed to the same list of hosts;
2429 . First host does not offer TLS;
2430 . First host accepts first address;
2431 . First host gives temporary error to second address;
2432 . Second host offers TLS and a TLS session is established;
2433 . Second host accepts second address.
2434 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2435 and peerdn, if requested) that were in fact used only for the second
2438 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2439 attention to any local part prefix or suffix that was matched by the
2440 router that accepted the address. It now behaves in the same way as it
2441 does for delivery: the affixes are removed from the local part unless
2442 rcpt_include_affixes is set on the transport.
2444 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2445 timeout during the DATA phase of an incoming message.
2447 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2448 applied a patch sanctioned by the Sieve maintainer.
2450 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2451 the uid or gid is negative. A case of a negative gid caused this to be
2452 noticed. The fix allows for either to be negative.
2454 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2455 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2458 PH/36 The change PH/12 above was broken. Fixed it.
2460 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2461 the grounds that routing the same address twice would always produce the
2462 same answer. This might have been true once, but it is certainly no
2463 longer true now. Routing a child address may depend on the previous
2464 routing that produced that child. Some complicated redirection strategies
2465 went wrong when messages had multiple recipients, and made Exim's
2466 behaviour dependent on the order in which the addresses were given.
2468 I have moved the duplicate checking until after the routing is complete.
2469 Exim scans the addresses that are assigned to local and remote
2470 transports, and removes any duplicates. This means that more work will be
2471 done, as duplicates will always all be routed, but duplicates are
2472 presumably rare, so I don't expect this is of any significance.
2474 For deliveries to pipes, files, and autoreplies, the duplicate checking
2475 still happens during the routing process, since they are not going to be
2478 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2479 It corrects a timeout issue with spamd. This is Ian's comment: "The
2480 background is that sometimes spamd either never reads data from a
2481 connection it has accepted, or it never writes response data. The exiscan
2482 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2483 blindly assumes that writes won't block so it may never time out."
2485 PH/39 Allow G after quota size as well as K and M.
2487 PH/40 The value set for $authenticated_id in an authenticator may not contain
2488 binary zeroes or newlines because the value is written to log lines and
2489 to spool files. There was no check on this. Now the value is run through
2490 the string_printing() function so that such characters are converted to
2491 printable escape sequences.
2493 PH/41 $message_linecount is a new variable that contains the total number of
2494 lines in the message. Compare $body_linecount, which is the count for the
2497 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2498 bounce and delay warning messages, except in certain special cases, which
2501 (a) An SMTP error message from a remote host;
2502 (b) A message specified in a :fail: redirection;
2503 (c) A message specified in a "fail" command in a system filter;
2504 (d) A message specified in a FAIL return from the queryprogram router;
2505 (e) A message specified by the cannot_route_message router option.
2507 In these cases only, Exim does include the error details in bounce and
2508 warning messages. There are also a few cases where bland messages such
2509 as "unrouteable address" or "local delivery error" are given.
2511 PH/43 $value is now also set for the "else" part of a ${run expansion.
2513 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2514 being worked on, but at least Exim now implements the latest version to
2517 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2518 process to complete was treated as a delivery failure, a timeout while
2519 writing the message to the pipe was logged, but erroneously treated as a
2520 successful delivery. Such timeouts include transport filter timeouts. For
2521 consistency with the overall process timeout, these timeouts are now
2522 treated as errors, giving rise to delivery failures by default. However,
2523 there is now a new Boolean option for the pipe transport called
2524 timeout_defer, which, if set TRUE, converts the failures into defers for
2525 both kinds of timeout. A transport filter timeout is now identified in
2528 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2529 systems where "make" and "gmake" are different, calling "gmake" at top
2530 level broke things. I've arranged for the value of $(MAKE) to be passed
2531 from the Makefile to this script so that it can call the same version of
2535 A note about Exim versions 4.44 and 4.50
2536 ----------------------------------------
2538 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2539 changes of various kinds. As a consequence, a big documentation update was
2540 needed. This delayed the release for rather longer than seemed good, especially
2541 in the light of a couple of (minor) security issues. Therefore, the changes
2542 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2543 release. So 4.44 and 4.50 are in effect two different branches that both start
2546 I have left the 4.50 change log unchanged; it contains all the changes since
2547 4.43. The change log for 4.44 is below; many of its items are identical to
2548 those for 4.50. This seems to be the most sensible way to preserve the
2549 historical information.
2555 1. Minor wording change to the doc/README.SIEVE file.
2557 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2558 computation of the current number of files was incorrect.
2560 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2561 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2564 4. Give more explanation in the error message when the command for a transport
2565 filter fails to execute.
2567 5. There are several places where Exim runs a non-Exim command in a
2568 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2569 was being done only for the command run by the queryprogram router. It is
2570 now done for all such subprocesses. The other cases are: ${run, transport
2571 filters, and the commands run by the lmtp and pipe transports.
2573 6. Added CONFIGURE_GROUP build-time option.
2575 7. Some older OS have a limit of 256 on the maximum number of file
2576 descriptors. Exim was using setrlimit() to set 1000 as a large value
2577 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2578 systems. I've change it so that if it can't get 1000, it tries for 256.
2580 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2581 was an oversight, and furthermore, ever since the addition of extra
2582 controls (e.g. 4.43/32), the checks on when to allow different forms of
2583 "control" were broken. There should now be diagnostics for all cases when a
2584 control that does not make sense is encountered.
2586 9. Added the /retain_sender option to "control=submission".
2588 10. $recipients is now available in the predata ACL (oversight).
2590 11. Tidy the search cache before the fork to do a delivery from a message
2591 received from the command line. Otherwise the child will trigger a lookup
2592 failure and thereby defer the delivery if it tries to use (for example) a
2593 cached ldap connection that the parent has called unbind on.
2595 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2596 of $address_data from the recipient verification was clobbered by the
2597 sender verification.
2599 13. The value of address_data from a sender verification is now available in
2600 $sender_address_data in subsequent conditions in the ACL statement.
2602 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2604 15. Added a new option "connect=<time>" to callout options, to set a different
2607 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2608 was its contents. (It was OK if the option was not defined at all.)
2610 17. A "Completed" log line is now written for messages that are removed from
2611 the spool by the -Mrm option.
2613 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2614 information about exactly what failed.
2616 19. Added -dd to debug only the daemon process.
2618 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2619 handles timeouts, both on the server side and network timeouts. Renamed the
2620 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2622 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2623 It is now set to "smtps".
2625 22. $host_address is now set to the target address during the checking of
2626 ignore_target_hosts.
2628 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2629 being passed; this would have caused $sender_host_name to have been used if
2630 matching the list had actually called for a host name (not very likely,
2631 since this list is usually IP addresses). A host name is now passed as
2634 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2635 code that specifies a non-restarting handler (typically sigaction() in
2636 modern systems) in an attempt to fix a rare and obscure crash bug.
2638 25. Narrowed the window for a race in the daemon that could cause it to ignore
2639 SIGCHLD signals. This is not a major problem, because they are used only to
2640 wake it up if nothing else does.
2642 26. A malformed maildirsize file could cause Exim to calculate negative values
2643 for the mailbox size or file count. Odd effects could occur as a result.
2644 The maildirsize information is now recalculated if the size or filecount
2647 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2648 support for a long time. Removed HAVE_SYS_VFS_H.
2650 28. Installed the latest version of exipick from John Jetmore.
2652 29. In an address list, if the pattern was not a regular expression, an empty
2653 subject address (from a bounce message) matched only if the pattern was an
2654 empty string. Non-empty patterns were not even tested. This was the wrong
2655 because it is perfectly reasonable to use an empty address as part of a
2656 database query. An empty address is now tested by patterns that are
2657 lookups. However, all the other forms of pattern expect the subject to
2658 contain a local part and a domain, and therefore, for them, an empty
2659 address still always fails if the pattern is not itself empty.
2661 30. Exim went into a mad DNS loop when attempting to do a callout where the
2662 host was specified on an smtp transport, and looking it up yielded more
2663 than one IP address.
2665 31. Re-factored the code for checking spool and log partition space into a
2666 function that finds that data and another that does the check. The former
2667 is then used to implement four new variables: $spool_space, $log_space,
2668 $spool_inodes, and $log_inodes.
2670 32. The RFC2047 encoding function was originally intended for short strings
2671 such as real names; it was not keeping to the 75-character limit for
2672 encoded words that the RFC imposes. It now respects the limit, and
2673 generates multiple encoded words if necessary. To be on the safe side, I
2674 have increased the buffer size for the ${rfc2047: expansion operator from
2677 33. It is now permitted to omit both strings after an "if" condition; if the
2678 condition is true, the result is "true". As before, when the second string
2679 is omitted, a false condition yields an empty string. This makes it less
2680 cumbersome to write custom ACL and router conditions.
2682 34. Failure to deliver a bounce message always caused it to be frozen, even if
2683 there was an errors_to setting on the router. The errors_to setting is now
2686 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2687 canonical form (fully expanded) before being placed in
2688 $sender_host_address.
2690 36. The table in the code that translates DNS record types into text (T_A to
2691 "A" for instance) was missing entries for NS and CNAME. It is just possible
2692 that this could have caused confusion if both these types were looked up
2693 for the same domain, because the text type is used as part of Exim's
2694 per-process caching. But the chance of anyone hitting this buglet seems
2697 37. The dnsdb lookup has been extended in a number of ways.
2699 (1) There is a new type, "zns", which walks up the domain tree until it
2700 finds some nameserver records. It should be used with care.
2702 (2) There is a new type, "mxh", which is like "mx" except that it returns
2703 just the host names, not the priorities.
2705 (3) It is now possible to give a list of domains (or IP addresses) to be
2706 looked up. The behaviour when one of the lookups defers can be
2707 controlled by a keyword.
2709 (4) It is now possible to specify the separator character for use when
2710 multiple records are returned.
2712 38. The dnslists ACL condition has been extended: it is now possible to supply
2713 a list of IP addresses and/or domains to be looked up in a particular DNS
2716 39. Added log_selector=+queue_time_overall.
2718 40. When running the queue in the test harness, wait just a tad after forking a
2719 delivery process, to get repeatability of debugging output.
2721 41. Include certificate and key file names in error message when GnuTLS fails
2722 to set them up, because the GnuTLS error message doesn't include the name
2723 of the failing file when there is a problem reading it.
2725 42. Allow both -bf and -bF in the same test run.
2727 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2729 44. The "Exiscan patch" is now merged into the mainline Exim source.
2731 45. Sometimes the final signoff response after QUIT could fail to get
2732 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2733 < 0 before doing a fflush(). This bug looks as though it goes back to the
2734 introduction of TLS in release 3.20, but "sometimes" must have been rare
2735 because the tests only now provoked it.
2737 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2738 (this can affect the format of dates).
2740 47. exim_tidydb, when checking for the continued existence of a message for
2741 which it has found a message-specific retry record, was not finding
2742 messages that were in split spool directories. Consequently, it was
2743 deleting retry records that should have stayed in existence.
2745 48. Steve fixed some bugs in eximstats.
2747 49. The SPA authentication driver was not abandoning authentication and moving
2748 on to the next authenticator when an expansion was forced to fail,
2749 contradicting the general specification for all authenticators. Instead it
2750 was generating a temporary error. It now behaves as specified.
2752 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2753 (the order specifies the preference for clients). The order is now AES256,
2754 AES128, 3DES, ARCFOUR128.
2756 51. Small patch to Sieve code - explicitly set From: when generating an
2759 52. Exim crashed if a remote delivery caused a very long error message to be
2760 recorded - for instance if somebody sent an entire SpamAssassin report back
2761 as a large number of 550 error lines. This bug was coincidentally fixed by
2762 increasing the size of one of Exim's internal buffers (big_buffer) that
2763 happened as part of the Exiscan merge. However, to be on the safe side, I
2764 have made the code more robust (and fixed the comments that describe what
2767 53. Now that there can be additional text after "Completed" in log lines (if
2768 the queue_time_overall log selector is set), a one-byte patch to exigrep
2769 was needed to allow it to recognize "Completed" as not the last thing in
2772 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2773 patch that reportedly fixes this has been added. I am not expert enough to
2774 create a test for it. This is what the patch creator wrote:
2776 "I found a little strange behaviour of ldap code when working with
2777 Windows 2003 AD Domain, where users was placed in more than one
2778 Organization Units. When I tried to give exim partial DN, the exit code
2779 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2780 But simultaneously result of request was absolutely normal ldap result,
2781 so I produce this patch..."
2783 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2784 REFERENCE, so I have modified the code to exclude the patch when that macro
2787 55. Some experimental protocols are using DNS PTR records for new purposes. The
2788 keys for these records are domain names, not reversed IP addresses. The
2789 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2790 leaves it alone. Component reversal etc. now happens only for IP addresses.
2793 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2795 57. Double the size of the debug message buffer (to 2048) so that more of very
2796 long debug lines gets shown.
2798 58. The exicyclog utility now does better if the number of log files to keep
2799 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2801 59. Two changes related to the smtp_active_hostname option:
2803 (1) $smtp_active_hostname is now available as a variable.
2804 (2) The default for smtp_banner uses $smtp_active_hostname instead
2805 of $primary_hostname.
2807 60. The host_aton() function is supposed to be passed a string that is known
2808 to be a valid IP address. However, in the case of IPv6 addresses, it was
2809 not checking this. This is a hostage to fortune. Exim now panics and dies
2810 if the condition is not met. A case was found where this could be provoked
2811 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2812 components; fortuitously, this particular loophole had already been fixed
2813 by change 4.50/55 above.
2815 If there are any other similar loopholes, the new check in host_aton()
2816 itself should stop them being exploited. The report I received stated that
2817 data on the command line could provoke the exploit when Exim was running as
2818 exim, but did not say which command line option was involved. All I could
2819 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2820 running as the user.
2823 61. There was a buffer overflow vulnerability in the SPA authentication code
2824 (which came originally from the Samba project). I have added a test to the
2825 spa_base64_to_bits() function which I hope fixes it.
2828 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2829 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2831 63. The daemon start-up calls getloadavg() while still root for those OS that
2832 need the first call to be done as root, but it missed one case: when
2833 deliver_queue_load_max is set with deliver_drop_privilege. This is
2834 necessary for the benefit of the queue runner, because there is no re-exec
2835 when deliver_drop_privilege is set.
2837 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2838 This has been fixed.
2840 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2841 was in use, was not putting the data itself into the right store pool;
2842 consequently, it could be overwritten for a subsequent message in the same
2843 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2846 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2848 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2849 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2850 the function were treating the return as a boolean value, which happened to
2851 work because 0=false and not-0=true, but is not correct code.
2853 68. The host_aton() function was not handling scoped IPv6 addresses (those
2854 with, for example, "%eth0" on the end) correctly.
2856 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2857 negated items (that is, ~something) in unsigned ints. Some compilers
2858 apparently mutter when there is no cast.
2860 70. If an address verification called from an ACL failed, and did not produce a
2861 user-specific message (i.e. there was only a "system" message), nothing was
2862 put in $acl_verify_message. In this situation, it now puts the system
2865 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2866 session; change 4.31/43 added the unwanted input to the log line - except
2867 that it did not do this in the start of session case. It now does.
2869 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2870 This is wrong and can cause the other end to generate a synchronization
2871 error if it is another Exim or anything else that does the synchronization
2872 check. A QUIT command is no longer sent after a timeout.
2874 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2875 during host lookups.
2877 74. The defer_ok option of callout verification was not working if it was used
2878 when verifying addresses in header lines, that is, for this case:
2880 verify = header_sender/callout=defer_ok
2882 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2883 those file descriptors could be used for SMTP connections. If anything
2884 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2885 could be sent to the SMTP client, causing chaos. The daemon now opens
2886 stdin, stdout, and stderr to /dev/null when it puts itself into the
2889 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2890 log by default. The user can override this with suitable Perl magic.
2892 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2893 the log message when discard triggers, was not working for the DATA ACL or
2894 for the non-SMTP ACL.
2896 78. Error message wording change in sieve.c.
2898 79. If smtp_accept_max_per_host was set, the number of connections could be
2899 restricted to fewer than expected, because the daemon was trying to set up
2900 a new connection before checking whether the processes handling previous
2901 connections had finished. The check for completed processes is now done
2902 earlier. On busy systems, this bug wouldn't be noticed because something
2903 else would have woken the daemon, and it would have reaped the completed
2906 80. If a message was submitted locally by a user whose login name contained one
2907 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2908 It caused a spool format error. I have fixed the spool reading code. A
2909 related problem was that the "from" clause in the Received: line became
2910 illegal because of the space(s). It is now covered by ${quote_local_part.
2912 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2915 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2917 83. Patch from Sieve maintainer to fix unterminated string problem in
2918 "vacation" handling.
2920 84. Some minor changes to the Linux configuration files to help with other
2921 OS variants using glibc.
2923 85. One more patch for Sieve to update vacation handling to latest spec.
2926 ----------------------------------------------------
2927 See the note above about the 4.44 and 4.50 releases.
2928 ----------------------------------------------------
2934 1. Change 4.43/35 introduced a bug that caused file counts to be
2935 incorrectly computed when quota_filecount was set in an appendfile
2938 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2939 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2942 3. Give more explanation in the error message when the command for a transport
2943 filter fails to execute.
2945 4. There are several places where Exim runs a non-Exim command in a
2946 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2947 was being done only for the command run by the queryprogram router. It is
2948 now done for all such subprocesses. The other cases are: ${run, transport
2949 filters, and the commands run by the lmtp and pipe transports.
2951 5. Some older OS have a limit of 256 on the maximum number of file
2952 descriptors. Exim was using setrlimit() to set 1000 as a large value
2953 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2954 systems. I've change it so that if it can't get 1000, it tries for 256.
2956 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2957 was an oversight, and furthermore, ever since the addition of extra
2958 controls (e.g. 4.43/32), the checks on when to allow different forms of
2959 "control" were broken. There should now be diagnostics for all cases when a
2960 control that does not make sense is encountered.
2962 7. $recipients is now available in the predata ACL (oversight).
2964 8. Tidy the search cache before the fork to do a delivery from a message
2965 received from the command line. Otherwise the child will trigger a lookup
2966 failure and thereby defer the delivery if it tries to use (for example) a
2967 cached ldap connection that the parent has called unbind on.
2969 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2970 of $address_data from the recipient verification was clobbered by the
2971 sender verification.
2973 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2974 was its contents. (It was OK if the option was not defined at all.)
2976 11. A "Completed" log line is now written for messages that are removed from
2977 the spool by the -Mrm option.
2979 12. $host_address is now set to the target address during the checking of
2980 ignore_target_hosts.
2982 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2983 being passed; this would have caused $sender_host_name to have been used if
2984 matching the list had actually called for a host name (not very likely,
2985 since this list is usually IP addresses). A host name is now passed as
2988 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2989 code that specifies a non-restarting handler (typically sigaction() in
2990 modern systems) in an attempt to fix a rare and obscure crash bug.
2992 15. Narrowed the window for a race in the daemon that could cause it to ignore
2993 SIGCHLD signals. This is not a major problem, because they are used only to
2994 wake it up if nothing else does.
2996 16. A malformed maildirsize file could cause Exim to calculate negative values
2997 for the mailbox size or file count. Odd effects could occur as a result.
2998 The maildirsize information is now recalculated if the size or filecount
3001 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3002 support for a long time. Removed HAVE_SYS_VFS_H.
3004 18. Updated exipick to current release from John Jetmore.
3006 19. Allow an empty sender to be matched against a lookup in an address list.
3007 Previously the only cases considered were a regular expression, or an
3010 20. Exim went into a mad DNS lookup loop when doing a callout where the
3011 host was specified on the transport, if the DNS lookup yielded more than
3014 21. The RFC2047 encoding function was originally intended for short strings
3015 such as real names; it was not keeping to the 75-character limit for
3016 encoded words that the RFC imposes. It now respects the limit, and
3017 generates multiple encoded words if necessary. To be on the safe side, I
3018 have increased the buffer size for the ${rfc2047: expansion operator from
3021 22. Failure to deliver a bounce message always caused it to be frozen, even if
3022 there was an errors_to setting on the router. The errors_to setting is now
3025 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3026 canonical form (fully expanded) before being placed in
3027 $sender_host_address.
3029 24. Updated eximstats to version 1.33
3031 25. Include certificate and key file names in error message when GnuTLS fails
3032 to set them up, because the GnuTLS error message doesn't include the name
3033 of the failing file when there is a problem reading it.
3035 26. Expand error message when OpenSSL has problems setting up cert/key files.
3038 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3039 (this can affect the format of dates).
3041 28. exim_tidydb, when checking for the continued existence of a message for
3042 which it has found a message-specific retry record, was not finding
3043 messages that were in split spool directories. Consequently, it was
3044 deleting retry records that should have stayed in existence.
3046 29. eximstats updated to version 1.35
3047 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3048 1.35 - bugfix such that pie charts by volume are generated correctly
3050 30. The SPA authentication driver was not abandoning authentication and moving
3051 on to the next authenticator when an expansion was forced to fail,
3052 contradicting the general specification for all authenticators. Instead it
3053 was generating a temporary error. It now behaves as specified.
3055 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3056 (the order specifies the preference for clients). The order is now AES256,
3057 AES128, 3DES, ARCFOUR128.
3059 31. Small patch to Sieve code - explicitly set From: when generating an
3062 32. Exim crashed if a remote delivery caused a very long error message to be
3063 recorded - for instance if somebody sent an entire SpamAssassin report back
3064 as a large number of 550 error lines. This bug was coincidentally fixed by
3065 increasing the size of one of Exim's internal buffers (big_buffer) that
3066 happened as part of the Exiscan merge. However, to be on the safe side, I
3067 have made the code more robust (and fixed the comments that describe what
3070 33. Some experimental protocols are using DNS PTR records for new purposes. The
3071 keys for these records are domain names, not reversed IP addresses. The
3072 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3073 leaves it alone. Component reversal etc. now happens only for IP addresses.
3076 34. The host_aton() function is supposed to be passed a string that is known
3077 to be a valid IP address. However, in the case of IPv6 addresses, it was
3078 not checking this. This is a hostage to fortune. Exim now panics and dies
3079 if the condition is not met. A case was found where this could be provoked
3080 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3081 components; fortuitously, this particular loophole had already been fixed
3082 by change 4.50/55 or 4.44/33 above.
3084 If there are any other similar loopholes, the new check in host_aton()
3085 itself should stop them being exploited. The report I received stated that
3086 data on the command line could provoke the exploit when Exim was running as
3087 exim, but did not say which command line option was involved. All I could
3088 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3089 running as the user.
3092 35. There was a buffer overflow vulnerability in the SPA authentication code
3093 (which came originally from the Samba project). I have added a test to the
3094 spa_base64_to_bits() function which I hope fixes it.
3097 36. The daemon start-up calls getloadavg() while still root for those OS that
3098 need the first call to be done as root, but it missed one case: when
3099 deliver_queue_load_max is set with deliver_drop_privilege. This is
3100 necessary for the benefit of the queue runner, because there is no re-exec
3101 when deliver_drop_privilege is set.
3103 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3104 was in use, was not putting the data itself into the right store pool;
3105 consequently, it could be overwritten for a subsequent message in the same
3106 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3109 38. Sometimes the final signoff response after QUIT could fail to get
3110 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3111 < 0 before doing a fflush(). This bug looks as though it goes back to the
3112 introduction of TLS in release 3.20, but "sometimes" must have been rare
3113 because the tests only now provoked it.
3119 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3120 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3121 Now it returns an integer. A number of calls were still expecting a T/F
3122 return. Fortuitously, in all cases, the tests worked in OK situations,
3123 which is the norm. However, things would have gone wrong on any write
3124 failures on the smtp file descriptor. This function is used when sending
3125 messages over SMTP and also when doing verify callouts.
3127 2. When Exim is called to do synchronous delivery of a locally submitted
3128 message (the -odf or -odi options), it no longer closes stderr before doing
3131 3. Implemented the mua_wrapper option.
3133 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3135 5. Implemented the functions header_remove(), header_testname(),
3136 header_add_at_position(), and receive_remove_recipient(), and exported them
3139 6. If an ACL "warn" statement specified the addition of headers, Exim already
3140 inserted X-ACL-Warn: at the start if there was no header name. However, it
3141 was not making this test for the second and subsequent header lines if
3142 there were newlines in the string. This meant that an invalid header could
3143 be inserted if Exim was badly configured.
3145 7. Allow an ACL "warn" statement to add header lines at the start or after all
3146 the Received: headers, as well as at the end.
3148 8. Added the rcpt_4xx retry error code.
3150 9. Added postmaster_mailfrom=xxx to callout verification option.
3152 10. Added mailfrom=xxxx to the callout verification option, for verify=
3155 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3156 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3158 12. Inserted some casts to stop certain compilers warning when using pointer
3159 differences as field lengths or precisions in printf-type calls (mostly
3160 affecting debugging statements).
3162 13. Added optional readline() support for -be (dynamically loaded).
3164 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3165 same clock tick as a message's arrival, so that its received time was the
3166 same as the "first fail" time on the retry record, and that message
3167 remained on the queue past the ultimate address timeout, every queue runner
3168 would try a delivery (because it was past the ultimate address timeout) but
3169 after another failure, the ultimate address timeout, which should have then
3170 bounced the address, did not kick in. This was a "< instead of <=" error;
3171 in most cases the first failure would have been in the next clock tick
3172 after the received time, and all would be well.
3174 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3175 being recognized when the domain list was tested by the match_domain
3176 condition in an expansion string.
3178 16. Added the ${str2b64: operator.
3180 17. Exim was always calling setrlimit() to set a large limit for the number of
3181 processes, without checking whether the existing limit was already
3182 adequate. (It did check for the limit on file descriptors.) Furthermore,
3183 errors from getrlimit() and setrlimit() were being ignored. Now they are
3184 logged to the main and panic logs, but Exim does carry on, to try to do its
3185 job under whatever limits there are.
3187 18. Imported PCRE 5.0.
3189 19. Trivial typo in log message " temporarily refused connection" (the leading
3192 20. If the log selector return_path_on_delivery was set and an address was
3193 redirected to /dev/null, the delivery process crashed because it assumed
3194 that a return path would always be set for a "successful" delivery. In this
3195 case, the whole delivery is bypassed as an optimization, and therefore no
3198 21. Internal re-arrangement: the function for sending a challenge and reading
3199 a response while authentication was assuming a zero-terminated challenge
3200 string. It's now changed to take a pointer and a length, to allow for
3201 binary data in such strings.
3203 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3205 23. Exim was not respecting finduser_retries when seeking the login of the
3206 uid under which it was called; it was always trying 10 times. (The default
3207 setting of finduser_retries is zero.) Also, it was sleeping after the final
3208 failure, which is pointless.
3210 24. Implemented tls_on_connect_ports.
3212 25. Implemented acl_smtp_predata.
3214 26. If the domain in control=submission is set empty, Exim assumes that the
3215 authenticated id is a complete email address when it generates From: or
3216 Sender: header lines.
3218 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3219 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3220 chown and chgrp in /bin and hostname in /usr/bin.
3222 28. Exim was keeping the "process log" file open after each use, just as it
3223 does for the main log. This opens the possibility of it remaining open for
3224 long periods when the USR1 signal hits a daemon. Occasional processlog
3225 errors were reported, that could have been caused by this. Anyway, it seems
3226 much more sensible not to leave this file open at all, so that is what now
3229 29. The long-running daemon process does not normally write to the log once it
3230 has entered its main loop, and it closes the log before doing so. This is
3231 so that log files can straightforwardly be renamed and moved. However,
3232 there are a couple of unusual error situations where the daemon does write
3233 log entries, and I had neglected to close the log afterwards.
3235 30. The text of an SMTP error response that was received during a remote
3236 delivery was being truncated at 512 bytes. This is too short for some of
3237 the long messages that one sometimes sees. I've increased the limit to
3240 31. It is now possible to make retry rules that apply only when a message has a
3241 specific sender, in particular, an empty sender.
3243 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3244 it possible to be selective about when SMTP synchronization is enforced.
3246 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3248 32. Implemented hosts_connection_nolog.
3250 33. Added an ACL for QUIT.
3252 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3255 35. Added mailbox_size and mailbox_filecount to appendfile.
3257 36. Added control = no_multiline_responses to ACLs.
3259 37. There was a bug in the logic of the code that waits for the clock to tick
3260 in the case where the clock went backwards by a substantial amount such
3261 that the microsecond fraction of "now" was more than the microsecond
3262 fraction of "then" (but the whole seconds number was less).
3264 38. Added support for the libradius Radius client library this is found on
3265 FreeBSD (previously only the radiusclient library was supported).
3271 1. When certain lookups returned multiple values in the form name=value, the
3272 quoting of the values was not always being done properly. Specifically:
3273 (a) If the value started with a double quote, but contained no whitespace,
3275 (b) If the value contained whitespace other than a space character (i.e.
3276 tabs or newlines or carriage returns) it was not quoted.
3277 This fix has been applied to the mysql and pgsql lookups by writing a
3278 separate quoting function and calling it from the lookup code. The fix
3279 should probably also be applied to nisplus, ibase and oracle lookups, but
3280 since I cannot test any of those, I have not disturbed their existing code.
3282 2. A hit in the callout cache for a specific address caused a log line with no
3283 reason for rejecting RCPT. Now it says "Previous (cached) callout
3284 verification failure".
3286 3. There was an off-by-one bug in the queryprogram router. An over-long
3287 return line was truncated at 256 instead of 255 characters, thereby
3288 overflowing its buffer with the terminating zero. As well as fixing this, I
3289 have increased the buffer size to 1024 (and made a note to document this).
3291 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3292 when Exim is waiting for an SMTP response from a remote server, Exim
3293 restarts its select() call on the socket, thereby resetting its timeout.
3294 This is not a problem when such interrupts are rare. Somebody set up a cron
3295 job to run exiwhat every 2 minutes, which is less than the normal select()
3296 timeout (5 or 10 minutes). This meant that the select() timeout never
3297 kicked in because it was always reset. I have fixed this by comparing the
3298 time when an interrupt arrives with the time at the start of the first call
3299 to select(). If more time than the timeout has elapsed, the interrupt is
3300 treated as a timeout.
3302 5. Some internal re-factoring in preparation for the addition of Sieve
3303 extensions (by MH). In particular, the "personal" test is moved to a
3304 separate function, and given an option for scanning Cc: and Bcc: (which is
3305 not set for Exim filters).
3307 6. When Exim created an email address using the login of the caller as the
3308 local part (e.g. when creating a From: or Sender: header line), it was not
3309 quoting the local part when it contained special characters such as @.
3311 7. Installed new OpenBSD configuration files.
3313 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3314 try to make them clearer.
3316 9. Callout options, other than the timeout value, were being ignored when
3317 verifying sender addresses in header lines. For example, when using
3319 verify = header_sender/callout=no_cache
3321 the cache was (incorrectly) being used.
3323 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3324 only the Cygwin environment.
3326 11. When return_path_on_delivery was set as a log selector, if different remote
3327 addresses in the same message used different return paths and parallel
3328 remote delivery occurred, the wrong values would sometimes be logged.
3329 (Whenever a remote delivery process finished, the return path value from
3330 the most recently started remote delivery process was logged.)
3332 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3333 lines when AUTH and/or TLS are in use. This is the "received protocol"
3334 field. Exim used to use "asmtp" for authenticated SMTP, without any
3335 indication (in the protocol name) for TLS use. Now it follows the RFC and
3336 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3337 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3338 names appear in log lines as well as in Received: header lines.
3340 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3341 extensions, and comparison tests, and to fix some bugs.
3343 14. Changes to the "personal" filter test:
3345 (1) The test was buggy in that it was just doing the equivalent of
3346 "contains" tests on header lines. For example, if a user's address was
3347 anne@some.where, the "personal" test would incorrectly be true for
3349 To: susanne@some.where
3351 This test is now done by extracting each address from the header in turn,
3352 and checking the entire address. Other tests that are part of "personal"
3353 are now done using regular expressions (for example, to check local parts
3354 of addresses in From: header lines).
3356 (2) The list of non-personal local parts in From: addresses has been
3357 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3358 taken from the Sieve specification recommendations.
3360 (3) If the message contains any header line starting with "List-" it is
3361 treated as non-personal.
3363 (4) The test for "circular" in the Subject: header line has been removed
3364 because it now seems ill-conceived.
3366 15. Minor typos in src/EDITME comments corrected.
3368 16. Installed latest exipick from John Jetmore.
3370 17. If headers_add on a router specified a text string that was too long for
3371 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3372 of string_sprintf() is now avoided.
3374 18. $message_body_size was not set (it was always zero) when running the DATA
3375 ACL and the local_scan() function.
3377 19. For the "mail" command in an Exim filter, no default was being set for
3378 the once_repeat time, causing a random time value to be used if "once" was
3379 specified. (If the value happened to be <= 0, no repeat happened.) The
3380 default is now 0s, meaning "never repeat". The "vacation" command was OK
3381 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3382 (I found it when inspecting the code).
3384 20. There is now an overall timeout for performing a callout verification. It
3385 defaults to 4 times the callout timeout, which applies to individual SMTP
3386 commands during the callout. The overall timeout applies when there is more
3387 than one host that can be tried. The timeout is checked before trying the
3388 next host. This prevents very long delays if there are a large number of
3389 hosts and all are timing out (e.g. when the network connections are timing
3390 out). The value of the overall timeout can be changed by specifying an
3391 additional sub-option for "callout", called "maxwait". For example:
3393 verify = sender/callout=5s,maxwait=20s
3395 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3396 to the end before writing, but this should make it even safer).
3398 22. Exim was forgetting that it had advertised PIPELINING for the second and
3399 subsequent messages on an SMTP connection. It was also not resetting its
3400 memory on STARTTLS and an internal HELO.
3402 23. When Exim logs an SMTP synchronization error within a session, it now
3403 records whether PIPELINING has been advertised or not.
3405 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3406 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3407 rather than long int.
3409 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3411 26. Added the never_mail option to autoreply.
3417 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3418 crash if the getsockname() call failed; this can happen if a connection is
3419 closed very soon after it is established. The problem was simply in the
3420 order in which certain operations were done, causing Exim to try to write
3421 to the SMTP stream before it had set up the file descriptor. The bug has
3422 been fixed by making things happen in the correct order.
3428 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3429 before the connection was closed, thus losing the rejection response.
3431 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3432 some early Solaris releases, but causes trouble in current releases where
3433 socklen_t is defined.
3435 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3438 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3441 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3442 that releasing the top of it at the end releases what was used for sub-
3443 expansions (unless the block got too big). However, discard this block if
3444 the first thing is a variable or header, so that we can use its block when
3445 it is dynamic (useful for very large $message_headers, for example).
3447 6. Lookups now cache *every* query, not just the most recent. A new, separate
3448 store pool is used for this. It can be recovered when all lookup caches are
3449 flushed. Lookups now release memory at the end of their result strings.
3450 This has involved some general refactoring of the lookup sources.
3452 7. Some code has been added to the store_xxx() functions to reduce the amount
3453 of flapping under certain conditions.
3455 8. log_incoming_interface used to affect only the <= reception log lines. Now
3456 it causes the local interface and port to be added to several more SMTP log
3457 lines, for example "SMTP connection from", and rejection lines.
3459 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3461 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3463 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3464 could be overwritten at the end of the current message (or the start of a
3465 new message if it was set in a HELO ACL). The value is now preserved for
3466 the duration of the SMTP connection.
3468 12. If a transport had a headers_rewrite setting, and a matching header line
3469 contained an unqualified address, that address was qualified, even if it
3470 did not match any rewriting rules. The underlying bug was that the values
3471 of the flags that permit the existence of unqualified sender and recipient
3472 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3473 non-local messages, and by -bnq for local messages) were not being
3474 preserved with the message after it was received.
3476 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3477 "next input=" as part of the text comprising the host identity instead of
3478 the correct text. The code was using the same buffer for two different
3479 strings. However, depending on which order the printing function evaluated
3480 its arguments, the bug did not always show up. Under Linux, for example, my
3481 test suite worked just fine.
3483 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3484 doesn't work with some older versions of Perl. It has been changed to "my",
3485 which in any case is probably the better facility to use.
3487 15. A really picky compiler found some instances of statements for creating
3488 error messages that either had too many or two few arguments for the format
3491 16. The size of the buffer for calls to the DNS resolver has been increased
3492 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3493 for addresses that have a lot of PTR records. This alleviates a problem; it
3494 does not fully solve it.
3496 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3497 buffer now truncates the list and logs the incident, which is the same
3498 action as happens when Exim is looking up a host name and its aliases.
3499 Previously in this situation something unpredictable would happen;
3500 sometimes it was "internal error: store_reset failed".
3502 18. If a server dropped the connection unexpectedly when an Exim client was
3503 using GnuTLS and trying to read a response, the client delivery process
3504 crashed while trying to generate an error log message.
3506 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3507 string, for example:
3509 warn message = H1: something\nH2: something
3511 the text was added as a single header line from Exim's point of view
3512 though it ended up OK in the delivered message. However, searching for the
3513 second and subsequent header lines using $h_h2: did not work. This has been
3514 fixed. Similarly, if a system filter added multiple headers in this way,
3515 the routers could not see them.
3517 20. Expanded the error message when iplsearch is called with an invalid key to
3518 suggest using net-iplsearch in a host list.
3520 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3521 ACLs are no longer actually imposed (and a message to that effect is
3524 22. If a "gecos" field in a passwd entry contained escaped characters, in
3525 particular, if it contained a \" sequence, Exim got it wrong when building
3526 a From: or a Sender: header from that name. A second bug also caused
3527 incorrect handling when an unquoted " was present following a character
3528 that needed quoting.
3530 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3531 was not being matched caselessly.
3533 24. Arranged for all hyphens in the exim.8 source to be escaped with
3536 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3537 properly. Recipient callout cache records were still being keyed to include
3538 the sender, even when use_sender was set false. This led to far more
3539 callouts that were necessary. The sender is no longer included in the key
3540 when use_sender is false.
3542 26. Added "control = submission" modifier to ACLs.
3544 27. Added the ${base62d: operator to decode base 62 numbers.
3546 28. dnsdb lookups can now access SRV records.
3548 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3549 the configuration file.
3551 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3552 (-v) mode. This makes the output for a verbose queue run more intelligible.
3554 31. Added a use_postmaster feature to recipient callouts.
3556 32. Added the $body_zerocount variable, containing the number of binary zero
3557 bytes in the message body.
3559 33. The time of last modification of the "new" subdirectory is now used as the
3560 "mailbox time last read" when there is a quota error for a maildir
3563 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3565 35. Added +ignore_unknown as a special item in host lists.
3567 36. Code for decoding IPv6 addresses in host lists is now included, even if
3568 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3569 address was recognized as an IP address, but was then not correctly decoded
3570 into binary, causing unexpected and incorrect effects when compared with
3577 1. Very minor rewording of debugging text in manualroute to say "list of
3578 hosts" instead of "hostlist".
3580 2. If verify=header_syntax was set, and a header line with an unqualified
3581 address (no domain) and a large number of spaces between the end of the
3582 name and the colon was received, the reception process suffered a buffer
3583 overflow, and (when I tested it) crashed. This was caused by some obsolete
3584 code that should have been removed. The fix is to remove it!
3586 3. When running in the test harness, delay a bit after writing a bounce
3587 message to get a bit more predictability in the log output.
3589 4. Added a call to search_tidyup() just before forking a reception process. In
3590 theory, someone could use a lookup in the expansion of smtp_accept_max_
3591 per_host which, without the tidyup, could leave open a database connection.
3593 5. Added the variables $recipient_data and $sender_data which get set from a
3594 lookup success in an ACL "recipients" or "senders" condition, or a router
3595 "senders" option, similar to $domain_data and $local_part_data.
3597 6. Moved the writing of debug_print from before to after the "senders" test
3600 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3601 problems for message scanning, either using a data ACL, or using
3602 local_scan() because the Received: header was not generated till after they
3603 were called (in order to set the time as the time of reception completion).
3604 I have revised the way this works. The header is now generated after the
3605 body is received, but before the ACL or local_scan() are called. After they
3606 are run, the timestamp in the header is updated.
3612 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3613 before starting a queue runner without re-exec. This happened only when
3614 deliver_drop_privilege was set or when the Exim user was set to root. The
3615 effect of the bug was that timeouts during subsequent deliveries caused
3616 crashes instead of being properly handled. The handler is now left at its
3617 default (and expected) setting.
3619 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3620 message, again when deliver_drop_privilege is set or Exim is run as root.
3621 The bug described in (1) was not present in this case, but the tidying up
3622 of the other signals was missing. I have made the two cases consistent.
3624 3. The ignore_target_hosts setting on a manualroute router was being ignored
3625 for hosts that were looked up using the /MX notation.
3627 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3630 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3631 operated on the sender address. After changing the $sender_address to <>
3632 for the sender address verify, Exim was re-instated it as the original
3633 (before rewriting) address, but remembering that it had rewritten it, so it
3634 wasn't rewriting it again. This bug also had the effect of breaking the
3635 sender address verification caching when the sender address was rewritten.
3637 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3638 This has been changed so that if the ip literal address matches
3639 ignore_target_hosts, the router declines.
3641 7. Added expansion conditions match_domain, match_address, and match_local_
3642 part (NOT match_host).
3644 8. The placeholder for the Received: header didn't have a length field set.
3646 9. Added code to Exim itself and to exim_lock to test for a specific race
3647 condition that could lead to file corruption when using MBX delivery. The
3648 issue is with the lockfile that is created in /tmp. If this file is removed
3649 after a process has opened it but before that process has acquired a lock,
3650 there is the potential for a second process to recreate the file and also
3651 acquire a lock. This could lead to two Exim processes writing to the file
3652 at the same time. The added code performs the same test as UW imapd; it
3653 checks after acquiring the lock that its file descriptor still refers to
3654 the same named file.
3656 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3657 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3660 11. Added the smtp_active_hostname option. If used, this will typically be made
3661 to depend on the incoming interface address. Because $interface_address is
3662 not set up until the daemon has forked a reception process, error responses
3663 that can happen earlier (such as "too many connections") no longer contain
3666 12. If an expansion in a condition on a "warn" statement fails because a lookup
3667 defers, the "warn" statement is abandoned, and the next ACL statement is
3668 processed. Previously this caused the whole ACL to be aborted.
3670 13. Added the iplsearch lookup type.
3672 14. Added ident_timeout as a log selector.
3674 15. Added tls_certificate_verified as a log selector.
3676 16. Added a global option tls_require_ciphers (compare the smtp transport
3677 option of the same name). This controls incoming TLS connections.
3679 17. I finally figured out how to make tls_require_ciphers do a similar thing
3680 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3681 before starting the TLS session.
3683 18. Tabs are now shown as \t in -bP output.
3685 19. If the log selector return_path_on_delivery was set, Exim crashed when
3686 bouncing a message because it had too many Received: header lines.
3688 20. If two routers both had headers_remove settings, and the first one included
3689 a superfluous trailing colon, the final name in the first list and the
3690 first name in the second list were incorrectly joined into one item (with a
3691 colon in the middle).
3697 1. Added -C and -D options to the exinext utility, mainly to make it easier
3698 to include in the automated testing, but these could be helpful when
3699 multiple configurations are in use.
3701 2. The exinext utility was not formatting the output nicely when there was
3702 an alternate port involved in the retry record key, nor when there was a
3703 message id as well (for retries that were specific to a specific message
3704 and a specific host). It was also confused by IPv6 addresses, because of
3705 the additional colons they contain. I have fixed the IPv4 problem, and
3706 patched it up to do a reasonable job for IPv6.
3708 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3709 delivery, the log line now contains "pipelined" if PIPELINING was used.
3711 4. An SMTP transport process used to panic and die if the bind() call to set
3712 an explicit outgoing interface failed. This has been changed; it is now
3713 treated in the same way as a connect() failure.
3715 5. A reference to $sender_host_name in the part of a conditional expansion
3716 that was being skipped was still causing a DNS lookup. This no longer
3719 6. The def: expansion condition was not recognizing references to header lines
3720 that used bh_ and bheader_.
3722 7. Added the _cache feature to named lists.
3724 8. The code for checking quota_filecount in the appendfile transport was
3725 allowing one more file than it should have been.
3727 9. For compatibility with Sendmail, the command line option
3735 and sets the incoming protocol and host name (for trusted callers). The
3736 host name and its colon can be omitted when only the protocol is to be set.
3737 Note the Exim already has two private options, -pd and -ps, that refer to
3738 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3739 "s", but I don't think that's a major issue.
3741 10. A number of refactoring changes to the code, none of which should affect
3744 (a) The number of logging options was getting close to filling up the
3745 32-bit word that was used as a bit map. I have split them into two classes:
3746 those that are passed in the argument to log_write(), and those that are
3747 only ever tested independently outside of that function. These are now in
3748 separate 32-bit words, so there is plenty of room for expansion again.
3749 There is no change in the user interface or the logging behaviour.
3751 (b) When building, for example, log lines, the code previously used a
3752 macro that called string_cat() twice, in order to add two strings. This is
3753 not really sufficiently general. Furthermore, there was one instance where
3754 it was actually wrong because one of the argument was used twice, and in
3755 one call a function was used. (As it happened, calling the function twice
3756 did not affect the overall behaviour.) The macro has been replaced by a
3757 function that can join an arbitrary number of extra strings onto a growing
3760 (c) The code for expansion conditions now uses a table and a binary chop
3761 instead of a serial search (which was left over from when there were very
3762 few conditions). Also, it now recognizes conditions like "pam" even when
3763 the relevant support is not compiled in: a suitably worded error message is
3764 given if an attempt is made to use such a condition.
3766 11. Added ${time_interval:xxxxx}.
3768 12. A bug was causing one of the ddress fields not to be passed back correctly
3769 from remote delivery subprocesses. The field in question was not being
3770 subsequently used, so this caused to problems in practice.
3772 13. Added new log selectors queue_time and deliver_time.
3774 14. Might have fixed a bug in maildirsizefile handling that threw up
3775 "unexpected character" debug warnings, and recalculated the data
3776 unnecessarily. In any case, I expanded the warning message to give more
3779 15. Added the message "Restricted characters in address" to the statements in
3780 the default ACL that block characters like @ and % in local parts.
3782 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3783 Three changes have been made:
3785 (a) There was a serious bug; a negative response to MAIL caused the whole
3786 recipient domain to be cached as invalid, thereby blocking all messages
3787 to all local parts at the same domain, from all senders. This bug has
3788 been fixed. The domain is no longer cached after a negative response to
3789 MAIL if the sender used is not empty.
3791 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3794 (c) A new callout option, "use_sender" has been added for people who want
3795 the modified behaviour.
3801 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3804 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3805 indeed breaks things for older releases.
3807 3. Added additional logging to the case where there is a problem reading data
3808 from a filter that is running in a subprocess using a pipe, in order to
3809 try to track down a specific problem.
3811 4. Testing facility fudge: when running in the test harness and attempting
3812 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3813 getting "No route to host". Convert this to a timeout.
3815 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3818 6. Some OS don't have socklen_t but use size_t instead. This affects the
3819 fifth argument of getsockopt() amongst other things. This is now
3820 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3821 can be set for individual OS. I have set it for SunOS5, OSF1, and
3822 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3823 some earlier ones do not.
3825 7. Change 4.30/15 was not doing the test caselessly.
3827 8. The standard form for an IPv6 address literal was being rejected by address
3828 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3829 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3830 this, as well as the form without the "IPv6" on the front (but only when
3831 address literals are enabled, of course).
3833 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3835 10. Exim crashed if a message with an empty sender address specified by -f
3836 encountered a router with an errors_to setting. This could be provoked only
3837 by a command such as
3841 where an empty string was supplied; "<>" did not hit this bug.
3843 11. Installed PCRE release 4.5.
3845 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3846 remained set. It is now erased.
3848 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3849 times from message ids (which are base 36 rather than the normal 62).
3851 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3852 were being counted as actual protocol errors, and logged if the log
3853 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3854 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3855 and DATA following a set of rejected RCPTs do not count as protocol errors.
3856 In other words, Exim assumes they were pipelined, though this may not
3857 actually be the case. Of course, in all cases the client gets an
3858 appropriate error code.
3860 15. If a lookup fails in an ACL condition, a message about the failure may
3861 be available; it is used if testing the ACL cannot continue, because most
3862 such messages specify what the cause of the deferral is. However, some
3863 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3864 that caused an old message to be retained and used if a later statement
3865 caused a defer, replacing the real cause of the deferral.
3867 16. If an IP address had so many PTR records that the DNS lookup buffer
3868 was not large enough to hold them, Exim could crash while trying to process
3869 the truncated data. It now detects and logs this case.
3871 17. Further to 4.21/58, another change has been made: if (and only if) the
3872 first line of a message (the first header line) ends with CRLF, a bare LF
3873 in a subsequent header line has a space inserted after it, so as not to
3874 terminate the header.
3876 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3877 unnecessarily, used atoi() instead of strtol(), and didn't check the
3878 termination when getting file sizes from file names by regex.
3880 19. Completely re-implemented the support for maildirsize files, in the light
3881 of a number of problems with the previous contributed implementation
3882 (4.30/29). In particular:
3884 . If the quota is zero, the maildirsize file is maintained, but no quota is
3887 . If the maildir directory does not exist, it is created before any attempt
3888 to write a maildirsize file.
3890 . The quota value in the file is just a cache; if the quota is changed in
3891 the transport, the new value overrides.
3893 . A regular expression is available for excluding directories from the
3896 20. The autoreply transport checks the characters in options that define the
3897 message's headers; it allows continued headers, but it was checking with
3898 isspace() after an embedded newline instead of explicitly looking for a
3901 21. If all the "regular" hosts to which an address was routed had passed their
3902 expiry times, and had not reached their retry times, the address was
3903 bounced, even if fallback hosts were defined. Now Exim should go on to try
3906 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3907 equivalent code in the SMTP transport. Some hosts send humungous responses
3908 to HELO/EHLO, more than 1024 it seems.
3910 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3911 gives compiler warnings in some environments. I've now done it "properly",
3914 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3915 (because of problems with the built-in one) was declared to return uschar *
3916 instead of char *, causing compiler failure.
3918 25. Fixed a file descriptor leak when processing alias/forward files.
3920 26. Fixed a minor format string issue in dbfn.c.
3922 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3924 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3925 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3928 29. When a sender address is verified, it is cached, to save repeating the test
3929 when there is more than one recipient in a message. However, when the
3930 verification involves a callout, it is possible for different callout
3931 options to be set for different recipients. It is too complicated to keep
3932 track of this in the cache, so now Exim always runs a verification when a
3933 callout is required, relying on the callout cache for the optimization.
3934 The overhead is duplication of the address routing, but this should not be
3937 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3938 to be verified with callout=postmaster, and the main callout worked but the
3939 postmaster check failed, the verification correctly failed. However, if a
3940 subsequent RCPT command asked for sender verification *without* the
3941 postmaster check, incorrect caching caused this verification also to fail,
3944 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3945 it was not caching the DNS options (qualify_single, search_parents) that
3946 were used when the lookup failed. A subsequent lookup with different
3947 options therefore always gave the same answer, though there were cases
3948 where it should not have. (Example: a "domains = !$mx_any" option on a
3949 dnslookup router: the "domains" option is always processed without any
3950 widening, but the router might have qualify_single set.) Now Exim uses the
3951 cached value only when the same options are set.
3953 32. Added John Jetmore's "exipick" utility to the distribution.
3955 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3956 than a timeout (e.g. a certificate is required, and is not provided), an
3957 Exim server now closes the connection immediately. Previously it waited for
3958 the client to close - but if the client is SSL, it seems that they each
3959 wait for each other, leading to a delay before one of them times out.
3961 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3962 maintained 0.8.x compatibility because I don't think many are using it, and
3963 it is clearly obsolete.
3965 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3968 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3969 client certificate was expired. A simple patch fixes this, though I don't
3970 understand the full logic of why the verify callback is called multiple
3973 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3974 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3975 which causes problems with some clients (such as the Certicom SSL Plus
3976 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3977 disables the coutermeasure allowing Eudora to connect."
3979 38. Exim was not checking that a write() to a log file succeeded. This could
3980 lead to Bad Things if a log got too big, in particular if it hit a file
3981 size limit. Exim now panics and dies if it cannot write to a log file, just
3982 as it does if it cannot open a log file.
3984 39. Modified OS/Makefile-Linux so that it now contains
3986 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3988 The two -D definitions ensure that Exim is compiled with large file
3989 support, which makes it possible to handle log files that are bigger than
3992 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3993 instance) a domain was checked against a named list that involved a lookup,
3994 causing $domain_data to be set, then another domain was checked against the
3995 same list, then the first domain was re-checked, the value of $domain_data
3996 after the final check could be wrong. In particular, if the second check
3997 failed, it could be set empty. This bug probably also applied to
4000 41. The strip_trailing_dot option was not being applied to the address given
4001 with the -f command-line option.
4003 42. The code for reading a message's header from the spool was incrementing
4004 $received_count, but never initializing it. This meant that the value was
4005 incorrect (doubled) while delivering a message in the same process in which
4006 it was received. In the most common configuration of Exim, this never
4007 happens - a fresh exec is done - but it can happen when
4008 deliver_drop_privilege is set.
4010 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4011 it now includes up to 150 characters of the unexpected data in the log
4014 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4015 and building data strings. The size of both of these buffers was 10 000
4016 bytes - far larger than anybody would *ever* want, thought I. Needless to
4017 say, somebody hit the limit. I have increased the maximum line length to
4018 20 000 and the maximum data length of concatenated lines to 100 000. I have
4019 also fixed two bugs, because there was no checking on these buffers. Tsk,
4020 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4021 buffer is too small.
4023 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4024 lsearch lookups. Now it does.
4026 46. When parsing a route_list item in a manualroute router, a fixed-length
4027 buffer was used for the list of hosts. I made this 1024 bytes long,
4028 thinking that nobody would ever have a list of hosts that long. Wrong.
4029 Somebody had a whole pile of complicated expansion conditions, and the
4030 string was silently truncated, leading to an expansion error. It turns out
4031 that it is easier to change to an unlimited length (owing to other changes
4032 that have happened since this code was originally written) than to build
4033 structure for giving a limitation error. The length of the item that
4034 expands into the list of hosts is now unlimited.
4036 47. The lsearch lookup could not handle data where the length of text line was
4037 more than 4095 characters. Such lines were truncated, leading to shortened
4038 data being returned. It should now handle lines of any length.
4040 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4041 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4044 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4045 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4046 no longer happens in comment lines. A list of replacements is now placed
4047 at the head of all of the source files, except those whose only change is
4048 to replace PERL_COMMAND in the very first #! line.
4050 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4051 messages on the queue, with a bottom-up merge sort, using code contributed
4052 by Michael Haardt. This should make operations like -bp somewhat faster on
4053 large queues. It won't affect queue runners, except when queue_run_in_order
4056 51. Installed eximstats 1.31 in the distribution.
4058 52. Added support for SRV lookups to the dnslookup router.
4060 53. If an ACL referred to $message_body or $message_body_end, the value was not
4061 reset for any messages that followed in the same SMTP session.
4063 54. The store-handling optimization for building very long strings was not
4064 differentiating between the different store pools. I don't think this
4065 actually made any difference in practice, but I've tidied it.
4067 55. While running the routers to verify a sender address, $sender_address
4068 was still set to the sender address. This is wrong, because when routing to
4069 send a bounce to the sender, it would be empty. Therefore, I have changed
4070 it so that, while verifying a sender address, $sender_address is set to <>.
4071 (There is no change to what happens when verifying a recipient address.)
4073 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4074 target A or AAAA records (if not already returned) without resetting the
4075 qualify_single or search_parents options of the DNS resolver. These are
4076 inappropriate in this case because the targets of MX and SRV records must
4077 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4078 target that, when qualified, matched something in the local domain. These
4079 two options are now turned off when doing these lookups.
4081 57. It seems that at least some releases of Reiserfs (which does not have the
4082 concept of a fixed number of inodes) returns zero and not -1 for the
4083 number of available inodes. This interacted badly with check_spool_inodes,
4084 which assumed that -1 was the "no such thing" setting. What I have done is
4085 to check that the total number of inodes is greater than zero before doing
4086 the test of how many are available.
4088 58. When a "warn" ACL statement has a log_message modifier, the message is
4089 remembered, and not repeated. This is to avoid a lot of repetition when a
4090 message has many recipients that cause the same warning to be written.
4091 Howewer, Exim was preserving the list of already written lines for an
4092 entire SMTP session, which doesn't seem right. The memory is now reset if a
4093 new message is started.
4095 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4096 debugging output unless log_rewrite was also set.
4098 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4099 of (int)(handle) when we know that handle contains (void *)(-1).
4101 61. The Exim daemon panic-logs an error return when it closes the incoming
4102 connection. However "connection reset by peer" seems to be common, and
4103 isn't really an error worthy of noting specially, so that particular error
4106 62. When Exim is trying to find all the local interfaces, it used to panic and
4107 die if the ioctl to get the interface flags failed. However, it seems that
4108 on at least one OS (Solaris 9) it is possible to have an interface that is
4109 included in the list of interfaces, but for which you get a failure error
4110 for this call. This happens when the interface is not "plumbed" into a
4111 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4112 failure of the "get flags" call assumes that the interface is down.
4114 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4115 makes life easier for people who are doing arithmetic on fields extracted
4116 from dates, where you often get leading zeros that should not be
4117 interpreted as octal.
4119 64. Added qualify_domain to the redirect router, to override the global
4122 65. If a pathologically long header line contained very many addresses (the
4123 report of this problem mentioned 10 000) and each of them was rewritten,
4124 Exim could use up a very large amount of memory. (It kept on making new
4125 copies of the header line as it rewrote, and never released the old ones.)
4126 At the expense of a bit more processing, the header rewriting function has
4127 been changed so that it no longer eats memory in this way.
4129 66. The generation of the Received: header has been moved from the time that a
4130 message starts to be received, to the time that it finishes. The timestamp
4131 in the Received: header should now be very close to that of the <= log
4132 line. There are two side-effects of this change:
4134 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4135 logged header lines no longer include the local Received: line, because
4136 it has not yet been created. The same applies to a copy of the message
4137 that is returned to a non-SMTP sender when a message is rejected.
4139 (b) When a filter file is tested using -bf, no additional Received: header
4140 is added to the test message. After some thought, I decided that this
4143 This change does not affect the value of $received_for. It is still set
4144 after address rewriting, but before local_scan() is called.
4146 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4148 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4149 gave an unhelpful panic error message, and a defer error. I have managed to
4150 change this behaviour so that it now rejects any supplied certificate,
4151 which seems right, as the list of acceptable certificates is empty.
4153 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4154 gave an unhelpful defer error. I have not managed to make this reject any
4155 supplied certificates, but the error message it gives is "no certificate
4156 supplied", which is not helpful.
4158 70. exigrep's output now also includes lines that are not associated with any
4159 message, but which match the given pattern. Implemented by a patch from
4160 Martin Sluka, which also tidied up the Perl a bit.
4162 71. Recipient callout verification, like sender verification, was using <> in
4163 the MAIL FROM command. This isn't really the right thing, since the actual
4164 sender may affect whether the remote host accepts the recipient or not. I
4165 have changed it to use the actual sender in the callout; this means that
4166 the cache record is now keyed on a recipient/sender pair, not just the
4167 recipient address. There doesn't seem to be a real danger of callout loops,
4168 since a callout by the remote host to check the sender would use <>.
4169 [SEE ABOVE: changed after hitting problems.]
4171 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4172 temporary errors. However, in the case of such a code being given after
4173 the end of a data transmission (i.e. after ".") Exim was failing to write
4174 a retry record for the message. (Yes, there was some broken host that was
4175 actually sending 8xx at this point.)
4177 73. An unknown lookup type in a host list could cause Exim to panic-die when
4178 the list was checked. (An example that provoked this was putting <; in the
4179 middle of a list instead of at the start.) If this happened during a DATA
4180 ACL check, a -D file could be left lying around. This kind of configuration
4181 error no longer causes Exim to die; instead it causes a defer errror. The
4182 incident is still logged to the main and panic logs.
4184 74. Buglet left over from Exim 3 conversion. The message "too many messages
4185 in one connection" was written to the rejectlog but not the mainlog, except
4186 when address rewriting (yes!) was being logged.
4188 75. Added write_rejectlog option.
4190 76. When a system filter was run not as root (that is, when system_filter_user
4191 was set), the values of the $n variables were not being returned to the
4192 main process; thus, they were not subsequently available in the $sn
4195 77. Added +return_path_on_delivery log selector.
4197 78. A connection timeout was being treated differently from recipients deferred
4198 when testing hosts_max_try with a message that was older than the host's
4199 retry timeout. (The host should not be counted, thus allowing all hosts to
4200 be tried at least once before bouncing.) This may have been the cause of an
4201 occasionally reported bug whereby a message would remain on the queue
4202 longer than the retry timeout, but would be bounced if a delivery was
4203 forced. I say "may" because I never totally pinned down the problem;
4204 setting up timeout/retry tests is difficult. See also the next item.
4206 79. The ultimate address timeout was not being applied to errors that involved
4207 a combination of host plus message (for example, a timeout on a MAIL
4208 command). When an address resolved to a number of possible hosts, and they
4209 were not all tried for each delivery (e.g. because of hosts_max_try), a
4210 message could remain on the queue longer than the retry timeout.
4212 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4215 81. Fixed an obscure SMTP outgoing bug which required at least the following
4216 conditions: (a) there was another message waiting for the same server;
4217 (b) the server returned 5xx to all RCPT commands in the first message so
4218 that the message was not completed; (c) the server dropped the connection
4219 or gave a negative response to the RSET that Exim sends to abort the
4220 transaction. The observed case was a dropped connection after DATA that had
4221 been sent in pipelining mode. That is, the server had advertised PIPELINING
4222 but was not implementing it correctly. The effect of the bug was incorrect
4223 behaviour, such as trying another host, and this could lead to a crash.
4229 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4230 and daemon.c were passed as pointers to ints; they should have been
4231 pointers to socklen_t variables (which are typically unsigned ints).
4233 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4236 3. Fixed a really odd bug that affected only the testing scheme; patching a
4237 certain fixed string in the binary changed the value of another string that
4238 happened to be identical to the end of the original first string.
4240 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4241 name", it returns that address as the IP address. On some operating
4242 systems (e.g. Solaris), it also passes back the IP address string as the
4243 "host name". However, on others (e.g. Linux), it passes back an empty
4244 string. Exim wasn't checking for this, and was changing the host name to an
4245 empty string, assuming it had been canonicized.
4247 5. Although rare, it is permitted to have more than one PTR record for a given
4248 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4249 all the names associated with an address, because they do in Solaris.
4250 However, it seems that they do not in Linux for data that comes from the
4251 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4252 I found this out when I moved to a new Linux workstation and tried to run
4253 the Exim test suite.
4255 To get round this problem I have changed the code so that it now does its
4256 own call to the DNS to look up PTR records when searching for a host name.
4257 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4258 addresses that are only in /etc/hosts are still found.
4260 This behaviour is, however, controlled by an option called host_lookup_
4261 order, which defaults to "bydns:byaddr". If people want to use the other
4262 order, or indeed, just use one or the other means of lookup, they can
4263 specify it in this variable.
4265 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4266 some operating systems, this comes back from gethostbyaddr() as an empty
4267 string, and this is what Exim used to test for. However, it seems that in
4268 other systems, "." is yielded. Exim now tests for this case too.
4270 7. The values of check_spool_space and check_log_space are now held internally
4271 as a number of kilobytes instead of an absolute number of bytes. If a
4272 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4273 kilobyte. This means that much larger values can be stored.
4275 8. Exim monitor: an attempt to get the action menu when not actually pointing
4276 at a message produces an empty menu entitled "No message selected". This
4277 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4278 no entries in it ("Shell widget menu has zero width and/or height"). So I
4279 have added a single, blank menu entry in this case.
4281 9. Added ${quote_local_part.
4283 10. MIME decoding is now applied to the contents of Subject: header lines when
4286 11. Now that a reference to $sender_host_address automatically causes a reverse
4287 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4288 host lookup before query-style lookups in lists that might use this
4289 variable. This has therefore been abolished, and the "net-" prefix is no
4290 longer necessary for query-style lookups.
4292 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4293 have been a typo for LFLAGS, so it has been changed.
4295 13. The install script calls Exim with "-C /dev/null" in order to find the
4296 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4297 to be output. Howeve, since Exim outputs its version number before the
4298 error, it didn't break the script. It just looked ugly. I fixed this by
4299 always allowing "-C /dev/null" if the caller is root.
4301 14. Ignore overlarge ACL variable number when reading spool file - insurance
4302 against a later release with more variables having written the file.
4304 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4305 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4306 this, as well as the form without the "IPv6" on the front.
4308 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4309 OS/Makefile-Darwin file.
4311 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4312 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4314 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4315 variable. However, some OS use ints for inodes. I've added cast to long int
4316 to get rid of the compiler warning.
4318 19. I had forgotten to lock out "/../" in configuration file names when
4319 ALT_CONFIG_PREFIX was set.
4321 20. Routers used for verification do not need to specify transports. However,
4322 if such a router generated a host list, and callout was configured, Exim
4323 crashed, because it could not find a port number from the (non-existent)
4324 transport. It now assumes port 25 in this circumstance.
4326 21. Added the -t option to exigrep.
4328 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4329 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4330 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4331 them is set, LOOKUP_LSEARCH is forced.
4333 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4335 24. Added sender and host information to the "rejected by local_scan()" log
4336 line; previously there was no indication of these.
4338 25. Added .include_if_exists.
4340 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4341 Linux. It turns out that not all file systems support this. Apparently some
4342 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4343 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4344 not supported on the file descriptor, is now ignored when Exim is trying to
4345 sync a directory. This applies only to Linux.
4347 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4349 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4350 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4353 29. Added support for maildirsize files from supplied patch (modified a bit).
4355 30. The use of :fail: followed by an empty string could lead Exim to respond to
4356 sender verification failures with (e.g.):
4358 550 Verification failed for <xxx>
4359 550 Sender verify failed
4361 where the first response line was missing the '-' that indicates it is not
4362 the final line of the response.
4364 31. The loop for finding the name of the user that called Exim had a hardwired
4365 limit of 10; it now uses the value of finduser_retries, which is used for
4366 all other user lookups.
4368 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4371 33. Exim was neglecting to zero errno before one call of strtol() when
4372 expanding a string and expecting an integer value. On some systems this
4373 resulted in spurious "integer overflow" errors. Also, it was casting the
4374 result into an int without checking.
4376 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4377 did not work. The code looks as if it has *never* worked, though it appears
4378 to have been documented since at least releast 1.62. I have made it work.
4380 35. The "timeout_DNS" error in retry rules, also documented since at least
4381 1.62, also never worked. As it isn't clear exactly what this means, and
4382 clearly it isn't a major issue, I have abolished the feature by treating it
4383 as "timeout", and writing a warning to the main and panic logs.
4385 36. The display of retry rules for -brt wasn't always showing the error code
4388 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4389 timeout_connect_A, timeout_connect_MX.
4391 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4392 to the empty sender.
4394 39. The daemon was not analysing the content of -oX till after it had closed
4395 stderr and disconnected from the controlling terminal. This meant that any
4396 syntax errors were only noted on the panic log, and the return code from
4397 the command was 0. By re-arranging the code a little, I've made the
4398 decoding happen first, so such errors now appear on stderr, and the return
4399 code is 1. However, the actual setting up of the sockets still happens in
4400 the disconnected process, so errors there are still only recorded on the
4403 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4404 connections (as happens on some IP stacks) was logged at start up time as
4405 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4406 it from "IPv6 and IPv4", which means that two separate sockets are being
4409 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4410 says whether AF_INET or AF_INET6 was passed as an argument.
4412 42. Exiwhat output was messed up when time zones were included in log
4415 43. Exiwhat now gives more information about the daemon's listening ports,
4416 and whether -tls-on-connect was used.
4418 44. The "port" option of the smtp transport is now expanded.
4420 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4421 silently ignored. Now an error message is written to the main and panic
4424 46. There's a new ACL modifier called "logwrite" which writes to a log file
4425 as soon as it is encountered.
4427 47. Added $local_user_uid and $local_user_gid at routing time.
4429 48. Exim crashed when trying to verify a sender address that was being
4432 49. Exim was recognizing only a space character after ".include". It now also
4433 recognizes a tab character.
4435 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4436 extracting the relevant information from the specification. The man page no
4437 longer contains scrambled data for the -d option, and I've added a section
4438 at the front about calling Exim under different names.
4440 51. Added "extra_headers" argument to the "mail" command in filter files.
4442 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4445 53. Installed eximstats 1.29.
4447 54. Added transport_filter_timeout as a generic transport option.
4449 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4450 Cc: header lines. This was required by RFC 822, but it not required by RFC
4453 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4454 incoming messages that did not have them. Now it does so only if the
4455 message originates locally, that is, if there is no associated remote host
4456 address. When Resent- header lines are present, this applies to the Resent-
4457 lines rather than the non-Resent- lines.
4459 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4460 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4462 58. Messages for configuration errors now include the name of the main
4463 configuration file - useful now that there may be more than one file in a
4464 list (.included file names were always shown).
4466 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4467 for those rare installations that do not start the daemon as root or run it
4468 setuid root. I've cut out the call to initgroups() if the daemon is not
4471 60. The Exim user and group can now be bound into the binary as text strings
4472 that are looked up at the start of Exim's processing.
4474 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4476 62. Added $mailstore_basename variable.
4478 63. Installed patch to sieve.c from Michael Haardt.
4480 64. When Exim failed to open the panic log after failing to open the main log,
4481 the original message it was trying to log was written to stderr and debug
4482 output, but if they were not available (the usual case in production), it
4483 was lost. Now it is written to syslog before the two lines that record the
4484 failures to open the logs.
4486 65. Users' Exim filters run in subprocesses under the user's uid. It is
4487 possible for a "deliver" command or an alias in a "personal" command to
4488 provoke an address rewrite. If logging of address rewriting is configured,
4489 this fails because the process is not running as root or exim. There may be
4490 a better way of dealing with this, but for the moment (because 4.30 needs
4491 to be released), I have disabled address rewrite logging when running a
4492 filter in a non-root, non-exim process.
4498 1. The buildconfig auxiliary program wasn't quoting the value set for
4499 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4500 not defined. This bug was masked in 4.22 by the effect that was fixed in
4503 2. Some messages that were rejected after a message id was allocated were
4504 shown as "incomplete" by exigrep. It no longer does this for messages that
4505 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4507 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4508 have allow_domain_literals set, the ID did not get logged in the <= line.
4509 Domain literals are now always recognized in Message-ID: header lines.
4511 4. The first argument for a ${extract expansion item is the key name or field
4512 number. Leading and trailing spaces in this item were not being ignored,
4513 causing some misleading effects.
4515 5. When deliver_drop_privilege was set, single queue runner processes started
4516 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4517 same command in the process it spins off) were not dropping privilege.
4519 6. When the daemon running as "exim" started a queue runner, it always
4520 re-executed Exim in the spun-off process. This is a waste of effort when
4521 deliver_drop_privilege is set. The new process now just calls the
4522 queue-runner function directly.
4528 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4531 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4532 ignored. Though the use of -oP was forcing the writing of a pid file, it
4533 was always written to the default place.
4535 3. If the message "no IP address found for host xxxx" is generated during
4536 incoming verification, it is now followed by identification of the incoming
4537 connection (so you can more easily find what provoked it).
4539 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4541 5. Added some features to "harden" Exim a bit more against certain attacks:
4543 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4544 be put in Local/Makefile. This is like the never_users runtime option,
4545 but it cannot be overridden. The default setting is "root".
4547 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4548 prefix string with which any file named in a -C command line option
4551 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4552 is retained for -C and -D only if the caller of Exim is root. Without
4553 it, the exim user may also use -C and -D and retain privilege.
4555 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4556 command line option is disabled.
4558 6. Macro names set by the -D option must start with an upper case letter, just
4559 like macro names defined in the configuration file.
4561 7. Added "dereference=" facility to LDAP.
4563 8. Two instances of the typo "uknown" in the source files are fixed.
4565 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4566 the Configure-Makefile script screwed up while processing it.
4568 10. Incorporated PCRE 4.4.
4570 11. The SMTP synchronization check was not operating right at the start of an
4571 SMTP session. For example, it could not catch a HELO sent before the client
4572 waited for the greeting. There is now a check for outstanding input at the
4573 point when the greeting is written. Because of the duplex, asynchronous
4574 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4575 way, but not yet received, when the check is performed.
4577 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4578 on TCP/IP sockets, because this apparently causes some broken clients to
4581 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4582 unchanged) from the Cygwin maintainer.
4584 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4585 was supported instead of testing for mbx. Effectively a typo.
4587 15. The spa authenticator server code was not checking that the input it
4588 received was valid base64.
4590 16. The debug output line for the "set" modifier in ACLs was not showing the
4591 name of the variable that was being set.
4593 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4595 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4596 lookup on its own. Something else was needed to trigger the lookup. For
4597 example, a match in host_lookup or the need for a host name in a host list.
4598 Now, if $sender_host_name is referenced and the host name has not yet been
4599 looked up, a lookup is performed. If the lookup fails, the variable remains
4600 empty, and $host_lookup_failed is set to "1".
4602 19. Added "eqi" as a case-independent comparison operator.
4604 20. The saslauthd authentication condition could segfault if neither service
4605 nor realm was specified.
4607 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4608 error message that was logged was misleading, and incoming SMTP
4609 connections were dropped. The message is now more accurate, and temporary
4610 errors are given to SMTP connections.
4612 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4613 (except RSET) with a 421 error, until QUIT is received. However, it was
4614 failing to send a response to QUIT.
4616 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4617 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4618 ACL is now run after the helo_try_verify_hosts code.
4620 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4621 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4622 case-independent, so other case variants are also recognized). Apparently
4623 some systems use these upper case variants.
4625 25. If more than two messages were waiting for the same host, and a transport
4626 filter was specified for the transport, Exim sent two messages over the
4627 same TCP/IP connection, and then failed with "socket operation on non-
4628 socket" when it tried to send the third.
4630 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4632 27. The extern definition of crypt16() in expand.c was not being excluded when
4633 the OS had its own crypt16() function.
4635 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4636 as a preferred synonym for return_size_limit, both as an option and as an
4639 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4641 30. Changed the default configuration ACL to relax the local part checking rule
4642 for addresses that are not in any local domains. For these addresses,
4643 slashes and pipe symbols are allowed within local parts, but the sequence
4644 /../ is explicitly forbidden.
4646 31. SPA server authentication was not clearing the challenge buffer before
4649 32. log_message in a "warn" ACL statement was writing to the reject log as
4650 well as to the main log, which contradicts the documentation and doesn't
4651 seem right (because no rejection is happening). So I have stopped it.
4653 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4654 However, I am unable to do any testing of this.
4656 34. Fixed an infelicity in the appendfile transport. When checking directories
4657 for a mailbox, to see if any needed to be created, it was accidentally
4658 using path names with one or more superfluous leading slashes; tracing
4659 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4661 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4662 contents are added to the log line that is written for every discarded
4663 recipient. (Previously a log_message setting was ignored.)
4665 36. The ${quote: operator now quotes the string if it is empty.
4667 37. The install script runs exim in order to find its version number. If for
4668 some reason other than non-existence or emptiness, which it checks, it
4669 could not run './exim', it was installing it with an empty version number,
4670 i.e. as "exim-". This error state is now caught, and the installation is
4673 38. An argument was missing from the function that creates an error message
4674 when Exim fails to connect to the socket for saslauthd authentication.
4675 This could cause Exim to crash, or give a corrupted message.
4677 39. Added isip, isip4, and isip6 to ${if conditions.
4679 40. The ACL variables $acl_xx are now saved with the message, and can be
4680 accessed later in routers, transports, and filters.
4682 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4683 strings in the file are not string-expanded.
4685 42. If a MAIL command specified a SIZE value that was too large to fit into an
4686 int variable, the check against message_size_limit failed. Such values are
4687 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4688 day this will have to be increased, but I don't think I want to be around
4689 when emails are that large.
4696 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4697 iconv() is not standard in FreeBSD.
4699 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4700 IPv6 enabled. The observed symptom was a segmentation fault on return from
4701 the function os_common_find_running_interfaces() in src/os.c.
4703 3. In the check_special_case() function in daemon.c I had used "errno" as an
4704 argument name, which causes warnings on some systems. This was basically a
4705 typo, since it was named "eno" in the comments!
4707 4. The code that waits for the clock to tick (at a resolution of some fraction
4708 of a second) so as to ensure message-id uniqueness was always waiting for
4709 at least one whole tick, when it could have waited for less. [This is
4710 almost certainly not relevant at current processor speeds, where it is
4711 unlikely to ever wait at all. But we try to future-proof.]
4713 5. The function that sleeps for a time interval that includes fractions of a
4714 second contained a race. It did not block SIGALRM between setting the
4715 timer, and suspending (a couple of lines later). If the interval was short
4716 and the sigsuspend() was delayed until after it had expired, the suspension
4717 never ended. On busy systems this could lead to processes getting stuck for
4720 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4721 process, before it forks any delivery processes. The open lookup caching
4722 mechanism meant that the open file or database connection was passed into
4723 the delivery process. The problem was that delivery processes always tidy
4724 up cached lookup data. This could cause a problem for the next delivery
4725 process started by the queue runner, because the external queue runner
4726 process does not know about the closure. So the next delivery process
4727 still has data in the lookup cache. In the case of a file lookup, there was
4728 no problem because closing a file descriptor in a subprocess doesn't affect
4729 the parent. However, if the lookup was caching a connection to a database,
4730 the connection was closed, and the second delivery process was likely to
4731 see errors such as "PGSQL: query failed: server closed the connection
4732 unexpectedly". The problem has been fixed by closing all cached lookups
4733 in a queue runner before running a delivery process.
4735 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4736 seem to have the "const" qualifier which it has on other OS. I've
4739 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4740 *of the same type* (client or server) with the same public name that an
4741 error should be diagnosed.
4743 9. When Exim looked up a host name for an IP address, but failed to find the
4744 original IP address when looking up the host name (a safety check), it
4745 output the message "<ip address> does not match any IP for NULL", which was
4746 confusing, to say the least. The bug was that the host name should have
4747 appeared instead of "NULL".
4749 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4750 user that is built into the binary, and the -C or -D options is used, root
4751 privilege is dropped before the configuration file is read. In addition,
4752 logging is switched to stderr instead of the normal log files. If the
4753 configuration then re-defines the Exim user, the unprivileged environment
4754 is probably not what is expected, so Exim logs a panic warning message (but
4757 However, if deliver_drop_privilege is set, the unprivileged state may well
4758 be exactly what is intended, so the warning has been cut out in that case,
4759 and Exim is allowed to try to write to its normal log files.
4765 1. smtp_return_error_details was not giving details for temporary sender
4766 or receiver verification errors.
4768 2. Diagnose a configuration error if two authenticators have the same public
4771 3. Exim used not to create the message log file for a message until the first
4772 delivery attempt. This could be confusing when incoming messages were held
4773 for policy or load reasons. The message log file is now created at the time
4774 the message is received, and an initial "Received" line is written to it.
4776 4. The automatically generated man page for command line options had a minor
4777 bug that caused no ill effects; however, a more serious problem was that
4778 the procedure for building the man page automatically didn't always
4779 operate. Consequently, release 4.20 contains an out-of-date version. This
4780 shouldn't happen again.
4782 5. When building Exim with embedded Perl support, the script that builds the
4783 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4784 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4786 6. The freeze_tell option was not being used for messages that were frozen on
4787 arrival, either by an ACL or by local_scan().
4789 7. Added the smtp_incomplete_transaction log selector.
4791 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4792 was accepting AUTH without a new EHLO.
4794 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4795 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4798 10. Logging of TCP/IP connections (when configured) now happens in the main
4799 daemon process instead of the child process, so that the TCP/IP connection
4800 count is more accurate (but it can never be perfect).
4802 11. The use of "drop" in a nested ACL was not being handled correctly in the
4803 outer ACL. Now, if condition failure induced by the nested "drop" causes
4804 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4805 or "require"), the connection is dropped.
4807 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4808 that yield "discard" can now be used with an "accept" or a "discard" verb,
4809 but an error is generated for any others (because I can't see a useful way
4810 to define what should happen).
4812 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4813 are now processed in the same way as lines in the Exim configuration file.
4814 In particular, continuation lines are supported.
4816 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4818 15. Added -ti meaning -t -i.
4820 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4821 domains, and warn by logging if others are found.
4823 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4824 returned from GIFCONF when Exim is trying to find the list of interfaces on
4825 a host. The code in os.c has been modified to copy each ifreq to an aligned
4826 structure in all cases.
4828 Also, in some cases, the returned ifreq's were being copied to a 'struct
4829 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4830 means the last couple of bytes of an IPv6 address could be chopped if the
4831 ifreq contained only a normal sockaddr (14 bytes storage).
4833 18. Named domain lists were not supported in the hosts_treat_as_local option.
4834 An entry such as +xxxx was not recognized, and was treated as a literal
4837 19. Ensure that header lines added by a DATA ACL are included in the reject log
4838 if the ACL subsequently rejects the message.
4840 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4841 MD5 (which is deprecated).
4843 21. When testing a filter file using -bf, Exim was writing a message when it
4844 took the sender from a "From " line in the message, but it was not doing so
4845 when it took $return_path from a Return-Path: header line. It now does.
4847 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4848 with a valid header line field name (a series of printing characters
4849 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4851 23. Changed "disc" in the source to "disk" to conform to the documentation and
4852 the book and for uniformity.
4854 24. Ignore Sendmail's -Ooption=value command line item.
4856 25. When execve() failed while trying to run a command in a pipe transport,
4857 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4858 could be confused with a return value of 69 from the command itself. This
4859 has been changed to 127, the value the shell returns if it is asked to run
4860 a non-existent command. The wording for the related log line suggests a
4861 non-existent command as the problem.
4863 26. If received_header_text expands to an empty string, do not add a Received:
4864 header line to the message. (Well, it adds a token one on the spool, but
4865 marks it "old" so that it doesn't get used or transmitted.)
4867 27. Installed eximstats 1.28 (addition of -nt option).
4869 28. There was no check for failure on the call to getsockname() in the daemon
4870 code. This can fail if there is a shortage of resources on the system, with
4871 ENOMEM, for example. A temporary error is now given on failure.
4873 29. Contrary to the C standard, it seems that in some environments, the
4874 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4875 program. Exim now does this explicitly; it affects the formatting of
4876 timestamps using strftime().
4878 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4879 complaints. I've now initialized all the variables, to avoid this.
4881 32. Header lines added by a system filter were not being "seen" during
4882 transport-time rewrites.
4884 33. The info_callback() function passed to OpenSSL is set up with type void
4885 (*)(SSL *, int, int), as described somewhere. However, when calling the
4886 function (actually a macro) that sets it up, the type void(*)() is
4887 expected. I've put in a cast to prevent warnings from picky compilers.
4889 34. If a DNS black list lookup found a CNAME record, but there were no A
4890 records associated with the domain it pointed at, Exim crashed.
4892 35. If a DNS black list lookup returned more than one A record, Exim ignored
4893 all but the first. It now scans all returned addresses if a particular IP
4894 value is being sought. In this situation, the contents of the
4895 $dnslist_value variable are a list of all the addresses, separated by a
4898 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4899 to accept a host name and all its aliases if the forward lookup for any of
4900 them yielded the IP address of the incoming connection. Now it accepts only
4901 those names whose forward lookup yields the correct IP address. Any other
4902 names are discarded. This closes a loophole whereby a rogue DNS
4903 administrator could create reverse DNS records to break through a
4904 wildcarded host restriction in an ACL.
4906 37. If a user filter or a system filter that ran in a subprocess used any of
4907 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4908 the wrong values were passed to the pipe command ($thisaddress had the
4909 value of $0, $0 had the value of $1, etc). This bug was introduced by
4910 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4912 38. Improved the line breaking for long SMTP error messages from ACLs.
4913 Previously, if there was no break point between 40 and 75 characters, Exim
4914 left the rest of the message alone. Two changes have been made: (a) I've
4915 reduced the minimum length to 35 characters; (b) if it can't find a break
4916 point between 35 and 75 characters, it looks ahead and uses the first one
4917 that it finds. This may give the occasional overlong line, but at least the
4918 remaining text gets split now.
4920 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4921 file descriptors might be low, and that setting 1000 would always raise it.
4922 It turns out that in some environments, the limit is already over 1000 and
4923 that lowering it causes trouble. So now Exim takes care not to decrease it.
4925 40. When delivering a message, the value of $return_path is set to $sender_
4926 address at the start of routing (routers may change the value). By an
4927 oversight, this default was not being set up when an address was tested by
4928 -bt or -bv, which affected the outcome if any router or filter referred to
4931 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4932 log only when "message" or "log_message" are set. However, if one of the
4933 conditions was an address verification, or a call to a nested ACL, the
4934 messages generated by the underlying test were being passed through. This
4935 no longer happens. The underlying message is available in $acl_verify_
4936 message for both "message" and "log_message" expansions, so it can be
4937 passed through if needed.
4939 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4940 new expansion $bh_ to give the encoded byte string without charset
4941 translation. Translation happens only if iconv() is available; HAVE_ICONV
4942 indicates this at build time. HEADERS_CHARSET gives the charset to
4943 translate to; headers_charset can change it in the configuration, and
4944 "headers charset" can change it in an individual filter file.
4946 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4947 that creates RFC 2047 encoded "words" labels them as that charset instead
4948 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4949 expansion operator; (ii) when Exim creates a From: line for a local
4950 message; (iii) when a header line is rewritten to include a "phrase" part.
4952 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4953 buggy, causing it to skip the first lines of messages whose message ID
4954 ended in 'D'. This would not have bitten before Exim release 4.14, because
4955 message IDs were unlikely to end in 'D' before then. The effect was to have
4956 incorrect size information for certain domains.
4958 45. #include "config.h" was missing at the start of the crypt16.c module. This
4959 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4962 46. If there was a timeout during a "random" callout check, Exim treated it as
4963 a failure of the random address, and carried on sending RSET and the real
4964 address. If the delay was just some slowness somewhere, the response to the
4965 original RCPT would be taken as a response to RSET and so on, causing
4966 mayhem of various kinds.
4968 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4969 when I implemented it. It didn't allow for the fact that some option values
4970 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4971 the right test for positive values.
4973 48. Domain names in DNS records are case-independent. Exim always looks them up
4974 in lower case. Some resolvers return domain names in exactly the case they
4975 appear in the zone file, that is, they may contain uppercase letters. Not
4976 all resolvers do this - some return always lower case. Exim was treating a
4977 change of case by a resolver as a change of domain, similar to a widening
4978 of a domain abbreviation. This triggered its re-routing code and so it was
4979 trying to route what was effectively the same domain again. This normally
4980 caused routing to fail (because the router wouldn't handle the domain
4981 twice). Now Exim checks for this case specially, and just changes the
4982 casing of the domain that it ultimately uses when it transmits the message
4985 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4988 50. If a filter generated a file delivery with a non-absolute name (possible if
4989 no home directory exists for the router), the forbid_file option was not
4992 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4993 the existing equality matching.
4995 52. Exim was using ints instead of ino_t variables in some places where it was
4996 dealing with inode numbers.
4998 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4999 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5000 called TMPDIR, and if it finds it is different, it changes its value.
5002 54. The smtp_printf() function is now made available to local_scan() so
5003 additional output lines can be written before returning. There is also an
5004 smtp_fflush() function to enable the detection of a dropped connection.
5005 The variables smtp_input and smtp_batched_input are exported to
5008 55. Changed the default runtime configuration: the message "Unknown user"
5009 has been removed from the ACL, and instead placed on the localuser router,
5010 using the cannot_route_message feature. This means that any verification
5011 failures that generate their own messages won't get overridden. Similarly,
5012 the "Unrouteable address" message that was in the ACL for unverifiable
5013 relay addresses has also been removed.
5015 56. Added hosts_avoid_esmtp to the smtp transport.
5017 57. The exicyclog script was not checking for the esoteric option
5018 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5019 will work only if exicyclog is run under the appropriate euid.
5021 58. Following a discussion on the list, the rules by which Exim recognises line
5022 endings on incoming messages have been changed. The -dropcr and drop_cr
5023 options are now no-ops, retained only for backwards compatibility. The
5024 following line terminators are recognized: LF CRLF CR. However, special
5025 processing applies to CR:
5027 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5028 nor a local message in the state where . is a terminator.
5030 (ii) If a bare CR is encountered in a header line, an extra space is added
5031 after the line terminator so as not to end the header. The reasoning
5032 behind this is that bare CRs in header lines are most likely either
5033 to be mistakes, or people trying to play silly games.
5035 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5036 was being incorrectly given as 18 bytes larger than it should have been.
5037 This is a VOB (very old bug).
5039 60. This may never have affected anything current, but just in case it has:
5040 When the local host is found other than at the start of a list of hosts,
5041 the local host, those with the same MX, and any that follow, are discarded.
5042 When the list in question was part of a longer list of hosts, the following
5043 hosts (not currently being processed) were also being discarded. This no
5044 longer happens. I'm not sure if this situation could ever has previously
5047 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5050 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5051 header. This is something that is recommended in a new Internet Draft, and
5052 is something that is documented as being done by Sendmail. There are two
5053 possible values. For messages generated by the autoreply transport, Exim
5056 Auto-Submitted: auto-replied
5058 whereas for all other generated messages (e.g. bounces) it adds
5060 Auto-Submitted: auto-generated
5062 63. The "personal" condition in filters now includes a test for the
5063 Auto-Submitted: header. If it contains the string "auto-" the message it
5064 not considered personal.
5066 64. Added rcpt_include_affixes as a generic transport option.
5068 65. Added queue_only_override (default true).
5070 66. Added the syslog_duplication option.
5072 67. If what should have been the first header line of a message consisted of
5073 a space followed by a colon, Exim was mis-interpreting it as a header line.
5074 It isn't of course - it is syntactically invalid and should therefore be
5075 treated as the start of the message body. The misbehaviour could have
5076 caused a number of strange effects, including loss of data in subsequent
5077 header lines, and spool format errors.
5079 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5080 client host had authenticated. This control can now be exercised by an ACL
5081 for more flexibility.
5083 69. By default, callouts do not happen when testing with -bh. There is now a
5084 variant, -bhc, which does actually run the callout code, including
5085 consulting and updating the callout cache.
5087 70. Added support for saslauthd authentication, courtesy of Alexander
5090 71. If statvfs() failed on the spool or log directories while checking their
5091 size for availability, Exim confusingly gave the error "space shortage".
5092 Furthermore, in debugging mode it crashed with a floating point exception.
5093 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5094 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5095 serious problem, Exim now writes to the main and panic logs when this
5096 happens, with details of the failure. It then refuses to accept the
5097 incoming message, giving the message "spool directory problem" or "log
5098 directory problem" with a 421 code for SMTP messages.
5100 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5101 0, 1, and 2 exist, because some OS complain for execs without them (see
5102 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5103 descriptors. However, the code omitted to check that the open succeeded,
5104 causing mysterious errors if for some reason the permissions on /dev/null
5105 got screwed. Now Exim writes a message to the main and panic logs, and
5106 bombs out if it can't open /dev/null.
5108 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5109 interact so that it is all more flexible. It is supposed to remain
5110 backwards compatible. Also added extra_local_interfaces.
5112 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5113 to bomb out with an assertion failure - to the client this appears as a
5114 connection drop. This problem occurs in the part of the code that was taken
5115 from the Samba project. Fortunately, the assertion is in a very simple
5116 function, so I have fixed this by reproducing the function inline in the
5117 one place where it is called, and arranging for authentication to fail
5118 instead of killing the process with assert().
5120 75. The SPA client code was not working when the server requested OEM rather
5121 than Unicode encoding.
5123 76. Added code to make require_files with a specific uid setting more usable in
5124 the case where statting the file as root fails - usually a non-root-mounted
5125 NFS file system. When this happens and the failure is EACCES, Exim now
5126 forks a subprocess and does the per-uid checking as the relevant uid.
5128 77. Added process_log_path.
5130 78. If log_file_path was not explicitly set, a setting of check_log_space or
5131 check_log_inodes was ignored.
5133 79. If a space check for the spool or log partitions fails, the incident is now
5134 logged. Of course, in the latter case the data may get lost...
5136 80. Added the %p formatting code to string_format() so that it can be used to
5137 print addresses in debug_print(). Adjusted all the address printing in the
5138 debugging in store.c to use %p rather than %d.
5140 81. There was a concern that a line of code in smtp_in.c could overflow a
5141 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5142 initially expressed, the concern was not well-founded, because trailing
5143 spaces are removed early. However, if the trailing spaces were followed by
5144 a NULL, they did not get removed, so the overflow was possible. Two fixes
5147 (a) I re-wrote the offending code in a cleaner fashion.
5148 (b) If an incoming SMTP command contains a NULL character, it is rejected
5151 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5152 runs initgroups(), so that if the Exim user is in any additional groups,
5153 they will be used during message reception.
5159 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob