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.
71 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
72 $data_path, and $header_path variables; fixed documentation bugs and
75 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
76 exipick to access non-standard spools, including the "frozen" queue
79 NM/01 Bugzilla 965: Support mysql stored procedures.
80 Patch from Alain Williams
82 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
84 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
85 Patch from Andreas Metzler
87 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
88 Patch from Kirill Miazine
90 NM/05 Bugzilla 671: Added umask to procmail example.
92 JJ/03 installed exipick 20100323.0, fixing doc bug
94 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
95 directory. Notification and patch from Dan Rosenberg.
97 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
99 TK/02 Improve log output when DKIM signing operation fails.
101 MH/01 Treat the transport option dkim_domain as a colon separated
102 list, not as a single string, and sign the message with each element,
103 omitting multiple occurences of the same signer.
105 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
106 Bugzilla 985, 986. Patch by Simon Arlott
108 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
109 Patch by Simon Arlott
111 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
112 MBX locking. Notification from Dan Rosenberg.
118 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
120 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
122 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
124 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
126 NM/04 Bugzilla 915: Flag broken perl installation during build.
132 TK/01 Added patch by Johannes Berg that expands the main option
133 "spamd_address" if it starts with a dollar sign.
135 TK/02 Write list of recipients to X-Envelope-Sender header when building
136 the mbox-format spool file for content scanning (suggested by Jakob
139 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
140 (http://www.dcc-servers.net/) support via dccifd. Activated by
141 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
143 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
144 by Mark Daniel Reidel <mr@df.eu>.
146 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
147 When building exim an external PCRE library is now needed -
148 PCRE is a system library on the majority of modern systems.
149 See entry on PCRE_LIBS in EDITME file.
151 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
152 conversation. Added nologin parameter to request.
153 Patch contributed by Kirill Miazine.
155 TF/01 Do not log submission mode rewrites if they do not change the address.
157 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
159 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
160 log files in place. Contributed by Roberto Lima.
162 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
164 TF/03 Bugzilla 615: When checking the local_parts router precondition
165 after a local_part_suffix or local_part_prefix option, Exim now
166 does not use the address's named list lookup cache, since this
167 contains cached lookups for the whole local part.
169 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
170 Robert Millan. Documentation is in experimental-spec.txt.
172 TF/04 Bugzilla 668: Fix parallel build (make -j).
174 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
176 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
177 Patch provided by Jan Srzednicki.
179 TF/05 Leading white space used to be stripped from $spam_report which
180 wrecked the formatting. Now it is preserved.
182 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
183 that they are available at delivery time.
185 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
187 TF/08 TLS error reporting now respects the incoming_interface and
188 incoming_port log selectors.
190 TF/09 Produce a more useful error message if an SMTP transport's hosts
191 setting expands to an empty string.
193 NM/06 Bugzilla 744: EXPN did not work under TLS.
194 Patch provided by Phil Pennock.
196 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
197 Patch provided by Richard Godbee.
199 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
200 acl_smtp_notquit, added index entry.
202 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
203 Patch provided by Eugene Bujak.
205 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
206 accept(). Patch provided by Maxim Dounin.
208 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
209 Patch provided by Phil Pennock.
211 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
213 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
214 Patch provided by Brad "anomie" Jorsch.
216 NM/14 Bugzilla 622: Added timeout setting to transport filter.
217 Patch provided by Dean Brooks.
219 TK/05 Add native DKIM support (does not depend on external libraries).
221 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
222 Patch provided by Graeme Fowler.
224 NM/16 Bugzilla 851: Documentation example syntax fix.
226 NM/17 Changed NOTICE file to remove references to embedded PCRE.
228 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
231 NM/19 Bugzilla 745: TLS version reporting.
232 Patch provided by Phil Pennock.
234 NM/20 Bugzilla 167: bool: condition support.
235 Patch provided by Phil Pennock.
237 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
238 clients. Patch provided by Phil Pennock.
240 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
241 Patch provided by Brad "anomie" Jorsch.
243 NM/23 Bugzilla 687: Fix misparses in eximstats.
244 Patch provided by Heiko Schlittermann.
246 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
247 Patch provided by Heiko Schlittermann.
249 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
250 plus update to original patch.
252 NM/26 Bugzilla 799: Documentation correction for ratelimit.
254 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
255 Patch provided by David Brownlee.
257 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
259 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
261 NM/30 Bugzilla 888: TLS documentation bugfixes.
263 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
265 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
266 Unlike the original bugzilla I have changed all shell scripts in src tree.
268 NM/33 Bugzilla 898: Transport filter timeout fix.
269 Patch by Todd Rinaldo.
271 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
272 Patch by Serge Demonchaux.
274 NM/35 Bugzilla 39: Base64 decode bug fixes.
275 Patch by Jakob Hirsch.
277 NM/36 Bugzilla 909: Correct connect() call in dcc code.
279 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
281 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
283 NM/39 Bugzilla 911: Fixed MakeLinks build script.
289 TK/01 Add preliminary DKIM support. Currently requires a forked version of
290 ALT-N's libdkim that I have put here:
291 http://duncanthrax.net/exim-experimental/
293 Note to Michael Haardt: I had to rename some vars in sieve.c. They
294 were called 'true' and it seems that C99 defines that as a reserved
295 keyword to be used with 'bool' variable types. That means you could
296 not include C99-style headers which use bools without triggering
297 build errors in sieve.c.
299 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
300 as mailq or other aliases. Changed the --help handling significantly
301 to do whats expected. exim_usage() emits usage/help information.
303 SC/01 Added the -bylocaldomain option to eximstats.
305 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
307 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
309 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
315 PH/01 Another patch from the Sieve maintainer.
317 PH/02 When an IPv6 address is converted to a string for single-key lookup
318 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
319 dots are used instead of colons so that keys in lsearch files need not
320 contain colons. This was done some time before quoting was made available
321 in lsearch files. However, iplsearch files do require colons in IPv6 keys
322 (notated using the quote facility) so as to distinguish them from IPv4
323 keys. This meant that lookups for IP addresses in host lists did not work
324 for iplsearch lookups.
326 This has been fixed by arranging for IPv6 addresses to be expressed with
327 colons if the lookup type is iplsearch. This is not incompatible, because
328 previously such lookups could never work.
330 The situation is now rather anomolous, since one *can* have colons in
331 ordinary lsearch keys. However, making the change in all cases is
332 incompatible and would probably break a number of configurations.
334 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
337 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
338 conversion specification without a maximum field width, thereby enabling
339 a rogue spamd server to cause a buffer overflow. While nobody in their
340 right mind would setup Exim to query an untrusted spamd server, an
341 attacker that gains access to a server running spamd could potentially
342 exploit this vulnerability to run arbitrary code as the Exim user.
344 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
345 $primary_hostname instead of what libspf2 thinks the hosts name is.
347 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
348 a directory entry by the name of the lookup key. Previously, if a
349 symlink pointed to a non-existing file or a file in a directory that
350 Exim lacked permissions to read, a lookup for a key matching that
351 symlink would fail. Now it is enough that a matching directory entry
352 exists, symlink or not. (Bugzilla 503.)
354 PH/03 The body_linecount and body_zerocount variables are now exported in the
357 PH/04 Added the $dnslist_matched variable.
359 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
360 This means they are set thereafter only if the connection becomes
363 PH/06 Added the client_condition to authenticators so that some can be skipped
364 by clients under certain conditions.
366 PH/07 The error message for a badly-placed control=no_multiline_responses left
367 "_responses" off the end of the name.
369 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
371 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
372 (without spaces) instead of just copying the configuration text.
374 PH/10 Added the /noupdate option to the ratelimit ACL condition.
376 PH/11 Added $max_received_linelength.
378 PH/12 Added +ignore_defer and +include_defer to host lists.
380 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
381 way in which PCRE > 7.0 is built.
383 PH/14 Implemented queue_only_load_latch.
385 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
386 MAIL command. The effect was to mangle the value on 64-bit systems.
388 PH/16 Another patch from the Sieve maintainer.
390 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
392 PH/18 If a system quota error occurred while trying to create the file for
393 a maildir delivery, the message "Mailbox is full" was not appended to the
394 bounce if the delivery eventually timed out. Change 4.67/27 below applied
395 only to a quota excession during the actual writing of the file.
397 PH/19 It seems that peer DN values may contain newlines (and other non-printing
398 characters?) which causes problems in log lines. The DN values are now
399 passed through string_printing() before being added to log lines.
401 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
402 and InterBase are left for another time.)
404 PH/21 Added message_body_newlines option.
406 PH/22 Guard against possible overflow in moan_check_errorcopy().
408 PH/23 POSIX allows open() to be a macro; guard against that.
410 PH/24 If the recipient of an error message contained an @ in the local part
411 (suitably quoted, of course), incorrect values were put in $domain and
412 $local_part during the evaluation of errors_copy.
418 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
419 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
422 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
423 issue a MAIL command.
425 PH/02 In an ACL statement such as
427 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
429 if a client was not listed at all, or was listed with a value other than
430 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
431 the condition was not true (as it should be), so access was not denied.
432 The bug was that the ! inversion was incorrectly passed on to the second
433 item. This has been fixed.
435 PH/03 Added additional dnslists conditions == and =& which are different from
436 = and & when the dns lookup returns more than one IP address.
438 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
439 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
441 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
442 FSYNC, which compiles an option called disable_fsync that allows for
443 bypassing fsync(). The documentation is heavily laced with warnings.
445 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
447 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
448 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
449 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
450 including adding "make clean"; (3) Added -fPIC when compiling the test
451 dynamically loaded module, to get rid of a warning.
453 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
454 message fails, move_frozen_messages = true and ignore_bounce_errors_after
455 = 0s. The bug is otherwise harmless.
457 PH/07 There was a bug in the dovecot authenticator such that the value of
458 $auth1 could be overwritten, and so not correctly preserved, after a
459 successful authentication. This usually meant that the value preserved by
460 the server_setid option was incorrect.
462 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
464 PH/09 Installed PCRE release 7.0.
466 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
467 run for batched SMTP input. It is now run at the start of every message
468 in the batch. While fixing this I discovered that the process information
469 (output by running exiwhat) was not always getting set for -bs and -bS
470 input. This is fixed, and it now also says "batched" for BSMTP.
472 PH/11 Added control=no_pipelining.
474 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
475 patch, slightly modified), and move the expansion of helo_data till after
476 the connection is made in the smtp transport (so it can use these
479 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
481 PH/14 Added log_selector = +pid.
483 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
485 PH/16 Add ${if forany and ${if forall.
487 PH/17 Added dsn_from option to vary the From: line in DSNs.
489 PH/18 Flush SMTP output before performing a callout, unless control =
490 no_callout_flush is set.
492 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
493 was true (the default) a successful delivery failed to delete the retry
494 item, thus causing premature timeout of the address. The bug is now
497 PH/20 Added hosts_avoid_pipelining to the smtp transport.
499 PH/21 Long custom messages for fakedefer and fakereject are now split up
500 into multiline reponses in the same way that messages for "deny" and
501 other ACL rejections are.
503 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
504 with slight modification.
506 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
507 draft, changing the syntax and factoring some duplicate code".
509 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
510 for deliveries of the second and subsequent messages over the same SMTP
513 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
514 ${reduce, with only minor "tidies".
516 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
518 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
519 expansion side effects.
521 PH/27 When a message times out after an over-quota error from an Exim-imposed
522 quota, the bounce message says "mailbox is full". This message was not
523 being given when it was a system quota that was exceeded. It now should
526 MH/03 Made $recipients available in local_scan(). local_scan() already has
527 better access to the recipient list through recipients_list[], but
528 $recipients can be useful in postmaster-provided expansion strings.
530 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
531 in the case of a MAIL command with additional options following the
532 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
533 were accidentally chopped off.
535 PH/29 SMTP synchronization checks are implemented when a command is read -
536 there is a check that no more input is waiting when there shouldn't be
537 any. However, for some commands, a delay in an ACL can mean that it is
538 some time before the response is written. In this time, more input might
539 arrive, invalidly. So now there are extra checks after an ACL has run for
540 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
541 pipelining has not been advertised.
543 PH/30 MH's patch to allow iscntrl() characters to be list separators.
545 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
546 returned in the SMTP response when smtp_return_error_details was false.
549 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
550 instead of the C I/O that was originally supplied, because problems were
553 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
554 Exim which did not show up earlier: it was assuming that a call to
555 SSL_CTX_set_info_callback() might give an error value. In fact, there is
556 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
557 was a macro that became an assignment, so it seemed to work. This has
558 changed to a proper function call with a void return, hence the compile
559 error. Exim's code has been fixed.
561 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
564 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
566 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
568 PH/37 If a message is not accepted after it has had an id assigned (e.g.
569 because it turns out to be too big or there is a timeout) there is no
570 "Completed" line in the log. When some messages of this type were
571 selected by exigrep, they were listed as "not completed". Others were
572 picked up by some special patterns. I have improved the selection
573 criteria to be more general.
575 PH/38 The host_find_failed option in the manualroute router can now be set
576 to "ignore", to completely ignore a host whose IP address cannot be
577 found. If all hosts are ignored, the behaviour is controlled by the new
578 host_all_ignored option.
580 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
581 homing or because of multiple MX records with /mx) generated more than
582 one IP address, and the following item turned out to be the local host,
583 all the secondary addresses of the first item were incorrectly removed
584 from the list, along with the local host and any following hosts (which
585 is what is supposed to happen).
587 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
588 whoever called Exim into the -H file. In the case of the daemon it was
589 behaving confusingly. When first started, it used values for whoever
590 started the daemon, but after a SIGHUP it used the Exim user (because it
591 calls itself on a restart). I have changed the code so that it now always
594 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
595 message are rejected with the same error (e.g. no authentication or bad
596 sender address), and a DATA command is nevertheless sent (as can happen
597 with PIPELINING or a stupid MUA), the error message that was given to the
598 RCPT commands is included in the rejection of the DATA command. This is
599 intended to be helpful for MUAs that show only the final error to their
602 PH/42 Another patch from the Sieve maintainer.
604 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
605 Eximstats - Fixed some broken HTML links and added missing column headers
607 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
608 columns for Rejects, Temp Rejects, Ham, and Spam rows.
610 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
612 PH/43 Yet another patch from the Sieve maintainer.
614 PH/44 I found a way to check for a TCP/IP connection going away before sending
615 the response to the final '.' that terminates a message, but only in the
616 case where the client has not sent further data following the '.'
617 (unfortunately, this is allowed). However, in many cases there won't be
618 any further data because there won't be any more messages to send. A call
619 to select() can be used: if it shows that the input is "ready", there is
620 either input waiting, or the socket has been closed. An attempt to read
621 the next input character can distinguish the two cases. Previously, Exim
622 would have sent an OK response which the client would never have see.
623 This could lead to message repetition. This fix should cure that, at
624 least in a lot of common cases.
626 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
627 advertised in response to EHLO.
633 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
634 fixed by 4.65/MH/01 (is this a record?) are fixed:
636 (i) An empty string was always treated as zero by the numeric comparison
637 operators. This behaviour has been restored.
639 (ii) It is documented that the numeric comparison operators always treat
640 their arguments as decimal numbers. This was broken in that numbers
641 starting with 0 were being interpreted as octal.
643 While fixing these problems I realized that there was another issue that
644 hadn't been noticed. Values of message_size_limit (both the global option
645 and the transport option) were treated as octal if they started with 0.
646 The documentation was vague. These values are now always treated as
647 decimal, and I will make that clear in the documentation.
653 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
654 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
657 MH/01 Don't check that the operands of numeric comparison operators are
658 integers when their expansion is in "skipping" mode (fixes bug
659 introduced by 4.64-PH/07).
661 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
662 child addresses, Exim now panics and dies. Previously, because the count
663 is held in a short int, deliveries were likely to be lost. As such a
664 large number of recipients for a single message is ridiculous
665 (performance will be very, very poor), I have chosen to impose a limit
666 rather than extend the field.
672 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
673 leftover -K file (the existence of which was triggered by #402).
674 While we were at it, introduced process PID as part of the -K
675 filename. This should rule out race conditions when creating
678 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
679 processing considerably. Previous code took too long for large mails,
680 triggering a timeout which in turn triggers #401.
682 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
683 in the DK code in transports.c. sendfile() is not really portable,
684 hence the _LINUX specificness.
686 TF/01 In the add_headers option to the mail command in an Exim filter,
687 there was a bug that Exim would claim a syntax error in any
688 header after the first one which had an odd number of characters
691 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
692 callout verification, Exim cached a "reject" for the entire domain. This
693 is correct for most verifications, but it is not correct for a recipient
694 verification with use_sender or use_postmaster set, because in that case
695 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
696 case of MAIL FROM:<> rejection from other early rejections (e.g.
697 rejection of HELO). When verifying a recipient using a non-null MAIL
698 address, the cache is ignored if it shows MAIL FROM:<> rejection.
699 Whatever the result of the callout, the value of the domain cache is
700 left unchanged (for any other kind of callout, getting as far as trying
701 RCPT means that the domain itself is ok).
703 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
706 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
707 manifest itself as EPIPE rather than ECONNECT. When tidying away a
708 session, the daemon ignores ECONNECT errors and logs others; it now
709 ignores EPIPE as well.
711 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
712 (quoted-printable decoding).
714 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
715 later the small subsequent patch to fix an introduced bug.
717 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
719 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
721 PH/08 An error is now given if message_size_limit is specified negative.
723 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
724 to be given (somewhat) arbitrary names.
726 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
729 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
730 miscellaneous code fixes
732 PH/10 Added the log_reject_target ACL modifier to specify where to log
735 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
736 hostname. This is wrong, because it relates to the incoming message (and
737 probably the interface on which it is arriving) and not to the outgoing
738 callout (which could be using a different interface). This has been
739 changed to use the value of the helo_data option from the smtp transport
740 instead - this is what is used when a message is actually being sent. If
741 there is no remote transport (possible with a router that sets up host
742 addresses), $smtp_active_hostname is used.
744 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
745 tweaks were necessary in order to get it to work (see also 21 below):
746 (a) The code assumed that strncpy() returns a negative number on buffer
747 overflow, which isn't the case. Replaced with Exim's string_format()
749 (b) There were several signed/unsigned issues. I just did the minimum
750 hacking in of casts. There is scope for a larger refactoring.
751 (c) The code used strcasecmp() which is not a standard C function.
752 Replaced with Exim's strcmpic() function.
753 (d) The code set only $1; it now sets $auth1 as well.
754 (e) A simple test gave the error "authentication client didn't specify
755 service in request". It would seem that Dovecot has changed its
756 interface. Fortunately there's a specification; I followed it and
757 changed what the client sends and it appears to be working now.
759 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
762 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
763 address A is aliased to B and C, where B exists and C does not. Without
764 -v the output is "A verified" because verification stops after a
765 successful redirection if more than one address is generated. However,
766 with -v the child addresses are also verified. Exim was outputting "A
767 failed to verify" and then showing the successful verification for C,
768 with its parentage. It now outputs "B failed to verify", showing B's
769 parentage before showing the successful verification of C.
771 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
772 look up a TXT record in a specific list after matching in a combined
775 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
776 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
777 they consult the DNS. I had assumed they would set it the way they
778 wanted; and indeed my experiments on Linux seem to show that in some
779 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
780 To be on the safe side, however, I have now made the interface to
781 host_find_byname() similar to host_find_bydns(), with an argument
782 containing the DNS resolver options. The host_find_byname() function now
783 sets these options at its start, just as host_find_bydns() does. The smtp
784 transport options dns_qualify_single and dns_search_parents are passed to
785 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
786 of host_find_byname() use the default settings of RES_DEFNAMES
787 (qualify_single) but not RES_DNSRCH (search_parents).
789 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
790 spool_read_header() do less string testing, by means of a preliminary
791 switch on the second character of optional "-foo" lines. (This is
792 overdue, caused by the large number of possibilities that now exist.
793 Originally there were few.) While I was there, I also converted the
794 str(n)cmp tests so they don't re-test the leading "-" and the first
795 character, in the hope this might squeeze out yet more improvement.
797 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
798 flag allowing group syntax was set by the header_syntax check but not
799 turned off, possible causing trouble later; (2) The flag was not being
800 set at all for the header_verify test, causing "group"-style headers to
801 be rejected. I have now set it in this case, and also caused header_
802 verify to ignore an empty address taken from a group. While doing this, I
803 came across some other cases where the code for allowing group syntax
804 while scanning a header line wasn't quite right (mostly, not resetting
805 the flag correctly in the right place). These bugs could have caused
806 trouble for malformed header lines. I hope it is now all correct.
808 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
809 with the "reply" argument non-NULL. The code, however (which originally
810 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
811 but it didn't always do it. This confused somebody who was copying the
812 code for some other use. I have removed all the tests.
814 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
815 feature that was used to support insecure browsers during the U.S. crypto
816 embargo. It requires special client support, and Exim is probably the
817 only MTA that supported it -- and would never use it because real RSA is
818 always available. This code has been removed, because it had the bad
819 effect of slowing Exim down by computing (never used) parameters for the
820 RSA_EXPORT functionality.
822 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
823 authenticator to fail if there's a tab character in the incoming data
824 (there should never be unless someone is messing about, as it's supposed
825 to be base64-encoded). Also added, on Timo's advice, the "secured" option
826 if the connection is using TLS or if the remote IP is the same as the
827 local IP, and the "valid-client-cert option" if a client certificate has
830 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
831 authenticators. This can be used for authorization after authentication
832 succeeds. (In the case of plaintext, it servers for both authentication
835 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
836 if any retry times were supplied.
838 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
839 connection, where there is no client IP address to check. In this
840 situation, the verify now always succeeds.
842 PH/25 Applied John Jetmore's -Mset patch.
844 PH/26 Added -bem to be like -Mset, but loading a message from a file.
846 PH/27 In a string expansion for a processed (not raw) header when multiple
847 headers of the same name were present, leading whitespace was being
848 removed from all of them, but trailing whitespace was being removed only
849 from the last one. Now trailing whitespace is removed from each header
850 before concatenation. Completely empty headers in a concatenation (as
853 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
854 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
856 PH/29 [Removed. This was a change that I later backed out, and forgot to
857 correct the ChangeLog entry (that I had efficiently created) before
858 committing the later change.]
860 PH/30 Exim was sometimes attempting to deliver messages that had suffered
861 address errors (4xx response to RCPT) over the same connection as other
862 messages routed to the same hosts. Such deliveries are always "forced",
863 so retry times are not inspected. This resulted in far too many retries
864 for the affected addresses. The effect occurred only when there were more
865 hosts than the hosts_max_try setting in the smtp transport when it had
866 the 4xx errors. Those hosts that it had tried were not added to the list
867 of hosts for which the message was waiting, so if all were tried, there
868 was no problem. Two fixes have been applied:
870 (i) If there are any address or message errors in an SMTP delivery, none
871 of the hosts (tried or untried) are now added to the list of hosts
872 for which the message is waiting, so the message should not be a
873 candidate for sending over the same connection that was used for a
874 successful delivery of some other message. This seems entirely
875 reasonable: after all the message is NOT "waiting for some host".
876 This is so "obvious" that I'm not sure why it wasn't done
877 previously. Hope I haven't missed anything, but it can't do any
878 harm, as the worst effect is to miss an optimization.
880 (ii) If, despite (i), such a delivery is accidentally attempted, the
881 routing retry time is respected, so at least it doesn't keep
882 hammering the server.
884 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
885 in ${readsocket because some servers need this prod.
887 PH/32 Added some extra debug output when updating a wait-xxx database.
889 PH/33 The hint "could be header name not terminated by colon", which has been
890 given for certain expansion errors for a long time, was not being given
891 for the ${if def:h_colon_omitted{... case.
893 PH/34 The spec says: "With one important exception, whenever a domain list is
894 being scanned, $domain contains the subject domain." There was at least
895 one case where this was not true.
897 PH/35 The error "getsockname() failed: connection reset by peer" was being
898 written to the panic log as well as the main log, but it isn't really
899 panic-worthy as it just means the connection died rather early on. I have
900 removed the panic log writing for the ECONNRESET error when getsockname()
903 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
904 runs only) independently of the message's sender address. This meant
905 that, if the 4xx error was in fact related to the sender, a different
906 message to the same recipient with a different sender could confuse
907 things. In particualar, this can happen when sending to a greylisting
908 server, but other circumstances could also provoke similar problems.
909 I have changed the default so that the retry time for these errors is now
910 based a combination of the sender and recipient addresses. This change
911 can be overridden by setting address_retry_include_sender=false in the
914 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
915 remote server are returned as part of bounce messages. This was not
916 happening for LMTP over a pipe (the lmtp transport), but now it is the
917 same for both kinds of LMTP.
919 PH/38 Despite being documented as not happening, Exim was rewriting addresses
920 in header lines that were in fact CNAMEs. This is no longer the case.
922 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
923 and queue runs started by the daemon processed all messages. This has
924 been fixed so that -R and -S can now usefully be given with -q<time>.
926 PH/40 Import PCRE release 6.7 (fixes some bugs).
928 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
930 PH/42 Give an error if -q is specified more than once.
932 PH/43 Renamed the variables $interface_address and $interface_port as
933 $received_ip_address and $received_port, to make it clear that these
934 values apply to message reception, and not to the outgoing interface when
935 a message is delivered. (The old names remain recognized, of course.)
937 PH/44 There was no timeout on the connect() call when using a Unix domain
938 socket in the ${readsocket expansion. There now is.
940 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
941 be meaningful with "accept".
943 SC/01 Eximstats V1.43
944 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
946 SC/02 Eximstats V1.44
947 Use a glob alias rather than an array ref in the generated
948 parser. This improves both readability and performance.
950 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
951 Collect SpamAssassin and rejection statistics.
952 Don't display local sender or destination tables unless
953 there is data to show.
954 Added average volumes into the top table text output.
956 SC/04 Eximstats V1.46
957 Collect data on the number of addresses (recipients)
958 as well as the number of messages.
960 SC/05 Eximstats V1.47
961 Added 'Message too big' to the list of mail rejection
962 reasons (thanks to Marco Gaiarin).
964 SC/06 Eximstats V1.48
965 Mainlog lines which have GMT offsets and are too short to
966 have a flag are now skipped.
968 SC/07 Eximstats V1.49 (Alain Williams)
969 Added the -emptyok flag.
971 SC/08 Eximstats V1.50
972 Fixes for obtaining the IP address from reject messages.
974 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
975 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
976 whitesspace changes from 4.64-PH/27
978 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
981 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
982 are found, allow negative numbers in numeric criteria)
984 JJ/06 exipick.20061117.2, added new $message_body_missing variable
986 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
987 to match changes made in 4.64-PH/43
989 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
991 PH/47 Put in an explicit test for a DNS lookup of an address record where the
992 "domain" is actually an IP address, and force a failure. This locks out
993 those revolvers/nameservers that support "A-for-A" lookups, in
994 contravention of the specifications.
996 PH/48 When a host name was looked up from an IP address, and the subsequent
997 forward lookup of the name timed out, the host name was left in
998 $sender_host_name, contrary to the specification.
1000 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1001 restricted to single-key lookups, Exim was not diagnosing an error if
1002 * or *@ was used with a query-style lookup.
1004 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1006 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1007 long ago, but noone interested enough thought of it. Let's just say that
1008 the "1.1" means that there are some new functions that weren't there at
1009 some point in the past.
1011 PH/51 Error processing for expansion failure of helo_data from an smtp
1012 transport during callout processing was broken.
1014 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1015 tested/used via the -bh/-bhc/-bs options.
1017 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1018 bug, fixed in subsequent PCRE releases).
1020 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1021 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1023 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1029 SC/01 Use a glob alias rather than an array ref in eximstats generated
1030 parser. This improves both readability and performance.
1032 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1033 Don't display local sender or destination tables in eximstats unless
1034 there is data to show.
1035 Added average volumes into the eximstats top table text output.
1037 SC/03 Collect data on the number of addresses (recipients) as well
1038 as the number of messages in eximstats.
1040 TF/01 Correct an error in the documentation for the redirect router. Exim
1041 does (usually) call initgroups() when daemonizing.
1043 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1044 with consistent privilege compared to when running as a daemon.
1046 TF/03 Note in the spec that $authenticated_id is not set for local
1047 submissions from trusted users.
1049 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1050 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1052 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1053 by adding some example configuration directives to the default
1054 configuration file. A little bit of work is required to uncomment the
1055 directives and define how usernames and passwords are checked, but
1056 there is now a framework to start from.
1058 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1059 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1060 without this. I don't know how relevant this is to other LDAP libraries.
1062 PH/02 Add the verb name to the "unknown ACL verb" error.
1064 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1066 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1068 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1069 directory not expanded when it should be if an expanded home directory
1070 was set for the address (which is overridden by the transport).
1072 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1075 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1076 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1077 because it is too late at that time, and has no effect.
1079 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1080 security issue with \' (bugzilla #107). I could not use the
1081 PQescapeStringConn() function, because it needs a PGconn value as one of
1084 PH/08 When testing addresses using -bt, indicate those final addresses that
1085 are duplicates that would not cause an additional delivery. At least one
1086 person was confused, thinking that -bt output corresponded to deliveries.
1087 (Suppressing duplicates isn't a good idea as you lose the information
1088 about possibly different redirections that led to the duplicates.)
1090 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1091 systems where poll() doesn't work, in particular OS X.
1093 PH/10 Added more information to debugging output for retry time not reached.
1095 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1096 operations in malware.c.
1098 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1101 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1102 syslog_duplication set false, log lines that would normally be written
1103 both the the main log and to the reject log were not written to syslog at
1106 PH/14 In the default configuration, change the use of "message" in ACL warn
1107 statements to "add_header".
1109 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1110 not followed by a command (e.g. "seen endif").
1112 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1113 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1116 PH/17 Added extra conditions to the default value of delay_warning_condition
1120 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1121 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1122 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1125 The Auto-Submitted: and various List- headers are standardised, whereas I
1126 don't think Precedence: ever was.
1128 PH/18 Refactored debugging code in route_finduser() to show more information,
1129 in particular, the error code if getpwnam() issues one.
1131 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1132 This is apparently needed in addition to the PH/07 change above to avoid
1133 any possible encoding problems.
1135 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1136 but not after initializing Perl.
1138 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1139 output them only if debugging. By default they are written stderr,
1140 apparently, which is not desirable.
1142 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1145 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1148 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1150 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1151 authenticated or an ident call has been made. Suppress the default
1152 values for $authenticated_id and $authenticated_sender (but permit -oMai
1153 and -oMas) when testing with -bh.
1155 PH/24 Re-jigged the order of the tests in the default configuration so that the
1156 tests for valid domains and recipients precede the DNS black list and CSA
1157 tests, on the grounds that those ones are more expensive.
1159 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1160 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1161 This bug exists in every version of Exim that I still have, right back to
1164 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1165 However, an attempt to turn on case-sensitivity in a regex key by
1166 including (?-i) didn't work because the subject string was already
1167 lowercased, and the effects were non-intuitive. It turns out that a
1168 one-line patch can be used to allow (?-i) to work as expected.
1174 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1175 other effects) broke the use of negated acl sub-conditions.
1177 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1180 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1181 "Deny" causes Exim to reject the incoming connection with a 554 error.
1182 Unfortunately, if there is a major crisis, such as a disk failure,
1183 tcp-wrappers gives "deny", whereas what one would like would be some
1184 kind of temporary error. A kludge has been added to help with this.
1185 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1186 554 error is used if errno is still zero or contains ENOENT (which occurs
1187 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1190 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1192 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1193 errors. Otherwise a message that provokes a temporary error (when other
1194 messages do not) can cause a whole host to time out.
1196 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1197 addresses were routed directly to files or pipes from a redirect router.
1198 File deliveries just didn't batch; pipe deliveries might have suffered
1201 PH/06 A failure to get a lock for a hints database would erroneously always say
1202 "Failed to get write lock", even when it was really a read lock.
1204 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1205 of 0600. This has been changed to use the value of the lockfile_mode
1206 option (which defaults to 0600).
1208 PH/08 Applied small patch from the Sieve maintainer.
1210 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1211 folder from quota calculations, a direct delivery into this folder messed
1212 up the contents of the maildirsize file. This was because the regex was
1213 used only to exclude .Trash (or whatever) when the size of the mailbox
1214 was calculated. There was no check that a delivery was happening into an
1215 excluded directory. This bug has been fixed by ignoring all quota
1216 processing for deliveries into excluded directories.
1218 PH/10 Added the maildirfolder_create_regex option to appendfile.
1224 PH/01 The code for finding all the local interface addresses on a FreeBSD
1225 system running IPv6 was broken. This may well have applied to all BSD
1226 systems, as well as to others that have similar system calls. The broken
1227 code found IPv4 interfaces correctly, but gave incorrect values for the
1228 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1229 that it would not match correctly against @[] and not recognize the IPv6
1232 PH/02 The ipliteral router was not recognizing addresses of the form user@
1233 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1235 PH/03 Added disable_ipv6.
1237 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1238 decoded form, because it is most often used to construct To: headers
1239 lines in autoreplies, and the decoded form may well be syntactically
1240 invalid. However, $reply_address has leading white space removed, and all
1241 newlines turned into spaces so that the autoreply transport does not
1244 PH/05 If group was specified without a user on a router, and no group or user
1245 was specified on a transport, the group from the router was ignored.
1247 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1248 for visible compile-time settings that can be used to change these
1249 numbers, for those that want even more. Backwards compatibility with old
1250 spool files has been maintained. However, going back to a previous Exim
1251 release will lost any variables that are in spool files.
1253 PH/07 Two small changes when running in the test harness: increase delay when
1254 passing a TCP/IP connection to a new process, in case the original
1255 process has to generate a bounce, and remove special handling of
1256 127.0.0.2 (sic), which is no longer necessary.
1258 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1259 be the same on different OS.
1261 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1264 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1265 whether --show-vars was specified or not
1267 JJ/02 exipick: Added support for new ACL variable spool format introduced
1270 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1271 syntactically invalid From: or Reply-to: line, and a filter used this to
1272 generate an autoreply, and therefore failed to obtain an address for the
1273 autoreply, Exim could try to deliver to a non-existent relative file
1274 name, causing unrelated and misleading errors. What now happens is that
1275 it logs this as a hard delivery error, but does not attempt to create a
1278 PH/11 The exinext utility has a -C option for testing purposes, but although
1279 the given file was scanned by exinext itself; it wasn't being passed on
1280 when Exim was called.
1282 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1283 an end-of-file indication when reading a command response.
1285 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1286 compiled. In many other places in Exim, IPv6 addresses are always
1287 recognized, so I have changed this. It also means that IPv4 domain
1288 literals of the form [IPV4:n.n.n.n] are now always recognized.
1290 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1291 used if the router is not running as root, for example, when verifying at
1292 ACL time, or when using -bh. The debugging output from this situation was
1293 non-existent - all you got was a failure to exec. I have made two
1296 (a) Failures to set uid/gid, the current directory, or a process leader
1297 in a subprocess such as that created by queryprogram now generate
1298 suitable debugging ouput when -d is set.
1300 (b) The queryprogram router detects when it is not running as root,
1301 outputs suitable debugging information if -d is set, and then runs
1302 the subprocess without attempting to change uid/gid.
1304 PH/15 Minor change to Makefile for building test_host (undocumented testing
1307 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1308 additional section of a DNS packet that returns MX or SRV records.
1309 Instead, it always explicitly searches for A/AAAA records. This avoids
1310 major problems that occur when a DNS server includes only records of one
1311 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1312 fixed another bug: if SRV records were looked up and the corresponding
1313 address records were *not* found in the additional section, the port
1314 values from the SRV records were lost.
1316 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1317 using the correct key (the original address) when searching the retry
1318 rules in order to find which one to use for generating the retry hint.
1320 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1321 adding the default one. Similarly, if it contains a Reply-To: header, the
1322 errors_reply_to option, if set, is not used.
1324 PH/19 When calculating a retry time, Exim used to measure the "time since
1325 failure" by looking at the "first failed" field in the retry record. Now
1326 it does not use this if it is later than than the arrival time of the
1327 message. Instead it uses the arrival time. This makes for better
1328 behaviour in cases where some deliveries succeed, thus re-setting the
1329 "first failed" field. An example is a quota failure for a huge message
1330 when small messages continue to be delivered. Without this change, the
1331 "time since failure" will always be short, possible causing more frequent
1332 delivery attempts for the huge message than are intended.
1333 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1335 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1336 $1, $2, $3) because the numerical variables can be reset during some
1337 expansion items (e.g. "match"), thereby losing the authentication data.
1339 PH/21 Make -bV show the size of off_t variables so that the test suite can
1340 decide whether to run tests for quotas > 2G.
1342 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1343 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1344 filecount value is greater than 2G or if a quota value is greater than 2G
1345 on a system where the size of off_t is not greater than 4, a panic error
1348 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1349 never match. The debug and -bh output now contains an explicit error
1350 message indicating a malformed IPv4 address or mask.
1352 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1353 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1354 PH/23 above applies.
1356 PH/25 Do not write to syslog when running in the test harness. The only
1357 occasion when this arises is a failure to open the main or panic logs
1358 (for which there is an explicit test).
1360 PH/26 Added the /no_tell option to "control=freeze".
1362 PH/27 If a host name lookup failed very early in a connection, for example, if
1363 the IP address matched host_lookup and the reverse lookup yielded a name
1364 that did not have a forward lookup, an error message of the form "no IP
1365 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1366 could be logged. Now it outputs the IP address instead of "NULL".
1368 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1369 allows the sender and the authenticated sender to be set when
1370 submitting a message from within Exim. Since child_open_exim() is
1371 documented for local_scan(), the new function should be too.
1373 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1374 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1375 results in an empty string is now treated as unset.
1377 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1379 PH/31 Added sender_verify_fail logging option.
1381 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1382 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1383 tidied the source and removed it altogether.
1385 PH/33 When a queue run was abandoned because the load average was too high, a
1386 log line was always written; now it is written only if the queue_run log
1387 selector is set. In addition, the log line for abandonment now contains
1388 information about the queue run such as the pid. This is always present
1389 in "start" and "stop" lines but was omitted from the "abandon" line.
1391 PH/34 Omit spaces between a header name and the colon in the error message that
1392 is given when verify = headers_syntax fails (if there are lots of them,
1393 the message gets confusing).
1395 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1396 names, as there are now some PTR records that contain slashes. This check
1397 is only to protect against broken name servers that fall over on strange
1398 characters, so the fact that it applies to all lookups doesn't matter.
1400 PH/36 Now that the new test suite is complete, we can remove some of the
1401 special code in Exim that was needed for the old test suite. For example,
1402 sorting DNS records because real resolvers return them in an arbitrary
1403 order. The new test suite's fake resolver always returns records in the
1406 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1407 bounces) except when queue_only is set, to avoid logging races between
1408 the different processes.
1410 PH/38 Panic-die if .include specifies a non-absolute path.
1412 PH/39 A tweak to the "H" retry rule from its user.
1414 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1415 a label. They prevented compilation on older perls.
1417 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1418 a warning to be raised on newish perls.
1420 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1421 on queue. Changes to match documented behaviour of showing count of
1422 messages matching specified criteria.
1424 PH/40 Changed the default ident timeout from 30s to 5s.
1426 PH/41 Added support for the use of login_cap features, on those BSD systems
1427 that have them, for controlling the resources used by pipe deliveries.
1429 PH/42 The content-scanning code uses fopen() to create files in which to put
1430 message data. Previously it was not paying any attention to the mode of
1431 the files. Exim runs with umask(0) because the rest of the code creates
1432 files with open(), and sets the required mode explicitly. Thus, these
1433 files were ending up world-writeable. This was not a big issue, because,
1434 being within the spool directory, they were not world-accessible. I have
1435 created a function called modefopen, which takes an additional mode
1436 argument. It sets umask(777), creates the file, chmods it to the required
1437 mode, then resets the umask. All the relevant calls to fopen() in the
1438 content scanning code have been changed to use this function.
1440 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1441 to 24 hours. This avoids potential overflow problems when processing G
1442 and H retry rules. I suspect nobody ever tinkers with this value.
1444 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1446 PH/45 When the plaintext authenticator is running as a client, the server's
1447 challenges are checked to ensure they are valid base64 strings. By
1448 default, the authentication attempt is cancelled if an invalid string is
1449 received. Setting client_ignore_invalid_base64 true ignores these errors.
1450 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1451 they are received. Thus, the responses can be made to depend on the
1452 challenges. If an invalid string is ignored, an empty string is placed in
1455 PH/46 Messages that are created by the autoreply transport now contains a
1456 References: header, in accordance with RFCs 2822 and 3834.
1458 PH/47 Added authenticated_sender_force to the smtp transport.
1460 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1462 PH/49 Installed latest patch from the Sieve maintainer.
1464 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1465 was also set, the appendfile transport was unnecessarily scanning a
1466 directory of message files (e.g. for maildir delivery) to find the count
1467 of files (along with the size), even though it did not need this
1468 information. It now does the scan only if it needs to find either the
1469 size of the count of files.
1471 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1473 PH/52 Two bugs concerned with error handling when the smtp transport is
1476 (i) Exim was not creating retry information for temporary errors given
1477 for individual recipients after the DATA command when the smtp transport
1478 was used in LMTP mode. This meant that they could be retried too
1479 frequently, and not timed out correctly.
1481 (ii) Exim was setting the flag that allows error details to be returned
1482 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1483 recipients that were returned after the DATA command.
1485 PH/53 This is related to PH/52, but is more general: for any failing address,
1486 when detailed error information was permitted to be returned to the
1487 sender, but the error was temporary, then after the final timeout, only
1488 "retry timeout exceeded" was returned. Now it returns the full error as
1489 well as "retry timeout exceeded".
1491 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1492 do this, and (what is worse) MTAs that accept it.
1494 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1495 will now be deprecated.
1497 PH/56 New os.c-cygwin from the Cygwin maintainer.
1499 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1500 formats (previously only available in exim formats via -bpr, -bpru,
1501 and -bpra. Now also available in native and exiqgrep formats)
1503 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1504 with very large, slow to parse queues
1506 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1508 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1510 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1511 responses to authentication challenges, though it was showing the
1512 challenges; (ii) I've removed the CR characters from the debug output for
1515 PH/58 Allow for the insertion of a newline as well as a space when a string
1516 is turned into more than one encoded-word during RFC 2047 encoding. The
1517 Sieve code now uses this.
1519 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1520 data_4xx, lost_connection, tls_required.
1522 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1523 message was being sent as an SMTP response.
1525 PH/61 Add -l and -k options to exicyclog.
1527 PH/62 When verifying, if an address was redirected to one new address, so that
1528 verification continued, and the new address failed or deferred after
1529 having set something in $address_data, the value of $address_data was not
1530 passed back to the ACL. This was different to the case when no
1531 redirection occurred. The value is now passed back in both cases.
1533 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1534 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1535 use login_cap.h, so on its own it isn't the distinguishing feature. The
1536 new name refers directly to the setclassresources() function.
1538 PH/65 Added configuration files for NetBSD3.
1540 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1542 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1543 is preferred over IPv4.
1545 PH/68 The bounce_return_message and bounce_return_body options were not being
1546 honoured for bounces generated during the reception of non-SMTP messages.
1547 In particular, this applied to messages rejected by the ACL. This bug has
1548 been fixed. However, if bounce_return_message is true and bounce_return_
1549 body is false, the headers that are returned for a non-SMTP message
1550 include only those that have been read before the error was detected.
1551 (In the case of an ACL rejection, they have all been read.)
1553 PH/69 The HTML version of the specification is now built in a directory called
1554 spec_html instead of spec.html, because the latter looks like a path with
1555 a MIME-type, and this confuses some software.
1557 PH/70 Catch two compiler warnings in sieve.c.
1559 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1560 function verify_get_ident() calls ip_connect() to connect a socket, but
1561 if the "connect()" function timed out, ip_connect() used to close the
1562 socket. However, verify_get_ident() also closes the socket later, and in
1563 between Exim writes to the log, which may get opened at this point. When
1564 the socket was closed in ip_connect(), the log could get the same file
1565 descriptor number as the socket. This naturally causes chaos. The fix is
1566 not to close the socket in ip_connect(); the socket should be closed by
1567 the function that creates it. There was only one place in the code where
1568 this was missing, in the iplookup router, which I don't think anybody now
1569 uses, but I've fixed it anyway.
1571 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1572 well as to direct DNS lookups. Otherwise the handling of names in host
1573 lists is inconsistent and therefore confusing.
1579 PH/01 Two changes to the default runtime configuration:
1581 (1) Move the checks for relay_from_hosts and authenticated clients from
1582 after to before the (commented out) DNS black list checks.
1584 (2) Add control=submission to the relay_from_hosts and authenticated
1585 clients checks, on the grounds that messages accepted by these
1586 statements are most likely to be submissions.
1588 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1590 (1) Generate an error if the third argument for the ${prvs expansion is
1593 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1596 (3) Reset the variables that are obtained from the first argument of
1597 ${prvscheck and used in the second argument before leaving the code,
1598 because their memory is reclaimed, so using them afterwards may do
1601 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1602 one (it's much easier than Tom thought :-).
1604 (5) Because of (4), we can now allow for the use of $prvscheck_result
1605 inside the third argument.
1607 PH/03 For some reason, the default setting of PATH when running a command from
1608 a pipe transport was just "/usr/bin". I have changed it to
1611 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1612 anything to be listed in the output from -bV.
1614 PH/05 When a filter generated an autoreply, the entire To: header line was
1615 quoted in the delivery log line, like this:
1617 => >A.N.Other <ano@some.domain> <original@ddress> ...
1619 This has been changed so that it extracts the operative address. There
1620 may be more than one such address. If so, they are comma-separated, like
1623 => >ano@some.domain,ona@other.domain <original@ddress> ...
1625 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1626 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1627 not being looked up in the rDNS to get a host name, Exim was showing the
1628 IP address twice in Received: lines, even though the IP addresses were
1629 identical. For example:
1631 Received: from [1.2.3.4] (helo=[1.2.3.4])
1633 However, if the real host name was known, it was omitting the HELO data
1634 if it matched the actual IP address. This has been tidied up so that it
1635 doesn't show the same IP address twice.
1637 PH/07 When both +timestamp and +memory debugging was on, the value given by
1638 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1639 called by the debug printing, thereby overwriting the timestamp buffer.
1640 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1642 PH/08 When the original message was included in an autoreply transport, it
1643 always said "this is a copy of the message, including all the headers",
1644 even if body_only or headers_only was set. It now gives an appropriate
1647 PH/09 Applied a patch from the Sieve maintainer which:
1649 o fixes some comments
1650 o adds the (disabled) notify extension core
1651 o adds some debug output for the result of if/elsif tests
1652 o points to the current vacation draft in the documentation
1653 and documents the missing references header update
1657 o fixes a bug in processing the envelope test (when testing
1658 multiple envelope elements, the last element determinted the
1661 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1662 Electronic Mail") by including:
1664 Auto-submitted: auto-generated
1666 in the messages that it generates (bounce messages and others, such as
1667 warnings). In the case of bounce messages for non-SMTP mesages, there was
1668 also a typo: it was using "Auto_submitted" (underscore instead of
1669 hyphen). Since every message generated by Exim is necessarily in response
1670 to another message, thes have all been changed to:
1672 Auto-Submitted: auto-replied
1674 in accordance with these statements in the RFC:
1676 The auto-replied keyword:
1678 - SHOULD be used on messages sent in direct response to another
1679 message by an automatic process,
1681 - MUST NOT be used on manually-generated messages,
1683 - MAY be used on Delivery Status Notifications (DSNs) and Message
1684 Disposition Notifications (MDNs),
1686 - MUST NOT be used on messages generated by automatic or periodic
1687 processes, except for messages which are automatic responses to
1690 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1691 to the default Received: header definition.
1693 PH/12 Added log selector acl_warn_skipped (default on).
1695 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1696 variables because (a) they are in the wrong storage pool and (b) even if
1697 they were copied, it wouldn't work properly because of the caching.
1699 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1700 checking when decoding. Apparently there are clients that generate
1701 overlong encoded strings. Why am I not surprised?
1703 PH/15 If the first argument of "${if match_address" was not empty, but did not
1704 contain an "@" character, Exim crashed. Now it writes a panic log message
1705 and treats the condition as false.
1707 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1709 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1710 extension "envelope-auth". The code is finished and in agreement with
1711 other implementations, but there is no documentation so far and in fact,
1712 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1713 not changing the active code.
1715 Print executed "if" and "elsif" statements when debugging is used. This
1716 helps a great deal to understand what a filter does.
1718 Document more things not specified clearly in RFC3028. I had all this
1719 sorted out, when out of a sudden new issues came to my mind. Oops."
1721 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1724 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1725 canonical form (as documented). However, after a host name lookup from
1726 the IP address, check_host() was doing a simple string comparison with
1727 addresses acquired from the DNS when checking that the found name did
1728 have the original IP as one of its addresses. Since any found IPv6
1729 addresses are likely to be in abbreviated form, the comparison could
1730 fail. Luckily, there already exists a function for doing the comparison
1731 by converting both addresses to binary, so now that is used instead of
1732 the text comparison.
1734 PH/20 There was another similar case to PH/19, when a complete host name was
1735 given in a host list; looking up its IP address could give an abbreviated
1736 form, whereas the current host's name might or might not be abbreviated.
1737 The same fix has been applied.
1743 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1744 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1747 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1748 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1750 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1752 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1753 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1754 Also, for "Auto-Submitted", treat anything other than "no" as
1755 non-personal, in accordance with RFC 3834. (Previously it treated
1756 anything starting "auto-" as non-personal.)
1758 TF/01 The control=submission/name=... option had a problem with syntax
1759 errors if the name included a slash character. The /name= option
1760 now slurps the rest of the string, so it can include any characters
1761 but it must come last in the list of options (after /sender_retain
1764 PH/05 Some modifications to the interface to the fake nameserver for the new
1772 TK/01 Added the "success_on_redirect" address verification option. See
1773 NewStuff for rationale and an example.
1775 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1777 PH/02 Patch to exigrep to allow it to work on syslog lines.
1779 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1780 fread() to skip over the body file's header line, because in Cygwin the
1781 header line is locked and is inaccessible.
1783 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1784 co-exist for some time) to make it clear that it is the Exim ID that is
1785 referenced, not the Message-ID: header line.
1787 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1788 string_format() function, because snprintf() does not exist on all
1791 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1792 ${stat: expansion item.
1794 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1795 protocol synchronization error", to keep the pedants happy.
1797 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1798 well as for IRIX systems, when gcc is being used. See the host.c source
1801 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1803 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1806 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1807 transport and to the smtp transport in LMTP mode.
1809 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1811 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1812 run a filter in a subprocess. This could lead to confusion in subsequent
1813 lookups in the parent process. There should also be a search_tidyup() at
1814 the end of the subprocess.
1816 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1817 only if the host matched helo_try_verify_hosts, which caused the
1818 verification to occur when the EHLO/HELO command was issued. The ACL just
1819 tested the remembered result. Now, if a previous verification attempt has
1820 not happened, "verify = helo" does it there and then.
1822 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1824 TK/03 Fix log output including CR from clamd.
1826 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1827 exist provoked a memory error which could cause a segfault.
1829 PH/15 Installed PCRE 6.2
1831 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1833 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1834 of the problem. Specifically, suggested +O2 rather than +O1 for the
1837 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1839 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1840 record, the port was not correctly logged when the outgoing_port log
1841 selector was set (it logged the transort's default port).
1843 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1844 fallback_hosts, and "hosts" in the smtp transport.
1846 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1847 host errors such as "Connection refused".
1849 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1850 authentication with radiusclient 0.4.9:
1852 - Error returned from rc_read_config was caught wrongly
1853 - Username/password not passed on to radius server due to wrong length.
1855 The presumption is that some radiusclient API changes for 4.51/PH/17
1856 were not taken care of correctly. The code is still untested by me (my
1857 Linux distribution still has 0.3.2 of radiusclient), but it was
1858 contributed by a Radius user.
1860 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1861 expanding the "port" option of the smtp transport.
1863 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1864 while reading a MIME header. Thanks to Tom Hughes for a patch.
1866 PH/24 Include config.h inside local_scan.h so that configuration settings are
1869 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1870 that in an ACL for RCPT (for example), you can examine exactly what was
1873 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1874 commands, but it was not correctly comparing the address with the actual
1875 client host address. Thus, it would show the EHLO address in Received:
1876 header lines when this was not necessary.
1878 PH/27 Added the % operator to ${eval:}.
1880 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1881 it should be ignoring failures (because with -C, for example, it has lost
1882 privilege). It wasn't ignoring creation failures other than "already
1885 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1888 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1889 because an input line was too long, either on its own, or by virtue of
1890 too many continuations, the temporary file was not being removed, and the
1891 return code was incorrect.
1893 PH/31 Missing "BOOL" in function definition in filtertest.c.
1895 PH/32 Applied Sieve patches from the maintainer.
1897 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1899 PH/33 Added "verify = not_blind".
1901 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1902 Local/Makefile (with some defaults set). These are used in built scripts
1903 such as exicyclog, but they have never been used in the exim_install
1904 script (though there are many overriding facilities there). I have
1905 arranged that the exim_install script now takes note of these two
1908 PH/35 Installed configuration files for Dragonfly.
1910 PH/36 When a locally submitted message by a trusted user did not contain a
1911 From: header, and the sender address was obtained from -f or from an SMTP
1912 MAIL command, and the trusted user did not use -F to supply a sender
1913 name, $originator_name was incorrectly used when constructing a From:
1914 header. Furthermore, $originator_name was used for submission mode
1915 messages from external hosts without From: headers in a similar way,
1916 which is clearly wrong.
1918 PH/37 Added control=suppress_local_fixups.
1920 PH/38 When log_selector = +received_sender was set, and the addition of the
1921 sender made the log line's construction buffer exactly full, or one byte
1922 less than full, an overflow happened when the terminating "\n" was
1925 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1926 when the result of a list match is failure because a DNS lookup failed.
1928 PH/40 RM_COMMAND is now used in the building process.
1930 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1931 the "build-* directories that it finds.
1933 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1934 address in a domain literal was a prefix of an interface address.
1936 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1937 when verifying a sender address, unless rewrite_headers is false.
1939 PH/44 Wrote a long comment about why errors_to addresses are verified as
1940 recipients, not senders.
1942 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1943 the ratelimit ACL was added.
1945 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1947 PH/46 Added extra information about PostgreSQL errors to the error string.
1949 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1950 avoiding the need to install special zones in a real server. This is
1951 backwards compatible; if it can't find the fake resolver, it drops back.
1952 Thus, both old and new test suites can be run.
1954 TF/02 Added util/ratelimit.pl
1956 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1957 clock is set back in time.
1959 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1960 Candler <B.Candler@pobox.com>.
1962 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1963 OK for addresses that are the result of redirections.
1965 PH/48 A number of further additions for the benefit of the new test suite,
1966 including a fake gethostbyname() that interfaces to the fake DNS resolver
1969 TF/06 The fix for widen_domains has also been applied to qualify_single and
1970 search_parents which are the other dnslookup options that can cause
1973 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1976 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1978 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1979 service" if it encountered an MX record with an empty target hostname.
1980 The message is now "an MX or SRV record indicated no SMTP service".
1982 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1983 if the DNS of the sending site is misconfigured. This is quite a
1984 common situation. This change restores the behaviour of treating a
1985 helo verification defer as a failure.
1987 PH/51 If self=fail was set on a router, the bounce message did not include the
1988 actual error message.
1994 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1996 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1997 command itself ended in error, the underlying message about the transport
1998 filter timeout was being overwritten with the pipe command error. Now the
1999 underlying error message should be appended to the second error message.
2001 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2003 PH/02 Reduce the amount of output that "make" produces by default. Full output
2004 can still be requested.
2006 PH/03 The warning log line about a condition test deferring for a "warn" verb
2007 was being output only once per connection, rather than after each
2008 occurrence (because it was using the same function as for successful
2009 "warn" verbs). This seems wrong, so I have changed it.
2011 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2012 it should not have, which might have caused a crash in the right
2013 circumstances, but probably never did.
2015 PH/04 Installed a modified version of Tony Finch's patch to make submission
2016 mode fix the return path as well as the Sender: header line, and to
2017 add a /name= option so that you can make the user's friendly name appear
2020 TF/03 Added the control = fakedefer ACL modifier.
2022 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2023 Mark Lowes for thorough testing.
2025 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2027 TK/03 Merged latest SRS patch from Miles Wilton.
2029 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2030 with the definition in sysexits.h (which is #included earlier).
2031 Fortunately, Exim does not actually use EX_OK. The code used to try to
2032 preserve the sysexits.h value, by assumimg that macro definitions were
2033 scanned for macro replacements. I have been disabused of this notion,
2034 so now the code just undefines EX_OK before #including unistd.h.
2036 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2037 in the smtp transport. When a block could not be written in a single
2038 write() function, the timeout was being re-applied to each part-write.
2039 This seems wrong - if the receiver was accepting one byte at a time it
2040 would take for ever. The timeout is now adjusted when this happens. It
2041 doesn't have to be particularly precise.
2043 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2044 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2046 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2047 without a domain if the check to <postmaster@domain> fails.
2049 SC/01 Eximstats: added -xls and the ability to specify output files
2050 (patch written by Frank Heydlauf).
2052 SC/02 Eximstats: use FileHandles for outputing results.
2054 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2056 SC/04 Eximstats: fixed display of large numbers with -nvr option
2058 SC/05 Eximstats: fixed merging of reports with empty tables.
2060 SC/06 Eximstats: added the -include_original_destination flag
2062 SC/07 Eximstats: removed tabs and trailing whitespace.
2064 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2066 TK/06 MBOX spool code: Add real "From " MBOX separator line
2067 so the .eml file is really in mbox format (even though
2068 most programs do not really care). Patch from Alex Miller.
2070 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2071 The latter is generated from $received_to and is only set if the
2072 message has one envelope recipient. SA can use these headers,
2073 obviously out-of-the-box. Patch from Alex Miller.
2075 PH/08 The ${def test on a variable was returning false if the variable's
2076 value was "0", contrary to what the specification has always said!
2077 The result should be true unless the variable is empty.
2079 PH/09 The syntax error of a character other than { following "${if
2080 def:variable_name" (after optional whitespace) was not being diagnosed.
2081 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2082 accidental colon was present, for example, could give incorrect results.
2084 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2085 result is used (not including appendfile, where other changes are about
2088 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2089 This involved changing a lot of size variables from int to off_t. It
2090 should work with maildirs and everything.
2092 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2093 spamd dying while we are connected to it.
2095 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2098 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2099 function for BDB 4.3.
2101 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2103 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2104 the sources for PCRE have been re-arranged and more files are now
2107 PH/15 The code I had for printing potentially long long variables in PH/11
2108 above was not the best (it lost precision). The length of off_t variables
2109 is now inspected at build time, and an appropriate printing format (%ld
2110 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2111 to be "long long int" or "long int". This is needed for the internal
2112 formatting function string_vformat().
2114 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2115 the configuration file to be ":syslog", then the script "guesses" where
2116 the logs files are, rather than using the compiled in default. In our
2117 case the guess is not the same as the compiled default, so the script
2118 suddenly stopped working when I started to use syslog. The patch checks
2119 to see if log_file_path is "". If so, it attempts to read it from exim
2120 with no configuration file to get the compiled in version, before it
2121 falls back to the previous guessing code."
2123 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2124 implementing BATV in an Exim configuration. See NewStuff for the gory
2127 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2128 Makefile that are specific to HP-UX.
2130 PH/18 If the "use_postmaster" option was set for a recipient callout together
2131 with the "random" option, the postmaster address was used as the MAIL
2132 FROM address for the random test, but not for the subsequent recipient
2133 test. It is now used for both.
2135 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2136 patch removes a few documentation additions to RFC 3028, because the
2137 latest draft now contains them. It adds the new en;ascii-case comparator
2138 and a new error check for 8bit text in MIME parts. Comparator and
2139 require names are now matched exactly. I enabled the subaddress
2140 extension, but it is not well tested yet (read: it works for me)."
2142 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2143 rework some of the code of TK/09 above to avoid the hardwired use of
2144 "%lld" and "long long". Replaced the call to snprintf() with a call to
2147 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2148 records point to non-existent hosts", "retry timeout exceeded", and
2149 "retry time not reached for any host after a long failure period".
2151 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2152 experimental DomainKeys support:
2154 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2155 (2) On an error such as an illegally used "control", the wrong name for
2156 the control was given.
2158 These problems did NOT occur unless DomainKeys support was compiled.
2160 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2162 PH/24 Added ${if match_ip condition.
2164 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2165 databases so that it will be absolutely obvious if a crash occurs in the
2166 DB library. This is a regular occurrence (often caused by mis-matched
2169 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2170 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2171 picked up on a user's system that detects such things. There doesn't seem
2172 to be a gcc warning option for this - only an attribute that has to be
2173 put on the function's prototype. It seems that in Fedora Core 4 they have
2174 set this on a number of new functions. No doubt there will be more in due
2177 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2178 specify a transport. However, if an address that was verified by such a
2179 router was the subject of a callout, Exim crashed because it tried to
2180 read the rcpt_include_affixes from the non-existent transport. Now it
2181 just assumes that the setting of that option is false. This bug was
2182 introduced by 4.51/PH/31.
2184 PH/28 Changed -d+all to exclude +memory, because that information is very
2185 rarely of interest, but it makes the output a lot bigger. People tend to
2186 do -d+all out of habit.
2188 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2189 code in os-type was giving problems when libc.so lives in lib64, like on
2192 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2193 aren't the modern standard, and it seems that some systems' include files
2194 don't always have them. Exim was already checking for some of the newer
2195 ones like T_AAAA, and defining it itself. I've added checks for all the
2196 record types that Exim uses.
2198 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2199 not automatically generating a new one, as it is supposed to. This
2200 prevented TLS from working. If the file did exist, but contained invalid
2201 data, a new version was generated, as expected. It was only the case of a
2202 non-existent file that was broken.
2204 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2205 with a change in libdomainkeys > 0.64.
2207 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2208 from DNS. If the selector record carries the flag, it now has
2209 precedence over the domain-wide flag.
2211 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2213 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2214 the use of a port name that isn't defined in /etc/services) occurred, the
2215 message was deferred as in a normal delivery, and thus remained on the
2216 spool, instead of being failed because of the mua_wrapper setting. This
2217 is now fixed, and I tidied up some of the mua_wrapper messages at the
2220 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2221 the messages in a hash of arrays rather than using individual hashes.
2222 This is a bit cleaner and results in dramatic memory savings, albeit
2223 at a slight CPU cost.
2225 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2226 as requested by Marc Sherman.
2228 SC/10 Eximstats: added histograms for user specified patterns as requested
2231 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2233 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2234 fopen() in the content-scanning modules that did not already have it.
2240 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2241 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2243 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2245 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2247 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2248 to test Sieve filters that use "vacation".
2250 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2251 that changes the way the GnuTLS parameters are stored in the cache file.
2252 The new format can be generated externally. For backward compatibility,
2253 if the data in the cache doesn't make sense, Exim assumes it has read an
2254 old-format file, and it generates new data and writes a new file. This
2255 means that you can't go back to an older release without removing the
2258 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2259 work if there are any delivery delays because "one_time" forces the
2260 parent to be marked "delivered", so its unseen clone is never tried
2261 again. For this reason, Exim now forbids the simultaneous setting of
2264 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2265 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2266 entry if you want to know the details. The fix, however, neglected to
2267 consider the case when local delivery batching is involved. The test for
2268 "previously delivered" was not happening when checking to see if an
2269 address could be batched with a previous (undelivered) one; under
2270 certain circumstances this could lead to multiple deliveries to the same
2273 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2274 in its include files, and this causes problems building Exim.
2276 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2277 header_syntax) but Exim was just ignoring anything given after a slash.
2278 In particular, this caused confusion with an attempt to use "verify =
2279 reverse_host_lookup/defer_ok". An error is now given when options are
2280 supplied for verify items that do not have them. (Maybe reverse_host_
2281 lookup should have a defer_ok option, but that's a different point.)
2283 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2284 defined by RFC 821) to 2048, because there were problems with some AUTH
2285 commands, and RFC 1869 says the size should be increased for extended
2286 SMTP commands that take arguments.
2288 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2291 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2292 "unknown" error; now it says that the functionality isn't in the binary.
2294 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2295 an address' error message when a string expansion fails (syntax or
2296 whatever). Otherwise the password may appear in the log. Following change
2297 PH/42 below, there is no longer a chance of it appearing in a bounce
2300 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2302 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2303 of its addresses was ever tried. (Bugzilla bug #2.)
2305 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2306 the result incorrectly in the debug output. (It correctly added a newline
2307 to what was transported.)
2309 TF/01 Added $received_time.
2311 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2312 commented out examples of how to interface to a virus scanner and to
2313 SpamAssassin. Also added commented examples of av_scanner and
2314 spamd_address settings.
2316 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2317 and controls are allowed in which ACLs. There were a couple of minor
2318 errors. Some of the entries in the conditions table (which is a table of
2319 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2320 negation of where the condition IS allowed.
2322 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2324 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2325 header file does not have a version number, so I've had to invent a new
2326 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2327 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2328 radiusclient), but it was contributed by a Radius user.
2330 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2331 files or directories, for OpenSSL.
2333 PH/19 When an Exim process that is running as root has to create an Exim log
2334 file, it does so in a subprocess that runs as exim:exim so as to get the
2335 ownership right at creation (otherwise, other Exim processes might see
2336 the file with the wrong ownership). There was no test for failure of this
2337 fork() call, which would lead to the process getting stuck as it waited
2338 for a non-existent subprocess. Forks do occasionally fail when resources
2339 run out. I reviewed all the other calls to fork(); they all seem to check
2342 PH/20 When checking for unexpected SMTP input at connect time (before writing
2343 the banner), Exim was not dealing correctly with a non-positive return
2344 from the read() function. If the client had disconnected by this time,
2345 the result was a log entry for a synchronization error with an empty
2346 string after "input=" when read() returned zero. If read() returned -1
2347 (an event I could not check), uninitialized data bytes were printed.
2348 There were reports of junk text (parts of files, etc) appearing after
2351 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2353 PH/22 Added support for macro redefinition, and (re)definition in between
2354 driver and ACL definitions.
2356 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2357 forgetting to use the resulting value; it was using the unexpanded value.
2359 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2360 hadn't been configured. The fix is from Juergen Kreileder, who
2361 understands it better than I do:
2363 "Here's what I see happening with three configured cyrus_sasl
2364 authenticators configured (plain, login, cram-md5):
2366 On startup auth_cyrus_sasl_init() gets called for each of these.
2367 This means three calls to sasl_listmech() without a specified mech_list.
2368 => SASL tests which mechs of all available mechs actually work
2369 => three warnings about OTP not working
2370 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2372 With the patch, sasl_listmech() also gets called three times. But now
2373 SASL's mech_list option is set to the server_mech specified in the the
2374 authenticator. Or in other words, the answer from sasl_listmech()
2375 gets limited to just the mech you're testing for (which is different
2377 => the return list contains just 'plain' or 'login', 'cram-md5' or
2378 nothing depending on the value of ob->server_mech.
2380 I've just tested the patch: Authentication still works fine,
2381 unavailable mechs specified in the exim configuration are still
2382 caught, and the auth.log warnings about OTP are gone."
2384 PH/25 When debugging is enabled, the contents of the command line are added
2385 to the debugging output, even when log_selector=+arguments is not
2388 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2389 answer is "GNU", and only if the return is "GNU/something" is the answer
2392 PH/27 $acl_verify_message is now set immediately after the failure of a
2393 verification in an ACL, and so is available in subsequent modifiers. In
2394 particular, the message can be preserved by coding like this:
2396 warn !verify = sender
2397 set acl_m0 = $acl_verify_message
2399 Previously, $acl_verify_message was set only while expanding "message"
2400 and "log_message" when a very denied access.
2402 PH/28 Modified OS/os.c-Linux with
2404 -#ifndef OS_LOAD_AVERAGE
2405 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2407 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2408 nomenclature these days.)
2410 PH/29 Installed patch from the Sieve maintainer that adds the options
2411 sieve_useraddress and sieve_subaddress to the redirect router.
2413 PH/30 In these circumstances:
2414 . Two addresses routed to the same list of hosts;
2415 . First host does not offer TLS;
2416 . First host accepts first address;
2417 . First host gives temporary error to second address;
2418 . Second host offers TLS and a TLS session is established;
2419 . Second host accepts second address.
2420 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2421 and peerdn, if requested) that were in fact used only for the second
2424 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2425 attention to any local part prefix or suffix that was matched by the
2426 router that accepted the address. It now behaves in the same way as it
2427 does for delivery: the affixes are removed from the local part unless
2428 rcpt_include_affixes is set on the transport.
2430 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2431 timeout during the DATA phase of an incoming message.
2433 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2434 applied a patch sanctioned by the Sieve maintainer.
2436 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2437 the uid or gid is negative. A case of a negative gid caused this to be
2438 noticed. The fix allows for either to be negative.
2440 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2441 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2444 PH/36 The change PH/12 above was broken. Fixed it.
2446 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2447 the grounds that routing the same address twice would always produce the
2448 same answer. This might have been true once, but it is certainly no
2449 longer true now. Routing a child address may depend on the previous
2450 routing that produced that child. Some complicated redirection strategies
2451 went wrong when messages had multiple recipients, and made Exim's
2452 behaviour dependent on the order in which the addresses were given.
2454 I have moved the duplicate checking until after the routing is complete.
2455 Exim scans the addresses that are assigned to local and remote
2456 transports, and removes any duplicates. This means that more work will be
2457 done, as duplicates will always all be routed, but duplicates are
2458 presumably rare, so I don't expect this is of any significance.
2460 For deliveries to pipes, files, and autoreplies, the duplicate checking
2461 still happens during the routing process, since they are not going to be
2464 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2465 It corrects a timeout issue with spamd. This is Ian's comment: "The
2466 background is that sometimes spamd either never reads data from a
2467 connection it has accepted, or it never writes response data. The exiscan
2468 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2469 blindly assumes that writes won't block so it may never time out."
2471 PH/39 Allow G after quota size as well as K and M.
2473 PH/40 The value set for $authenticated_id in an authenticator may not contain
2474 binary zeroes or newlines because the value is written to log lines and
2475 to spool files. There was no check on this. Now the value is run through
2476 the string_printing() function so that such characters are converted to
2477 printable escape sequences.
2479 PH/41 $message_linecount is a new variable that contains the total number of
2480 lines in the message. Compare $body_linecount, which is the count for the
2483 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2484 bounce and delay warning messages, except in certain special cases, which
2487 (a) An SMTP error message from a remote host;
2488 (b) A message specified in a :fail: redirection;
2489 (c) A message specified in a "fail" command in a system filter;
2490 (d) A message specified in a FAIL return from the queryprogram router;
2491 (e) A message specified by the cannot_route_message router option.
2493 In these cases only, Exim does include the error details in bounce and
2494 warning messages. There are also a few cases where bland messages such
2495 as "unrouteable address" or "local delivery error" are given.
2497 PH/43 $value is now also set for the "else" part of a ${run expansion.
2499 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2500 being worked on, but at least Exim now implements the latest version to
2503 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2504 process to complete was treated as a delivery failure, a timeout while
2505 writing the message to the pipe was logged, but erroneously treated as a
2506 successful delivery. Such timeouts include transport filter timeouts. For
2507 consistency with the overall process timeout, these timeouts are now
2508 treated as errors, giving rise to delivery failures by default. However,
2509 there is now a new Boolean option for the pipe transport called
2510 timeout_defer, which, if set TRUE, converts the failures into defers for
2511 both kinds of timeout. A transport filter timeout is now identified in
2514 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2515 systems where "make" and "gmake" are different, calling "gmake" at top
2516 level broke things. I've arranged for the value of $(MAKE) to be passed
2517 from the Makefile to this script so that it can call the same version of
2521 A note about Exim versions 4.44 and 4.50
2522 ----------------------------------------
2524 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2525 changes of various kinds. As a consequence, a big documentation update was
2526 needed. This delayed the release for rather longer than seemed good, especially
2527 in the light of a couple of (minor) security issues. Therefore, the changes
2528 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2529 release. So 4.44 and 4.50 are in effect two different branches that both start
2532 I have left the 4.50 change log unchanged; it contains all the changes since
2533 4.43. The change log for 4.44 is below; many of its items are identical to
2534 those for 4.50. This seems to be the most sensible way to preserve the
2535 historical information.
2541 1. Minor wording change to the doc/README.SIEVE file.
2543 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2544 computation of the current number of files was incorrect.
2546 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2547 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2550 4. Give more explanation in the error message when the command for a transport
2551 filter fails to execute.
2553 5. There are several places where Exim runs a non-Exim command in a
2554 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2555 was being done only for the command run by the queryprogram router. It is
2556 now done for all such subprocesses. The other cases are: ${run, transport
2557 filters, and the commands run by the lmtp and pipe transports.
2559 6. Added CONFIGURE_GROUP build-time option.
2561 7. Some older OS have a limit of 256 on the maximum number of file
2562 descriptors. Exim was using setrlimit() to set 1000 as a large value
2563 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2564 systems. I've change it so that if it can't get 1000, it tries for 256.
2566 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2567 was an oversight, and furthermore, ever since the addition of extra
2568 controls (e.g. 4.43/32), the checks on when to allow different forms of
2569 "control" were broken. There should now be diagnostics for all cases when a
2570 control that does not make sense is encountered.
2572 9. Added the /retain_sender option to "control=submission".
2574 10. $recipients is now available in the predata ACL (oversight).
2576 11. Tidy the search cache before the fork to do a delivery from a message
2577 received from the command line. Otherwise the child will trigger a lookup
2578 failure and thereby defer the delivery if it tries to use (for example) a
2579 cached ldap connection that the parent has called unbind on.
2581 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2582 of $address_data from the recipient verification was clobbered by the
2583 sender verification.
2585 13. The value of address_data from a sender verification is now available in
2586 $sender_address_data in subsequent conditions in the ACL statement.
2588 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2590 15. Added a new option "connect=<time>" to callout options, to set a different
2593 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2594 was its contents. (It was OK if the option was not defined at all.)
2596 17. A "Completed" log line is now written for messages that are removed from
2597 the spool by the -Mrm option.
2599 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2600 information about exactly what failed.
2602 19. Added -dd to debug only the daemon process.
2604 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2605 handles timeouts, both on the server side and network timeouts. Renamed the
2606 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2608 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2609 It is now set to "smtps".
2611 22. $host_address is now set to the target address during the checking of
2612 ignore_target_hosts.
2614 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2615 being passed; this would have caused $sender_host_name to have been used if
2616 matching the list had actually called for a host name (not very likely,
2617 since this list is usually IP addresses). A host name is now passed as
2620 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2621 code that specifies a non-restarting handler (typically sigaction() in
2622 modern systems) in an attempt to fix a rare and obscure crash bug.
2624 25. Narrowed the window for a race in the daemon that could cause it to ignore
2625 SIGCHLD signals. This is not a major problem, because they are used only to
2626 wake it up if nothing else does.
2628 26. A malformed maildirsize file could cause Exim to calculate negative values
2629 for the mailbox size or file count. Odd effects could occur as a result.
2630 The maildirsize information is now recalculated if the size or filecount
2633 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2634 support for a long time. Removed HAVE_SYS_VFS_H.
2636 28. Installed the latest version of exipick from John Jetmore.
2638 29. In an address list, if the pattern was not a regular expression, an empty
2639 subject address (from a bounce message) matched only if the pattern was an
2640 empty string. Non-empty patterns were not even tested. This was the wrong
2641 because it is perfectly reasonable to use an empty address as part of a
2642 database query. An empty address is now tested by patterns that are
2643 lookups. However, all the other forms of pattern expect the subject to
2644 contain a local part and a domain, and therefore, for them, an empty
2645 address still always fails if the pattern is not itself empty.
2647 30. Exim went into a mad DNS loop when attempting to do a callout where the
2648 host was specified on an smtp transport, and looking it up yielded more
2649 than one IP address.
2651 31. Re-factored the code for checking spool and log partition space into a
2652 function that finds that data and another that does the check. The former
2653 is then used to implement four new variables: $spool_space, $log_space,
2654 $spool_inodes, and $log_inodes.
2656 32. The RFC2047 encoding function was originally intended for short strings
2657 such as real names; it was not keeping to the 75-character limit for
2658 encoded words that the RFC imposes. It now respects the limit, and
2659 generates multiple encoded words if necessary. To be on the safe side, I
2660 have increased the buffer size for the ${rfc2047: expansion operator from
2663 33. It is now permitted to omit both strings after an "if" condition; if the
2664 condition is true, the result is "true". As before, when the second string
2665 is omitted, a false condition yields an empty string. This makes it less
2666 cumbersome to write custom ACL and router conditions.
2668 34. Failure to deliver a bounce message always caused it to be frozen, even if
2669 there was an errors_to setting on the router. The errors_to setting is now
2672 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2673 canonical form (fully expanded) before being placed in
2674 $sender_host_address.
2676 36. The table in the code that translates DNS record types into text (T_A to
2677 "A" for instance) was missing entries for NS and CNAME. It is just possible
2678 that this could have caused confusion if both these types were looked up
2679 for the same domain, because the text type is used as part of Exim's
2680 per-process caching. But the chance of anyone hitting this buglet seems
2683 37. The dnsdb lookup has been extended in a number of ways.
2685 (1) There is a new type, "zns", which walks up the domain tree until it
2686 finds some nameserver records. It should be used with care.
2688 (2) There is a new type, "mxh", which is like "mx" except that it returns
2689 just the host names, not the priorities.
2691 (3) It is now possible to give a list of domains (or IP addresses) to be
2692 looked up. The behaviour when one of the lookups defers can be
2693 controlled by a keyword.
2695 (4) It is now possible to specify the separator character for use when
2696 multiple records are returned.
2698 38. The dnslists ACL condition has been extended: it is now possible to supply
2699 a list of IP addresses and/or domains to be looked up in a particular DNS
2702 39. Added log_selector=+queue_time_overall.
2704 40. When running the queue in the test harness, wait just a tad after forking a
2705 delivery process, to get repeatability of debugging output.
2707 41. Include certificate and key file names in error message when GnuTLS fails
2708 to set them up, because the GnuTLS error message doesn't include the name
2709 of the failing file when there is a problem reading it.
2711 42. Allow both -bf and -bF in the same test run.
2713 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2715 44. The "Exiscan patch" is now merged into the mainline Exim source.
2717 45. Sometimes the final signoff response after QUIT could fail to get
2718 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2719 < 0 before doing a fflush(). This bug looks as though it goes back to the
2720 introduction of TLS in release 3.20, but "sometimes" must have been rare
2721 because the tests only now provoked it.
2723 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2724 (this can affect the format of dates).
2726 47. exim_tidydb, when checking for the continued existence of a message for
2727 which it has found a message-specific retry record, was not finding
2728 messages that were in split spool directories. Consequently, it was
2729 deleting retry records that should have stayed in existence.
2731 48. Steve fixed some bugs in eximstats.
2733 49. The SPA authentication driver was not abandoning authentication and moving
2734 on to the next authenticator when an expansion was forced to fail,
2735 contradicting the general specification for all authenticators. Instead it
2736 was generating a temporary error. It now behaves as specified.
2738 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2739 (the order specifies the preference for clients). The order is now AES256,
2740 AES128, 3DES, ARCFOUR128.
2742 51. Small patch to Sieve code - explicitly set From: when generating an
2745 52. Exim crashed if a remote delivery caused a very long error message to be
2746 recorded - for instance if somebody sent an entire SpamAssassin report back
2747 as a large number of 550 error lines. This bug was coincidentally fixed by
2748 increasing the size of one of Exim's internal buffers (big_buffer) that
2749 happened as part of the Exiscan merge. However, to be on the safe side, I
2750 have made the code more robust (and fixed the comments that describe what
2753 53. Now that there can be additional text after "Completed" in log lines (if
2754 the queue_time_overall log selector is set), a one-byte patch to exigrep
2755 was needed to allow it to recognize "Completed" as not the last thing in
2758 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2759 patch that reportedly fixes this has been added. I am not expert enough to
2760 create a test for it. This is what the patch creator wrote:
2762 "I found a little strange behaviour of ldap code when working with
2763 Windows 2003 AD Domain, where users was placed in more than one
2764 Organization Units. When I tried to give exim partial DN, the exit code
2765 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2766 But simultaneously result of request was absolutely normal ldap result,
2767 so I produce this patch..."
2769 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2770 REFERENCE, so I have modified the code to exclude the patch when that macro
2773 55. Some experimental protocols are using DNS PTR records for new purposes. The
2774 keys for these records are domain names, not reversed IP addresses. The
2775 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2776 leaves it alone. Component reversal etc. now happens only for IP addresses.
2779 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2781 57. Double the size of the debug message buffer (to 2048) so that more of very
2782 long debug lines gets shown.
2784 58. The exicyclog utility now does better if the number of log files to keep
2785 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2787 59. Two changes related to the smtp_active_hostname option:
2789 (1) $smtp_active_hostname is now available as a variable.
2790 (2) The default for smtp_banner uses $smtp_active_hostname instead
2791 of $primary_hostname.
2793 60. The host_aton() function is supposed to be passed a string that is known
2794 to be a valid IP address. However, in the case of IPv6 addresses, it was
2795 not checking this. This is a hostage to fortune. Exim now panics and dies
2796 if the condition is not met. A case was found where this could be provoked
2797 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2798 components; fortuitously, this particular loophole had already been fixed
2799 by change 4.50/55 above.
2801 If there are any other similar loopholes, the new check in host_aton()
2802 itself should stop them being exploited. The report I received stated that
2803 data on the command line could provoke the exploit when Exim was running as
2804 exim, but did not say which command line option was involved. All I could
2805 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2806 running as the user.
2809 61. There was a buffer overflow vulnerability in the SPA authentication code
2810 (which came originally from the Samba project). I have added a test to the
2811 spa_base64_to_bits() function which I hope fixes it.
2814 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2815 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2817 63. The daemon start-up calls getloadavg() while still root for those OS that
2818 need the first call to be done as root, but it missed one case: when
2819 deliver_queue_load_max is set with deliver_drop_privilege. This is
2820 necessary for the benefit of the queue runner, because there is no re-exec
2821 when deliver_drop_privilege is set.
2823 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2824 This has been fixed.
2826 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2827 was in use, was not putting the data itself into the right store pool;
2828 consequently, it could be overwritten for a subsequent message in the same
2829 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2832 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2834 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2835 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2836 the function were treating the return as a boolean value, which happened to
2837 work because 0=false and not-0=true, but is not correct code.
2839 68. The host_aton() function was not handling scoped IPv6 addresses (those
2840 with, for example, "%eth0" on the end) correctly.
2842 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2843 negated items (that is, ~something) in unsigned ints. Some compilers
2844 apparently mutter when there is no cast.
2846 70. If an address verification called from an ACL failed, and did not produce a
2847 user-specific message (i.e. there was only a "system" message), nothing was
2848 put in $acl_verify_message. In this situation, it now puts the system
2851 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2852 session; change 4.31/43 added the unwanted input to the log line - except
2853 that it did not do this in the start of session case. It now does.
2855 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2856 This is wrong and can cause the other end to generate a synchronization
2857 error if it is another Exim or anything else that does the synchronization
2858 check. A QUIT command is no longer sent after a timeout.
2860 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2861 during host lookups.
2863 74. The defer_ok option of callout verification was not working if it was used
2864 when verifying addresses in header lines, that is, for this case:
2866 verify = header_sender/callout=defer_ok
2868 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2869 those file descriptors could be used for SMTP connections. If anything
2870 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2871 could be sent to the SMTP client, causing chaos. The daemon now opens
2872 stdin, stdout, and stderr to /dev/null when it puts itself into the
2875 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2876 log by default. The user can override this with suitable Perl magic.
2878 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2879 the log message when discard triggers, was not working for the DATA ACL or
2880 for the non-SMTP ACL.
2882 78. Error message wording change in sieve.c.
2884 79. If smtp_accept_max_per_host was set, the number of connections could be
2885 restricted to fewer than expected, because the daemon was trying to set up
2886 a new connection before checking whether the processes handling previous
2887 connections had finished. The check for completed processes is now done
2888 earlier. On busy systems, this bug wouldn't be noticed because something
2889 else would have woken the daemon, and it would have reaped the completed
2892 80. If a message was submitted locally by a user whose login name contained one
2893 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2894 It caused a spool format error. I have fixed the spool reading code. A
2895 related problem was that the "from" clause in the Received: line became
2896 illegal because of the space(s). It is now covered by ${quote_local_part.
2898 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2901 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2903 83. Patch from Sieve maintainer to fix unterminated string problem in
2904 "vacation" handling.
2906 84. Some minor changes to the Linux configuration files to help with other
2907 OS variants using glibc.
2909 85. One more patch for Sieve to update vacation handling to latest spec.
2912 ----------------------------------------------------
2913 See the note above about the 4.44 and 4.50 releases.
2914 ----------------------------------------------------
2920 1. Change 4.43/35 introduced a bug that caused file counts to be
2921 incorrectly computed when quota_filecount was set in an appendfile
2924 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2925 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2928 3. Give more explanation in the error message when the command for a transport
2929 filter fails to execute.
2931 4. There are several places where Exim runs a non-Exim command in a
2932 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2933 was being done only for the command run by the queryprogram router. It is
2934 now done for all such subprocesses. The other cases are: ${run, transport
2935 filters, and the commands run by the lmtp and pipe transports.
2937 5. Some older OS have a limit of 256 on the maximum number of file
2938 descriptors. Exim was using setrlimit() to set 1000 as a large value
2939 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2940 systems. I've change it so that if it can't get 1000, it tries for 256.
2942 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2943 was an oversight, and furthermore, ever since the addition of extra
2944 controls (e.g. 4.43/32), the checks on when to allow different forms of
2945 "control" were broken. There should now be diagnostics for all cases when a
2946 control that does not make sense is encountered.
2948 7. $recipients is now available in the predata ACL (oversight).
2950 8. Tidy the search cache before the fork to do a delivery from a message
2951 received from the command line. Otherwise the child will trigger a lookup
2952 failure and thereby defer the delivery if it tries to use (for example) a
2953 cached ldap connection that the parent has called unbind on.
2955 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2956 of $address_data from the recipient verification was clobbered by the
2957 sender verification.
2959 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2960 was its contents. (It was OK if the option was not defined at all.)
2962 11. A "Completed" log line is now written for messages that are removed from
2963 the spool by the -Mrm option.
2965 12. $host_address is now set to the target address during the checking of
2966 ignore_target_hosts.
2968 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2969 being passed; this would have caused $sender_host_name to have been used if
2970 matching the list had actually called for a host name (not very likely,
2971 since this list is usually IP addresses). A host name is now passed as
2974 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2975 code that specifies a non-restarting handler (typically sigaction() in
2976 modern systems) in an attempt to fix a rare and obscure crash bug.
2978 15. Narrowed the window for a race in the daemon that could cause it to ignore
2979 SIGCHLD signals. This is not a major problem, because they are used only to
2980 wake it up if nothing else does.
2982 16. A malformed maildirsize file could cause Exim to calculate negative values
2983 for the mailbox size or file count. Odd effects could occur as a result.
2984 The maildirsize information is now recalculated if the size or filecount
2987 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2988 support for a long time. Removed HAVE_SYS_VFS_H.
2990 18. Updated exipick to current release from John Jetmore.
2992 19. Allow an empty sender to be matched against a lookup in an address list.
2993 Previously the only cases considered were a regular expression, or an
2996 20. Exim went into a mad DNS lookup loop when doing a callout where the
2997 host was specified on the transport, if the DNS lookup yielded more than
3000 21. The RFC2047 encoding function was originally intended for short strings
3001 such as real names; it was not keeping to the 75-character limit for
3002 encoded words that the RFC imposes. It now respects the limit, and
3003 generates multiple encoded words if necessary. To be on the safe side, I
3004 have increased the buffer size for the ${rfc2047: expansion operator from
3007 22. Failure to deliver a bounce message always caused it to be frozen, even if
3008 there was an errors_to setting on the router. The errors_to setting is now
3011 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3012 canonical form (fully expanded) before being placed in
3013 $sender_host_address.
3015 24. Updated eximstats to version 1.33
3017 25. Include certificate and key file names in error message when GnuTLS fails
3018 to set them up, because the GnuTLS error message doesn't include the name
3019 of the failing file when there is a problem reading it.
3021 26. Expand error message when OpenSSL has problems setting up cert/key files.
3024 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3025 (this can affect the format of dates).
3027 28. exim_tidydb, when checking for the continued existence of a message for
3028 which it has found a message-specific retry record, was not finding
3029 messages that were in split spool directories. Consequently, it was
3030 deleting retry records that should have stayed in existence.
3032 29. eximstats updated to version 1.35
3033 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3034 1.35 - bugfix such that pie charts by volume are generated correctly
3036 30. The SPA authentication driver was not abandoning authentication and moving
3037 on to the next authenticator when an expansion was forced to fail,
3038 contradicting the general specification for all authenticators. Instead it
3039 was generating a temporary error. It now behaves as specified.
3041 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3042 (the order specifies the preference for clients). The order is now AES256,
3043 AES128, 3DES, ARCFOUR128.
3045 31. Small patch to Sieve code - explicitly set From: when generating an
3048 32. Exim crashed if a remote delivery caused a very long error message to be
3049 recorded - for instance if somebody sent an entire SpamAssassin report back
3050 as a large number of 550 error lines. This bug was coincidentally fixed by
3051 increasing the size of one of Exim's internal buffers (big_buffer) that
3052 happened as part of the Exiscan merge. However, to be on the safe side, I
3053 have made the code more robust (and fixed the comments that describe what
3056 33. Some experimental protocols are using DNS PTR records for new purposes. The
3057 keys for these records are domain names, not reversed IP addresses. The
3058 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3059 leaves it alone. Component reversal etc. now happens only for IP addresses.
3062 34. The host_aton() function is supposed to be passed a string that is known
3063 to be a valid IP address. However, in the case of IPv6 addresses, it was
3064 not checking this. This is a hostage to fortune. Exim now panics and dies
3065 if the condition is not met. A case was found where this could be provoked
3066 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3067 components; fortuitously, this particular loophole had already been fixed
3068 by change 4.50/55 or 4.44/33 above.
3070 If there are any other similar loopholes, the new check in host_aton()
3071 itself should stop them being exploited. The report I received stated that
3072 data on the command line could provoke the exploit when Exim was running as
3073 exim, but did not say which command line option was involved. All I could
3074 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3075 running as the user.
3078 35. There was a buffer overflow vulnerability in the SPA authentication code
3079 (which came originally from the Samba project). I have added a test to the
3080 spa_base64_to_bits() function which I hope fixes it.
3083 36. The daemon start-up calls getloadavg() while still root for those OS that
3084 need the first call to be done as root, but it missed one case: when
3085 deliver_queue_load_max is set with deliver_drop_privilege. This is
3086 necessary for the benefit of the queue runner, because there is no re-exec
3087 when deliver_drop_privilege is set.
3089 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3090 was in use, was not putting the data itself into the right store pool;
3091 consequently, it could be overwritten for a subsequent message in the same
3092 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3095 38. Sometimes the final signoff response after QUIT could fail to get
3096 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3097 < 0 before doing a fflush(). This bug looks as though it goes back to the
3098 introduction of TLS in release 3.20, but "sometimes" must have been rare
3099 because the tests only now provoked it.
3105 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3106 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3107 Now it returns an integer. A number of calls were still expecting a T/F
3108 return. Fortuitously, in all cases, the tests worked in OK situations,
3109 which is the norm. However, things would have gone wrong on any write
3110 failures on the smtp file descriptor. This function is used when sending
3111 messages over SMTP and also when doing verify callouts.
3113 2. When Exim is called to do synchronous delivery of a locally submitted
3114 message (the -odf or -odi options), it no longer closes stderr before doing
3117 3. Implemented the mua_wrapper option.
3119 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3121 5. Implemented the functions header_remove(), header_testname(),
3122 header_add_at_position(), and receive_remove_recipient(), and exported them
3125 6. If an ACL "warn" statement specified the addition of headers, Exim already
3126 inserted X-ACL-Warn: at the start if there was no header name. However, it
3127 was not making this test for the second and subsequent header lines if
3128 there were newlines in the string. This meant that an invalid header could
3129 be inserted if Exim was badly configured.
3131 7. Allow an ACL "warn" statement to add header lines at the start or after all
3132 the Received: headers, as well as at the end.
3134 8. Added the rcpt_4xx retry error code.
3136 9. Added postmaster_mailfrom=xxx to callout verification option.
3138 10. Added mailfrom=xxxx to the callout verification option, for verify=
3141 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3142 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3144 12. Inserted some casts to stop certain compilers warning when using pointer
3145 differences as field lengths or precisions in printf-type calls (mostly
3146 affecting debugging statements).
3148 13. Added optional readline() support for -be (dynamically loaded).
3150 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3151 same clock tick as a message's arrival, so that its received time was the
3152 same as the "first fail" time on the retry record, and that message
3153 remained on the queue past the ultimate address timeout, every queue runner
3154 would try a delivery (because it was past the ultimate address timeout) but
3155 after another failure, the ultimate address timeout, which should have then
3156 bounced the address, did not kick in. This was a "< instead of <=" error;
3157 in most cases the first failure would have been in the next clock tick
3158 after the received time, and all would be well.
3160 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3161 being recognized when the domain list was tested by the match_domain
3162 condition in an expansion string.
3164 16. Added the ${str2b64: operator.
3166 17. Exim was always calling setrlimit() to set a large limit for the number of
3167 processes, without checking whether the existing limit was already
3168 adequate. (It did check for the limit on file descriptors.) Furthermore,
3169 errors from getrlimit() and setrlimit() were being ignored. Now they are
3170 logged to the main and panic logs, but Exim does carry on, to try to do its
3171 job under whatever limits there are.
3173 18. Imported PCRE 5.0.
3175 19. Trivial typo in log message " temporarily refused connection" (the leading
3178 20. If the log selector return_path_on_delivery was set and an address was
3179 redirected to /dev/null, the delivery process crashed because it assumed
3180 that a return path would always be set for a "successful" delivery. In this
3181 case, the whole delivery is bypassed as an optimization, and therefore no
3184 21. Internal re-arrangement: the function for sending a challenge and reading
3185 a response while authentication was assuming a zero-terminated challenge
3186 string. It's now changed to take a pointer and a length, to allow for
3187 binary data in such strings.
3189 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3191 23. Exim was not respecting finduser_retries when seeking the login of the
3192 uid under which it was called; it was always trying 10 times. (The default
3193 setting of finduser_retries is zero.) Also, it was sleeping after the final
3194 failure, which is pointless.
3196 24. Implemented tls_on_connect_ports.
3198 25. Implemented acl_smtp_predata.
3200 26. If the domain in control=submission is set empty, Exim assumes that the
3201 authenticated id is a complete email address when it generates From: or
3202 Sender: header lines.
3204 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3205 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3206 chown and chgrp in /bin and hostname in /usr/bin.
3208 28. Exim was keeping the "process log" file open after each use, just as it
3209 does for the main log. This opens the possibility of it remaining open for
3210 long periods when the USR1 signal hits a daemon. Occasional processlog
3211 errors were reported, that could have been caused by this. Anyway, it seems
3212 much more sensible not to leave this file open at all, so that is what now
3215 29. The long-running daemon process does not normally write to the log once it
3216 has entered its main loop, and it closes the log before doing so. This is
3217 so that log files can straightforwardly be renamed and moved. However,
3218 there are a couple of unusual error situations where the daemon does write
3219 log entries, and I had neglected to close the log afterwards.
3221 30. The text of an SMTP error response that was received during a remote
3222 delivery was being truncated at 512 bytes. This is too short for some of
3223 the long messages that one sometimes sees. I've increased the limit to
3226 31. It is now possible to make retry rules that apply only when a message has a
3227 specific sender, in particular, an empty sender.
3229 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3230 it possible to be selective about when SMTP synchronization is enforced.
3232 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3234 32. Implemented hosts_connection_nolog.
3236 33. Added an ACL for QUIT.
3238 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3241 35. Added mailbox_size and mailbox_filecount to appendfile.
3243 36. Added control = no_multiline_responses to ACLs.
3245 37. There was a bug in the logic of the code that waits for the clock to tick
3246 in the case where the clock went backwards by a substantial amount such
3247 that the microsecond fraction of "now" was more than the microsecond
3248 fraction of "then" (but the whole seconds number was less).
3250 38. Added support for the libradius Radius client library this is found on
3251 FreeBSD (previously only the radiusclient library was supported).
3257 1. When certain lookups returned multiple values in the form name=value, the
3258 quoting of the values was not always being done properly. Specifically:
3259 (a) If the value started with a double quote, but contained no whitespace,
3261 (b) If the value contained whitespace other than a space character (i.e.
3262 tabs or newlines or carriage returns) it was not quoted.
3263 This fix has been applied to the mysql and pgsql lookups by writing a
3264 separate quoting function and calling it from the lookup code. The fix
3265 should probably also be applied to nisplus, ibase and oracle lookups, but
3266 since I cannot test any of those, I have not disturbed their existing code.
3268 2. A hit in the callout cache for a specific address caused a log line with no
3269 reason for rejecting RCPT. Now it says "Previous (cached) callout
3270 verification failure".
3272 3. There was an off-by-one bug in the queryprogram router. An over-long
3273 return line was truncated at 256 instead of 255 characters, thereby
3274 overflowing its buffer with the terminating zero. As well as fixing this, I
3275 have increased the buffer size to 1024 (and made a note to document this).
3277 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3278 when Exim is waiting for an SMTP response from a remote server, Exim
3279 restarts its select() call on the socket, thereby resetting its timeout.
3280 This is not a problem when such interrupts are rare. Somebody set up a cron
3281 job to run exiwhat every 2 minutes, which is less than the normal select()
3282 timeout (5 or 10 minutes). This meant that the select() timeout never
3283 kicked in because it was always reset. I have fixed this by comparing the
3284 time when an interrupt arrives with the time at the start of the first call
3285 to select(). If more time than the timeout has elapsed, the interrupt is
3286 treated as a timeout.
3288 5. Some internal re-factoring in preparation for the addition of Sieve
3289 extensions (by MH). In particular, the "personal" test is moved to a
3290 separate function, and given an option for scanning Cc: and Bcc: (which is
3291 not set for Exim filters).
3293 6. When Exim created an email address using the login of the caller as the
3294 local part (e.g. when creating a From: or Sender: header line), it was not
3295 quoting the local part when it contained special characters such as @.
3297 7. Installed new OpenBSD configuration files.
3299 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3300 try to make them clearer.
3302 9. Callout options, other than the timeout value, were being ignored when
3303 verifying sender addresses in header lines. For example, when using
3305 verify = header_sender/callout=no_cache
3307 the cache was (incorrectly) being used.
3309 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3310 only the Cygwin environment.
3312 11. When return_path_on_delivery was set as a log selector, if different remote
3313 addresses in the same message used different return paths and parallel
3314 remote delivery occurred, the wrong values would sometimes be logged.
3315 (Whenever a remote delivery process finished, the return path value from
3316 the most recently started remote delivery process was logged.)
3318 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3319 lines when AUTH and/or TLS are in use. This is the "received protocol"
3320 field. Exim used to use "asmtp" for authenticated SMTP, without any
3321 indication (in the protocol name) for TLS use. Now it follows the RFC and
3322 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3323 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3324 names appear in log lines as well as in Received: header lines.
3326 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3327 extensions, and comparison tests, and to fix some bugs.
3329 14. Changes to the "personal" filter test:
3331 (1) The test was buggy in that it was just doing the equivalent of
3332 "contains" tests on header lines. For example, if a user's address was
3333 anne@some.where, the "personal" test would incorrectly be true for
3335 To: susanne@some.where
3337 This test is now done by extracting each address from the header in turn,
3338 and checking the entire address. Other tests that are part of "personal"
3339 are now done using regular expressions (for example, to check local parts
3340 of addresses in From: header lines).
3342 (2) The list of non-personal local parts in From: addresses has been
3343 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3344 taken from the Sieve specification recommendations.
3346 (3) If the message contains any header line starting with "List-" it is
3347 treated as non-personal.
3349 (4) The test for "circular" in the Subject: header line has been removed
3350 because it now seems ill-conceived.
3352 15. Minor typos in src/EDITME comments corrected.
3354 16. Installed latest exipick from John Jetmore.
3356 17. If headers_add on a router specified a text string that was too long for
3357 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3358 of string_sprintf() is now avoided.
3360 18. $message_body_size was not set (it was always zero) when running the DATA
3361 ACL and the local_scan() function.
3363 19. For the "mail" command in an Exim filter, no default was being set for
3364 the once_repeat time, causing a random time value to be used if "once" was
3365 specified. (If the value happened to be <= 0, no repeat happened.) The
3366 default is now 0s, meaning "never repeat". The "vacation" command was OK
3367 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3368 (I found it when inspecting the code).
3370 20. There is now an overall timeout for performing a callout verification. It
3371 defaults to 4 times the callout timeout, which applies to individual SMTP
3372 commands during the callout. The overall timeout applies when there is more
3373 than one host that can be tried. The timeout is checked before trying the
3374 next host. This prevents very long delays if there are a large number of
3375 hosts and all are timing out (e.g. when the network connections are timing
3376 out). The value of the overall timeout can be changed by specifying an
3377 additional sub-option for "callout", called "maxwait". For example:
3379 verify = sender/callout=5s,maxwait=20s
3381 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3382 to the end before writing, but this should make it even safer).
3384 22. Exim was forgetting that it had advertised PIPELINING for the second and
3385 subsequent messages on an SMTP connection. It was also not resetting its
3386 memory on STARTTLS and an internal HELO.
3388 23. When Exim logs an SMTP synchronization error within a session, it now
3389 records whether PIPELINING has been advertised or not.
3391 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3392 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3393 rather than long int.
3395 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3397 26. Added the never_mail option to autoreply.
3403 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3404 crash if the getsockname() call failed; this can happen if a connection is
3405 closed very soon after it is established. The problem was simply in the
3406 order in which certain operations were done, causing Exim to try to write
3407 to the SMTP stream before it had set up the file descriptor. The bug has
3408 been fixed by making things happen in the correct order.
3414 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3415 before the connection was closed, thus losing the rejection response.
3417 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3418 some early Solaris releases, but causes trouble in current releases where
3419 socklen_t is defined.
3421 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3424 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3427 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3428 that releasing the top of it at the end releases what was used for sub-
3429 expansions (unless the block got too big). However, discard this block if
3430 the first thing is a variable or header, so that we can use its block when
3431 it is dynamic (useful for very large $message_headers, for example).
3433 6. Lookups now cache *every* query, not just the most recent. A new, separate
3434 store pool is used for this. It can be recovered when all lookup caches are
3435 flushed. Lookups now release memory at the end of their result strings.
3436 This has involved some general refactoring of the lookup sources.
3438 7. Some code has been added to the store_xxx() functions to reduce the amount
3439 of flapping under certain conditions.
3441 8. log_incoming_interface used to affect only the <= reception log lines. Now
3442 it causes the local interface and port to be added to several more SMTP log
3443 lines, for example "SMTP connection from", and rejection lines.
3445 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3447 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3449 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3450 could be overwritten at the end of the current message (or the start of a
3451 new message if it was set in a HELO ACL). The value is now preserved for
3452 the duration of the SMTP connection.
3454 12. If a transport had a headers_rewrite setting, and a matching header line
3455 contained an unqualified address, that address was qualified, even if it
3456 did not match any rewriting rules. The underlying bug was that the values
3457 of the flags that permit the existence of unqualified sender and recipient
3458 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3459 non-local messages, and by -bnq for local messages) were not being
3460 preserved with the message after it was received.
3462 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3463 "next input=" as part of the text comprising the host identity instead of
3464 the correct text. The code was using the same buffer for two different
3465 strings. However, depending on which order the printing function evaluated
3466 its arguments, the bug did not always show up. Under Linux, for example, my
3467 test suite worked just fine.
3469 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3470 doesn't work with some older versions of Perl. It has been changed to "my",
3471 which in any case is probably the better facility to use.
3473 15. A really picky compiler found some instances of statements for creating
3474 error messages that either had too many or two few arguments for the format
3477 16. The size of the buffer for calls to the DNS resolver has been increased
3478 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3479 for addresses that have a lot of PTR records. This alleviates a problem; it
3480 does not fully solve it.
3482 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3483 buffer now truncates the list and logs the incident, which is the same
3484 action as happens when Exim is looking up a host name and its aliases.
3485 Previously in this situation something unpredictable would happen;
3486 sometimes it was "internal error: store_reset failed".
3488 18. If a server dropped the connection unexpectedly when an Exim client was
3489 using GnuTLS and trying to read a response, the client delivery process
3490 crashed while trying to generate an error log message.
3492 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3493 string, for example:
3495 warn message = H1: something\nH2: something
3497 the text was added as a single header line from Exim's point of view
3498 though it ended up OK in the delivered message. However, searching for the
3499 second and subsequent header lines using $h_h2: did not work. This has been
3500 fixed. Similarly, if a system filter added multiple headers in this way,
3501 the routers could not see them.
3503 20. Expanded the error message when iplsearch is called with an invalid key to
3504 suggest using net-iplsearch in a host list.
3506 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3507 ACLs are no longer actually imposed (and a message to that effect is
3510 22. If a "gecos" field in a passwd entry contained escaped characters, in
3511 particular, if it contained a \" sequence, Exim got it wrong when building
3512 a From: or a Sender: header from that name. A second bug also caused
3513 incorrect handling when an unquoted " was present following a character
3514 that needed quoting.
3516 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3517 was not being matched caselessly.
3519 24. Arranged for all hyphens in the exim.8 source to be escaped with
3522 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3523 properly. Recipient callout cache records were still being keyed to include
3524 the sender, even when use_sender was set false. This led to far more
3525 callouts that were necessary. The sender is no longer included in the key
3526 when use_sender is false.
3528 26. Added "control = submission" modifier to ACLs.
3530 27. Added the ${base62d: operator to decode base 62 numbers.
3532 28. dnsdb lookups can now access SRV records.
3534 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3535 the configuration file.
3537 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3538 (-v) mode. This makes the output for a verbose queue run more intelligible.
3540 31. Added a use_postmaster feature to recipient callouts.
3542 32. Added the $body_zerocount variable, containing the number of binary zero
3543 bytes in the message body.
3545 33. The time of last modification of the "new" subdirectory is now used as the
3546 "mailbox time last read" when there is a quota error for a maildir
3549 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3551 35. Added +ignore_unknown as a special item in host lists.
3553 36. Code for decoding IPv6 addresses in host lists is now included, even if
3554 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3555 address was recognized as an IP address, but was then not correctly decoded
3556 into binary, causing unexpected and incorrect effects when compared with
3563 1. Very minor rewording of debugging text in manualroute to say "list of
3564 hosts" instead of "hostlist".
3566 2. If verify=header_syntax was set, and a header line with an unqualified
3567 address (no domain) and a large number of spaces between the end of the
3568 name and the colon was received, the reception process suffered a buffer
3569 overflow, and (when I tested it) crashed. This was caused by some obsolete
3570 code that should have been removed. The fix is to remove it!
3572 3. When running in the test harness, delay a bit after writing a bounce
3573 message to get a bit more predictability in the log output.
3575 4. Added a call to search_tidyup() just before forking a reception process. In
3576 theory, someone could use a lookup in the expansion of smtp_accept_max_
3577 per_host which, without the tidyup, could leave open a database connection.
3579 5. Added the variables $recipient_data and $sender_data which get set from a
3580 lookup success in an ACL "recipients" or "senders" condition, or a router
3581 "senders" option, similar to $domain_data and $local_part_data.
3583 6. Moved the writing of debug_print from before to after the "senders" test
3586 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3587 problems for message scanning, either using a data ACL, or using
3588 local_scan() because the Received: header was not generated till after they
3589 were called (in order to set the time as the time of reception completion).
3590 I have revised the way this works. The header is now generated after the
3591 body is received, but before the ACL or local_scan() are called. After they
3592 are run, the timestamp in the header is updated.
3598 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3599 before starting a queue runner without re-exec. This happened only when
3600 deliver_drop_privilege was set or when the Exim user was set to root. The
3601 effect of the bug was that timeouts during subsequent deliveries caused
3602 crashes instead of being properly handled. The handler is now left at its
3603 default (and expected) setting.
3605 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3606 message, again when deliver_drop_privilege is set or Exim is run as root.
3607 The bug described in (1) was not present in this case, but the tidying up
3608 of the other signals was missing. I have made the two cases consistent.
3610 3. The ignore_target_hosts setting on a manualroute router was being ignored
3611 for hosts that were looked up using the /MX notation.
3613 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3616 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3617 operated on the sender address. After changing the $sender_address to <>
3618 for the sender address verify, Exim was re-instated it as the original
3619 (before rewriting) address, but remembering that it had rewritten it, so it
3620 wasn't rewriting it again. This bug also had the effect of breaking the
3621 sender address verification caching when the sender address was rewritten.
3623 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3624 This has been changed so that if the ip literal address matches
3625 ignore_target_hosts, the router declines.
3627 7. Added expansion conditions match_domain, match_address, and match_local_
3628 part (NOT match_host).
3630 8. The placeholder for the Received: header didn't have a length field set.
3632 9. Added code to Exim itself and to exim_lock to test for a specific race
3633 condition that could lead to file corruption when using MBX delivery. The
3634 issue is with the lockfile that is created in /tmp. If this file is removed
3635 after a process has opened it but before that process has acquired a lock,
3636 there is the potential for a second process to recreate the file and also
3637 acquire a lock. This could lead to two Exim processes writing to the file
3638 at the same time. The added code performs the same test as UW imapd; it
3639 checks after acquiring the lock that its file descriptor still refers to
3640 the same named file.
3642 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3643 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3646 11. Added the smtp_active_hostname option. If used, this will typically be made
3647 to depend on the incoming interface address. Because $interface_address is
3648 not set up until the daemon has forked a reception process, error responses
3649 that can happen earlier (such as "too many connections") no longer contain
3652 12. If an expansion in a condition on a "warn" statement fails because a lookup
3653 defers, the "warn" statement is abandoned, and the next ACL statement is
3654 processed. Previously this caused the whole ACL to be aborted.
3656 13. Added the iplsearch lookup type.
3658 14. Added ident_timeout as a log selector.
3660 15. Added tls_certificate_verified as a log selector.
3662 16. Added a global option tls_require_ciphers (compare the smtp transport
3663 option of the same name). This controls incoming TLS connections.
3665 17. I finally figured out how to make tls_require_ciphers do a similar thing
3666 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3667 before starting the TLS session.
3669 18. Tabs are now shown as \t in -bP output.
3671 19. If the log selector return_path_on_delivery was set, Exim crashed when
3672 bouncing a message because it had too many Received: header lines.
3674 20. If two routers both had headers_remove settings, and the first one included
3675 a superfluous trailing colon, the final name in the first list and the
3676 first name in the second list were incorrectly joined into one item (with a
3677 colon in the middle).
3683 1. Added -C and -D options to the exinext utility, mainly to make it easier
3684 to include in the automated testing, but these could be helpful when
3685 multiple configurations are in use.
3687 2. The exinext utility was not formatting the output nicely when there was
3688 an alternate port involved in the retry record key, nor when there was a
3689 message id as well (for retries that were specific to a specific message
3690 and a specific host). It was also confused by IPv6 addresses, because of
3691 the additional colons they contain. I have fixed the IPv4 problem, and
3692 patched it up to do a reasonable job for IPv6.
3694 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3695 delivery, the log line now contains "pipelined" if PIPELINING was used.
3697 4. An SMTP transport process used to panic and die if the bind() call to set
3698 an explicit outgoing interface failed. This has been changed; it is now
3699 treated in the same way as a connect() failure.
3701 5. A reference to $sender_host_name in the part of a conditional expansion
3702 that was being skipped was still causing a DNS lookup. This no longer
3705 6. The def: expansion condition was not recognizing references to header lines
3706 that used bh_ and bheader_.
3708 7. Added the _cache feature to named lists.
3710 8. The code for checking quota_filecount in the appendfile transport was
3711 allowing one more file than it should have been.
3713 9. For compatibility with Sendmail, the command line option
3721 and sets the incoming protocol and host name (for trusted callers). The
3722 host name and its colon can be omitted when only the protocol is to be set.
3723 Note the Exim already has two private options, -pd and -ps, that refer to
3724 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3725 "s", but I don't think that's a major issue.
3727 10. A number of refactoring changes to the code, none of which should affect
3730 (a) The number of logging options was getting close to filling up the
3731 32-bit word that was used as a bit map. I have split them into two classes:
3732 those that are passed in the argument to log_write(), and those that are
3733 only ever tested independently outside of that function. These are now in
3734 separate 32-bit words, so there is plenty of room for expansion again.
3735 There is no change in the user interface or the logging behaviour.
3737 (b) When building, for example, log lines, the code previously used a
3738 macro that called string_cat() twice, in order to add two strings. This is
3739 not really sufficiently general. Furthermore, there was one instance where
3740 it was actually wrong because one of the argument was used twice, and in
3741 one call a function was used. (As it happened, calling the function twice
3742 did not affect the overall behaviour.) The macro has been replaced by a
3743 function that can join an arbitrary number of extra strings onto a growing
3746 (c) The code for expansion conditions now uses a table and a binary chop
3747 instead of a serial search (which was left over from when there were very
3748 few conditions). Also, it now recognizes conditions like "pam" even when
3749 the relevant support is not compiled in: a suitably worded error message is
3750 given if an attempt is made to use such a condition.
3752 11. Added ${time_interval:xxxxx}.
3754 12. A bug was causing one of the ddress fields not to be passed back correctly
3755 from remote delivery subprocesses. The field in question was not being
3756 subsequently used, so this caused to problems in practice.
3758 13. Added new log selectors queue_time and deliver_time.
3760 14. Might have fixed a bug in maildirsizefile handling that threw up
3761 "unexpected character" debug warnings, and recalculated the data
3762 unnecessarily. In any case, I expanded the warning message to give more
3765 15. Added the message "Restricted characters in address" to the statements in
3766 the default ACL that block characters like @ and % in local parts.
3768 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3769 Three changes have been made:
3771 (a) There was a serious bug; a negative response to MAIL caused the whole
3772 recipient domain to be cached as invalid, thereby blocking all messages
3773 to all local parts at the same domain, from all senders. This bug has
3774 been fixed. The domain is no longer cached after a negative response to
3775 MAIL if the sender used is not empty.
3777 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3780 (c) A new callout option, "use_sender" has been added for people who want
3781 the modified behaviour.
3787 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3790 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3791 indeed breaks things for older releases.
3793 3. Added additional logging to the case where there is a problem reading data
3794 from a filter that is running in a subprocess using a pipe, in order to
3795 try to track down a specific problem.
3797 4. Testing facility fudge: when running in the test harness and attempting
3798 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3799 getting "No route to host". Convert this to a timeout.
3801 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3804 6. Some OS don't have socklen_t but use size_t instead. This affects the
3805 fifth argument of getsockopt() amongst other things. This is now
3806 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3807 can be set for individual OS. I have set it for SunOS5, OSF1, and
3808 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3809 some earlier ones do not.
3811 7. Change 4.30/15 was not doing the test caselessly.
3813 8. The standard form for an IPv6 address literal was being rejected by address
3814 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3815 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3816 this, as well as the form without the "IPv6" on the front (but only when
3817 address literals are enabled, of course).
3819 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3821 10. Exim crashed if a message with an empty sender address specified by -f
3822 encountered a router with an errors_to setting. This could be provoked only
3823 by a command such as
3827 where an empty string was supplied; "<>" did not hit this bug.
3829 11. Installed PCRE release 4.5.
3831 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3832 remained set. It is now erased.
3834 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3835 times from message ids (which are base 36 rather than the normal 62).
3837 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3838 were being counted as actual protocol errors, and logged if the log
3839 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3840 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3841 and DATA following a set of rejected RCPTs do not count as protocol errors.
3842 In other words, Exim assumes they were pipelined, though this may not
3843 actually be the case. Of course, in all cases the client gets an
3844 appropriate error code.
3846 15. If a lookup fails in an ACL condition, a message about the failure may
3847 be available; it is used if testing the ACL cannot continue, because most
3848 such messages specify what the cause of the deferral is. However, some
3849 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3850 that caused an old message to be retained and used if a later statement
3851 caused a defer, replacing the real cause of the deferral.
3853 16. If an IP address had so many PTR records that the DNS lookup buffer
3854 was not large enough to hold them, Exim could crash while trying to process
3855 the truncated data. It now detects and logs this case.
3857 17. Further to 4.21/58, another change has been made: if (and only if) the
3858 first line of a message (the first header line) ends with CRLF, a bare LF
3859 in a subsequent header line has a space inserted after it, so as not to
3860 terminate the header.
3862 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3863 unnecessarily, used atoi() instead of strtol(), and didn't check the
3864 termination when getting file sizes from file names by regex.
3866 19. Completely re-implemented the support for maildirsize files, in the light
3867 of a number of problems with the previous contributed implementation
3868 (4.30/29). In particular:
3870 . If the quota is zero, the maildirsize file is maintained, but no quota is
3873 . If the maildir directory does not exist, it is created before any attempt
3874 to write a maildirsize file.
3876 . The quota value in the file is just a cache; if the quota is changed in
3877 the transport, the new value overrides.
3879 . A regular expression is available for excluding directories from the
3882 20. The autoreply transport checks the characters in options that define the
3883 message's headers; it allows continued headers, but it was checking with
3884 isspace() after an embedded newline instead of explicitly looking for a
3887 21. If all the "regular" hosts to which an address was routed had passed their
3888 expiry times, and had not reached their retry times, the address was
3889 bounced, even if fallback hosts were defined. Now Exim should go on to try
3892 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3893 equivalent code in the SMTP transport. Some hosts send humungous responses
3894 to HELO/EHLO, more than 1024 it seems.
3896 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3897 gives compiler warnings in some environments. I've now done it "properly",
3900 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3901 (because of problems with the built-in one) was declared to return uschar *
3902 instead of char *, causing compiler failure.
3904 25. Fixed a file descriptor leak when processing alias/forward files.
3906 26. Fixed a minor format string issue in dbfn.c.
3908 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3910 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3911 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3914 29. When a sender address is verified, it is cached, to save repeating the test
3915 when there is more than one recipient in a message. However, when the
3916 verification involves a callout, it is possible for different callout
3917 options to be set for different recipients. It is too complicated to keep
3918 track of this in the cache, so now Exim always runs a verification when a
3919 callout is required, relying on the callout cache for the optimization.
3920 The overhead is duplication of the address routing, but this should not be
3923 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3924 to be verified with callout=postmaster, and the main callout worked but the
3925 postmaster check failed, the verification correctly failed. However, if a
3926 subsequent RCPT command asked for sender verification *without* the
3927 postmaster check, incorrect caching caused this verification also to fail,
3930 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3931 it was not caching the DNS options (qualify_single, search_parents) that
3932 were used when the lookup failed. A subsequent lookup with different
3933 options therefore always gave the same answer, though there were cases
3934 where it should not have. (Example: a "domains = !$mx_any" option on a
3935 dnslookup router: the "domains" option is always processed without any
3936 widening, but the router might have qualify_single set.) Now Exim uses the
3937 cached value only when the same options are set.
3939 32. Added John Jetmore's "exipick" utility to the distribution.
3941 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3942 than a timeout (e.g. a certificate is required, and is not provided), an
3943 Exim server now closes the connection immediately. Previously it waited for
3944 the client to close - but if the client is SSL, it seems that they each
3945 wait for each other, leading to a delay before one of them times out.
3947 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3948 maintained 0.8.x compatibility because I don't think many are using it, and
3949 it is clearly obsolete.
3951 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3954 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3955 client certificate was expired. A simple patch fixes this, though I don't
3956 understand the full logic of why the verify callback is called multiple
3959 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3960 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3961 which causes problems with some clients (such as the Certicom SSL Plus
3962 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3963 disables the coutermeasure allowing Eudora to connect."
3965 38. Exim was not checking that a write() to a log file succeeded. This could
3966 lead to Bad Things if a log got too big, in particular if it hit a file
3967 size limit. Exim now panics and dies if it cannot write to a log file, just
3968 as it does if it cannot open a log file.
3970 39. Modified OS/Makefile-Linux so that it now contains
3972 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3974 The two -D definitions ensure that Exim is compiled with large file
3975 support, which makes it possible to handle log files that are bigger than
3978 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3979 instance) a domain was checked against a named list that involved a lookup,
3980 causing $domain_data to be set, then another domain was checked against the
3981 same list, then the first domain was re-checked, the value of $domain_data
3982 after the final check could be wrong. In particular, if the second check
3983 failed, it could be set empty. This bug probably also applied to
3986 41. The strip_trailing_dot option was not being applied to the address given
3987 with the -f command-line option.
3989 42. The code for reading a message's header from the spool was incrementing
3990 $received_count, but never initializing it. This meant that the value was
3991 incorrect (doubled) while delivering a message in the same process in which
3992 it was received. In the most common configuration of Exim, this never
3993 happens - a fresh exec is done - but it can happen when
3994 deliver_drop_privilege is set.
3996 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3997 it now includes up to 150 characters of the unexpected data in the log
4000 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4001 and building data strings. The size of both of these buffers was 10 000
4002 bytes - far larger than anybody would *ever* want, thought I. Needless to
4003 say, somebody hit the limit. I have increased the maximum line length to
4004 20 000 and the maximum data length of concatenated lines to 100 000. I have
4005 also fixed two bugs, because there was no checking on these buffers. Tsk,
4006 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4007 buffer is too small.
4009 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4010 lsearch lookups. Now it does.
4012 46. When parsing a route_list item in a manualroute router, a fixed-length
4013 buffer was used for the list of hosts. I made this 1024 bytes long,
4014 thinking that nobody would ever have a list of hosts that long. Wrong.
4015 Somebody had a whole pile of complicated expansion conditions, and the
4016 string was silently truncated, leading to an expansion error. It turns out
4017 that it is easier to change to an unlimited length (owing to other changes
4018 that have happened since this code was originally written) than to build
4019 structure for giving a limitation error. The length of the item that
4020 expands into the list of hosts is now unlimited.
4022 47. The lsearch lookup could not handle data where the length of text line was
4023 more than 4095 characters. Such lines were truncated, leading to shortened
4024 data being returned. It should now handle lines of any length.
4026 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4027 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4030 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4031 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4032 no longer happens in comment lines. A list of replacements is now placed
4033 at the head of all of the source files, except those whose only change is
4034 to replace PERL_COMMAND in the very first #! line.
4036 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4037 messages on the queue, with a bottom-up merge sort, using code contributed
4038 by Michael Haardt. This should make operations like -bp somewhat faster on
4039 large queues. It won't affect queue runners, except when queue_run_in_order
4042 51. Installed eximstats 1.31 in the distribution.
4044 52. Added support for SRV lookups to the dnslookup router.
4046 53. If an ACL referred to $message_body or $message_body_end, the value was not
4047 reset for any messages that followed in the same SMTP session.
4049 54. The store-handling optimization for building very long strings was not
4050 differentiating between the different store pools. I don't think this
4051 actually made any difference in practice, but I've tidied it.
4053 55. While running the routers to verify a sender address, $sender_address
4054 was still set to the sender address. This is wrong, because when routing to
4055 send a bounce to the sender, it would be empty. Therefore, I have changed
4056 it so that, while verifying a sender address, $sender_address is set to <>.
4057 (There is no change to what happens when verifying a recipient address.)
4059 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4060 target A or AAAA records (if not already returned) without resetting the
4061 qualify_single or search_parents options of the DNS resolver. These are
4062 inappropriate in this case because the targets of MX and SRV records must
4063 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4064 target that, when qualified, matched something in the local domain. These
4065 two options are now turned off when doing these lookups.
4067 57. It seems that at least some releases of Reiserfs (which does not have the
4068 concept of a fixed number of inodes) returns zero and not -1 for the
4069 number of available inodes. This interacted badly with check_spool_inodes,
4070 which assumed that -1 was the "no such thing" setting. What I have done is
4071 to check that the total number of inodes is greater than zero before doing
4072 the test of how many are available.
4074 58. When a "warn" ACL statement has a log_message modifier, the message is
4075 remembered, and not repeated. This is to avoid a lot of repetition when a
4076 message has many recipients that cause the same warning to be written.
4077 Howewer, Exim was preserving the list of already written lines for an
4078 entire SMTP session, which doesn't seem right. The memory is now reset if a
4079 new message is started.
4081 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4082 debugging output unless log_rewrite was also set.
4084 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4085 of (int)(handle) when we know that handle contains (void *)(-1).
4087 61. The Exim daemon panic-logs an error return when it closes the incoming
4088 connection. However "connection reset by peer" seems to be common, and
4089 isn't really an error worthy of noting specially, so that particular error
4092 62. When Exim is trying to find all the local interfaces, it used to panic and
4093 die if the ioctl to get the interface flags failed. However, it seems that
4094 on at least one OS (Solaris 9) it is possible to have an interface that is
4095 included in the list of interfaces, but for which you get a failure error
4096 for this call. This happens when the interface is not "plumbed" into a
4097 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4098 failure of the "get flags" call assumes that the interface is down.
4100 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4101 makes life easier for people who are doing arithmetic on fields extracted
4102 from dates, where you often get leading zeros that should not be
4103 interpreted as octal.
4105 64. Added qualify_domain to the redirect router, to override the global
4108 65. If a pathologically long header line contained very many addresses (the
4109 report of this problem mentioned 10 000) and each of them was rewritten,
4110 Exim could use up a very large amount of memory. (It kept on making new
4111 copies of the header line as it rewrote, and never released the old ones.)
4112 At the expense of a bit more processing, the header rewriting function has
4113 been changed so that it no longer eats memory in this way.
4115 66. The generation of the Received: header has been moved from the time that a
4116 message starts to be received, to the time that it finishes. The timestamp
4117 in the Received: header should now be very close to that of the <= log
4118 line. There are two side-effects of this change:
4120 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4121 logged header lines no longer include the local Received: line, because
4122 it has not yet been created. The same applies to a copy of the message
4123 that is returned to a non-SMTP sender when a message is rejected.
4125 (b) When a filter file is tested using -bf, no additional Received: header
4126 is added to the test message. After some thought, I decided that this
4129 This change does not affect the value of $received_for. It is still set
4130 after address rewriting, but before local_scan() is called.
4132 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4134 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4135 gave an unhelpful panic error message, and a defer error. I have managed to
4136 change this behaviour so that it now rejects any supplied certificate,
4137 which seems right, as the list of acceptable certificates is empty.
4139 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4140 gave an unhelpful defer error. I have not managed to make this reject any
4141 supplied certificates, but the error message it gives is "no certificate
4142 supplied", which is not helpful.
4144 70. exigrep's output now also includes lines that are not associated with any
4145 message, but which match the given pattern. Implemented by a patch from
4146 Martin Sluka, which also tidied up the Perl a bit.
4148 71. Recipient callout verification, like sender verification, was using <> in
4149 the MAIL FROM command. This isn't really the right thing, since the actual
4150 sender may affect whether the remote host accepts the recipient or not. I
4151 have changed it to use the actual sender in the callout; this means that
4152 the cache record is now keyed on a recipient/sender pair, not just the
4153 recipient address. There doesn't seem to be a real danger of callout loops,
4154 since a callout by the remote host to check the sender would use <>.
4155 [SEE ABOVE: changed after hitting problems.]
4157 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4158 temporary errors. However, in the case of such a code being given after
4159 the end of a data transmission (i.e. after ".") Exim was failing to write
4160 a retry record for the message. (Yes, there was some broken host that was
4161 actually sending 8xx at this point.)
4163 73. An unknown lookup type in a host list could cause Exim to panic-die when
4164 the list was checked. (An example that provoked this was putting <; in the
4165 middle of a list instead of at the start.) If this happened during a DATA
4166 ACL check, a -D file could be left lying around. This kind of configuration
4167 error no longer causes Exim to die; instead it causes a defer errror. The
4168 incident is still logged to the main and panic logs.
4170 74. Buglet left over from Exim 3 conversion. The message "too many messages
4171 in one connection" was written to the rejectlog but not the mainlog, except
4172 when address rewriting (yes!) was being logged.
4174 75. Added write_rejectlog option.
4176 76. When a system filter was run not as root (that is, when system_filter_user
4177 was set), the values of the $n variables were not being returned to the
4178 main process; thus, they were not subsequently available in the $sn
4181 77. Added +return_path_on_delivery log selector.
4183 78. A connection timeout was being treated differently from recipients deferred
4184 when testing hosts_max_try with a message that was older than the host's
4185 retry timeout. (The host should not be counted, thus allowing all hosts to
4186 be tried at least once before bouncing.) This may have been the cause of an
4187 occasionally reported bug whereby a message would remain on the queue
4188 longer than the retry timeout, but would be bounced if a delivery was
4189 forced. I say "may" because I never totally pinned down the problem;
4190 setting up timeout/retry tests is difficult. See also the next item.
4192 79. The ultimate address timeout was not being applied to errors that involved
4193 a combination of host plus message (for example, a timeout on a MAIL
4194 command). When an address resolved to a number of possible hosts, and they
4195 were not all tried for each delivery (e.g. because of hosts_max_try), a
4196 message could remain on the queue longer than the retry timeout.
4198 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4201 81. Fixed an obscure SMTP outgoing bug which required at least the following
4202 conditions: (a) there was another message waiting for the same server;
4203 (b) the server returned 5xx to all RCPT commands in the first message so
4204 that the message was not completed; (c) the server dropped the connection
4205 or gave a negative response to the RSET that Exim sends to abort the
4206 transaction. The observed case was a dropped connection after DATA that had
4207 been sent in pipelining mode. That is, the server had advertised PIPELINING
4208 but was not implementing it correctly. The effect of the bug was incorrect
4209 behaviour, such as trying another host, and this could lead to a crash.
4215 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4216 and daemon.c were passed as pointers to ints; they should have been
4217 pointers to socklen_t variables (which are typically unsigned ints).
4219 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4222 3. Fixed a really odd bug that affected only the testing scheme; patching a
4223 certain fixed string in the binary changed the value of another string that
4224 happened to be identical to the end of the original first string.
4226 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4227 name", it returns that address as the IP address. On some operating
4228 systems (e.g. Solaris), it also passes back the IP address string as the
4229 "host name". However, on others (e.g. Linux), it passes back an empty
4230 string. Exim wasn't checking for this, and was changing the host name to an
4231 empty string, assuming it had been canonicized.
4233 5. Although rare, it is permitted to have more than one PTR record for a given
4234 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4235 all the names associated with an address, because they do in Solaris.
4236 However, it seems that they do not in Linux for data that comes from the
4237 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4238 I found this out when I moved to a new Linux workstation and tried to run
4239 the Exim test suite.
4241 To get round this problem I have changed the code so that it now does its
4242 own call to the DNS to look up PTR records when searching for a host name.
4243 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4244 addresses that are only in /etc/hosts are still found.
4246 This behaviour is, however, controlled by an option called host_lookup_
4247 order, which defaults to "bydns:byaddr". If people want to use the other
4248 order, or indeed, just use one or the other means of lookup, they can
4249 specify it in this variable.
4251 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4252 some operating systems, this comes back from gethostbyaddr() as an empty
4253 string, and this is what Exim used to test for. However, it seems that in
4254 other systems, "." is yielded. Exim now tests for this case too.
4256 7. The values of check_spool_space and check_log_space are now held internally
4257 as a number of kilobytes instead of an absolute number of bytes. If a
4258 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4259 kilobyte. This means that much larger values can be stored.
4261 8. Exim monitor: an attempt to get the action menu when not actually pointing
4262 at a message produces an empty menu entitled "No message selected". This
4263 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4264 no entries in it ("Shell widget menu has zero width and/or height"). So I
4265 have added a single, blank menu entry in this case.
4267 9. Added ${quote_local_part.
4269 10. MIME decoding is now applied to the contents of Subject: header lines when
4272 11. Now that a reference to $sender_host_address automatically causes a reverse
4273 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4274 host lookup before query-style lookups in lists that might use this
4275 variable. This has therefore been abolished, and the "net-" prefix is no
4276 longer necessary for query-style lookups.
4278 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4279 have been a typo for LFLAGS, so it has been changed.
4281 13. The install script calls Exim with "-C /dev/null" in order to find the
4282 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4283 to be output. Howeve, since Exim outputs its version number before the
4284 error, it didn't break the script. It just looked ugly. I fixed this by
4285 always allowing "-C /dev/null" if the caller is root.
4287 14. Ignore overlarge ACL variable number when reading spool file - insurance
4288 against a later release with more variables having written the file.
4290 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4291 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4292 this, as well as the form without the "IPv6" on the front.
4294 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4295 OS/Makefile-Darwin file.
4297 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4298 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4300 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4301 variable. However, some OS use ints for inodes. I've added cast to long int
4302 to get rid of the compiler warning.
4304 19. I had forgotten to lock out "/../" in configuration file names when
4305 ALT_CONFIG_PREFIX was set.
4307 20. Routers used for verification do not need to specify transports. However,
4308 if such a router generated a host list, and callout was configured, Exim
4309 crashed, because it could not find a port number from the (non-existent)
4310 transport. It now assumes port 25 in this circumstance.
4312 21. Added the -t option to exigrep.
4314 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4315 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4316 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4317 them is set, LOOKUP_LSEARCH is forced.
4319 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4321 24. Added sender and host information to the "rejected by local_scan()" log
4322 line; previously there was no indication of these.
4324 25. Added .include_if_exists.
4326 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4327 Linux. It turns out that not all file systems support this. Apparently some
4328 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4329 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4330 not supported on the file descriptor, is now ignored when Exim is trying to
4331 sync a directory. This applies only to Linux.
4333 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4335 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4336 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4339 29. Added support for maildirsize files from supplied patch (modified a bit).
4341 30. The use of :fail: followed by an empty string could lead Exim to respond to
4342 sender verification failures with (e.g.):
4344 550 Verification failed for <xxx>
4345 550 Sender verify failed
4347 where the first response line was missing the '-' that indicates it is not
4348 the final line of the response.
4350 31. The loop for finding the name of the user that called Exim had a hardwired
4351 limit of 10; it now uses the value of finduser_retries, which is used for
4352 all other user lookups.
4354 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4357 33. Exim was neglecting to zero errno before one call of strtol() when
4358 expanding a string and expecting an integer value. On some systems this
4359 resulted in spurious "integer overflow" errors. Also, it was casting the
4360 result into an int without checking.
4362 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4363 did not work. The code looks as if it has *never* worked, though it appears
4364 to have been documented since at least releast 1.62. I have made it work.
4366 35. The "timeout_DNS" error in retry rules, also documented since at least
4367 1.62, also never worked. As it isn't clear exactly what this means, and
4368 clearly it isn't a major issue, I have abolished the feature by treating it
4369 as "timeout", and writing a warning to the main and panic logs.
4371 36. The display of retry rules for -brt wasn't always showing the error code
4374 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4375 timeout_connect_A, timeout_connect_MX.
4377 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4378 to the empty sender.
4380 39. The daemon was not analysing the content of -oX till after it had closed
4381 stderr and disconnected from the controlling terminal. This meant that any
4382 syntax errors were only noted on the panic log, and the return code from
4383 the command was 0. By re-arranging the code a little, I've made the
4384 decoding happen first, so such errors now appear on stderr, and the return
4385 code is 1. However, the actual setting up of the sockets still happens in
4386 the disconnected process, so errors there are still only recorded on the
4389 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4390 connections (as happens on some IP stacks) was logged at start up time as
4391 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4392 it from "IPv6 and IPv4", which means that two separate sockets are being
4395 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4396 says whether AF_INET or AF_INET6 was passed as an argument.
4398 42. Exiwhat output was messed up when time zones were included in log
4401 43. Exiwhat now gives more information about the daemon's listening ports,
4402 and whether -tls-on-connect was used.
4404 44. The "port" option of the smtp transport is now expanded.
4406 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4407 silently ignored. Now an error message is written to the main and panic
4410 46. There's a new ACL modifier called "logwrite" which writes to a log file
4411 as soon as it is encountered.
4413 47. Added $local_user_uid and $local_user_gid at routing time.
4415 48. Exim crashed when trying to verify a sender address that was being
4418 49. Exim was recognizing only a space character after ".include". It now also
4419 recognizes a tab character.
4421 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4422 extracting the relevant information from the specification. The man page no
4423 longer contains scrambled data for the -d option, and I've added a section
4424 at the front about calling Exim under different names.
4426 51. Added "extra_headers" argument to the "mail" command in filter files.
4428 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4431 53. Installed eximstats 1.29.
4433 54. Added transport_filter_timeout as a generic transport option.
4435 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4436 Cc: header lines. This was required by RFC 822, but it not required by RFC
4439 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4440 incoming messages that did not have them. Now it does so only if the
4441 message originates locally, that is, if there is no associated remote host
4442 address. When Resent- header lines are present, this applies to the Resent-
4443 lines rather than the non-Resent- lines.
4445 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4446 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4448 58. Messages for configuration errors now include the name of the main
4449 configuration file - useful now that there may be more than one file in a
4450 list (.included file names were always shown).
4452 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4453 for those rare installations that do not start the daemon as root or run it
4454 setuid root. I've cut out the call to initgroups() if the daemon is not
4457 60. The Exim user and group can now be bound into the binary as text strings
4458 that are looked up at the start of Exim's processing.
4460 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4462 62. Added $mailstore_basename variable.
4464 63. Installed patch to sieve.c from Michael Haardt.
4466 64. When Exim failed to open the panic log after failing to open the main log,
4467 the original message it was trying to log was written to stderr and debug
4468 output, but if they were not available (the usual case in production), it
4469 was lost. Now it is written to syslog before the two lines that record the
4470 failures to open the logs.
4472 65. Users' Exim filters run in subprocesses under the user's uid. It is
4473 possible for a "deliver" command or an alias in a "personal" command to
4474 provoke an address rewrite. If logging of address rewriting is configured,
4475 this fails because the process is not running as root or exim. There may be
4476 a better way of dealing with this, but for the moment (because 4.30 needs
4477 to be released), I have disabled address rewrite logging when running a
4478 filter in a non-root, non-exim process.
4484 1. The buildconfig auxiliary program wasn't quoting the value set for
4485 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4486 not defined. This bug was masked in 4.22 by the effect that was fixed in
4489 2. Some messages that were rejected after a message id was allocated were
4490 shown as "incomplete" by exigrep. It no longer does this for messages that
4491 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4493 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4494 have allow_domain_literals set, the ID did not get logged in the <= line.
4495 Domain literals are now always recognized in Message-ID: header lines.
4497 4. The first argument for a ${extract expansion item is the key name or field
4498 number. Leading and trailing spaces in this item were not being ignored,
4499 causing some misleading effects.
4501 5. When deliver_drop_privilege was set, single queue runner processes started
4502 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4503 same command in the process it spins off) were not dropping privilege.
4505 6. When the daemon running as "exim" started a queue runner, it always
4506 re-executed Exim in the spun-off process. This is a waste of effort when
4507 deliver_drop_privilege is set. The new process now just calls the
4508 queue-runner function directly.
4514 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4517 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4518 ignored. Though the use of -oP was forcing the writing of a pid file, it
4519 was always written to the default place.
4521 3. If the message "no IP address found for host xxxx" is generated during
4522 incoming verification, it is now followed by identification of the incoming
4523 connection (so you can more easily find what provoked it).
4525 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4527 5. Added some features to "harden" Exim a bit more against certain attacks:
4529 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4530 be put in Local/Makefile. This is like the never_users runtime option,
4531 but it cannot be overridden. The default setting is "root".
4533 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4534 prefix string with which any file named in a -C command line option
4537 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4538 is retained for -C and -D only if the caller of Exim is root. Without
4539 it, the exim user may also use -C and -D and retain privilege.
4541 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4542 command line option is disabled.
4544 6. Macro names set by the -D option must start with an upper case letter, just
4545 like macro names defined in the configuration file.
4547 7. Added "dereference=" facility to LDAP.
4549 8. Two instances of the typo "uknown" in the source files are fixed.
4551 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4552 the Configure-Makefile script screwed up while processing it.
4554 10. Incorporated PCRE 4.4.
4556 11. The SMTP synchronization check was not operating right at the start of an
4557 SMTP session. For example, it could not catch a HELO sent before the client
4558 waited for the greeting. There is now a check for outstanding input at the
4559 point when the greeting is written. Because of the duplex, asynchronous
4560 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4561 way, but not yet received, when the check is performed.
4563 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4564 on TCP/IP sockets, because this apparently causes some broken clients to
4567 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4568 unchanged) from the Cygwin maintainer.
4570 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4571 was supported instead of testing for mbx. Effectively a typo.
4573 15. The spa authenticator server code was not checking that the input it
4574 received was valid base64.
4576 16. The debug output line for the "set" modifier in ACLs was not showing the
4577 name of the variable that was being set.
4579 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4581 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4582 lookup on its own. Something else was needed to trigger the lookup. For
4583 example, a match in host_lookup or the need for a host name in a host list.
4584 Now, if $sender_host_name is referenced and the host name has not yet been
4585 looked up, a lookup is performed. If the lookup fails, the variable remains
4586 empty, and $host_lookup_failed is set to "1".
4588 19. Added "eqi" as a case-independent comparison operator.
4590 20. The saslauthd authentication condition could segfault if neither service
4591 nor realm was specified.
4593 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4594 error message that was logged was misleading, and incoming SMTP
4595 connections were dropped. The message is now more accurate, and temporary
4596 errors are given to SMTP connections.
4598 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4599 (except RSET) with a 421 error, until QUIT is received. However, it was
4600 failing to send a response to QUIT.
4602 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4603 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4604 ACL is now run after the helo_try_verify_hosts code.
4606 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4607 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4608 case-independent, so other case variants are also recognized). Apparently
4609 some systems use these upper case variants.
4611 25. If more than two messages were waiting for the same host, and a transport
4612 filter was specified for the transport, Exim sent two messages over the
4613 same TCP/IP connection, and then failed with "socket operation on non-
4614 socket" when it tried to send the third.
4616 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4618 27. The extern definition of crypt16() in expand.c was not being excluded when
4619 the OS had its own crypt16() function.
4621 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4622 as a preferred synonym for return_size_limit, both as an option and as an
4625 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4627 30. Changed the default configuration ACL to relax the local part checking rule
4628 for addresses that are not in any local domains. For these addresses,
4629 slashes and pipe symbols are allowed within local parts, but the sequence
4630 /../ is explicitly forbidden.
4632 31. SPA server authentication was not clearing the challenge buffer before
4635 32. log_message in a "warn" ACL statement was writing to the reject log as
4636 well as to the main log, which contradicts the documentation and doesn't
4637 seem right (because no rejection is happening). So I have stopped it.
4639 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4640 However, I am unable to do any testing of this.
4642 34. Fixed an infelicity in the appendfile transport. When checking directories
4643 for a mailbox, to see if any needed to be created, it was accidentally
4644 using path names with one or more superfluous leading slashes; tracing
4645 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4647 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4648 contents are added to the log line that is written for every discarded
4649 recipient. (Previously a log_message setting was ignored.)
4651 36. The ${quote: operator now quotes the string if it is empty.
4653 37. The install script runs exim in order to find its version number. If for
4654 some reason other than non-existence or emptiness, which it checks, it
4655 could not run './exim', it was installing it with an empty version number,
4656 i.e. as "exim-". This error state is now caught, and the installation is
4659 38. An argument was missing from the function that creates an error message
4660 when Exim fails to connect to the socket for saslauthd authentication.
4661 This could cause Exim to crash, or give a corrupted message.
4663 39. Added isip, isip4, and isip6 to ${if conditions.
4665 40. The ACL variables $acl_xx are now saved with the message, and can be
4666 accessed later in routers, transports, and filters.
4668 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4669 strings in the file are not string-expanded.
4671 42. If a MAIL command specified a SIZE value that was too large to fit into an
4672 int variable, the check against message_size_limit failed. Such values are
4673 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4674 day this will have to be increased, but I don't think I want to be around
4675 when emails are that large.
4682 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4683 iconv() is not standard in FreeBSD.
4685 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4686 IPv6 enabled. The observed symptom was a segmentation fault on return from
4687 the function os_common_find_running_interfaces() in src/os.c.
4689 3. In the check_special_case() function in daemon.c I had used "errno" as an
4690 argument name, which causes warnings on some systems. This was basically a
4691 typo, since it was named "eno" in the comments!
4693 4. The code that waits for the clock to tick (at a resolution of some fraction
4694 of a second) so as to ensure message-id uniqueness was always waiting for
4695 at least one whole tick, when it could have waited for less. [This is
4696 almost certainly not relevant at current processor speeds, where it is
4697 unlikely to ever wait at all. But we try to future-proof.]
4699 5. The function that sleeps for a time interval that includes fractions of a
4700 second contained a race. It did not block SIGALRM between setting the
4701 timer, and suspending (a couple of lines later). If the interval was short
4702 and the sigsuspend() was delayed until after it had expired, the suspension
4703 never ended. On busy systems this could lead to processes getting stuck for
4706 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4707 process, before it forks any delivery processes. The open lookup caching
4708 mechanism meant that the open file or database connection was passed into
4709 the delivery process. The problem was that delivery processes always tidy
4710 up cached lookup data. This could cause a problem for the next delivery
4711 process started by the queue runner, because the external queue runner
4712 process does not know about the closure. So the next delivery process
4713 still has data in the lookup cache. In the case of a file lookup, there was
4714 no problem because closing a file descriptor in a subprocess doesn't affect
4715 the parent. However, if the lookup was caching a connection to a database,
4716 the connection was closed, and the second delivery process was likely to
4717 see errors such as "PGSQL: query failed: server closed the connection
4718 unexpectedly". The problem has been fixed by closing all cached lookups
4719 in a queue runner before running a delivery process.
4721 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4722 seem to have the "const" qualifier which it has on other OS. I've
4725 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4726 *of the same type* (client or server) with the same public name that an
4727 error should be diagnosed.
4729 9. When Exim looked up a host name for an IP address, but failed to find the
4730 original IP address when looking up the host name (a safety check), it
4731 output the message "<ip address> does not match any IP for NULL", which was
4732 confusing, to say the least. The bug was that the host name should have
4733 appeared instead of "NULL".
4735 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4736 user that is built into the binary, and the -C or -D options is used, root
4737 privilege is dropped before the configuration file is read. In addition,
4738 logging is switched to stderr instead of the normal log files. If the
4739 configuration then re-defines the Exim user, the unprivileged environment
4740 is probably not what is expected, so Exim logs a panic warning message (but
4743 However, if deliver_drop_privilege is set, the unprivileged state may well
4744 be exactly what is intended, so the warning has been cut out in that case,
4745 and Exim is allowed to try to write to its normal log files.
4751 1. smtp_return_error_details was not giving details for temporary sender
4752 or receiver verification errors.
4754 2. Diagnose a configuration error if two authenticators have the same public
4757 3. Exim used not to create the message log file for a message until the first
4758 delivery attempt. This could be confusing when incoming messages were held
4759 for policy or load reasons. The message log file is now created at the time
4760 the message is received, and an initial "Received" line is written to it.
4762 4. The automatically generated man page for command line options had a minor
4763 bug that caused no ill effects; however, a more serious problem was that
4764 the procedure for building the man page automatically didn't always
4765 operate. Consequently, release 4.20 contains an out-of-date version. This
4766 shouldn't happen again.
4768 5. When building Exim with embedded Perl support, the script that builds the
4769 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4770 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4772 6. The freeze_tell option was not being used for messages that were frozen on
4773 arrival, either by an ACL or by local_scan().
4775 7. Added the smtp_incomplete_transaction log selector.
4777 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4778 was accepting AUTH without a new EHLO.
4780 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4781 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4784 10. Logging of TCP/IP connections (when configured) now happens in the main
4785 daemon process instead of the child process, so that the TCP/IP connection
4786 count is more accurate (but it can never be perfect).
4788 11. The use of "drop" in a nested ACL was not being handled correctly in the
4789 outer ACL. Now, if condition failure induced by the nested "drop" causes
4790 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4791 or "require"), the connection is dropped.
4793 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4794 that yield "discard" can now be used with an "accept" or a "discard" verb,
4795 but an error is generated for any others (because I can't see a useful way
4796 to define what should happen).
4798 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4799 are now processed in the same way as lines in the Exim configuration file.
4800 In particular, continuation lines are supported.
4802 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4804 15. Added -ti meaning -t -i.
4806 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4807 domains, and warn by logging if others are found.
4809 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4810 returned from GIFCONF when Exim is trying to find the list of interfaces on
4811 a host. The code in os.c has been modified to copy each ifreq to an aligned
4812 structure in all cases.
4814 Also, in some cases, the returned ifreq's were being copied to a 'struct
4815 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4816 means the last couple of bytes of an IPv6 address could be chopped if the
4817 ifreq contained only a normal sockaddr (14 bytes storage).
4819 18. Named domain lists were not supported in the hosts_treat_as_local option.
4820 An entry such as +xxxx was not recognized, and was treated as a literal
4823 19. Ensure that header lines added by a DATA ACL are included in the reject log
4824 if the ACL subsequently rejects the message.
4826 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4827 MD5 (which is deprecated).
4829 21. When testing a filter file using -bf, Exim was writing a message when it
4830 took the sender from a "From " line in the message, but it was not doing so
4831 when it took $return_path from a Return-Path: header line. It now does.
4833 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4834 with a valid header line field name (a series of printing characters
4835 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4837 23. Changed "disc" in the source to "disk" to conform to the documentation and
4838 the book and for uniformity.
4840 24. Ignore Sendmail's -Ooption=value command line item.
4842 25. When execve() failed while trying to run a command in a pipe transport,
4843 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4844 could be confused with a return value of 69 from the command itself. This
4845 has been changed to 127, the value the shell returns if it is asked to run
4846 a non-existent command. The wording for the related log line suggests a
4847 non-existent command as the problem.
4849 26. If received_header_text expands to an empty string, do not add a Received:
4850 header line to the message. (Well, it adds a token one on the spool, but
4851 marks it "old" so that it doesn't get used or transmitted.)
4853 27. Installed eximstats 1.28 (addition of -nt option).
4855 28. There was no check for failure on the call to getsockname() in the daemon
4856 code. This can fail if there is a shortage of resources on the system, with
4857 ENOMEM, for example. A temporary error is now given on failure.
4859 29. Contrary to the C standard, it seems that in some environments, the
4860 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4861 program. Exim now does this explicitly; it affects the formatting of
4862 timestamps using strftime().
4864 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4865 complaints. I've now initialized all the variables, to avoid this.
4867 32. Header lines added by a system filter were not being "seen" during
4868 transport-time rewrites.
4870 33. The info_callback() function passed to OpenSSL is set up with type void
4871 (*)(SSL *, int, int), as described somewhere. However, when calling the
4872 function (actually a macro) that sets it up, the type void(*)() is
4873 expected. I've put in a cast to prevent warnings from picky compilers.
4875 34. If a DNS black list lookup found a CNAME record, but there were no A
4876 records associated with the domain it pointed at, Exim crashed.
4878 35. If a DNS black list lookup returned more than one A record, Exim ignored
4879 all but the first. It now scans all returned addresses if a particular IP
4880 value is being sought. In this situation, the contents of the
4881 $dnslist_value variable are a list of all the addresses, separated by a
4884 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4885 to accept a host name and all its aliases if the forward lookup for any of
4886 them yielded the IP address of the incoming connection. Now it accepts only
4887 those names whose forward lookup yields the correct IP address. Any other
4888 names are discarded. This closes a loophole whereby a rogue DNS
4889 administrator could create reverse DNS records to break through a
4890 wildcarded host restriction in an ACL.
4892 37. If a user filter or a system filter that ran in a subprocess used any of
4893 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4894 the wrong values were passed to the pipe command ($thisaddress had the
4895 value of $0, $0 had the value of $1, etc). This bug was introduced by
4896 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4898 38. Improved the line breaking for long SMTP error messages from ACLs.
4899 Previously, if there was no break point between 40 and 75 characters, Exim
4900 left the rest of the message alone. Two changes have been made: (a) I've
4901 reduced the minimum length to 35 characters; (b) if it can't find a break
4902 point between 35 and 75 characters, it looks ahead and uses the first one
4903 that it finds. This may give the occasional overlong line, but at least the
4904 remaining text gets split now.
4906 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4907 file descriptors might be low, and that setting 1000 would always raise it.
4908 It turns out that in some environments, the limit is already over 1000 and
4909 that lowering it causes trouble. So now Exim takes care not to decrease it.
4911 40. When delivering a message, the value of $return_path is set to $sender_
4912 address at the start of routing (routers may change the value). By an
4913 oversight, this default was not being set up when an address was tested by
4914 -bt or -bv, which affected the outcome if any router or filter referred to
4917 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4918 log only when "message" or "log_message" are set. However, if one of the
4919 conditions was an address verification, or a call to a nested ACL, the
4920 messages generated by the underlying test were being passed through. This
4921 no longer happens. The underlying message is available in $acl_verify_
4922 message for both "message" and "log_message" expansions, so it can be
4923 passed through if needed.
4925 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4926 new expansion $bh_ to give the encoded byte string without charset
4927 translation. Translation happens only if iconv() is available; HAVE_ICONV
4928 indicates this at build time. HEADERS_CHARSET gives the charset to
4929 translate to; headers_charset can change it in the configuration, and
4930 "headers charset" can change it in an individual filter file.
4932 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4933 that creates RFC 2047 encoded "words" labels them as that charset instead
4934 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4935 expansion operator; (ii) when Exim creates a From: line for a local
4936 message; (iii) when a header line is rewritten to include a "phrase" part.
4938 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4939 buggy, causing it to skip the first lines of messages whose message ID
4940 ended in 'D'. This would not have bitten before Exim release 4.14, because
4941 message IDs were unlikely to end in 'D' before then. The effect was to have
4942 incorrect size information for certain domains.
4944 45. #include "config.h" was missing at the start of the crypt16.c module. This
4945 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4948 46. If there was a timeout during a "random" callout check, Exim treated it as
4949 a failure of the random address, and carried on sending RSET and the real
4950 address. If the delay was just some slowness somewhere, the response to the
4951 original RCPT would be taken as a response to RSET and so on, causing
4952 mayhem of various kinds.
4954 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4955 when I implemented it. It didn't allow for the fact that some option values
4956 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4957 the right test for positive values.
4959 48. Domain names in DNS records are case-independent. Exim always looks them up
4960 in lower case. Some resolvers return domain names in exactly the case they
4961 appear in the zone file, that is, they may contain uppercase letters. Not
4962 all resolvers do this - some return always lower case. Exim was treating a
4963 change of case by a resolver as a change of domain, similar to a widening
4964 of a domain abbreviation. This triggered its re-routing code and so it was
4965 trying to route what was effectively the same domain again. This normally
4966 caused routing to fail (because the router wouldn't handle the domain
4967 twice). Now Exim checks for this case specially, and just changes the
4968 casing of the domain that it ultimately uses when it transmits the message
4971 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4974 50. If a filter generated a file delivery with a non-absolute name (possible if
4975 no home directory exists for the router), the forbid_file option was not
4978 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4979 the existing equality matching.
4981 52. Exim was using ints instead of ino_t variables in some places where it was
4982 dealing with inode numbers.
4984 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4985 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4986 called TMPDIR, and if it finds it is different, it changes its value.
4988 54. The smtp_printf() function is now made available to local_scan() so
4989 additional output lines can be written before returning. There is also an
4990 smtp_fflush() function to enable the detection of a dropped connection.
4991 The variables smtp_input and smtp_batched_input are exported to
4994 55. Changed the default runtime configuration: the message "Unknown user"
4995 has been removed from the ACL, and instead placed on the localuser router,
4996 using the cannot_route_message feature. This means that any verification
4997 failures that generate their own messages won't get overridden. Similarly,
4998 the "Unrouteable address" message that was in the ACL for unverifiable
4999 relay addresses has also been removed.
5001 56. Added hosts_avoid_esmtp to the smtp transport.
5003 57. The exicyclog script was not checking for the esoteric option
5004 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5005 will work only if exicyclog is run under the appropriate euid.
5007 58. Following a discussion on the list, the rules by which Exim recognises line
5008 endings on incoming messages have been changed. The -dropcr and drop_cr
5009 options are now no-ops, retained only for backwards compatibility. The
5010 following line terminators are recognized: LF CRLF CR. However, special
5011 processing applies to CR:
5013 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5014 nor a local message in the state where . is a terminator.
5016 (ii) If a bare CR is encountered in a header line, an extra space is added
5017 after the line terminator so as not to end the header. The reasoning
5018 behind this is that bare CRs in header lines are most likely either
5019 to be mistakes, or people trying to play silly games.
5021 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5022 was being incorrectly given as 18 bytes larger than it should have been.
5023 This is a VOB (very old bug).
5025 60. This may never have affected anything current, but just in case it has:
5026 When the local host is found other than at the start of a list of hosts,
5027 the local host, those with the same MX, and any that follow, are discarded.
5028 When the list in question was part of a longer list of hosts, the following
5029 hosts (not currently being processed) were also being discarded. This no
5030 longer happens. I'm not sure if this situation could ever has previously
5033 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5036 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5037 header. This is something that is recommended in a new Internet Draft, and
5038 is something that is documented as being done by Sendmail. There are two
5039 possible values. For messages generated by the autoreply transport, Exim
5042 Auto-Submitted: auto-replied
5044 whereas for all other generated messages (e.g. bounces) it adds
5046 Auto-Submitted: auto-generated
5048 63. The "personal" condition in filters now includes a test for the
5049 Auto-Submitted: header. If it contains the string "auto-" the message it
5050 not considered personal.
5052 64. Added rcpt_include_affixes as a generic transport option.
5054 65. Added queue_only_override (default true).
5056 66. Added the syslog_duplication option.
5058 67. If what should have been the first header line of a message consisted of
5059 a space followed by a colon, Exim was mis-interpreting it as a header line.
5060 It isn't of course - it is syntactically invalid and should therefore be
5061 treated as the start of the message body. The misbehaviour could have
5062 caused a number of strange effects, including loss of data in subsequent
5063 header lines, and spool format errors.
5065 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5066 client host had authenticated. This control can now be exercised by an ACL
5067 for more flexibility.
5069 69. By default, callouts do not happen when testing with -bh. There is now a
5070 variant, -bhc, which does actually run the callout code, including
5071 consulting and updating the callout cache.
5073 70. Added support for saslauthd authentication, courtesy of Alexander
5076 71. If statvfs() failed on the spool or log directories while checking their
5077 size for availability, Exim confusingly gave the error "space shortage".
5078 Furthermore, in debugging mode it crashed with a floating point exception.
5079 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5080 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5081 serious problem, Exim now writes to the main and panic logs when this
5082 happens, with details of the failure. It then refuses to accept the
5083 incoming message, giving the message "spool directory problem" or "log
5084 directory problem" with a 421 code for SMTP messages.
5086 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5087 0, 1, and 2 exist, because some OS complain for execs without them (see
5088 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5089 descriptors. However, the code omitted to check that the open succeeded,
5090 causing mysterious errors if for some reason the permissions on /dev/null
5091 got screwed. Now Exim writes a message to the main and panic logs, and
5092 bombs out if it can't open /dev/null.
5094 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5095 interact so that it is all more flexible. It is supposed to remain
5096 backwards compatible. Also added extra_local_interfaces.
5098 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5099 to bomb out with an assertion failure - to the client this appears as a
5100 connection drop. This problem occurs in the part of the code that was taken
5101 from the Samba project. Fortunately, the assertion is in a very simple
5102 function, so I have fixed this by reproducing the function inline in the
5103 one place where it is called, and arranging for authentication to fail
5104 instead of killing the process with assert().
5106 75. The SPA client code was not working when the server requested OEM rather
5107 than Unicode encoding.
5109 76. Added code to make require_files with a specific uid setting more usable in
5110 the case where statting the file as root fails - usually a non-root-mounted
5111 NFS file system. When this happens and the failure is EACCES, Exim now
5112 forks a subprocess and does the per-uid checking as the relevant uid.
5114 77. Added process_log_path.
5116 78. If log_file_path was not explicitly set, a setting of check_log_space or
5117 check_log_inodes was ignored.
5119 79. If a space check for the spool or log partitions fails, the incident is now
5120 logged. Of course, in the latter case the data may get lost...
5122 80. Added the %p formatting code to string_format() so that it can be used to
5123 print addresses in debug_print(). Adjusted all the address printing in the
5124 debugging in store.c to use %p rather than %d.
5126 81. There was a concern that a line of code in smtp_in.c could overflow a
5127 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5128 initially expressed, the concern was not well-founded, because trailing
5129 spaces are removed early. However, if the trailing spaces were followed by
5130 a NULL, they did not get removed, so the overflow was possible. Two fixes
5133 (a) I re-wrote the offending code in a cleaner fashion.
5134 (b) If an incoming SMTP command contains a NULL character, it is rejected
5137 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5138 runs initgroups(), so that if the Exim user is in any additional groups,
5139 they will be used during message reception.
5145 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob