1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.627 2010/06/06 22:46:33 pdp Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PP/01 Date: & Message-Id: revert to normally being appended to a message,
10 only prepend for the Resent-* case. Fixes regression introduced in
11 Exim 4.70 by NM/22 for Bugzilla 607.
13 PP/02 Include check_rfc2047_length in configure.default because we're seeing
14 increasing numbers of administrators be bitten by this.
16 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
18 PP/03 Bugzilla 994: added openssl_options main configuration option.
20 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
22 PP/05 Bugzilla 834: provide a permit_codedump option for pipe transports.
24 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
26 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
27 without a peer certificate (I believe), leading to a segfault because of
28 an assumption that peers always have certificates. Be a little more
29 paranoid. Problem reported by Martin Tscholak.
31 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
32 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
33 NB: ClamAV planning to remove STREAM in "middle of 2010".
35 PP/09 Implemented reverse_ip expansion operator.
37 PP/10 Bugzilla 937: provide a "debug" ACL control.
39 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
41 PP/12 Bugzilla 973: Implement --version.
43 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
45 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
51 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
52 $data_path, and $header_path variables; fixed documentation bugs and
55 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
56 exipick to access non-standard spools, including the "frozen" queue
59 NM/01 Bugzilla 965: Support mysql stored procedures.
60 Patch from Alain Williams
62 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
64 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
65 Patch from Andreas Metzler
67 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
68 Patch from Kirill Miazine
70 NM/05 Bugzilla 671: Added umask to procmail example.
72 JJ/03 installed exipick 20100323.0, fixing doc bug
74 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
75 directory. Notification and patch from Dan Rosenberg.
77 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
79 TK/02 Improve log output when DKIM signing operation fails.
81 MH/01 Treat the transport option dkim_domain as a colon separated
82 list, not as a single string, and sign the message with each element,
83 omitting multiple occurences of the same signer.
85 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
86 Bugzilla 985, 986. Patch by Simon Arlott
88 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
91 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
92 MBX locking. Notification from Dan Rosenberg.
98 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
100 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
102 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
104 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
106 NM/04 Bugzilla 915: Flag broken perl installation during build.
112 TK/01 Added patch by Johannes Berg that expands the main option
113 "spamd_address" if it starts with a dollar sign.
115 TK/02 Write list of recipients to X-Envelope-Sender header when building
116 the mbox-format spool file for content scanning (suggested by Jakob
119 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
120 (http://www.dcc-servers.net/) support via dccifd. Activated by
121 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
123 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
124 by Mark Daniel Reidel <mr@df.eu>.
126 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
127 When building exim an external PCRE library is now needed -
128 PCRE is a system library on the majority of modern systems.
129 See entry on PCRE_LIBS in EDITME file.
131 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
132 conversation. Added nologin parameter to request.
133 Patch contributed by Kirill Miazine.
135 TF/01 Do not log submission mode rewrites if they do not change the address.
137 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
139 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
140 log files in place. Contributed by Roberto Lima.
142 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
144 TF/03 Bugzilla 615: When checking the local_parts router precondition
145 after a local_part_suffix or local_part_prefix option, Exim now
146 does not use the address's named list lookup cache, since this
147 contains cached lookups for the whole local part.
149 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
150 Robert Millan. Documentation is in experimental-spec.txt.
152 TF/04 Bugzilla 668: Fix parallel build (make -j).
154 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
156 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
157 Patch provided by Jan Srzednicki.
159 TF/05 Leading white space used to be stripped from $spam_report which
160 wrecked the formatting. Now it is preserved.
162 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
163 that they are available at delivery time.
165 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
167 TF/08 TLS error reporting now respects the incoming_interface and
168 incoming_port log selectors.
170 TF/09 Produce a more useful error message if an SMTP transport's hosts
171 setting expands to an empty string.
173 NM/06 Bugzilla 744: EXPN did not work under TLS.
174 Patch provided by Phil Pennock.
176 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
177 Patch provided by Richard Godbee.
179 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
180 acl_smtp_notquit, added index entry.
182 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
183 Patch provided by Eugene Bujak.
185 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
186 accept(). Patch provided by Maxim Dounin.
188 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
189 Patch provided by Phil Pennock.
191 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
193 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
194 Patch provided by Brad "anomie" Jorsch.
196 NM/14 Bugzilla 622: Added timeout setting to transport filter.
197 Patch provided by Dean Brooks.
199 TK/05 Add native DKIM support (does not depend on external libraries).
201 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
202 Patch provided by Graeme Fowler.
204 NM/16 Bugzilla 851: Documentation example syntax fix.
206 NM/17 Changed NOTICE file to remove references to embedded PCRE.
208 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
211 NM/19 Bugzilla 745: TLS version reporting.
212 Patch provided by Phil Pennock.
214 NM/20 Bugzilla 167: bool: condition support.
215 Patch provided by Phil Pennock.
217 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
218 clients. Patch provided by Phil Pennock.
220 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
221 Patch provided by Brad "anomie" Jorsch.
223 NM/23 Bugzilla 687: Fix misparses in eximstats.
224 Patch provided by Heiko Schlittermann.
226 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
227 Patch provided by Heiko Schlittermann.
229 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
230 plus update to original patch.
232 NM/26 Bugzilla 799: Documentation correction for ratelimit.
234 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
235 Patch provided by David Brownlee.
237 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
239 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
241 NM/30 Bugzilla 888: TLS documentation bugfixes.
243 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
245 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
246 Unlike the original bugzilla I have changed all shell scripts in src tree.
248 NM/33 Bugzilla 898: Transport filter timeout fix.
249 Patch by Todd Rinaldo.
251 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
252 Patch by Serge Demonchaux.
254 NM/35 Bugzilla 39: Base64 decode bug fixes.
255 Patch by Jakob Hirsch.
257 NM/36 Bugzilla 909: Correct connect() call in dcc code.
259 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
261 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
263 NM/39 Bugzilla 911: Fixed MakeLinks build script.
269 TK/01 Add preliminary DKIM support. Currently requires a forked version of
270 ALT-N's libdkim that I have put here:
271 http://duncanthrax.net/exim-experimental/
273 Note to Michael Haardt: I had to rename some vars in sieve.c. They
274 were called 'true' and it seems that C99 defines that as a reserved
275 keyword to be used with 'bool' variable types. That means you could
276 not include C99-style headers which use bools without triggering
277 build errors in sieve.c.
279 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
280 as mailq or other aliases. Changed the --help handling significantly
281 to do whats expected. exim_usage() emits usage/help information.
283 SC/01 Added the -bylocaldomain option to eximstats.
285 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
287 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
289 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
295 PH/01 Another patch from the Sieve maintainer.
297 PH/02 When an IPv6 address is converted to a string for single-key lookup
298 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
299 dots are used instead of colons so that keys in lsearch files need not
300 contain colons. This was done some time before quoting was made available
301 in lsearch files. However, iplsearch files do require colons in IPv6 keys
302 (notated using the quote facility) so as to distinguish them from IPv4
303 keys. This meant that lookups for IP addresses in host lists did not work
304 for iplsearch lookups.
306 This has been fixed by arranging for IPv6 addresses to be expressed with
307 colons if the lookup type is iplsearch. This is not incompatible, because
308 previously such lookups could never work.
310 The situation is now rather anomolous, since one *can* have colons in
311 ordinary lsearch keys. However, making the change in all cases is
312 incompatible and would probably break a number of configurations.
314 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
317 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
318 conversion specification without a maximum field width, thereby enabling
319 a rogue spamd server to cause a buffer overflow. While nobody in their
320 right mind would setup Exim to query an untrusted spamd server, an
321 attacker that gains access to a server running spamd could potentially
322 exploit this vulnerability to run arbitrary code as the Exim user.
324 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
325 $primary_hostname instead of what libspf2 thinks the hosts name is.
327 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
328 a directory entry by the name of the lookup key. Previously, if a
329 symlink pointed to a non-existing file or a file in a directory that
330 Exim lacked permissions to read, a lookup for a key matching that
331 symlink would fail. Now it is enough that a matching directory entry
332 exists, symlink or not. (Bugzilla 503.)
334 PH/03 The body_linecount and body_zerocount variables are now exported in the
337 PH/04 Added the $dnslist_matched variable.
339 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
340 This means they are set thereafter only if the connection becomes
343 PH/06 Added the client_condition to authenticators so that some can be skipped
344 by clients under certain conditions.
346 PH/07 The error message for a badly-placed control=no_multiline_responses left
347 "_responses" off the end of the name.
349 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
351 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
352 (without spaces) instead of just copying the configuration text.
354 PH/10 Added the /noupdate option to the ratelimit ACL condition.
356 PH/11 Added $max_received_linelength.
358 PH/12 Added +ignore_defer and +include_defer to host lists.
360 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
361 way in which PCRE > 7.0 is built.
363 PH/14 Implemented queue_only_load_latch.
365 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
366 MAIL command. The effect was to mangle the value on 64-bit systems.
368 PH/16 Another patch from the Sieve maintainer.
370 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
372 PH/18 If a system quota error occurred while trying to create the file for
373 a maildir delivery, the message "Mailbox is full" was not appended to the
374 bounce if the delivery eventually timed out. Change 4.67/27 below applied
375 only to a quota excession during the actual writing of the file.
377 PH/19 It seems that peer DN values may contain newlines (and other non-printing
378 characters?) which causes problems in log lines. The DN values are now
379 passed through string_printing() before being added to log lines.
381 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
382 and InterBase are left for another time.)
384 PH/21 Added message_body_newlines option.
386 PH/22 Guard against possible overflow in moan_check_errorcopy().
388 PH/23 POSIX allows open() to be a macro; guard against that.
390 PH/24 If the recipient of an error message contained an @ in the local part
391 (suitably quoted, of course), incorrect values were put in $domain and
392 $local_part during the evaluation of errors_copy.
398 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
399 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
402 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
403 issue a MAIL command.
405 PH/02 In an ACL statement such as
407 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
409 if a client was not listed at all, or was listed with a value other than
410 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
411 the condition was not true (as it should be), so access was not denied.
412 The bug was that the ! inversion was incorrectly passed on to the second
413 item. This has been fixed.
415 PH/03 Added additional dnslists conditions == and =& which are different from
416 = and & when the dns lookup returns more than one IP address.
418 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
419 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
421 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
422 FSYNC, which compiles an option called disable_fsync that allows for
423 bypassing fsync(). The documentation is heavily laced with warnings.
425 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
427 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
428 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
429 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
430 including adding "make clean"; (3) Added -fPIC when compiling the test
431 dynamically loaded module, to get rid of a warning.
433 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
434 message fails, move_frozen_messages = true and ignore_bounce_errors_after
435 = 0s. The bug is otherwise harmless.
437 PH/07 There was a bug in the dovecot authenticator such that the value of
438 $auth1 could be overwritten, and so not correctly preserved, after a
439 successful authentication. This usually meant that the value preserved by
440 the server_setid option was incorrect.
442 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
444 PH/09 Installed PCRE release 7.0.
446 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
447 run for batched SMTP input. It is now run at the start of every message
448 in the batch. While fixing this I discovered that the process information
449 (output by running exiwhat) was not always getting set for -bs and -bS
450 input. This is fixed, and it now also says "batched" for BSMTP.
452 PH/11 Added control=no_pipelining.
454 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
455 patch, slightly modified), and move the expansion of helo_data till after
456 the connection is made in the smtp transport (so it can use these
459 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
461 PH/14 Added log_selector = +pid.
463 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
465 PH/16 Add ${if forany and ${if forall.
467 PH/17 Added dsn_from option to vary the From: line in DSNs.
469 PH/18 Flush SMTP output before performing a callout, unless control =
470 no_callout_flush is set.
472 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
473 was true (the default) a successful delivery failed to delete the retry
474 item, thus causing premature timeout of the address. The bug is now
477 PH/20 Added hosts_avoid_pipelining to the smtp transport.
479 PH/21 Long custom messages for fakedefer and fakereject are now split up
480 into multiline reponses in the same way that messages for "deny" and
481 other ACL rejections are.
483 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
484 with slight modification.
486 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
487 draft, changing the syntax and factoring some duplicate code".
489 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
490 for deliveries of the second and subsequent messages over the same SMTP
493 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
494 ${reduce, with only minor "tidies".
496 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
498 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
499 expansion side effects.
501 PH/27 When a message times out after an over-quota error from an Exim-imposed
502 quota, the bounce message says "mailbox is full". This message was not
503 being given when it was a system quota that was exceeded. It now should
506 MH/03 Made $recipients available in local_scan(). local_scan() already has
507 better access to the recipient list through recipients_list[], but
508 $recipients can be useful in postmaster-provided expansion strings.
510 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
511 in the case of a MAIL command with additional options following the
512 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
513 were accidentally chopped off.
515 PH/29 SMTP synchronization checks are implemented when a command is read -
516 there is a check that no more input is waiting when there shouldn't be
517 any. However, for some commands, a delay in an ACL can mean that it is
518 some time before the response is written. In this time, more input might
519 arrive, invalidly. So now there are extra checks after an ACL has run for
520 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
521 pipelining has not been advertised.
523 PH/30 MH's patch to allow iscntrl() characters to be list separators.
525 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
526 returned in the SMTP response when smtp_return_error_details was false.
529 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
530 instead of the C I/O that was originally supplied, because problems were
533 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
534 Exim which did not show up earlier: it was assuming that a call to
535 SSL_CTX_set_info_callback() might give an error value. In fact, there is
536 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
537 was a macro that became an assignment, so it seemed to work. This has
538 changed to a proper function call with a void return, hence the compile
539 error. Exim's code has been fixed.
541 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
544 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
546 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
548 PH/37 If a message is not accepted after it has had an id assigned (e.g.
549 because it turns out to be too big or there is a timeout) there is no
550 "Completed" line in the log. When some messages of this type were
551 selected by exigrep, they were listed as "not completed". Others were
552 picked up by some special patterns. I have improved the selection
553 criteria to be more general.
555 PH/38 The host_find_failed option in the manualroute router can now be set
556 to "ignore", to completely ignore a host whose IP address cannot be
557 found. If all hosts are ignored, the behaviour is controlled by the new
558 host_all_ignored option.
560 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
561 homing or because of multiple MX records with /mx) generated more than
562 one IP address, and the following item turned out to be the local host,
563 all the secondary addresses of the first item were incorrectly removed
564 from the list, along with the local host and any following hosts (which
565 is what is supposed to happen).
567 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
568 whoever called Exim into the -H file. In the case of the daemon it was
569 behaving confusingly. When first started, it used values for whoever
570 started the daemon, but after a SIGHUP it used the Exim user (because it
571 calls itself on a restart). I have changed the code so that it now always
574 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
575 message are rejected with the same error (e.g. no authentication or bad
576 sender address), and a DATA command is nevertheless sent (as can happen
577 with PIPELINING or a stupid MUA), the error message that was given to the
578 RCPT commands is included in the rejection of the DATA command. This is
579 intended to be helpful for MUAs that show only the final error to their
582 PH/42 Another patch from the Sieve maintainer.
584 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
585 Eximstats - Fixed some broken HTML links and added missing column headers
587 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
588 columns for Rejects, Temp Rejects, Ham, and Spam rows.
590 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
592 PH/43 Yet another patch from the Sieve maintainer.
594 PH/44 I found a way to check for a TCP/IP connection going away before sending
595 the response to the final '.' that terminates a message, but only in the
596 case where the client has not sent further data following the '.'
597 (unfortunately, this is allowed). However, in many cases there won't be
598 any further data because there won't be any more messages to send. A call
599 to select() can be used: if it shows that the input is "ready", there is
600 either input waiting, or the socket has been closed. An attempt to read
601 the next input character can distinguish the two cases. Previously, Exim
602 would have sent an OK response which the client would never have see.
603 This could lead to message repetition. This fix should cure that, at
604 least in a lot of common cases.
606 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
607 advertised in response to EHLO.
613 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
614 fixed by 4.65/MH/01 (is this a record?) are fixed:
616 (i) An empty string was always treated as zero by the numeric comparison
617 operators. This behaviour has been restored.
619 (ii) It is documented that the numeric comparison operators always treat
620 their arguments as decimal numbers. This was broken in that numbers
621 starting with 0 were being interpreted as octal.
623 While fixing these problems I realized that there was another issue that
624 hadn't been noticed. Values of message_size_limit (both the global option
625 and the transport option) were treated as octal if they started with 0.
626 The documentation was vague. These values are now always treated as
627 decimal, and I will make that clear in the documentation.
633 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
634 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
637 MH/01 Don't check that the operands of numeric comparison operators are
638 integers when their expansion is in "skipping" mode (fixes bug
639 introduced by 4.64-PH/07).
641 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
642 child addresses, Exim now panics and dies. Previously, because the count
643 is held in a short int, deliveries were likely to be lost. As such a
644 large number of recipients for a single message is ridiculous
645 (performance will be very, very poor), I have chosen to impose a limit
646 rather than extend the field.
652 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
653 leftover -K file (the existence of which was triggered by #402).
654 While we were at it, introduced process PID as part of the -K
655 filename. This should rule out race conditions when creating
658 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
659 processing considerably. Previous code took too long for large mails,
660 triggering a timeout which in turn triggers #401.
662 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
663 in the DK code in transports.c. sendfile() is not really portable,
664 hence the _LINUX specificness.
666 TF/01 In the add_headers option to the mail command in an Exim filter,
667 there was a bug that Exim would claim a syntax error in any
668 header after the first one which had an odd number of characters
671 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
672 callout verification, Exim cached a "reject" for the entire domain. This
673 is correct for most verifications, but it is not correct for a recipient
674 verification with use_sender or use_postmaster set, because in that case
675 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
676 case of MAIL FROM:<> rejection from other early rejections (e.g.
677 rejection of HELO). When verifying a recipient using a non-null MAIL
678 address, the cache is ignored if it shows MAIL FROM:<> rejection.
679 Whatever the result of the callout, the value of the domain cache is
680 left unchanged (for any other kind of callout, getting as far as trying
681 RCPT means that the domain itself is ok).
683 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
686 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
687 manifest itself as EPIPE rather than ECONNECT. When tidying away a
688 session, the daemon ignores ECONNECT errors and logs others; it now
689 ignores EPIPE as well.
691 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
692 (quoted-printable decoding).
694 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
695 later the small subsequent patch to fix an introduced bug.
697 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
699 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
701 PH/08 An error is now given if message_size_limit is specified negative.
703 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
704 to be given (somewhat) arbitrary names.
706 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
709 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
710 miscellaneous code fixes
712 PH/10 Added the log_reject_target ACL modifier to specify where to log
715 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
716 hostname. This is wrong, because it relates to the incoming message (and
717 probably the interface on which it is arriving) and not to the outgoing
718 callout (which could be using a different interface). This has been
719 changed to use the value of the helo_data option from the smtp transport
720 instead - this is what is used when a message is actually being sent. If
721 there is no remote transport (possible with a router that sets up host
722 addresses), $smtp_active_hostname is used.
724 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
725 tweaks were necessary in order to get it to work (see also 21 below):
726 (a) The code assumed that strncpy() returns a negative number on buffer
727 overflow, which isn't the case. Replaced with Exim's string_format()
729 (b) There were several signed/unsigned issues. I just did the minimum
730 hacking in of casts. There is scope for a larger refactoring.
731 (c) The code used strcasecmp() which is not a standard C function.
732 Replaced with Exim's strcmpic() function.
733 (d) The code set only $1; it now sets $auth1 as well.
734 (e) A simple test gave the error "authentication client didn't specify
735 service in request". It would seem that Dovecot has changed its
736 interface. Fortunately there's a specification; I followed it and
737 changed what the client sends and it appears to be working now.
739 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
742 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
743 address A is aliased to B and C, where B exists and C does not. Without
744 -v the output is "A verified" because verification stops after a
745 successful redirection if more than one address is generated. However,
746 with -v the child addresses are also verified. Exim was outputting "A
747 failed to verify" and then showing the successful verification for C,
748 with its parentage. It now outputs "B failed to verify", showing B's
749 parentage before showing the successful verification of C.
751 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
752 look up a TXT record in a specific list after matching in a combined
755 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
756 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
757 they consult the DNS. I had assumed they would set it the way they
758 wanted; and indeed my experiments on Linux seem to show that in some
759 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
760 To be on the safe side, however, I have now made the interface to
761 host_find_byname() similar to host_find_bydns(), with an argument
762 containing the DNS resolver options. The host_find_byname() function now
763 sets these options at its start, just as host_find_bydns() does. The smtp
764 transport options dns_qualify_single and dns_search_parents are passed to
765 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
766 of host_find_byname() use the default settings of RES_DEFNAMES
767 (qualify_single) but not RES_DNSRCH (search_parents).
769 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
770 spool_read_header() do less string testing, by means of a preliminary
771 switch on the second character of optional "-foo" lines. (This is
772 overdue, caused by the large number of possibilities that now exist.
773 Originally there were few.) While I was there, I also converted the
774 str(n)cmp tests so they don't re-test the leading "-" and the first
775 character, in the hope this might squeeze out yet more improvement.
777 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
778 flag allowing group syntax was set by the header_syntax check but not
779 turned off, possible causing trouble later; (2) The flag was not being
780 set at all for the header_verify test, causing "group"-style headers to
781 be rejected. I have now set it in this case, and also caused header_
782 verify to ignore an empty address taken from a group. While doing this, I
783 came across some other cases where the code for allowing group syntax
784 while scanning a header line wasn't quite right (mostly, not resetting
785 the flag correctly in the right place). These bugs could have caused
786 trouble for malformed header lines. I hope it is now all correct.
788 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
789 with the "reply" argument non-NULL. The code, however (which originally
790 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
791 but it didn't always do it. This confused somebody who was copying the
792 code for some other use. I have removed all the tests.
794 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
795 feature that was used to support insecure browsers during the U.S. crypto
796 embargo. It requires special client support, and Exim is probably the
797 only MTA that supported it -- and would never use it because real RSA is
798 always available. This code has been removed, because it had the bad
799 effect of slowing Exim down by computing (never used) parameters for the
800 RSA_EXPORT functionality.
802 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
803 authenticator to fail if there's a tab character in the incoming data
804 (there should never be unless someone is messing about, as it's supposed
805 to be base64-encoded). Also added, on Timo's advice, the "secured" option
806 if the connection is using TLS or if the remote IP is the same as the
807 local IP, and the "valid-client-cert option" if a client certificate has
810 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
811 authenticators. This can be used for authorization after authentication
812 succeeds. (In the case of plaintext, it servers for both authentication
815 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
816 if any retry times were supplied.
818 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
819 connection, where there is no client IP address to check. In this
820 situation, the verify now always succeeds.
822 PH/25 Applied John Jetmore's -Mset patch.
824 PH/26 Added -bem to be like -Mset, but loading a message from a file.
826 PH/27 In a string expansion for a processed (not raw) header when multiple
827 headers of the same name were present, leading whitespace was being
828 removed from all of them, but trailing whitespace was being removed only
829 from the last one. Now trailing whitespace is removed from each header
830 before concatenation. Completely empty headers in a concatenation (as
833 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
834 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
836 PH/29 [Removed. This was a change that I later backed out, and forgot to
837 correct the ChangeLog entry (that I had efficiently created) before
838 committing the later change.]
840 PH/30 Exim was sometimes attempting to deliver messages that had suffered
841 address errors (4xx response to RCPT) over the same connection as other
842 messages routed to the same hosts. Such deliveries are always "forced",
843 so retry times are not inspected. This resulted in far too many retries
844 for the affected addresses. The effect occurred only when there were more
845 hosts than the hosts_max_try setting in the smtp transport when it had
846 the 4xx errors. Those hosts that it had tried were not added to the list
847 of hosts for which the message was waiting, so if all were tried, there
848 was no problem. Two fixes have been applied:
850 (i) If there are any address or message errors in an SMTP delivery, none
851 of the hosts (tried or untried) are now added to the list of hosts
852 for which the message is waiting, so the message should not be a
853 candidate for sending over the same connection that was used for a
854 successful delivery of some other message. This seems entirely
855 reasonable: after all the message is NOT "waiting for some host".
856 This is so "obvious" that I'm not sure why it wasn't done
857 previously. Hope I haven't missed anything, but it can't do any
858 harm, as the worst effect is to miss an optimization.
860 (ii) If, despite (i), such a delivery is accidentally attempted, the
861 routing retry time is respected, so at least it doesn't keep
862 hammering the server.
864 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
865 in ${readsocket because some servers need this prod.
867 PH/32 Added some extra debug output when updating a wait-xxx database.
869 PH/33 The hint "could be header name not terminated by colon", which has been
870 given for certain expansion errors for a long time, was not being given
871 for the ${if def:h_colon_omitted{... case.
873 PH/34 The spec says: "With one important exception, whenever a domain list is
874 being scanned, $domain contains the subject domain." There was at least
875 one case where this was not true.
877 PH/35 The error "getsockname() failed: connection reset by peer" was being
878 written to the panic log as well as the main log, but it isn't really
879 panic-worthy as it just means the connection died rather early on. I have
880 removed the panic log writing for the ECONNRESET error when getsockname()
883 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
884 runs only) independently of the message's sender address. This meant
885 that, if the 4xx error was in fact related to the sender, a different
886 message to the same recipient with a different sender could confuse
887 things. In particualar, this can happen when sending to a greylisting
888 server, but other circumstances could also provoke similar problems.
889 I have changed the default so that the retry time for these errors is now
890 based a combination of the sender and recipient addresses. This change
891 can be overridden by setting address_retry_include_sender=false in the
894 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
895 remote server are returned as part of bounce messages. This was not
896 happening for LMTP over a pipe (the lmtp transport), but now it is the
897 same for both kinds of LMTP.
899 PH/38 Despite being documented as not happening, Exim was rewriting addresses
900 in header lines that were in fact CNAMEs. This is no longer the case.
902 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
903 and queue runs started by the daemon processed all messages. This has
904 been fixed so that -R and -S can now usefully be given with -q<time>.
906 PH/40 Import PCRE release 6.7 (fixes some bugs).
908 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
910 PH/42 Give an error if -q is specified more than once.
912 PH/43 Renamed the variables $interface_address and $interface_port as
913 $received_ip_address and $received_port, to make it clear that these
914 values apply to message reception, and not to the outgoing interface when
915 a message is delivered. (The old names remain recognized, of course.)
917 PH/44 There was no timeout on the connect() call when using a Unix domain
918 socket in the ${readsocket expansion. There now is.
920 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
921 be meaningful with "accept".
923 SC/01 Eximstats V1.43
924 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
926 SC/02 Eximstats V1.44
927 Use a glob alias rather than an array ref in the generated
928 parser. This improves both readability and performance.
930 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
931 Collect SpamAssassin and rejection statistics.
932 Don't display local sender or destination tables unless
933 there is data to show.
934 Added average volumes into the top table text output.
936 SC/04 Eximstats V1.46
937 Collect data on the number of addresses (recipients)
938 as well as the number of messages.
940 SC/05 Eximstats V1.47
941 Added 'Message too big' to the list of mail rejection
942 reasons (thanks to Marco Gaiarin).
944 SC/06 Eximstats V1.48
945 Mainlog lines which have GMT offsets and are too short to
946 have a flag are now skipped.
948 SC/07 Eximstats V1.49 (Alain Williams)
949 Added the -emptyok flag.
951 SC/08 Eximstats V1.50
952 Fixes for obtaining the IP address from reject messages.
954 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
955 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
956 whitesspace changes from 4.64-PH/27
958 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
961 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
962 are found, allow negative numbers in numeric criteria)
964 JJ/06 exipick.20061117.2, added new $message_body_missing variable
966 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
967 to match changes made in 4.64-PH/43
969 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
971 PH/47 Put in an explicit test for a DNS lookup of an address record where the
972 "domain" is actually an IP address, and force a failure. This locks out
973 those revolvers/nameservers that support "A-for-A" lookups, in
974 contravention of the specifications.
976 PH/48 When a host name was looked up from an IP address, and the subsequent
977 forward lookup of the name timed out, the host name was left in
978 $sender_host_name, contrary to the specification.
980 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
981 restricted to single-key lookups, Exim was not diagnosing an error if
982 * or *@ was used with a query-style lookup.
984 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
986 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
987 long ago, but noone interested enough thought of it. Let's just say that
988 the "1.1" means that there are some new functions that weren't there at
989 some point in the past.
991 PH/51 Error processing for expansion failure of helo_data from an smtp
992 transport during callout processing was broken.
994 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
995 tested/used via the -bh/-bhc/-bs options.
997 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
998 bug, fixed in subsequent PCRE releases).
1000 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1001 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1003 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1009 SC/01 Use a glob alias rather than an array ref in eximstats generated
1010 parser. This improves both readability and performance.
1012 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1013 Don't display local sender or destination tables in eximstats unless
1014 there is data to show.
1015 Added average volumes into the eximstats top table text output.
1017 SC/03 Collect data on the number of addresses (recipients) as well
1018 as the number of messages in eximstats.
1020 TF/01 Correct an error in the documentation for the redirect router. Exim
1021 does (usually) call initgroups() when daemonizing.
1023 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1024 with consistent privilege compared to when running as a daemon.
1026 TF/03 Note in the spec that $authenticated_id is not set for local
1027 submissions from trusted users.
1029 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1030 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1032 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1033 by adding some example configuration directives to the default
1034 configuration file. A little bit of work is required to uncomment the
1035 directives and define how usernames and passwords are checked, but
1036 there is now a framework to start from.
1038 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1039 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1040 without this. I don't know how relevant this is to other LDAP libraries.
1042 PH/02 Add the verb name to the "unknown ACL verb" error.
1044 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1046 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1048 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1049 directory not expanded when it should be if an expanded home directory
1050 was set for the address (which is overridden by the transport).
1052 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1055 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1056 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1057 because it is too late at that time, and has no effect.
1059 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1060 security issue with \' (bugzilla #107). I could not use the
1061 PQescapeStringConn() function, because it needs a PGconn value as one of
1064 PH/08 When testing addresses using -bt, indicate those final addresses that
1065 are duplicates that would not cause an additional delivery. At least one
1066 person was confused, thinking that -bt output corresponded to deliveries.
1067 (Suppressing duplicates isn't a good idea as you lose the information
1068 about possibly different redirections that led to the duplicates.)
1070 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1071 systems where poll() doesn't work, in particular OS X.
1073 PH/10 Added more information to debugging output for retry time not reached.
1075 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1076 operations in malware.c.
1078 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1081 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1082 syslog_duplication set false, log lines that would normally be written
1083 both the the main log and to the reject log were not written to syslog at
1086 PH/14 In the default configuration, change the use of "message" in ACL warn
1087 statements to "add_header".
1089 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1090 not followed by a command (e.g. "seen endif").
1092 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1093 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1096 PH/17 Added extra conditions to the default value of delay_warning_condition
1100 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1101 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1102 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1105 The Auto-Submitted: and various List- headers are standardised, whereas I
1106 don't think Precedence: ever was.
1108 PH/18 Refactored debugging code in route_finduser() to show more information,
1109 in particular, the error code if getpwnam() issues one.
1111 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1112 This is apparently needed in addition to the PH/07 change above to avoid
1113 any possible encoding problems.
1115 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1116 but not after initializing Perl.
1118 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1119 output them only if debugging. By default they are written stderr,
1120 apparently, which is not desirable.
1122 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1125 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1128 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1130 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1131 authenticated or an ident call has been made. Suppress the default
1132 values for $authenticated_id and $authenticated_sender (but permit -oMai
1133 and -oMas) when testing with -bh.
1135 PH/24 Re-jigged the order of the tests in the default configuration so that the
1136 tests for valid domains and recipients precede the DNS black list and CSA
1137 tests, on the grounds that those ones are more expensive.
1139 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1140 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1141 This bug exists in every version of Exim that I still have, right back to
1144 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1145 However, an attempt to turn on case-sensitivity in a regex key by
1146 including (?-i) didn't work because the subject string was already
1147 lowercased, and the effects were non-intuitive. It turns out that a
1148 one-line patch can be used to allow (?-i) to work as expected.
1154 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1155 other effects) broke the use of negated acl sub-conditions.
1157 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1160 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1161 "Deny" causes Exim to reject the incoming connection with a 554 error.
1162 Unfortunately, if there is a major crisis, such as a disk failure,
1163 tcp-wrappers gives "deny", whereas what one would like would be some
1164 kind of temporary error. A kludge has been added to help with this.
1165 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1166 554 error is used if errno is still zero or contains ENOENT (which occurs
1167 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1170 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1172 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1173 errors. Otherwise a message that provokes a temporary error (when other
1174 messages do not) can cause a whole host to time out.
1176 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1177 addresses were routed directly to files or pipes from a redirect router.
1178 File deliveries just didn't batch; pipe deliveries might have suffered
1181 PH/06 A failure to get a lock for a hints database would erroneously always say
1182 "Failed to get write lock", even when it was really a read lock.
1184 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1185 of 0600. This has been changed to use the value of the lockfile_mode
1186 option (which defaults to 0600).
1188 PH/08 Applied small patch from the Sieve maintainer.
1190 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1191 folder from quota calculations, a direct delivery into this folder messed
1192 up the contents of the maildirsize file. This was because the regex was
1193 used only to exclude .Trash (or whatever) when the size of the mailbox
1194 was calculated. There was no check that a delivery was happening into an
1195 excluded directory. This bug has been fixed by ignoring all quota
1196 processing for deliveries into excluded directories.
1198 PH/10 Added the maildirfolder_create_regex option to appendfile.
1204 PH/01 The code for finding all the local interface addresses on a FreeBSD
1205 system running IPv6 was broken. This may well have applied to all BSD
1206 systems, as well as to others that have similar system calls. The broken
1207 code found IPv4 interfaces correctly, but gave incorrect values for the
1208 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1209 that it would not match correctly against @[] and not recognize the IPv6
1212 PH/02 The ipliteral router was not recognizing addresses of the form user@
1213 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1215 PH/03 Added disable_ipv6.
1217 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1218 decoded form, because it is most often used to construct To: headers
1219 lines in autoreplies, and the decoded form may well be syntactically
1220 invalid. However, $reply_address has leading white space removed, and all
1221 newlines turned into spaces so that the autoreply transport does not
1224 PH/05 If group was specified without a user on a router, and no group or user
1225 was specified on a transport, the group from the router was ignored.
1227 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1228 for visible compile-time settings that can be used to change these
1229 numbers, for those that want even more. Backwards compatibility with old
1230 spool files has been maintained. However, going back to a previous Exim
1231 release will lost any variables that are in spool files.
1233 PH/07 Two small changes when running in the test harness: increase delay when
1234 passing a TCP/IP connection to a new process, in case the original
1235 process has to generate a bounce, and remove special handling of
1236 127.0.0.2 (sic), which is no longer necessary.
1238 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1239 be the same on different OS.
1241 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1244 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1245 whether --show-vars was specified or not
1247 JJ/02 exipick: Added support for new ACL variable spool format introduced
1250 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1251 syntactically invalid From: or Reply-to: line, and a filter used this to
1252 generate an autoreply, and therefore failed to obtain an address for the
1253 autoreply, Exim could try to deliver to a non-existent relative file
1254 name, causing unrelated and misleading errors. What now happens is that
1255 it logs this as a hard delivery error, but does not attempt to create a
1258 PH/11 The exinext utility has a -C option for testing purposes, but although
1259 the given file was scanned by exinext itself; it wasn't being passed on
1260 when Exim was called.
1262 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1263 an end-of-file indication when reading a command response.
1265 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1266 compiled. In many other places in Exim, IPv6 addresses are always
1267 recognized, so I have changed this. It also means that IPv4 domain
1268 literals of the form [IPV4:n.n.n.n] are now always recognized.
1270 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1271 used if the router is not running as root, for example, when verifying at
1272 ACL time, or when using -bh. The debugging output from this situation was
1273 non-existent - all you got was a failure to exec. I have made two
1276 (a) Failures to set uid/gid, the current directory, or a process leader
1277 in a subprocess such as that created by queryprogram now generate
1278 suitable debugging ouput when -d is set.
1280 (b) The queryprogram router detects when it is not running as root,
1281 outputs suitable debugging information if -d is set, and then runs
1282 the subprocess without attempting to change uid/gid.
1284 PH/15 Minor change to Makefile for building test_host (undocumented testing
1287 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1288 additional section of a DNS packet that returns MX or SRV records.
1289 Instead, it always explicitly searches for A/AAAA records. This avoids
1290 major problems that occur when a DNS server includes only records of one
1291 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1292 fixed another bug: if SRV records were looked up and the corresponding
1293 address records were *not* found in the additional section, the port
1294 values from the SRV records were lost.
1296 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1297 using the correct key (the original address) when searching the retry
1298 rules in order to find which one to use for generating the retry hint.
1300 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1301 adding the default one. Similarly, if it contains a Reply-To: header, the
1302 errors_reply_to option, if set, is not used.
1304 PH/19 When calculating a retry time, Exim used to measure the "time since
1305 failure" by looking at the "first failed" field in the retry record. Now
1306 it does not use this if it is later than than the arrival time of the
1307 message. Instead it uses the arrival time. This makes for better
1308 behaviour in cases where some deliveries succeed, thus re-setting the
1309 "first failed" field. An example is a quota failure for a huge message
1310 when small messages continue to be delivered. Without this change, the
1311 "time since failure" will always be short, possible causing more frequent
1312 delivery attempts for the huge message than are intended.
1313 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1315 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1316 $1, $2, $3) because the numerical variables can be reset during some
1317 expansion items (e.g. "match"), thereby losing the authentication data.
1319 PH/21 Make -bV show the size of off_t variables so that the test suite can
1320 decide whether to run tests for quotas > 2G.
1322 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1323 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1324 filecount value is greater than 2G or if a quota value is greater than 2G
1325 on a system where the size of off_t is not greater than 4, a panic error
1328 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1329 never match. The debug and -bh output now contains an explicit error
1330 message indicating a malformed IPv4 address or mask.
1332 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1333 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1334 PH/23 above applies.
1336 PH/25 Do not write to syslog when running in the test harness. The only
1337 occasion when this arises is a failure to open the main or panic logs
1338 (for which there is an explicit test).
1340 PH/26 Added the /no_tell option to "control=freeze".
1342 PH/27 If a host name lookup failed very early in a connection, for example, if
1343 the IP address matched host_lookup and the reverse lookup yielded a name
1344 that did not have a forward lookup, an error message of the form "no IP
1345 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1346 could be logged. Now it outputs the IP address instead of "NULL".
1348 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1349 allows the sender and the authenticated sender to be set when
1350 submitting a message from within Exim. Since child_open_exim() is
1351 documented for local_scan(), the new function should be too.
1353 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1354 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1355 results in an empty string is now treated as unset.
1357 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1359 PH/31 Added sender_verify_fail logging option.
1361 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1362 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1363 tidied the source and removed it altogether.
1365 PH/33 When a queue run was abandoned because the load average was too high, a
1366 log line was always written; now it is written only if the queue_run log
1367 selector is set. In addition, the log line for abandonment now contains
1368 information about the queue run such as the pid. This is always present
1369 in "start" and "stop" lines but was omitted from the "abandon" line.
1371 PH/34 Omit spaces between a header name and the colon in the error message that
1372 is given when verify = headers_syntax fails (if there are lots of them,
1373 the message gets confusing).
1375 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1376 names, as there are now some PTR records that contain slashes. This check
1377 is only to protect against broken name servers that fall over on strange
1378 characters, so the fact that it applies to all lookups doesn't matter.
1380 PH/36 Now that the new test suite is complete, we can remove some of the
1381 special code in Exim that was needed for the old test suite. For example,
1382 sorting DNS records because real resolvers return them in an arbitrary
1383 order. The new test suite's fake resolver always returns records in the
1386 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1387 bounces) except when queue_only is set, to avoid logging races between
1388 the different processes.
1390 PH/38 Panic-die if .include specifies a non-absolute path.
1392 PH/39 A tweak to the "H" retry rule from its user.
1394 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1395 a label. They prevented compilation on older perls.
1397 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1398 a warning to be raised on newish perls.
1400 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1401 on queue. Changes to match documented behaviour of showing count of
1402 messages matching specified criteria.
1404 PH/40 Changed the default ident timeout from 30s to 5s.
1406 PH/41 Added support for the use of login_cap features, on those BSD systems
1407 that have them, for controlling the resources used by pipe deliveries.
1409 PH/42 The content-scanning code uses fopen() to create files in which to put
1410 message data. Previously it was not paying any attention to the mode of
1411 the files. Exim runs with umask(0) because the rest of the code creates
1412 files with open(), and sets the required mode explicitly. Thus, these
1413 files were ending up world-writeable. This was not a big issue, because,
1414 being within the spool directory, they were not world-accessible. I have
1415 created a function called modefopen, which takes an additional mode
1416 argument. It sets umask(777), creates the file, chmods it to the required
1417 mode, then resets the umask. All the relevant calls to fopen() in the
1418 content scanning code have been changed to use this function.
1420 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1421 to 24 hours. This avoids potential overflow problems when processing G
1422 and H retry rules. I suspect nobody ever tinkers with this value.
1424 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1426 PH/45 When the plaintext authenticator is running as a client, the server's
1427 challenges are checked to ensure they are valid base64 strings. By
1428 default, the authentication attempt is cancelled if an invalid string is
1429 received. Setting client_ignore_invalid_base64 true ignores these errors.
1430 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1431 they are received. Thus, the responses can be made to depend on the
1432 challenges. If an invalid string is ignored, an empty string is placed in
1435 PH/46 Messages that are created by the autoreply transport now contains a
1436 References: header, in accordance with RFCs 2822 and 3834.
1438 PH/47 Added authenticated_sender_force to the smtp transport.
1440 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1442 PH/49 Installed latest patch from the Sieve maintainer.
1444 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1445 was also set, the appendfile transport was unnecessarily scanning a
1446 directory of message files (e.g. for maildir delivery) to find the count
1447 of files (along with the size), even though it did not need this
1448 information. It now does the scan only if it needs to find either the
1449 size of the count of files.
1451 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1453 PH/52 Two bugs concerned with error handling when the smtp transport is
1456 (i) Exim was not creating retry information for temporary errors given
1457 for individual recipients after the DATA command when the smtp transport
1458 was used in LMTP mode. This meant that they could be retried too
1459 frequently, and not timed out correctly.
1461 (ii) Exim was setting the flag that allows error details to be returned
1462 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1463 recipients that were returned after the DATA command.
1465 PH/53 This is related to PH/52, but is more general: for any failing address,
1466 when detailed error information was permitted to be returned to the
1467 sender, but the error was temporary, then after the final timeout, only
1468 "retry timeout exceeded" was returned. Now it returns the full error as
1469 well as "retry timeout exceeded".
1471 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1472 do this, and (what is worse) MTAs that accept it.
1474 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1475 will now be deprecated.
1477 PH/56 New os.c-cygwin from the Cygwin maintainer.
1479 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1480 formats (previously only available in exim formats via -bpr, -bpru,
1481 and -bpra. Now also available in native and exiqgrep formats)
1483 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1484 with very large, slow to parse queues
1486 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1488 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1490 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1491 responses to authentication challenges, though it was showing the
1492 challenges; (ii) I've removed the CR characters from the debug output for
1495 PH/58 Allow for the insertion of a newline as well as a space when a string
1496 is turned into more than one encoded-word during RFC 2047 encoding. The
1497 Sieve code now uses this.
1499 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1500 data_4xx, lost_connection, tls_required.
1502 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1503 message was being sent as an SMTP response.
1505 PH/61 Add -l and -k options to exicyclog.
1507 PH/62 When verifying, if an address was redirected to one new address, so that
1508 verification continued, and the new address failed or deferred after
1509 having set something in $address_data, the value of $address_data was not
1510 passed back to the ACL. This was different to the case when no
1511 redirection occurred. The value is now passed back in both cases.
1513 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1514 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1515 use login_cap.h, so on its own it isn't the distinguishing feature. The
1516 new name refers directly to the setclassresources() function.
1518 PH/65 Added configuration files for NetBSD3.
1520 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1522 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1523 is preferred over IPv4.
1525 PH/68 The bounce_return_message and bounce_return_body options were not being
1526 honoured for bounces generated during the reception of non-SMTP messages.
1527 In particular, this applied to messages rejected by the ACL. This bug has
1528 been fixed. However, if bounce_return_message is true and bounce_return_
1529 body is false, the headers that are returned for a non-SMTP message
1530 include only those that have been read before the error was detected.
1531 (In the case of an ACL rejection, they have all been read.)
1533 PH/69 The HTML version of the specification is now built in a directory called
1534 spec_html instead of spec.html, because the latter looks like a path with
1535 a MIME-type, and this confuses some software.
1537 PH/70 Catch two compiler warnings in sieve.c.
1539 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1540 function verify_get_ident() calls ip_connect() to connect a socket, but
1541 if the "connect()" function timed out, ip_connect() used to close the
1542 socket. However, verify_get_ident() also closes the socket later, and in
1543 between Exim writes to the log, which may get opened at this point. When
1544 the socket was closed in ip_connect(), the log could get the same file
1545 descriptor number as the socket. This naturally causes chaos. The fix is
1546 not to close the socket in ip_connect(); the socket should be closed by
1547 the function that creates it. There was only one place in the code where
1548 this was missing, in the iplookup router, which I don't think anybody now
1549 uses, but I've fixed it anyway.
1551 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1552 well as to direct DNS lookups. Otherwise the handling of names in host
1553 lists is inconsistent and therefore confusing.
1559 PH/01 Two changes to the default runtime configuration:
1561 (1) Move the checks for relay_from_hosts and authenticated clients from
1562 after to before the (commented out) DNS black list checks.
1564 (2) Add control=submission to the relay_from_hosts and authenticated
1565 clients checks, on the grounds that messages accepted by these
1566 statements are most likely to be submissions.
1568 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1570 (1) Generate an error if the third argument for the ${prvs expansion is
1573 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1576 (3) Reset the variables that are obtained from the first argument of
1577 ${prvscheck and used in the second argument before leaving the code,
1578 because their memory is reclaimed, so using them afterwards may do
1581 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1582 one (it's much easier than Tom thought :-).
1584 (5) Because of (4), we can now allow for the use of $prvscheck_result
1585 inside the third argument.
1587 PH/03 For some reason, the default setting of PATH when running a command from
1588 a pipe transport was just "/usr/bin". I have changed it to
1591 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1592 anything to be listed in the output from -bV.
1594 PH/05 When a filter generated an autoreply, the entire To: header line was
1595 quoted in the delivery log line, like this:
1597 => >A.N.Other <ano@some.domain> <original@ddress> ...
1599 This has been changed so that it extracts the operative address. There
1600 may be more than one such address. If so, they are comma-separated, like
1603 => >ano@some.domain,ona@other.domain <original@ddress> ...
1605 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1606 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1607 not being looked up in the rDNS to get a host name, Exim was showing the
1608 IP address twice in Received: lines, even though the IP addresses were
1609 identical. For example:
1611 Received: from [1.2.3.4] (helo=[1.2.3.4])
1613 However, if the real host name was known, it was omitting the HELO data
1614 if it matched the actual IP address. This has been tidied up so that it
1615 doesn't show the same IP address twice.
1617 PH/07 When both +timestamp and +memory debugging was on, the value given by
1618 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1619 called by the debug printing, thereby overwriting the timestamp buffer.
1620 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1622 PH/08 When the original message was included in an autoreply transport, it
1623 always said "this is a copy of the message, including all the headers",
1624 even if body_only or headers_only was set. It now gives an appropriate
1627 PH/09 Applied a patch from the Sieve maintainer which:
1629 o fixes some comments
1630 o adds the (disabled) notify extension core
1631 o adds some debug output for the result of if/elsif tests
1632 o points to the current vacation draft in the documentation
1633 and documents the missing references header update
1637 o fixes a bug in processing the envelope test (when testing
1638 multiple envelope elements, the last element determinted the
1641 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1642 Electronic Mail") by including:
1644 Auto-submitted: auto-generated
1646 in the messages that it generates (bounce messages and others, such as
1647 warnings). In the case of bounce messages for non-SMTP mesages, there was
1648 also a typo: it was using "Auto_submitted" (underscore instead of
1649 hyphen). Since every message generated by Exim is necessarily in response
1650 to another message, thes have all been changed to:
1652 Auto-Submitted: auto-replied
1654 in accordance with these statements in the RFC:
1656 The auto-replied keyword:
1658 - SHOULD be used on messages sent in direct response to another
1659 message by an automatic process,
1661 - MUST NOT be used on manually-generated messages,
1663 - MAY be used on Delivery Status Notifications (DSNs) and Message
1664 Disposition Notifications (MDNs),
1666 - MUST NOT be used on messages generated by automatic or periodic
1667 processes, except for messages which are automatic responses to
1670 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1671 to the default Received: header definition.
1673 PH/12 Added log selector acl_warn_skipped (default on).
1675 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1676 variables because (a) they are in the wrong storage pool and (b) even if
1677 they were copied, it wouldn't work properly because of the caching.
1679 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1680 checking when decoding. Apparently there are clients that generate
1681 overlong encoded strings. Why am I not surprised?
1683 PH/15 If the first argument of "${if match_address" was not empty, but did not
1684 contain an "@" character, Exim crashed. Now it writes a panic log message
1685 and treats the condition as false.
1687 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1689 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1690 extension "envelope-auth". The code is finished and in agreement with
1691 other implementations, but there is no documentation so far and in fact,
1692 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1693 not changing the active code.
1695 Print executed "if" and "elsif" statements when debugging is used. This
1696 helps a great deal to understand what a filter does.
1698 Document more things not specified clearly in RFC3028. I had all this
1699 sorted out, when out of a sudden new issues came to my mind. Oops."
1701 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1704 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1705 canonical form (as documented). However, after a host name lookup from
1706 the IP address, check_host() was doing a simple string comparison with
1707 addresses acquired from the DNS when checking that the found name did
1708 have the original IP as one of its addresses. Since any found IPv6
1709 addresses are likely to be in abbreviated form, the comparison could
1710 fail. Luckily, there already exists a function for doing the comparison
1711 by converting both addresses to binary, so now that is used instead of
1712 the text comparison.
1714 PH/20 There was another similar case to PH/19, when a complete host name was
1715 given in a host list; looking up its IP address could give an abbreviated
1716 form, whereas the current host's name might or might not be abbreviated.
1717 The same fix has been applied.
1723 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1724 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1727 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1728 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1730 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1732 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1733 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1734 Also, for "Auto-Submitted", treat anything other than "no" as
1735 non-personal, in accordance with RFC 3834. (Previously it treated
1736 anything starting "auto-" as non-personal.)
1738 TF/01 The control=submission/name=... option had a problem with syntax
1739 errors if the name included a slash character. The /name= option
1740 now slurps the rest of the string, so it can include any characters
1741 but it must come last in the list of options (after /sender_retain
1744 PH/05 Some modifications to the interface to the fake nameserver for the new
1752 TK/01 Added the "success_on_redirect" address verification option. See
1753 NewStuff for rationale and an example.
1755 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1757 PH/02 Patch to exigrep to allow it to work on syslog lines.
1759 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1760 fread() to skip over the body file's header line, because in Cygwin the
1761 header line is locked and is inaccessible.
1763 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1764 co-exist for some time) to make it clear that it is the Exim ID that is
1765 referenced, not the Message-ID: header line.
1767 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1768 string_format() function, because snprintf() does not exist on all
1771 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1772 ${stat: expansion item.
1774 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1775 protocol synchronization error", to keep the pedants happy.
1777 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1778 well as for IRIX systems, when gcc is being used. See the host.c source
1781 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1783 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1786 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1787 transport and to the smtp transport in LMTP mode.
1789 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1791 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1792 run a filter in a subprocess. This could lead to confusion in subsequent
1793 lookups in the parent process. There should also be a search_tidyup() at
1794 the end of the subprocess.
1796 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1797 only if the host matched helo_try_verify_hosts, which caused the
1798 verification to occur when the EHLO/HELO command was issued. The ACL just
1799 tested the remembered result. Now, if a previous verification attempt has
1800 not happened, "verify = helo" does it there and then.
1802 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1804 TK/03 Fix log output including CR from clamd.
1806 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1807 exist provoked a memory error which could cause a segfault.
1809 PH/15 Installed PCRE 6.2
1811 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1813 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1814 of the problem. Specifically, suggested +O2 rather than +O1 for the
1817 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1819 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1820 record, the port was not correctly logged when the outgoing_port log
1821 selector was set (it logged the transort's default port).
1823 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1824 fallback_hosts, and "hosts" in the smtp transport.
1826 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1827 host errors such as "Connection refused".
1829 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1830 authentication with radiusclient 0.4.9:
1832 - Error returned from rc_read_config was caught wrongly
1833 - Username/password not passed on to radius server due to wrong length.
1835 The presumption is that some radiusclient API changes for 4.51/PH/17
1836 were not taken care of correctly. The code is still untested by me (my
1837 Linux distribution still has 0.3.2 of radiusclient), but it was
1838 contributed by a Radius user.
1840 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1841 expanding the "port" option of the smtp transport.
1843 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1844 while reading a MIME header. Thanks to Tom Hughes for a patch.
1846 PH/24 Include config.h inside local_scan.h so that configuration settings are
1849 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1850 that in an ACL for RCPT (for example), you can examine exactly what was
1853 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1854 commands, but it was not correctly comparing the address with the actual
1855 client host address. Thus, it would show the EHLO address in Received:
1856 header lines when this was not necessary.
1858 PH/27 Added the % operator to ${eval:}.
1860 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1861 it should be ignoring failures (because with -C, for example, it has lost
1862 privilege). It wasn't ignoring creation failures other than "already
1865 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1868 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1869 because an input line was too long, either on its own, or by virtue of
1870 too many continuations, the temporary file was not being removed, and the
1871 return code was incorrect.
1873 PH/31 Missing "BOOL" in function definition in filtertest.c.
1875 PH/32 Applied Sieve patches from the maintainer.
1877 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1879 PH/33 Added "verify = not_blind".
1881 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1882 Local/Makefile (with some defaults set). These are used in built scripts
1883 such as exicyclog, but they have never been used in the exim_install
1884 script (though there are many overriding facilities there). I have
1885 arranged that the exim_install script now takes note of these two
1888 PH/35 Installed configuration files for Dragonfly.
1890 PH/36 When a locally submitted message by a trusted user did not contain a
1891 From: header, and the sender address was obtained from -f or from an SMTP
1892 MAIL command, and the trusted user did not use -F to supply a sender
1893 name, $originator_name was incorrectly used when constructing a From:
1894 header. Furthermore, $originator_name was used for submission mode
1895 messages from external hosts without From: headers in a similar way,
1896 which is clearly wrong.
1898 PH/37 Added control=suppress_local_fixups.
1900 PH/38 When log_selector = +received_sender was set, and the addition of the
1901 sender made the log line's construction buffer exactly full, or one byte
1902 less than full, an overflow happened when the terminating "\n" was
1905 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1906 when the result of a list match is failure because a DNS lookup failed.
1908 PH/40 RM_COMMAND is now used in the building process.
1910 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1911 the "build-* directories that it finds.
1913 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1914 address in a domain literal was a prefix of an interface address.
1916 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1917 when verifying a sender address, unless rewrite_headers is false.
1919 PH/44 Wrote a long comment about why errors_to addresses are verified as
1920 recipients, not senders.
1922 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1923 the ratelimit ACL was added.
1925 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1927 PH/46 Added extra information about PostgreSQL errors to the error string.
1929 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1930 avoiding the need to install special zones in a real server. This is
1931 backwards compatible; if it can't find the fake resolver, it drops back.
1932 Thus, both old and new test suites can be run.
1934 TF/02 Added util/ratelimit.pl
1936 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1937 clock is set back in time.
1939 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1940 Candler <B.Candler@pobox.com>.
1942 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1943 OK for addresses that are the result of redirections.
1945 PH/48 A number of further additions for the benefit of the new test suite,
1946 including a fake gethostbyname() that interfaces to the fake DNS resolver
1949 TF/06 The fix for widen_domains has also been applied to qualify_single and
1950 search_parents which are the other dnslookup options that can cause
1953 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1956 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1958 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1959 service" if it encountered an MX record with an empty target hostname.
1960 The message is now "an MX or SRV record indicated no SMTP service".
1962 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1963 if the DNS of the sending site is misconfigured. This is quite a
1964 common situation. This change restores the behaviour of treating a
1965 helo verification defer as a failure.
1967 PH/51 If self=fail was set on a router, the bounce message did not include the
1968 actual error message.
1974 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1976 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1977 command itself ended in error, the underlying message about the transport
1978 filter timeout was being overwritten with the pipe command error. Now the
1979 underlying error message should be appended to the second error message.
1981 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1983 PH/02 Reduce the amount of output that "make" produces by default. Full output
1984 can still be requested.
1986 PH/03 The warning log line about a condition test deferring for a "warn" verb
1987 was being output only once per connection, rather than after each
1988 occurrence (because it was using the same function as for successful
1989 "warn" verbs). This seems wrong, so I have changed it.
1991 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1992 it should not have, which might have caused a crash in the right
1993 circumstances, but probably never did.
1995 PH/04 Installed a modified version of Tony Finch's patch to make submission
1996 mode fix the return path as well as the Sender: header line, and to
1997 add a /name= option so that you can make the user's friendly name appear
2000 TF/03 Added the control = fakedefer ACL modifier.
2002 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2003 Mark Lowes for thorough testing.
2005 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2007 TK/03 Merged latest SRS patch from Miles Wilton.
2009 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2010 with the definition in sysexits.h (which is #included earlier).
2011 Fortunately, Exim does not actually use EX_OK. The code used to try to
2012 preserve the sysexits.h value, by assumimg that macro definitions were
2013 scanned for macro replacements. I have been disabused of this notion,
2014 so now the code just undefines EX_OK before #including unistd.h.
2016 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2017 in the smtp transport. When a block could not be written in a single
2018 write() function, the timeout was being re-applied to each part-write.
2019 This seems wrong - if the receiver was accepting one byte at a time it
2020 would take for ever. The timeout is now adjusted when this happens. It
2021 doesn't have to be particularly precise.
2023 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2024 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2026 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2027 without a domain if the check to <postmaster@domain> fails.
2029 SC/01 Eximstats: added -xls and the ability to specify output files
2030 (patch written by Frank Heydlauf).
2032 SC/02 Eximstats: use FileHandles for outputing results.
2034 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2036 SC/04 Eximstats: fixed display of large numbers with -nvr option
2038 SC/05 Eximstats: fixed merging of reports with empty tables.
2040 SC/06 Eximstats: added the -include_original_destination flag
2042 SC/07 Eximstats: removed tabs and trailing whitespace.
2044 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2046 TK/06 MBOX spool code: Add real "From " MBOX separator line
2047 so the .eml file is really in mbox format (even though
2048 most programs do not really care). Patch from Alex Miller.
2050 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2051 The latter is generated from $received_to and is only set if the
2052 message has one envelope recipient. SA can use these headers,
2053 obviously out-of-the-box. Patch from Alex Miller.
2055 PH/08 The ${def test on a variable was returning false if the variable's
2056 value was "0", contrary to what the specification has always said!
2057 The result should be true unless the variable is empty.
2059 PH/09 The syntax error of a character other than { following "${if
2060 def:variable_name" (after optional whitespace) was not being diagnosed.
2061 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2062 accidental colon was present, for example, could give incorrect results.
2064 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2065 result is used (not including appendfile, where other changes are about
2068 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2069 This involved changing a lot of size variables from int to off_t. It
2070 should work with maildirs and everything.
2072 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2073 spamd dying while we are connected to it.
2075 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2078 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2079 function for BDB 4.3.
2081 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2083 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2084 the sources for PCRE have been re-arranged and more files are now
2087 PH/15 The code I had for printing potentially long long variables in PH/11
2088 above was not the best (it lost precision). The length of off_t variables
2089 is now inspected at build time, and an appropriate printing format (%ld
2090 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2091 to be "long long int" or "long int". This is needed for the internal
2092 formatting function string_vformat().
2094 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2095 the configuration file to be ":syslog", then the script "guesses" where
2096 the logs files are, rather than using the compiled in default. In our
2097 case the guess is not the same as the compiled default, so the script
2098 suddenly stopped working when I started to use syslog. The patch checks
2099 to see if log_file_path is "". If so, it attempts to read it from exim
2100 with no configuration file to get the compiled in version, before it
2101 falls back to the previous guessing code."
2103 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2104 implementing BATV in an Exim configuration. See NewStuff for the gory
2107 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2108 Makefile that are specific to HP-UX.
2110 PH/18 If the "use_postmaster" option was set for a recipient callout together
2111 with the "random" option, the postmaster address was used as the MAIL
2112 FROM address for the random test, but not for the subsequent recipient
2113 test. It is now used for both.
2115 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2116 patch removes a few documentation additions to RFC 3028, because the
2117 latest draft now contains them. It adds the new en;ascii-case comparator
2118 and a new error check for 8bit text in MIME parts. Comparator and
2119 require names are now matched exactly. I enabled the subaddress
2120 extension, but it is not well tested yet (read: it works for me)."
2122 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2123 rework some of the code of TK/09 above to avoid the hardwired use of
2124 "%lld" and "long long". Replaced the call to snprintf() with a call to
2127 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2128 records point to non-existent hosts", "retry timeout exceeded", and
2129 "retry time not reached for any host after a long failure period".
2131 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2132 experimental DomainKeys support:
2134 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2135 (2) On an error such as an illegally used "control", the wrong name for
2136 the control was given.
2138 These problems did NOT occur unless DomainKeys support was compiled.
2140 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2142 PH/24 Added ${if match_ip condition.
2144 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2145 databases so that it will be absolutely obvious if a crash occurs in the
2146 DB library. This is a regular occurrence (often caused by mis-matched
2149 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2150 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2151 picked up on a user's system that detects such things. There doesn't seem
2152 to be a gcc warning option for this - only an attribute that has to be
2153 put on the function's prototype. It seems that in Fedora Core 4 they have
2154 set this on a number of new functions. No doubt there will be more in due
2157 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2158 specify a transport. However, if an address that was verified by such a
2159 router was the subject of a callout, Exim crashed because it tried to
2160 read the rcpt_include_affixes from the non-existent transport. Now it
2161 just assumes that the setting of that option is false. This bug was
2162 introduced by 4.51/PH/31.
2164 PH/28 Changed -d+all to exclude +memory, because that information is very
2165 rarely of interest, but it makes the output a lot bigger. People tend to
2166 do -d+all out of habit.
2168 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2169 code in os-type was giving problems when libc.so lives in lib64, like on
2172 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2173 aren't the modern standard, and it seems that some systems' include files
2174 don't always have them. Exim was already checking for some of the newer
2175 ones like T_AAAA, and defining it itself. I've added checks for all the
2176 record types that Exim uses.
2178 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2179 not automatically generating a new one, as it is supposed to. This
2180 prevented TLS from working. If the file did exist, but contained invalid
2181 data, a new version was generated, as expected. It was only the case of a
2182 non-existent file that was broken.
2184 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2185 with a change in libdomainkeys > 0.64.
2187 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2188 from DNS. If the selector record carries the flag, it now has
2189 precedence over the domain-wide flag.
2191 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2193 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2194 the use of a port name that isn't defined in /etc/services) occurred, the
2195 message was deferred as in a normal delivery, and thus remained on the
2196 spool, instead of being failed because of the mua_wrapper setting. This
2197 is now fixed, and I tidied up some of the mua_wrapper messages at the
2200 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2201 the messages in a hash of arrays rather than using individual hashes.
2202 This is a bit cleaner and results in dramatic memory savings, albeit
2203 at a slight CPU cost.
2205 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2206 as requested by Marc Sherman.
2208 SC/10 Eximstats: added histograms for user specified patterns as requested
2211 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2213 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2214 fopen() in the content-scanning modules that did not already have it.
2220 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2221 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2223 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2225 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2227 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2228 to test Sieve filters that use "vacation".
2230 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2231 that changes the way the GnuTLS parameters are stored in the cache file.
2232 The new format can be generated externally. For backward compatibility,
2233 if the data in the cache doesn't make sense, Exim assumes it has read an
2234 old-format file, and it generates new data and writes a new file. This
2235 means that you can't go back to an older release without removing the
2238 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2239 work if there are any delivery delays because "one_time" forces the
2240 parent to be marked "delivered", so its unseen clone is never tried
2241 again. For this reason, Exim now forbids the simultaneous setting of
2244 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2245 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2246 entry if you want to know the details. The fix, however, neglected to
2247 consider the case when local delivery batching is involved. The test for
2248 "previously delivered" was not happening when checking to see if an
2249 address could be batched with a previous (undelivered) one; under
2250 certain circumstances this could lead to multiple deliveries to the same
2253 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2254 in its include files, and this causes problems building Exim.
2256 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2257 header_syntax) but Exim was just ignoring anything given after a slash.
2258 In particular, this caused confusion with an attempt to use "verify =
2259 reverse_host_lookup/defer_ok". An error is now given when options are
2260 supplied for verify items that do not have them. (Maybe reverse_host_
2261 lookup should have a defer_ok option, but that's a different point.)
2263 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2264 defined by RFC 821) to 2048, because there were problems with some AUTH
2265 commands, and RFC 1869 says the size should be increased for extended
2266 SMTP commands that take arguments.
2268 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2271 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2272 "unknown" error; now it says that the functionality isn't in the binary.
2274 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2275 an address' error message when a string expansion fails (syntax or
2276 whatever). Otherwise the password may appear in the log. Following change
2277 PH/42 below, there is no longer a chance of it appearing in a bounce
2280 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2282 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2283 of its addresses was ever tried. (Bugzilla bug #2.)
2285 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2286 the result incorrectly in the debug output. (It correctly added a newline
2287 to what was transported.)
2289 TF/01 Added $received_time.
2291 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2292 commented out examples of how to interface to a virus scanner and to
2293 SpamAssassin. Also added commented examples of av_scanner and
2294 spamd_address settings.
2296 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2297 and controls are allowed in which ACLs. There were a couple of minor
2298 errors. Some of the entries in the conditions table (which is a table of
2299 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2300 negation of where the condition IS allowed.
2302 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2304 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2305 header file does not have a version number, so I've had to invent a new
2306 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2307 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2308 radiusclient), but it was contributed by a Radius user.
2310 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2311 files or directories, for OpenSSL.
2313 PH/19 When an Exim process that is running as root has to create an Exim log
2314 file, it does so in a subprocess that runs as exim:exim so as to get the
2315 ownership right at creation (otherwise, other Exim processes might see
2316 the file with the wrong ownership). There was no test for failure of this
2317 fork() call, which would lead to the process getting stuck as it waited
2318 for a non-existent subprocess. Forks do occasionally fail when resources
2319 run out. I reviewed all the other calls to fork(); they all seem to check
2322 PH/20 When checking for unexpected SMTP input at connect time (before writing
2323 the banner), Exim was not dealing correctly with a non-positive return
2324 from the read() function. If the client had disconnected by this time,
2325 the result was a log entry for a synchronization error with an empty
2326 string after "input=" when read() returned zero. If read() returned -1
2327 (an event I could not check), uninitialized data bytes were printed.
2328 There were reports of junk text (parts of files, etc) appearing after
2331 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2333 PH/22 Added support for macro redefinition, and (re)definition in between
2334 driver and ACL definitions.
2336 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2337 forgetting to use the resulting value; it was using the unexpanded value.
2339 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2340 hadn't been configured. The fix is from Juergen Kreileder, who
2341 understands it better than I do:
2343 "Here's what I see happening with three configured cyrus_sasl
2344 authenticators configured (plain, login, cram-md5):
2346 On startup auth_cyrus_sasl_init() gets called for each of these.
2347 This means three calls to sasl_listmech() without a specified mech_list.
2348 => SASL tests which mechs of all available mechs actually work
2349 => three warnings about OTP not working
2350 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2352 With the patch, sasl_listmech() also gets called three times. But now
2353 SASL's mech_list option is set to the server_mech specified in the the
2354 authenticator. Or in other words, the answer from sasl_listmech()
2355 gets limited to just the mech you're testing for (which is different
2357 => the return list contains just 'plain' or 'login', 'cram-md5' or
2358 nothing depending on the value of ob->server_mech.
2360 I've just tested the patch: Authentication still works fine,
2361 unavailable mechs specified in the exim configuration are still
2362 caught, and the auth.log warnings about OTP are gone."
2364 PH/25 When debugging is enabled, the contents of the command line are added
2365 to the debugging output, even when log_selector=+arguments is not
2368 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2369 answer is "GNU", and only if the return is "GNU/something" is the answer
2372 PH/27 $acl_verify_message is now set immediately after the failure of a
2373 verification in an ACL, and so is available in subsequent modifiers. In
2374 particular, the message can be preserved by coding like this:
2376 warn !verify = sender
2377 set acl_m0 = $acl_verify_message
2379 Previously, $acl_verify_message was set only while expanding "message"
2380 and "log_message" when a very denied access.
2382 PH/28 Modified OS/os.c-Linux with
2384 -#ifndef OS_LOAD_AVERAGE
2385 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2387 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2388 nomenclature these days.)
2390 PH/29 Installed patch from the Sieve maintainer that adds the options
2391 sieve_useraddress and sieve_subaddress to the redirect router.
2393 PH/30 In these circumstances:
2394 . Two addresses routed to the same list of hosts;
2395 . First host does not offer TLS;
2396 . First host accepts first address;
2397 . First host gives temporary error to second address;
2398 . Second host offers TLS and a TLS session is established;
2399 . Second host accepts second address.
2400 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2401 and peerdn, if requested) that were in fact used only for the second
2404 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2405 attention to any local part prefix or suffix that was matched by the
2406 router that accepted the address. It now behaves in the same way as it
2407 does for delivery: the affixes are removed from the local part unless
2408 rcpt_include_affixes is set on the transport.
2410 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2411 timeout during the DATA phase of an incoming message.
2413 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2414 applied a patch sanctioned by the Sieve maintainer.
2416 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2417 the uid or gid is negative. A case of a negative gid caused this to be
2418 noticed. The fix allows for either to be negative.
2420 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2421 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2424 PH/36 The change PH/12 above was broken. Fixed it.
2426 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2427 the grounds that routing the same address twice would always produce the
2428 same answer. This might have been true once, but it is certainly no
2429 longer true now. Routing a child address may depend on the previous
2430 routing that produced that child. Some complicated redirection strategies
2431 went wrong when messages had multiple recipients, and made Exim's
2432 behaviour dependent on the order in which the addresses were given.
2434 I have moved the duplicate checking until after the routing is complete.
2435 Exim scans the addresses that are assigned to local and remote
2436 transports, and removes any duplicates. This means that more work will be
2437 done, as duplicates will always all be routed, but duplicates are
2438 presumably rare, so I don't expect this is of any significance.
2440 For deliveries to pipes, files, and autoreplies, the duplicate checking
2441 still happens during the routing process, since they are not going to be
2444 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2445 It corrects a timeout issue with spamd. This is Ian's comment: "The
2446 background is that sometimes spamd either never reads data from a
2447 connection it has accepted, or it never writes response data. The exiscan
2448 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2449 blindly assumes that writes won't block so it may never time out."
2451 PH/39 Allow G after quota size as well as K and M.
2453 PH/40 The value set for $authenticated_id in an authenticator may not contain
2454 binary zeroes or newlines because the value is written to log lines and
2455 to spool files. There was no check on this. Now the value is run through
2456 the string_printing() function so that such characters are converted to
2457 printable escape sequences.
2459 PH/41 $message_linecount is a new variable that contains the total number of
2460 lines in the message. Compare $body_linecount, which is the count for the
2463 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2464 bounce and delay warning messages, except in certain special cases, which
2467 (a) An SMTP error message from a remote host;
2468 (b) A message specified in a :fail: redirection;
2469 (c) A message specified in a "fail" command in a system filter;
2470 (d) A message specified in a FAIL return from the queryprogram router;
2471 (e) A message specified by the cannot_route_message router option.
2473 In these cases only, Exim does include the error details in bounce and
2474 warning messages. There are also a few cases where bland messages such
2475 as "unrouteable address" or "local delivery error" are given.
2477 PH/43 $value is now also set for the "else" part of a ${run expansion.
2479 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2480 being worked on, but at least Exim now implements the latest version to
2483 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2484 process to complete was treated as a delivery failure, a timeout while
2485 writing the message to the pipe was logged, but erroneously treated as a
2486 successful delivery. Such timeouts include transport filter timeouts. For
2487 consistency with the overall process timeout, these timeouts are now
2488 treated as errors, giving rise to delivery failures by default. However,
2489 there is now a new Boolean option for the pipe transport called
2490 timeout_defer, which, if set TRUE, converts the failures into defers for
2491 both kinds of timeout. A transport filter timeout is now identified in
2494 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2495 systems where "make" and "gmake" are different, calling "gmake" at top
2496 level broke things. I've arranged for the value of $(MAKE) to be passed
2497 from the Makefile to this script so that it can call the same version of
2501 A note about Exim versions 4.44 and 4.50
2502 ----------------------------------------
2504 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2505 changes of various kinds. As a consequence, a big documentation update was
2506 needed. This delayed the release for rather longer than seemed good, especially
2507 in the light of a couple of (minor) security issues. Therefore, the changes
2508 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2509 release. So 4.44 and 4.50 are in effect two different branches that both start
2512 I have left the 4.50 change log unchanged; it contains all the changes since
2513 4.43. The change log for 4.44 is below; many of its items are identical to
2514 those for 4.50. This seems to be the most sensible way to preserve the
2515 historical information.
2521 1. Minor wording change to the doc/README.SIEVE file.
2523 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2524 computation of the current number of files was incorrect.
2526 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2527 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2530 4. Give more explanation in the error message when the command for a transport
2531 filter fails to execute.
2533 5. There are several places where Exim runs a non-Exim command in a
2534 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2535 was being done only for the command run by the queryprogram router. It is
2536 now done for all such subprocesses. The other cases are: ${run, transport
2537 filters, and the commands run by the lmtp and pipe transports.
2539 6. Added CONFIGURE_GROUP build-time option.
2541 7. Some older OS have a limit of 256 on the maximum number of file
2542 descriptors. Exim was using setrlimit() to set 1000 as a large value
2543 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2544 systems. I've change it so that if it can't get 1000, it tries for 256.
2546 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2547 was an oversight, and furthermore, ever since the addition of extra
2548 controls (e.g. 4.43/32), the checks on when to allow different forms of
2549 "control" were broken. There should now be diagnostics for all cases when a
2550 control that does not make sense is encountered.
2552 9. Added the /retain_sender option to "control=submission".
2554 10. $recipients is now available in the predata ACL (oversight).
2556 11. Tidy the search cache before the fork to do a delivery from a message
2557 received from the command line. Otherwise the child will trigger a lookup
2558 failure and thereby defer the delivery if it tries to use (for example) a
2559 cached ldap connection that the parent has called unbind on.
2561 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2562 of $address_data from the recipient verification was clobbered by the
2563 sender verification.
2565 13. The value of address_data from a sender verification is now available in
2566 $sender_address_data in subsequent conditions in the ACL statement.
2568 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2570 15. Added a new option "connect=<time>" to callout options, to set a different
2573 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2574 was its contents. (It was OK if the option was not defined at all.)
2576 17. A "Completed" log line is now written for messages that are removed from
2577 the spool by the -Mrm option.
2579 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2580 information about exactly what failed.
2582 19. Added -dd to debug only the daemon process.
2584 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2585 handles timeouts, both on the server side and network timeouts. Renamed the
2586 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2588 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2589 It is now set to "smtps".
2591 22. $host_address is now set to the target address during the checking of
2592 ignore_target_hosts.
2594 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2595 being passed; this would have caused $sender_host_name to have been used if
2596 matching the list had actually called for a host name (not very likely,
2597 since this list is usually IP addresses). A host name is now passed as
2600 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2601 code that specifies a non-restarting handler (typically sigaction() in
2602 modern systems) in an attempt to fix a rare and obscure crash bug.
2604 25. Narrowed the window for a race in the daemon that could cause it to ignore
2605 SIGCHLD signals. This is not a major problem, because they are used only to
2606 wake it up if nothing else does.
2608 26. A malformed maildirsize file could cause Exim to calculate negative values
2609 for the mailbox size or file count. Odd effects could occur as a result.
2610 The maildirsize information is now recalculated if the size or filecount
2613 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2614 support for a long time. Removed HAVE_SYS_VFS_H.
2616 28. Installed the latest version of exipick from John Jetmore.
2618 29. In an address list, if the pattern was not a regular expression, an empty
2619 subject address (from a bounce message) matched only if the pattern was an
2620 empty string. Non-empty patterns were not even tested. This was the wrong
2621 because it is perfectly reasonable to use an empty address as part of a
2622 database query. An empty address is now tested by patterns that are
2623 lookups. However, all the other forms of pattern expect the subject to
2624 contain a local part and a domain, and therefore, for them, an empty
2625 address still always fails if the pattern is not itself empty.
2627 30. Exim went into a mad DNS loop when attempting to do a callout where the
2628 host was specified on an smtp transport, and looking it up yielded more
2629 than one IP address.
2631 31. Re-factored the code for checking spool and log partition space into a
2632 function that finds that data and another that does the check. The former
2633 is then used to implement four new variables: $spool_space, $log_space,
2634 $spool_inodes, and $log_inodes.
2636 32. The RFC2047 encoding function was originally intended for short strings
2637 such as real names; it was not keeping to the 75-character limit for
2638 encoded words that the RFC imposes. It now respects the limit, and
2639 generates multiple encoded words if necessary. To be on the safe side, I
2640 have increased the buffer size for the ${rfc2047: expansion operator from
2643 33. It is now permitted to omit both strings after an "if" condition; if the
2644 condition is true, the result is "true". As before, when the second string
2645 is omitted, a false condition yields an empty string. This makes it less
2646 cumbersome to write custom ACL and router conditions.
2648 34. Failure to deliver a bounce message always caused it to be frozen, even if
2649 there was an errors_to setting on the router. The errors_to setting is now
2652 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2653 canonical form (fully expanded) before being placed in
2654 $sender_host_address.
2656 36. The table in the code that translates DNS record types into text (T_A to
2657 "A" for instance) was missing entries for NS and CNAME. It is just possible
2658 that this could have caused confusion if both these types were looked up
2659 for the same domain, because the text type is used as part of Exim's
2660 per-process caching. But the chance of anyone hitting this buglet seems
2663 37. The dnsdb lookup has been extended in a number of ways.
2665 (1) There is a new type, "zns", which walks up the domain tree until it
2666 finds some nameserver records. It should be used with care.
2668 (2) There is a new type, "mxh", which is like "mx" except that it returns
2669 just the host names, not the priorities.
2671 (3) It is now possible to give a list of domains (or IP addresses) to be
2672 looked up. The behaviour when one of the lookups defers can be
2673 controlled by a keyword.
2675 (4) It is now possible to specify the separator character for use when
2676 multiple records are returned.
2678 38. The dnslists ACL condition has been extended: it is now possible to supply
2679 a list of IP addresses and/or domains to be looked up in a particular DNS
2682 39. Added log_selector=+queue_time_overall.
2684 40. When running the queue in the test harness, wait just a tad after forking a
2685 delivery process, to get repeatability of debugging output.
2687 41. Include certificate and key file names in error message when GnuTLS fails
2688 to set them up, because the GnuTLS error message doesn't include the name
2689 of the failing file when there is a problem reading it.
2691 42. Allow both -bf and -bF in the same test run.
2693 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2695 44. The "Exiscan patch" is now merged into the mainline Exim source.
2697 45. Sometimes the final signoff response after QUIT could fail to get
2698 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2699 < 0 before doing a fflush(). This bug looks as though it goes back to the
2700 introduction of TLS in release 3.20, but "sometimes" must have been rare
2701 because the tests only now provoked it.
2703 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2704 (this can affect the format of dates).
2706 47. exim_tidydb, when checking for the continued existence of a message for
2707 which it has found a message-specific retry record, was not finding
2708 messages that were in split spool directories. Consequently, it was
2709 deleting retry records that should have stayed in existence.
2711 48. Steve fixed some bugs in eximstats.
2713 49. The SPA authentication driver was not abandoning authentication and moving
2714 on to the next authenticator when an expansion was forced to fail,
2715 contradicting the general specification for all authenticators. Instead it
2716 was generating a temporary error. It now behaves as specified.
2718 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2719 (the order specifies the preference for clients). The order is now AES256,
2720 AES128, 3DES, ARCFOUR128.
2722 51. Small patch to Sieve code - explicitly set From: when generating an
2725 52. Exim crashed if a remote delivery caused a very long error message to be
2726 recorded - for instance if somebody sent an entire SpamAssassin report back
2727 as a large number of 550 error lines. This bug was coincidentally fixed by
2728 increasing the size of one of Exim's internal buffers (big_buffer) that
2729 happened as part of the Exiscan merge. However, to be on the safe side, I
2730 have made the code more robust (and fixed the comments that describe what
2733 53. Now that there can be additional text after "Completed" in log lines (if
2734 the queue_time_overall log selector is set), a one-byte patch to exigrep
2735 was needed to allow it to recognize "Completed" as not the last thing in
2738 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2739 patch that reportedly fixes this has been added. I am not expert enough to
2740 create a test for it. This is what the patch creator wrote:
2742 "I found a little strange behaviour of ldap code when working with
2743 Windows 2003 AD Domain, where users was placed in more than one
2744 Organization Units. When I tried to give exim partial DN, the exit code
2745 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2746 But simultaneously result of request was absolutely normal ldap result,
2747 so I produce this patch..."
2749 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2750 REFERENCE, so I have modified the code to exclude the patch when that macro
2753 55. Some experimental protocols are using DNS PTR records for new purposes. The
2754 keys for these records are domain names, not reversed IP addresses. The
2755 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2756 leaves it alone. Component reversal etc. now happens only for IP addresses.
2759 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2761 57. Double the size of the debug message buffer (to 2048) so that more of very
2762 long debug lines gets shown.
2764 58. The exicyclog utility now does better if the number of log files to keep
2765 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2767 59. Two changes related to the smtp_active_hostname option:
2769 (1) $smtp_active_hostname is now available as a variable.
2770 (2) The default for smtp_banner uses $smtp_active_hostname instead
2771 of $primary_hostname.
2773 60. The host_aton() function is supposed to be passed a string that is known
2774 to be a valid IP address. However, in the case of IPv6 addresses, it was
2775 not checking this. This is a hostage to fortune. Exim now panics and dies
2776 if the condition is not met. A case was found where this could be provoked
2777 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2778 components; fortuitously, this particular loophole had already been fixed
2779 by change 4.50/55 above.
2781 If there are any other similar loopholes, the new check in host_aton()
2782 itself should stop them being exploited. The report I received stated that
2783 data on the command line could provoke the exploit when Exim was running as
2784 exim, but did not say which command line option was involved. All I could
2785 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2786 running as the user.
2789 61. There was a buffer overflow vulnerability in the SPA authentication code
2790 (which came originally from the Samba project). I have added a test to the
2791 spa_base64_to_bits() function which I hope fixes it.
2794 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2795 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2797 63. The daemon start-up calls getloadavg() while still root for those OS that
2798 need the first call to be done as root, but it missed one case: when
2799 deliver_queue_load_max is set with deliver_drop_privilege. This is
2800 necessary for the benefit of the queue runner, because there is no re-exec
2801 when deliver_drop_privilege is set.
2803 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2804 This has been fixed.
2806 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2807 was in use, was not putting the data itself into the right store pool;
2808 consequently, it could be overwritten for a subsequent message in the same
2809 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2812 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2814 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2815 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2816 the function were treating the return as a boolean value, which happened to
2817 work because 0=false and not-0=true, but is not correct code.
2819 68. The host_aton() function was not handling scoped IPv6 addresses (those
2820 with, for example, "%eth0" on the end) correctly.
2822 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2823 negated items (that is, ~something) in unsigned ints. Some compilers
2824 apparently mutter when there is no cast.
2826 70. If an address verification called from an ACL failed, and did not produce a
2827 user-specific message (i.e. there was only a "system" message), nothing was
2828 put in $acl_verify_message. In this situation, it now puts the system
2831 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2832 session; change 4.31/43 added the unwanted input to the log line - except
2833 that it did not do this in the start of session case. It now does.
2835 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2836 This is wrong and can cause the other end to generate a synchronization
2837 error if it is another Exim or anything else that does the synchronization
2838 check. A QUIT command is no longer sent after a timeout.
2840 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2841 during host lookups.
2843 74. The defer_ok option of callout verification was not working if it was used
2844 when verifying addresses in header lines, that is, for this case:
2846 verify = header_sender/callout=defer_ok
2848 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2849 those file descriptors could be used for SMTP connections. If anything
2850 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2851 could be sent to the SMTP client, causing chaos. The daemon now opens
2852 stdin, stdout, and stderr to /dev/null when it puts itself into the
2855 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2856 log by default. The user can override this with suitable Perl magic.
2858 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2859 the log message when discard triggers, was not working for the DATA ACL or
2860 for the non-SMTP ACL.
2862 78. Error message wording change in sieve.c.
2864 79. If smtp_accept_max_per_host was set, the number of connections could be
2865 restricted to fewer than expected, because the daemon was trying to set up
2866 a new connection before checking whether the processes handling previous
2867 connections had finished. The check for completed processes is now done
2868 earlier. On busy systems, this bug wouldn't be noticed because something
2869 else would have woken the daemon, and it would have reaped the completed
2872 80. If a message was submitted locally by a user whose login name contained one
2873 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2874 It caused a spool format error. I have fixed the spool reading code. A
2875 related problem was that the "from" clause in the Received: line became
2876 illegal because of the space(s). It is now covered by ${quote_local_part.
2878 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2881 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2883 83. Patch from Sieve maintainer to fix unterminated string problem in
2884 "vacation" handling.
2886 84. Some minor changes to the Linux configuration files to help with other
2887 OS variants using glibc.
2889 85. One more patch for Sieve to update vacation handling to latest spec.
2892 ----------------------------------------------------
2893 See the note above about the 4.44 and 4.50 releases.
2894 ----------------------------------------------------
2900 1. Change 4.43/35 introduced a bug that caused file counts to be
2901 incorrectly computed when quota_filecount was set in an appendfile
2904 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2905 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2908 3. Give more explanation in the error message when the command for a transport
2909 filter fails to execute.
2911 4. There are several places where Exim runs a non-Exim command in a
2912 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2913 was being done only for the command run by the queryprogram router. It is
2914 now done for all such subprocesses. The other cases are: ${run, transport
2915 filters, and the commands run by the lmtp and pipe transports.
2917 5. Some older OS have a limit of 256 on the maximum number of file
2918 descriptors. Exim was using setrlimit() to set 1000 as a large value
2919 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2920 systems. I've change it so that if it can't get 1000, it tries for 256.
2922 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2923 was an oversight, and furthermore, ever since the addition of extra
2924 controls (e.g. 4.43/32), the checks on when to allow different forms of
2925 "control" were broken. There should now be diagnostics for all cases when a
2926 control that does not make sense is encountered.
2928 7. $recipients is now available in the predata ACL (oversight).
2930 8. Tidy the search cache before the fork to do a delivery from a message
2931 received from the command line. Otherwise the child will trigger a lookup
2932 failure and thereby defer the delivery if it tries to use (for example) a
2933 cached ldap connection that the parent has called unbind on.
2935 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2936 of $address_data from the recipient verification was clobbered by the
2937 sender verification.
2939 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2940 was its contents. (It was OK if the option was not defined at all.)
2942 11. A "Completed" log line is now written for messages that are removed from
2943 the spool by the -Mrm option.
2945 12. $host_address is now set to the target address during the checking of
2946 ignore_target_hosts.
2948 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2949 being passed; this would have caused $sender_host_name to have been used if
2950 matching the list had actually called for a host name (not very likely,
2951 since this list is usually IP addresses). A host name is now passed as
2954 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2955 code that specifies a non-restarting handler (typically sigaction() in
2956 modern systems) in an attempt to fix a rare and obscure crash bug.
2958 15. Narrowed the window for a race in the daemon that could cause it to ignore
2959 SIGCHLD signals. This is not a major problem, because they are used only to
2960 wake it up if nothing else does.
2962 16. A malformed maildirsize file could cause Exim to calculate negative values
2963 for the mailbox size or file count. Odd effects could occur as a result.
2964 The maildirsize information is now recalculated if the size or filecount
2967 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2968 support for a long time. Removed HAVE_SYS_VFS_H.
2970 18. Updated exipick to current release from John Jetmore.
2972 19. Allow an empty sender to be matched against a lookup in an address list.
2973 Previously the only cases considered were a regular expression, or an
2976 20. Exim went into a mad DNS lookup loop when doing a callout where the
2977 host was specified on the transport, if the DNS lookup yielded more than
2980 21. The RFC2047 encoding function was originally intended for short strings
2981 such as real names; it was not keeping to the 75-character limit for
2982 encoded words that the RFC imposes. It now respects the limit, and
2983 generates multiple encoded words if necessary. To be on the safe side, I
2984 have increased the buffer size for the ${rfc2047: expansion operator from
2987 22. Failure to deliver a bounce message always caused it to be frozen, even if
2988 there was an errors_to setting on the router. The errors_to setting is now
2991 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2992 canonical form (fully expanded) before being placed in
2993 $sender_host_address.
2995 24. Updated eximstats to version 1.33
2997 25. Include certificate and key file names in error message when GnuTLS fails
2998 to set them up, because the GnuTLS error message doesn't include the name
2999 of the failing file when there is a problem reading it.
3001 26. Expand error message when OpenSSL has problems setting up cert/key files.
3004 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3005 (this can affect the format of dates).
3007 28. exim_tidydb, when checking for the continued existence of a message for
3008 which it has found a message-specific retry record, was not finding
3009 messages that were in split spool directories. Consequently, it was
3010 deleting retry records that should have stayed in existence.
3012 29. eximstats updated to version 1.35
3013 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3014 1.35 - bugfix such that pie charts by volume are generated correctly
3016 30. The SPA authentication driver was not abandoning authentication and moving
3017 on to the next authenticator when an expansion was forced to fail,
3018 contradicting the general specification for all authenticators. Instead it
3019 was generating a temporary error. It now behaves as specified.
3021 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3022 (the order specifies the preference for clients). The order is now AES256,
3023 AES128, 3DES, ARCFOUR128.
3025 31. Small patch to Sieve code - explicitly set From: when generating an
3028 32. Exim crashed if a remote delivery caused a very long error message to be
3029 recorded - for instance if somebody sent an entire SpamAssassin report back
3030 as a large number of 550 error lines. This bug was coincidentally fixed by
3031 increasing the size of one of Exim's internal buffers (big_buffer) that
3032 happened as part of the Exiscan merge. However, to be on the safe side, I
3033 have made the code more robust (and fixed the comments that describe what
3036 33. Some experimental protocols are using DNS PTR records for new purposes. The
3037 keys for these records are domain names, not reversed IP addresses. The
3038 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3039 leaves it alone. Component reversal etc. now happens only for IP addresses.
3042 34. The host_aton() function is supposed to be passed a string that is known
3043 to be a valid IP address. However, in the case of IPv6 addresses, it was
3044 not checking this. This is a hostage to fortune. Exim now panics and dies
3045 if the condition is not met. A case was found where this could be provoked
3046 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3047 components; fortuitously, this particular loophole had already been fixed
3048 by change 4.50/55 or 4.44/33 above.
3050 If there are any other similar loopholes, the new check in host_aton()
3051 itself should stop them being exploited. The report I received stated that
3052 data on the command line could provoke the exploit when Exim was running as
3053 exim, but did not say which command line option was involved. All I could
3054 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3055 running as the user.
3058 35. There was a buffer overflow vulnerability in the SPA authentication code
3059 (which came originally from the Samba project). I have added a test to the
3060 spa_base64_to_bits() function which I hope fixes it.
3063 36. The daemon start-up calls getloadavg() while still root for those OS that
3064 need the first call to be done as root, but it missed one case: when
3065 deliver_queue_load_max is set with deliver_drop_privilege. This is
3066 necessary for the benefit of the queue runner, because there is no re-exec
3067 when deliver_drop_privilege is set.
3069 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3070 was in use, was not putting the data itself into the right store pool;
3071 consequently, it could be overwritten for a subsequent message in the same
3072 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3075 38. Sometimes the final signoff response after QUIT could fail to get
3076 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3077 < 0 before doing a fflush(). This bug looks as though it goes back to the
3078 introduction of TLS in release 3.20, but "sometimes" must have been rare
3079 because the tests only now provoked it.
3085 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3086 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3087 Now it returns an integer. A number of calls were still expecting a T/F
3088 return. Fortuitously, in all cases, the tests worked in OK situations,
3089 which is the norm. However, things would have gone wrong on any write
3090 failures on the smtp file descriptor. This function is used when sending
3091 messages over SMTP and also when doing verify callouts.
3093 2. When Exim is called to do synchronous delivery of a locally submitted
3094 message (the -odf or -odi options), it no longer closes stderr before doing
3097 3. Implemented the mua_wrapper option.
3099 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3101 5. Implemented the functions header_remove(), header_testname(),
3102 header_add_at_position(), and receive_remove_recipient(), and exported them
3105 6. If an ACL "warn" statement specified the addition of headers, Exim already
3106 inserted X-ACL-Warn: at the start if there was no header name. However, it
3107 was not making this test for the second and subsequent header lines if
3108 there were newlines in the string. This meant that an invalid header could
3109 be inserted if Exim was badly configured.
3111 7. Allow an ACL "warn" statement to add header lines at the start or after all
3112 the Received: headers, as well as at the end.
3114 8. Added the rcpt_4xx retry error code.
3116 9. Added postmaster_mailfrom=xxx to callout verification option.
3118 10. Added mailfrom=xxxx to the callout verification option, for verify=
3121 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3122 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3124 12. Inserted some casts to stop certain compilers warning when using pointer
3125 differences as field lengths or precisions in printf-type calls (mostly
3126 affecting debugging statements).
3128 13. Added optional readline() support for -be (dynamically loaded).
3130 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3131 same clock tick as a message's arrival, so that its received time was the
3132 same as the "first fail" time on the retry record, and that message
3133 remained on the queue past the ultimate address timeout, every queue runner
3134 would try a delivery (because it was past the ultimate address timeout) but
3135 after another failure, the ultimate address timeout, which should have then
3136 bounced the address, did not kick in. This was a "< instead of <=" error;
3137 in most cases the first failure would have been in the next clock tick
3138 after the received time, and all would be well.
3140 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3141 being recognized when the domain list was tested by the match_domain
3142 condition in an expansion string.
3144 16. Added the ${str2b64: operator.
3146 17. Exim was always calling setrlimit() to set a large limit for the number of
3147 processes, without checking whether the existing limit was already
3148 adequate. (It did check for the limit on file descriptors.) Furthermore,
3149 errors from getrlimit() and setrlimit() were being ignored. Now they are
3150 logged to the main and panic logs, but Exim does carry on, to try to do its
3151 job under whatever limits there are.
3153 18. Imported PCRE 5.0.
3155 19. Trivial typo in log message " temporarily refused connection" (the leading
3158 20. If the log selector return_path_on_delivery was set and an address was
3159 redirected to /dev/null, the delivery process crashed because it assumed
3160 that a return path would always be set for a "successful" delivery. In this
3161 case, the whole delivery is bypassed as an optimization, and therefore no
3164 21. Internal re-arrangement: the function for sending a challenge and reading
3165 a response while authentication was assuming a zero-terminated challenge
3166 string. It's now changed to take a pointer and a length, to allow for
3167 binary data in such strings.
3169 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3171 23. Exim was not respecting finduser_retries when seeking the login of the
3172 uid under which it was called; it was always trying 10 times. (The default
3173 setting of finduser_retries is zero.) Also, it was sleeping after the final
3174 failure, which is pointless.
3176 24. Implemented tls_on_connect_ports.
3178 25. Implemented acl_smtp_predata.
3180 26. If the domain in control=submission is set empty, Exim assumes that the
3181 authenticated id is a complete email address when it generates From: or
3182 Sender: header lines.
3184 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3185 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3186 chown and chgrp in /bin and hostname in /usr/bin.
3188 28. Exim was keeping the "process log" file open after each use, just as it
3189 does for the main log. This opens the possibility of it remaining open for
3190 long periods when the USR1 signal hits a daemon. Occasional processlog
3191 errors were reported, that could have been caused by this. Anyway, it seems
3192 much more sensible not to leave this file open at all, so that is what now
3195 29. The long-running daemon process does not normally write to the log once it
3196 has entered its main loop, and it closes the log before doing so. This is
3197 so that log files can straightforwardly be renamed and moved. However,
3198 there are a couple of unusual error situations where the daemon does write
3199 log entries, and I had neglected to close the log afterwards.
3201 30. The text of an SMTP error response that was received during a remote
3202 delivery was being truncated at 512 bytes. This is too short for some of
3203 the long messages that one sometimes sees. I've increased the limit to
3206 31. It is now possible to make retry rules that apply only when a message has a
3207 specific sender, in particular, an empty sender.
3209 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3210 it possible to be selective about when SMTP synchronization is enforced.
3212 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3214 32. Implemented hosts_connection_nolog.
3216 33. Added an ACL for QUIT.
3218 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3221 35. Added mailbox_size and mailbox_filecount to appendfile.
3223 36. Added control = no_multiline_responses to ACLs.
3225 37. There was a bug in the logic of the code that waits for the clock to tick
3226 in the case where the clock went backwards by a substantial amount such
3227 that the microsecond fraction of "now" was more than the microsecond
3228 fraction of "then" (but the whole seconds number was less).
3230 38. Added support for the libradius Radius client library this is found on
3231 FreeBSD (previously only the radiusclient library was supported).
3237 1. When certain lookups returned multiple values in the form name=value, the
3238 quoting of the values was not always being done properly. Specifically:
3239 (a) If the value started with a double quote, but contained no whitespace,
3241 (b) If the value contained whitespace other than a space character (i.e.
3242 tabs or newlines or carriage returns) it was not quoted.
3243 This fix has been applied to the mysql and pgsql lookups by writing a
3244 separate quoting function and calling it from the lookup code. The fix
3245 should probably also be applied to nisplus, ibase and oracle lookups, but
3246 since I cannot test any of those, I have not disturbed their existing code.
3248 2. A hit in the callout cache for a specific address caused a log line with no
3249 reason for rejecting RCPT. Now it says "Previous (cached) callout
3250 verification failure".
3252 3. There was an off-by-one bug in the queryprogram router. An over-long
3253 return line was truncated at 256 instead of 255 characters, thereby
3254 overflowing its buffer with the terminating zero. As well as fixing this, I
3255 have increased the buffer size to 1024 (and made a note to document this).
3257 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3258 when Exim is waiting for an SMTP response from a remote server, Exim
3259 restarts its select() call on the socket, thereby resetting its timeout.
3260 This is not a problem when such interrupts are rare. Somebody set up a cron
3261 job to run exiwhat every 2 minutes, which is less than the normal select()
3262 timeout (5 or 10 minutes). This meant that the select() timeout never
3263 kicked in because it was always reset. I have fixed this by comparing the
3264 time when an interrupt arrives with the time at the start of the first call
3265 to select(). If more time than the timeout has elapsed, the interrupt is
3266 treated as a timeout.
3268 5. Some internal re-factoring in preparation for the addition of Sieve
3269 extensions (by MH). In particular, the "personal" test is moved to a
3270 separate function, and given an option for scanning Cc: and Bcc: (which is
3271 not set for Exim filters).
3273 6. When Exim created an email address using the login of the caller as the
3274 local part (e.g. when creating a From: or Sender: header line), it was not
3275 quoting the local part when it contained special characters such as @.
3277 7. Installed new OpenBSD configuration files.
3279 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3280 try to make them clearer.
3282 9. Callout options, other than the timeout value, were being ignored when
3283 verifying sender addresses in header lines. For example, when using
3285 verify = header_sender/callout=no_cache
3287 the cache was (incorrectly) being used.
3289 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3290 only the Cygwin environment.
3292 11. When return_path_on_delivery was set as a log selector, if different remote
3293 addresses in the same message used different return paths and parallel
3294 remote delivery occurred, the wrong values would sometimes be logged.
3295 (Whenever a remote delivery process finished, the return path value from
3296 the most recently started remote delivery process was logged.)
3298 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3299 lines when AUTH and/or TLS are in use. This is the "received protocol"
3300 field. Exim used to use "asmtp" for authenticated SMTP, without any
3301 indication (in the protocol name) for TLS use. Now it follows the RFC and
3302 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3303 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3304 names appear in log lines as well as in Received: header lines.
3306 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3307 extensions, and comparison tests, and to fix some bugs.
3309 14. Changes to the "personal" filter test:
3311 (1) The test was buggy in that it was just doing the equivalent of
3312 "contains" tests on header lines. For example, if a user's address was
3313 anne@some.where, the "personal" test would incorrectly be true for
3315 To: susanne@some.where
3317 This test is now done by extracting each address from the header in turn,
3318 and checking the entire address. Other tests that are part of "personal"
3319 are now done using regular expressions (for example, to check local parts
3320 of addresses in From: header lines).
3322 (2) The list of non-personal local parts in From: addresses has been
3323 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3324 taken from the Sieve specification recommendations.
3326 (3) If the message contains any header line starting with "List-" it is
3327 treated as non-personal.
3329 (4) The test for "circular" in the Subject: header line has been removed
3330 because it now seems ill-conceived.
3332 15. Minor typos in src/EDITME comments corrected.
3334 16. Installed latest exipick from John Jetmore.
3336 17. If headers_add on a router specified a text string that was too long for
3337 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3338 of string_sprintf() is now avoided.
3340 18. $message_body_size was not set (it was always zero) when running the DATA
3341 ACL and the local_scan() function.
3343 19. For the "mail" command in an Exim filter, no default was being set for
3344 the once_repeat time, causing a random time value to be used if "once" was
3345 specified. (If the value happened to be <= 0, no repeat happened.) The
3346 default is now 0s, meaning "never repeat". The "vacation" command was OK
3347 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3348 (I found it when inspecting the code).
3350 20. There is now an overall timeout for performing a callout verification. It
3351 defaults to 4 times the callout timeout, which applies to individual SMTP
3352 commands during the callout. The overall timeout applies when there is more
3353 than one host that can be tried. The timeout is checked before trying the
3354 next host. This prevents very long delays if there are a large number of
3355 hosts and all are timing out (e.g. when the network connections are timing
3356 out). The value of the overall timeout can be changed by specifying an
3357 additional sub-option for "callout", called "maxwait". For example:
3359 verify = sender/callout=5s,maxwait=20s
3361 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3362 to the end before writing, but this should make it even safer).
3364 22. Exim was forgetting that it had advertised PIPELINING for the second and
3365 subsequent messages on an SMTP connection. It was also not resetting its
3366 memory on STARTTLS and an internal HELO.
3368 23. When Exim logs an SMTP synchronization error within a session, it now
3369 records whether PIPELINING has been advertised or not.
3371 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3372 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3373 rather than long int.
3375 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3377 26. Added the never_mail option to autoreply.
3383 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3384 crash if the getsockname() call failed; this can happen if a connection is
3385 closed very soon after it is established. The problem was simply in the
3386 order in which certain operations were done, causing Exim to try to write
3387 to the SMTP stream before it had set up the file descriptor. The bug has
3388 been fixed by making things happen in the correct order.
3394 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3395 before the connection was closed, thus losing the rejection response.
3397 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3398 some early Solaris releases, but causes trouble in current releases where
3399 socklen_t is defined.
3401 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3404 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3407 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3408 that releasing the top of it at the end releases what was used for sub-
3409 expansions (unless the block got too big). However, discard this block if
3410 the first thing is a variable or header, so that we can use its block when
3411 it is dynamic (useful for very large $message_headers, for example).
3413 6. Lookups now cache *every* query, not just the most recent. A new, separate
3414 store pool is used for this. It can be recovered when all lookup caches are
3415 flushed. Lookups now release memory at the end of their result strings.
3416 This has involved some general refactoring of the lookup sources.
3418 7. Some code has been added to the store_xxx() functions to reduce the amount
3419 of flapping under certain conditions.
3421 8. log_incoming_interface used to affect only the <= reception log lines. Now
3422 it causes the local interface and port to be added to several more SMTP log
3423 lines, for example "SMTP connection from", and rejection lines.
3425 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3427 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3429 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3430 could be overwritten at the end of the current message (or the start of a
3431 new message if it was set in a HELO ACL). The value is now preserved for
3432 the duration of the SMTP connection.
3434 12. If a transport had a headers_rewrite setting, and a matching header line
3435 contained an unqualified address, that address was qualified, even if it
3436 did not match any rewriting rules. The underlying bug was that the values
3437 of the flags that permit the existence of unqualified sender and recipient
3438 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3439 non-local messages, and by -bnq for local messages) were not being
3440 preserved with the message after it was received.
3442 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3443 "next input=" as part of the text comprising the host identity instead of
3444 the correct text. The code was using the same buffer for two different
3445 strings. However, depending on which order the printing function evaluated
3446 its arguments, the bug did not always show up. Under Linux, for example, my
3447 test suite worked just fine.
3449 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3450 doesn't work with some older versions of Perl. It has been changed to "my",
3451 which in any case is probably the better facility to use.
3453 15. A really picky compiler found some instances of statements for creating
3454 error messages that either had too many or two few arguments for the format
3457 16. The size of the buffer for calls to the DNS resolver has been increased
3458 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3459 for addresses that have a lot of PTR records. This alleviates a problem; it
3460 does not fully solve it.
3462 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3463 buffer now truncates the list and logs the incident, which is the same
3464 action as happens when Exim is looking up a host name and its aliases.
3465 Previously in this situation something unpredictable would happen;
3466 sometimes it was "internal error: store_reset failed".
3468 18. If a server dropped the connection unexpectedly when an Exim client was
3469 using GnuTLS and trying to read a response, the client delivery process
3470 crashed while trying to generate an error log message.
3472 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3473 string, for example:
3475 warn message = H1: something\nH2: something
3477 the text was added as a single header line from Exim's point of view
3478 though it ended up OK in the delivered message. However, searching for the
3479 second and subsequent header lines using $h_h2: did not work. This has been
3480 fixed. Similarly, if a system filter added multiple headers in this way,
3481 the routers could not see them.
3483 20. Expanded the error message when iplsearch is called with an invalid key to
3484 suggest using net-iplsearch in a host list.
3486 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3487 ACLs are no longer actually imposed (and a message to that effect is
3490 22. If a "gecos" field in a passwd entry contained escaped characters, in
3491 particular, if it contained a \" sequence, Exim got it wrong when building
3492 a From: or a Sender: header from that name. A second bug also caused
3493 incorrect handling when an unquoted " was present following a character
3494 that needed quoting.
3496 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3497 was not being matched caselessly.
3499 24. Arranged for all hyphens in the exim.8 source to be escaped with
3502 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3503 properly. Recipient callout cache records were still being keyed to include
3504 the sender, even when use_sender was set false. This led to far more
3505 callouts that were necessary. The sender is no longer included in the key
3506 when use_sender is false.
3508 26. Added "control = submission" modifier to ACLs.
3510 27. Added the ${base62d: operator to decode base 62 numbers.
3512 28. dnsdb lookups can now access SRV records.
3514 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3515 the configuration file.
3517 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3518 (-v) mode. This makes the output for a verbose queue run more intelligible.
3520 31. Added a use_postmaster feature to recipient callouts.
3522 32. Added the $body_zerocount variable, containing the number of binary zero
3523 bytes in the message body.
3525 33. The time of last modification of the "new" subdirectory is now used as the
3526 "mailbox time last read" when there is a quota error for a maildir
3529 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3531 35. Added +ignore_unknown as a special item in host lists.
3533 36. Code for decoding IPv6 addresses in host lists is now included, even if
3534 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3535 address was recognized as an IP address, but was then not correctly decoded
3536 into binary, causing unexpected and incorrect effects when compared with
3543 1. Very minor rewording of debugging text in manualroute to say "list of
3544 hosts" instead of "hostlist".
3546 2. If verify=header_syntax was set, and a header line with an unqualified
3547 address (no domain) and a large number of spaces between the end of the
3548 name and the colon was received, the reception process suffered a buffer
3549 overflow, and (when I tested it) crashed. This was caused by some obsolete
3550 code that should have been removed. The fix is to remove it!
3552 3. When running in the test harness, delay a bit after writing a bounce
3553 message to get a bit more predictability in the log output.
3555 4. Added a call to search_tidyup() just before forking a reception process. In
3556 theory, someone could use a lookup in the expansion of smtp_accept_max_
3557 per_host which, without the tidyup, could leave open a database connection.
3559 5. Added the variables $recipient_data and $sender_data which get set from a
3560 lookup success in an ACL "recipients" or "senders" condition, or a router
3561 "senders" option, similar to $domain_data and $local_part_data.
3563 6. Moved the writing of debug_print from before to after the "senders" test
3566 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3567 problems for message scanning, either using a data ACL, or using
3568 local_scan() because the Received: header was not generated till after they
3569 were called (in order to set the time as the time of reception completion).
3570 I have revised the way this works. The header is now generated after the
3571 body is received, but before the ACL or local_scan() are called. After they
3572 are run, the timestamp in the header is updated.
3578 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3579 before starting a queue runner without re-exec. This happened only when
3580 deliver_drop_privilege was set or when the Exim user was set to root. The
3581 effect of the bug was that timeouts during subsequent deliveries caused
3582 crashes instead of being properly handled. The handler is now left at its
3583 default (and expected) setting.
3585 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3586 message, again when deliver_drop_privilege is set or Exim is run as root.
3587 The bug described in (1) was not present in this case, but the tidying up
3588 of the other signals was missing. I have made the two cases consistent.
3590 3. The ignore_target_hosts setting on a manualroute router was being ignored
3591 for hosts that were looked up using the /MX notation.
3593 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3596 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3597 operated on the sender address. After changing the $sender_address to <>
3598 for the sender address verify, Exim was re-instated it as the original
3599 (before rewriting) address, but remembering that it had rewritten it, so it
3600 wasn't rewriting it again. This bug also had the effect of breaking the
3601 sender address verification caching when the sender address was rewritten.
3603 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3604 This has been changed so that if the ip literal address matches
3605 ignore_target_hosts, the router declines.
3607 7. Added expansion conditions match_domain, match_address, and match_local_
3608 part (NOT match_host).
3610 8. The placeholder for the Received: header didn't have a length field set.
3612 9. Added code to Exim itself and to exim_lock to test for a specific race
3613 condition that could lead to file corruption when using MBX delivery. The
3614 issue is with the lockfile that is created in /tmp. If this file is removed
3615 after a process has opened it but before that process has acquired a lock,
3616 there is the potential for a second process to recreate the file and also
3617 acquire a lock. This could lead to two Exim processes writing to the file
3618 at the same time. The added code performs the same test as UW imapd; it
3619 checks after acquiring the lock that its file descriptor still refers to
3620 the same named file.
3622 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3623 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3626 11. Added the smtp_active_hostname option. If used, this will typically be made
3627 to depend on the incoming interface address. Because $interface_address is
3628 not set up until the daemon has forked a reception process, error responses
3629 that can happen earlier (such as "too many connections") no longer contain
3632 12. If an expansion in a condition on a "warn" statement fails because a lookup
3633 defers, the "warn" statement is abandoned, and the next ACL statement is
3634 processed. Previously this caused the whole ACL to be aborted.
3636 13. Added the iplsearch lookup type.
3638 14. Added ident_timeout as a log selector.
3640 15. Added tls_certificate_verified as a log selector.
3642 16. Added a global option tls_require_ciphers (compare the smtp transport
3643 option of the same name). This controls incoming TLS connections.
3645 17. I finally figured out how to make tls_require_ciphers do a similar thing
3646 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3647 before starting the TLS session.
3649 18. Tabs are now shown as \t in -bP output.
3651 19. If the log selector return_path_on_delivery was set, Exim crashed when
3652 bouncing a message because it had too many Received: header lines.
3654 20. If two routers both had headers_remove settings, and the first one included
3655 a superfluous trailing colon, the final name in the first list and the
3656 first name in the second list were incorrectly joined into one item (with a
3657 colon in the middle).
3663 1. Added -C and -D options to the exinext utility, mainly to make it easier
3664 to include in the automated testing, but these could be helpful when
3665 multiple configurations are in use.
3667 2. The exinext utility was not formatting the output nicely when there was
3668 an alternate port involved in the retry record key, nor when there was a
3669 message id as well (for retries that were specific to a specific message
3670 and a specific host). It was also confused by IPv6 addresses, because of
3671 the additional colons they contain. I have fixed the IPv4 problem, and
3672 patched it up to do a reasonable job for IPv6.
3674 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3675 delivery, the log line now contains "pipelined" if PIPELINING was used.
3677 4. An SMTP transport process used to panic and die if the bind() call to set
3678 an explicit outgoing interface failed. This has been changed; it is now
3679 treated in the same way as a connect() failure.
3681 5. A reference to $sender_host_name in the part of a conditional expansion
3682 that was being skipped was still causing a DNS lookup. This no longer
3685 6. The def: expansion condition was not recognizing references to header lines
3686 that used bh_ and bheader_.
3688 7. Added the _cache feature to named lists.
3690 8. The code for checking quota_filecount in the appendfile transport was
3691 allowing one more file than it should have been.
3693 9. For compatibility with Sendmail, the command line option
3701 and sets the incoming protocol and host name (for trusted callers). The
3702 host name and its colon can be omitted when only the protocol is to be set.
3703 Note the Exim already has two private options, -pd and -ps, that refer to
3704 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3705 "s", but I don't think that's a major issue.
3707 10. A number of refactoring changes to the code, none of which should affect
3710 (a) The number of logging options was getting close to filling up the
3711 32-bit word that was used as a bit map. I have split them into two classes:
3712 those that are passed in the argument to log_write(), and those that are
3713 only ever tested independently outside of that function. These are now in
3714 separate 32-bit words, so there is plenty of room for expansion again.
3715 There is no change in the user interface or the logging behaviour.
3717 (b) When building, for example, log lines, the code previously used a
3718 macro that called string_cat() twice, in order to add two strings. This is
3719 not really sufficiently general. Furthermore, there was one instance where
3720 it was actually wrong because one of the argument was used twice, and in
3721 one call a function was used. (As it happened, calling the function twice
3722 did not affect the overall behaviour.) The macro has been replaced by a
3723 function that can join an arbitrary number of extra strings onto a growing
3726 (c) The code for expansion conditions now uses a table and a binary chop
3727 instead of a serial search (which was left over from when there were very
3728 few conditions). Also, it now recognizes conditions like "pam" even when
3729 the relevant support is not compiled in: a suitably worded error message is
3730 given if an attempt is made to use such a condition.
3732 11. Added ${time_interval:xxxxx}.
3734 12. A bug was causing one of the ddress fields not to be passed back correctly
3735 from remote delivery subprocesses. The field in question was not being
3736 subsequently used, so this caused to problems in practice.
3738 13. Added new log selectors queue_time and deliver_time.
3740 14. Might have fixed a bug in maildirsizefile handling that threw up
3741 "unexpected character" debug warnings, and recalculated the data
3742 unnecessarily. In any case, I expanded the warning message to give more
3745 15. Added the message "Restricted characters in address" to the statements in
3746 the default ACL that block characters like @ and % in local parts.
3748 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3749 Three changes have been made:
3751 (a) There was a serious bug; a negative response to MAIL caused the whole
3752 recipient domain to be cached as invalid, thereby blocking all messages
3753 to all local parts at the same domain, from all senders. This bug has
3754 been fixed. The domain is no longer cached after a negative response to
3755 MAIL if the sender used is not empty.
3757 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3760 (c) A new callout option, "use_sender" has been added for people who want
3761 the modified behaviour.
3767 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3770 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3771 indeed breaks things for older releases.
3773 3. Added additional logging to the case where there is a problem reading data
3774 from a filter that is running in a subprocess using a pipe, in order to
3775 try to track down a specific problem.
3777 4. Testing facility fudge: when running in the test harness and attempting
3778 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3779 getting "No route to host". Convert this to a timeout.
3781 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3784 6. Some OS don't have socklen_t but use size_t instead. This affects the
3785 fifth argument of getsockopt() amongst other things. This is now
3786 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3787 can be set for individual OS. I have set it for SunOS5, OSF1, and
3788 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3789 some earlier ones do not.
3791 7. Change 4.30/15 was not doing the test caselessly.
3793 8. The standard form for an IPv6 address literal was being rejected by address
3794 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3795 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3796 this, as well as the form without the "IPv6" on the front (but only when
3797 address literals are enabled, of course).
3799 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3801 10. Exim crashed if a message with an empty sender address specified by -f
3802 encountered a router with an errors_to setting. This could be provoked only
3803 by a command such as
3807 where an empty string was supplied; "<>" did not hit this bug.
3809 11. Installed PCRE release 4.5.
3811 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3812 remained set. It is now erased.
3814 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3815 times from message ids (which are base 36 rather than the normal 62).
3817 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3818 were being counted as actual protocol errors, and logged if the log
3819 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3820 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3821 and DATA following a set of rejected RCPTs do not count as protocol errors.
3822 In other words, Exim assumes they were pipelined, though this may not
3823 actually be the case. Of course, in all cases the client gets an
3824 appropriate error code.
3826 15. If a lookup fails in an ACL condition, a message about the failure may
3827 be available; it is used if testing the ACL cannot continue, because most
3828 such messages specify what the cause of the deferral is. However, some
3829 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3830 that caused an old message to be retained and used if a later statement
3831 caused a defer, replacing the real cause of the deferral.
3833 16. If an IP address had so many PTR records that the DNS lookup buffer
3834 was not large enough to hold them, Exim could crash while trying to process
3835 the truncated data. It now detects and logs this case.
3837 17. Further to 4.21/58, another change has been made: if (and only if) the
3838 first line of a message (the first header line) ends with CRLF, a bare LF
3839 in a subsequent header line has a space inserted after it, so as not to
3840 terminate the header.
3842 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3843 unnecessarily, used atoi() instead of strtol(), and didn't check the
3844 termination when getting file sizes from file names by regex.
3846 19. Completely re-implemented the support for maildirsize files, in the light
3847 of a number of problems with the previous contributed implementation
3848 (4.30/29). In particular:
3850 . If the quota is zero, the maildirsize file is maintained, but no quota is
3853 . If the maildir directory does not exist, it is created before any attempt
3854 to write a maildirsize file.
3856 . The quota value in the file is just a cache; if the quota is changed in
3857 the transport, the new value overrides.
3859 . A regular expression is available for excluding directories from the
3862 20. The autoreply transport checks the characters in options that define the
3863 message's headers; it allows continued headers, but it was checking with
3864 isspace() after an embedded newline instead of explicitly looking for a
3867 21. If all the "regular" hosts to which an address was routed had passed their
3868 expiry times, and had not reached their retry times, the address was
3869 bounced, even if fallback hosts were defined. Now Exim should go on to try
3872 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3873 equivalent code in the SMTP transport. Some hosts send humungous responses
3874 to HELO/EHLO, more than 1024 it seems.
3876 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3877 gives compiler warnings in some environments. I've now done it "properly",
3880 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3881 (because of problems with the built-in one) was declared to return uschar *
3882 instead of char *, causing compiler failure.
3884 25. Fixed a file descriptor leak when processing alias/forward files.
3886 26. Fixed a minor format string issue in dbfn.c.
3888 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3890 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3891 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3894 29. When a sender address is verified, it is cached, to save repeating the test
3895 when there is more than one recipient in a message. However, when the
3896 verification involves a callout, it is possible for different callout
3897 options to be set for different recipients. It is too complicated to keep
3898 track of this in the cache, so now Exim always runs a verification when a
3899 callout is required, relying on the callout cache for the optimization.
3900 The overhead is duplication of the address routing, but this should not be
3903 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3904 to be verified with callout=postmaster, and the main callout worked but the
3905 postmaster check failed, the verification correctly failed. However, if a
3906 subsequent RCPT command asked for sender verification *without* the
3907 postmaster check, incorrect caching caused this verification also to fail,
3910 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3911 it was not caching the DNS options (qualify_single, search_parents) that
3912 were used when the lookup failed. A subsequent lookup with different
3913 options therefore always gave the same answer, though there were cases
3914 where it should not have. (Example: a "domains = !$mx_any" option on a
3915 dnslookup router: the "domains" option is always processed without any
3916 widening, but the router might have qualify_single set.) Now Exim uses the
3917 cached value only when the same options are set.
3919 32. Added John Jetmore's "exipick" utility to the distribution.
3921 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3922 than a timeout (e.g. a certificate is required, and is not provided), an
3923 Exim server now closes the connection immediately. Previously it waited for
3924 the client to close - but if the client is SSL, it seems that they each
3925 wait for each other, leading to a delay before one of them times out.
3927 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3928 maintained 0.8.x compatibility because I don't think many are using it, and
3929 it is clearly obsolete.
3931 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3934 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3935 client certificate was expired. A simple patch fixes this, though I don't
3936 understand the full logic of why the verify callback is called multiple
3939 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3940 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3941 which causes problems with some clients (such as the Certicom SSL Plus
3942 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3943 disables the coutermeasure allowing Eudora to connect."
3945 38. Exim was not checking that a write() to a log file succeeded. This could
3946 lead to Bad Things if a log got too big, in particular if it hit a file
3947 size limit. Exim now panics and dies if it cannot write to a log file, just
3948 as it does if it cannot open a log file.
3950 39. Modified OS/Makefile-Linux so that it now contains
3952 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3954 The two -D definitions ensure that Exim is compiled with large file
3955 support, which makes it possible to handle log files that are bigger than
3958 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3959 instance) a domain was checked against a named list that involved a lookup,
3960 causing $domain_data to be set, then another domain was checked against the
3961 same list, then the first domain was re-checked, the value of $domain_data
3962 after the final check could be wrong. In particular, if the second check
3963 failed, it could be set empty. This bug probably also applied to
3966 41. The strip_trailing_dot option was not being applied to the address given
3967 with the -f command-line option.
3969 42. The code for reading a message's header from the spool was incrementing
3970 $received_count, but never initializing it. This meant that the value was
3971 incorrect (doubled) while delivering a message in the same process in which
3972 it was received. In the most common configuration of Exim, this never
3973 happens - a fresh exec is done - but it can happen when
3974 deliver_drop_privilege is set.
3976 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3977 it now includes up to 150 characters of the unexpected data in the log
3980 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3981 and building data strings. The size of both of these buffers was 10 000
3982 bytes - far larger than anybody would *ever* want, thought I. Needless to
3983 say, somebody hit the limit. I have increased the maximum line length to
3984 20 000 and the maximum data length of concatenated lines to 100 000. I have
3985 also fixed two bugs, because there was no checking on these buffers. Tsk,
3986 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3987 buffer is too small.
3989 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3990 lsearch lookups. Now it does.
3992 46. When parsing a route_list item in a manualroute router, a fixed-length
3993 buffer was used for the list of hosts. I made this 1024 bytes long,
3994 thinking that nobody would ever have a list of hosts that long. Wrong.
3995 Somebody had a whole pile of complicated expansion conditions, and the
3996 string was silently truncated, leading to an expansion error. It turns out
3997 that it is easier to change to an unlimited length (owing to other changes
3998 that have happened since this code was originally written) than to build
3999 structure for giving a limitation error. The length of the item that
4000 expands into the list of hosts is now unlimited.
4002 47. The lsearch lookup could not handle data where the length of text line was
4003 more than 4095 characters. Such lines were truncated, leading to shortened
4004 data being returned. It should now handle lines of any length.
4006 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4007 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4010 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4011 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4012 no longer happens in comment lines. A list of replacements is now placed
4013 at the head of all of the source files, except those whose only change is
4014 to replace PERL_COMMAND in the very first #! line.
4016 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4017 messages on the queue, with a bottom-up merge sort, using code contributed
4018 by Michael Haardt. This should make operations like -bp somewhat faster on
4019 large queues. It won't affect queue runners, except when queue_run_in_order
4022 51. Installed eximstats 1.31 in the distribution.
4024 52. Added support for SRV lookups to the dnslookup router.
4026 53. If an ACL referred to $message_body or $message_body_end, the value was not
4027 reset for any messages that followed in the same SMTP session.
4029 54. The store-handling optimization for building very long strings was not
4030 differentiating between the different store pools. I don't think this
4031 actually made any difference in practice, but I've tidied it.
4033 55. While running the routers to verify a sender address, $sender_address
4034 was still set to the sender address. This is wrong, because when routing to
4035 send a bounce to the sender, it would be empty. Therefore, I have changed
4036 it so that, while verifying a sender address, $sender_address is set to <>.
4037 (There is no change to what happens when verifying a recipient address.)
4039 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4040 target A or AAAA records (if not already returned) without resetting the
4041 qualify_single or search_parents options of the DNS resolver. These are
4042 inappropriate in this case because the targets of MX and SRV records must
4043 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4044 target that, when qualified, matched something in the local domain. These
4045 two options are now turned off when doing these lookups.
4047 57. It seems that at least some releases of Reiserfs (which does not have the
4048 concept of a fixed number of inodes) returns zero and not -1 for the
4049 number of available inodes. This interacted badly with check_spool_inodes,
4050 which assumed that -1 was the "no such thing" setting. What I have done is
4051 to check that the total number of inodes is greater than zero before doing
4052 the test of how many are available.
4054 58. When a "warn" ACL statement has a log_message modifier, the message is
4055 remembered, and not repeated. This is to avoid a lot of repetition when a
4056 message has many recipients that cause the same warning to be written.
4057 Howewer, Exim was preserving the list of already written lines for an
4058 entire SMTP session, which doesn't seem right. The memory is now reset if a
4059 new message is started.
4061 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4062 debugging output unless log_rewrite was also set.
4064 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4065 of (int)(handle) when we know that handle contains (void *)(-1).
4067 61. The Exim daemon panic-logs an error return when it closes the incoming
4068 connection. However "connection reset by peer" seems to be common, and
4069 isn't really an error worthy of noting specially, so that particular error
4072 62. When Exim is trying to find all the local interfaces, it used to panic and
4073 die if the ioctl to get the interface flags failed. However, it seems that
4074 on at least one OS (Solaris 9) it is possible to have an interface that is
4075 included in the list of interfaces, but for which you get a failure error
4076 for this call. This happens when the interface is not "plumbed" into a
4077 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4078 failure of the "get flags" call assumes that the interface is down.
4080 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4081 makes life easier for people who are doing arithmetic on fields extracted
4082 from dates, where you often get leading zeros that should not be
4083 interpreted as octal.
4085 64. Added qualify_domain to the redirect router, to override the global
4088 65. If a pathologically long header line contained very many addresses (the
4089 report of this problem mentioned 10 000) and each of them was rewritten,
4090 Exim could use up a very large amount of memory. (It kept on making new
4091 copies of the header line as it rewrote, and never released the old ones.)
4092 At the expense of a bit more processing, the header rewriting function has
4093 been changed so that it no longer eats memory in this way.
4095 66. The generation of the Received: header has been moved from the time that a
4096 message starts to be received, to the time that it finishes. The timestamp
4097 in the Received: header should now be very close to that of the <= log
4098 line. There are two side-effects of this change:
4100 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4101 logged header lines no longer include the local Received: line, because
4102 it has not yet been created. The same applies to a copy of the message
4103 that is returned to a non-SMTP sender when a message is rejected.
4105 (b) When a filter file is tested using -bf, no additional Received: header
4106 is added to the test message. After some thought, I decided that this
4109 This change does not affect the value of $received_for. It is still set
4110 after address rewriting, but before local_scan() is called.
4112 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4114 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4115 gave an unhelpful panic error message, and a defer error. I have managed to
4116 change this behaviour so that it now rejects any supplied certificate,
4117 which seems right, as the list of acceptable certificates is empty.
4119 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4120 gave an unhelpful defer error. I have not managed to make this reject any
4121 supplied certificates, but the error message it gives is "no certificate
4122 supplied", which is not helpful.
4124 70. exigrep's output now also includes lines that are not associated with any
4125 message, but which match the given pattern. Implemented by a patch from
4126 Martin Sluka, which also tidied up the Perl a bit.
4128 71. Recipient callout verification, like sender verification, was using <> in
4129 the MAIL FROM command. This isn't really the right thing, since the actual
4130 sender may affect whether the remote host accepts the recipient or not. I
4131 have changed it to use the actual sender in the callout; this means that
4132 the cache record is now keyed on a recipient/sender pair, not just the
4133 recipient address. There doesn't seem to be a real danger of callout loops,
4134 since a callout by the remote host to check the sender would use <>.
4135 [SEE ABOVE: changed after hitting problems.]
4137 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4138 temporary errors. However, in the case of such a code being given after
4139 the end of a data transmission (i.e. after ".") Exim was failing to write
4140 a retry record for the message. (Yes, there was some broken host that was
4141 actually sending 8xx at this point.)
4143 73. An unknown lookup type in a host list could cause Exim to panic-die when
4144 the list was checked. (An example that provoked this was putting <; in the
4145 middle of a list instead of at the start.) If this happened during a DATA
4146 ACL check, a -D file could be left lying around. This kind of configuration
4147 error no longer causes Exim to die; instead it causes a defer errror. The
4148 incident is still logged to the main and panic logs.
4150 74. Buglet left over from Exim 3 conversion. The message "too many messages
4151 in one connection" was written to the rejectlog but not the mainlog, except
4152 when address rewriting (yes!) was being logged.
4154 75. Added write_rejectlog option.
4156 76. When a system filter was run not as root (that is, when system_filter_user
4157 was set), the values of the $n variables were not being returned to the
4158 main process; thus, they were not subsequently available in the $sn
4161 77. Added +return_path_on_delivery log selector.
4163 78. A connection timeout was being treated differently from recipients deferred
4164 when testing hosts_max_try with a message that was older than the host's
4165 retry timeout. (The host should not be counted, thus allowing all hosts to
4166 be tried at least once before bouncing.) This may have been the cause of an
4167 occasionally reported bug whereby a message would remain on the queue
4168 longer than the retry timeout, but would be bounced if a delivery was
4169 forced. I say "may" because I never totally pinned down the problem;
4170 setting up timeout/retry tests is difficult. See also the next item.
4172 79. The ultimate address timeout was not being applied to errors that involved
4173 a combination of host plus message (for example, a timeout on a MAIL
4174 command). When an address resolved to a number of possible hosts, and they
4175 were not all tried for each delivery (e.g. because of hosts_max_try), a
4176 message could remain on the queue longer than the retry timeout.
4178 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4181 81. Fixed an obscure SMTP outgoing bug which required at least the following
4182 conditions: (a) there was another message waiting for the same server;
4183 (b) the server returned 5xx to all RCPT commands in the first message so
4184 that the message was not completed; (c) the server dropped the connection
4185 or gave a negative response to the RSET that Exim sends to abort the
4186 transaction. The observed case was a dropped connection after DATA that had
4187 been sent in pipelining mode. That is, the server had advertised PIPELINING
4188 but was not implementing it correctly. The effect of the bug was incorrect
4189 behaviour, such as trying another host, and this could lead to a crash.
4195 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4196 and daemon.c were passed as pointers to ints; they should have been
4197 pointers to socklen_t variables (which are typically unsigned ints).
4199 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4202 3. Fixed a really odd bug that affected only the testing scheme; patching a
4203 certain fixed string in the binary changed the value of another string that
4204 happened to be identical to the end of the original first string.
4206 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4207 name", it returns that address as the IP address. On some operating
4208 systems (e.g. Solaris), it also passes back the IP address string as the
4209 "host name". However, on others (e.g. Linux), it passes back an empty
4210 string. Exim wasn't checking for this, and was changing the host name to an
4211 empty string, assuming it had been canonicized.
4213 5. Although rare, it is permitted to have more than one PTR record for a given
4214 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4215 all the names associated with an address, because they do in Solaris.
4216 However, it seems that they do not in Linux for data that comes from the
4217 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4218 I found this out when I moved to a new Linux workstation and tried to run
4219 the Exim test suite.
4221 To get round this problem I have changed the code so that it now does its
4222 own call to the DNS to look up PTR records when searching for a host name.
4223 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4224 addresses that are only in /etc/hosts are still found.
4226 This behaviour is, however, controlled by an option called host_lookup_
4227 order, which defaults to "bydns:byaddr". If people want to use the other
4228 order, or indeed, just use one or the other means of lookup, they can
4229 specify it in this variable.
4231 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4232 some operating systems, this comes back from gethostbyaddr() as an empty
4233 string, and this is what Exim used to test for. However, it seems that in
4234 other systems, "." is yielded. Exim now tests for this case too.
4236 7. The values of check_spool_space and check_log_space are now held internally
4237 as a number of kilobytes instead of an absolute number of bytes. If a
4238 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4239 kilobyte. This means that much larger values can be stored.
4241 8. Exim monitor: an attempt to get the action menu when not actually pointing
4242 at a message produces an empty menu entitled "No message selected". This
4243 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4244 no entries in it ("Shell widget menu has zero width and/or height"). So I
4245 have added a single, blank menu entry in this case.
4247 9. Added ${quote_local_part.
4249 10. MIME decoding is now applied to the contents of Subject: header lines when
4252 11. Now that a reference to $sender_host_address automatically causes a reverse
4253 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4254 host lookup before query-style lookups in lists that might use this
4255 variable. This has therefore been abolished, and the "net-" prefix is no
4256 longer necessary for query-style lookups.
4258 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4259 have been a typo for LFLAGS, so it has been changed.
4261 13. The install script calls Exim with "-C /dev/null" in order to find the
4262 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4263 to be output. Howeve, since Exim outputs its version number before the
4264 error, it didn't break the script. It just looked ugly. I fixed this by
4265 always allowing "-C /dev/null" if the caller is root.
4267 14. Ignore overlarge ACL variable number when reading spool file - insurance
4268 against a later release with more variables having written the file.
4270 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4271 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4272 this, as well as the form without the "IPv6" on the front.
4274 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4275 OS/Makefile-Darwin file.
4277 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4278 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4280 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4281 variable. However, some OS use ints for inodes. I've added cast to long int
4282 to get rid of the compiler warning.
4284 19. I had forgotten to lock out "/../" in configuration file names when
4285 ALT_CONFIG_PREFIX was set.
4287 20. Routers used for verification do not need to specify transports. However,
4288 if such a router generated a host list, and callout was configured, Exim
4289 crashed, because it could not find a port number from the (non-existent)
4290 transport. It now assumes port 25 in this circumstance.
4292 21. Added the -t option to exigrep.
4294 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4295 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4296 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4297 them is set, LOOKUP_LSEARCH is forced.
4299 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4301 24. Added sender and host information to the "rejected by local_scan()" log
4302 line; previously there was no indication of these.
4304 25. Added .include_if_exists.
4306 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4307 Linux. It turns out that not all file systems support this. Apparently some
4308 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4309 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4310 not supported on the file descriptor, is now ignored when Exim is trying to
4311 sync a directory. This applies only to Linux.
4313 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4315 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4316 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4319 29. Added support for maildirsize files from supplied patch (modified a bit).
4321 30. The use of :fail: followed by an empty string could lead Exim to respond to
4322 sender verification failures with (e.g.):
4324 550 Verification failed for <xxx>
4325 550 Sender verify failed
4327 where the first response line was missing the '-' that indicates it is not
4328 the final line of the response.
4330 31. The loop for finding the name of the user that called Exim had a hardwired
4331 limit of 10; it now uses the value of finduser_retries, which is used for
4332 all other user lookups.
4334 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4337 33. Exim was neglecting to zero errno before one call of strtol() when
4338 expanding a string and expecting an integer value. On some systems this
4339 resulted in spurious "integer overflow" errors. Also, it was casting the
4340 result into an int without checking.
4342 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4343 did not work. The code looks as if it has *never* worked, though it appears
4344 to have been documented since at least releast 1.62. I have made it work.
4346 35. The "timeout_DNS" error in retry rules, also documented since at least
4347 1.62, also never worked. As it isn't clear exactly what this means, and
4348 clearly it isn't a major issue, I have abolished the feature by treating it
4349 as "timeout", and writing a warning to the main and panic logs.
4351 36. The display of retry rules for -brt wasn't always showing the error code
4354 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4355 timeout_connect_A, timeout_connect_MX.
4357 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4358 to the empty sender.
4360 39. The daemon was not analysing the content of -oX till after it had closed
4361 stderr and disconnected from the controlling terminal. This meant that any
4362 syntax errors were only noted on the panic log, and the return code from
4363 the command was 0. By re-arranging the code a little, I've made the
4364 decoding happen first, so such errors now appear on stderr, and the return
4365 code is 1. However, the actual setting up of the sockets still happens in
4366 the disconnected process, so errors there are still only recorded on the
4369 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4370 connections (as happens on some IP stacks) was logged at start up time as
4371 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4372 it from "IPv6 and IPv4", which means that two separate sockets are being
4375 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4376 says whether AF_INET or AF_INET6 was passed as an argument.
4378 42. Exiwhat output was messed up when time zones were included in log
4381 43. Exiwhat now gives more information about the daemon's listening ports,
4382 and whether -tls-on-connect was used.
4384 44. The "port" option of the smtp transport is now expanded.
4386 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4387 silently ignored. Now an error message is written to the main and panic
4390 46. There's a new ACL modifier called "logwrite" which writes to a log file
4391 as soon as it is encountered.
4393 47. Added $local_user_uid and $local_user_gid at routing time.
4395 48. Exim crashed when trying to verify a sender address that was being
4398 49. Exim was recognizing only a space character after ".include". It now also
4399 recognizes a tab character.
4401 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4402 extracting the relevant information from the specification. The man page no
4403 longer contains scrambled data for the -d option, and I've added a section
4404 at the front about calling Exim under different names.
4406 51. Added "extra_headers" argument to the "mail" command in filter files.
4408 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4411 53. Installed eximstats 1.29.
4413 54. Added transport_filter_timeout as a generic transport option.
4415 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4416 Cc: header lines. This was required by RFC 822, but it not required by RFC
4419 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4420 incoming messages that did not have them. Now it does so only if the
4421 message originates locally, that is, if there is no associated remote host
4422 address. When Resent- header lines are present, this applies to the Resent-
4423 lines rather than the non-Resent- lines.
4425 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4426 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4428 58. Messages for configuration errors now include the name of the main
4429 configuration file - useful now that there may be more than one file in a
4430 list (.included file names were always shown).
4432 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4433 for those rare installations that do not start the daemon as root or run it
4434 setuid root. I've cut out the call to initgroups() if the daemon is not
4437 60. The Exim user and group can now be bound into the binary as text strings
4438 that are looked up at the start of Exim's processing.
4440 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4442 62. Added $mailstore_basename variable.
4444 63. Installed patch to sieve.c from Michael Haardt.
4446 64. When Exim failed to open the panic log after failing to open the main log,
4447 the original message it was trying to log was written to stderr and debug
4448 output, but if they were not available (the usual case in production), it
4449 was lost. Now it is written to syslog before the two lines that record the
4450 failures to open the logs.
4452 65. Users' Exim filters run in subprocesses under the user's uid. It is
4453 possible for a "deliver" command or an alias in a "personal" command to
4454 provoke an address rewrite. If logging of address rewriting is configured,
4455 this fails because the process is not running as root or exim. There may be
4456 a better way of dealing with this, but for the moment (because 4.30 needs
4457 to be released), I have disabled address rewrite logging when running a
4458 filter in a non-root, non-exim process.
4464 1. The buildconfig auxiliary program wasn't quoting the value set for
4465 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4466 not defined. This bug was masked in 4.22 by the effect that was fixed in
4469 2. Some messages that were rejected after a message id was allocated were
4470 shown as "incomplete" by exigrep. It no longer does this for messages that
4471 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4473 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4474 have allow_domain_literals set, the ID did not get logged in the <= line.
4475 Domain literals are now always recognized in Message-ID: header lines.
4477 4. The first argument for a ${extract expansion item is the key name or field
4478 number. Leading and trailing spaces in this item were not being ignored,
4479 causing some misleading effects.
4481 5. When deliver_drop_privilege was set, single queue runner processes started
4482 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4483 same command in the process it spins off) were not dropping privilege.
4485 6. When the daemon running as "exim" started a queue runner, it always
4486 re-executed Exim in the spun-off process. This is a waste of effort when
4487 deliver_drop_privilege is set. The new process now just calls the
4488 queue-runner function directly.
4494 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4497 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4498 ignored. Though the use of -oP was forcing the writing of a pid file, it
4499 was always written to the default place.
4501 3. If the message "no IP address found for host xxxx" is generated during
4502 incoming verification, it is now followed by identification of the incoming
4503 connection (so you can more easily find what provoked it).
4505 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4507 5. Added some features to "harden" Exim a bit more against certain attacks:
4509 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4510 be put in Local/Makefile. This is like the never_users runtime option,
4511 but it cannot be overridden. The default setting is "root".
4513 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4514 prefix string with which any file named in a -C command line option
4517 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4518 is retained for -C and -D only if the caller of Exim is root. Without
4519 it, the exim user may also use -C and -D and retain privilege.
4521 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4522 command line option is disabled.
4524 6. Macro names set by the -D option must start with an upper case letter, just
4525 like macro names defined in the configuration file.
4527 7. Added "dereference=" facility to LDAP.
4529 8. Two instances of the typo "uknown" in the source files are fixed.
4531 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4532 the Configure-Makefile script screwed up while processing it.
4534 10. Incorporated PCRE 4.4.
4536 11. The SMTP synchronization check was not operating right at the start of an
4537 SMTP session. For example, it could not catch a HELO sent before the client
4538 waited for the greeting. There is now a check for outstanding input at the
4539 point when the greeting is written. Because of the duplex, asynchronous
4540 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4541 way, but not yet received, when the check is performed.
4543 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4544 on TCP/IP sockets, because this apparently causes some broken clients to
4547 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4548 unchanged) from the Cygwin maintainer.
4550 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4551 was supported instead of testing for mbx. Effectively a typo.
4553 15. The spa authenticator server code was not checking that the input it
4554 received was valid base64.
4556 16. The debug output line for the "set" modifier in ACLs was not showing the
4557 name of the variable that was being set.
4559 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4561 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4562 lookup on its own. Something else was needed to trigger the lookup. For
4563 example, a match in host_lookup or the need for a host name in a host list.
4564 Now, if $sender_host_name is referenced and the host name has not yet been
4565 looked up, a lookup is performed. If the lookup fails, the variable remains
4566 empty, and $host_lookup_failed is set to "1".
4568 19. Added "eqi" as a case-independent comparison operator.
4570 20. The saslauthd authentication condition could segfault if neither service
4571 nor realm was specified.
4573 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4574 error message that was logged was misleading, and incoming SMTP
4575 connections were dropped. The message is now more accurate, and temporary
4576 errors are given to SMTP connections.
4578 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4579 (except RSET) with a 421 error, until QUIT is received. However, it was
4580 failing to send a response to QUIT.
4582 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4583 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4584 ACL is now run after the helo_try_verify_hosts code.
4586 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4587 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4588 case-independent, so other case variants are also recognized). Apparently
4589 some systems use these upper case variants.
4591 25. If more than two messages were waiting for the same host, and a transport
4592 filter was specified for the transport, Exim sent two messages over the
4593 same TCP/IP connection, and then failed with "socket operation on non-
4594 socket" when it tried to send the third.
4596 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4598 27. The extern definition of crypt16() in expand.c was not being excluded when
4599 the OS had its own crypt16() function.
4601 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4602 as a preferred synonym for return_size_limit, both as an option and as an
4605 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4607 30. Changed the default configuration ACL to relax the local part checking rule
4608 for addresses that are not in any local domains. For these addresses,
4609 slashes and pipe symbols are allowed within local parts, but the sequence
4610 /../ is explicitly forbidden.
4612 31. SPA server authentication was not clearing the challenge buffer before
4615 32. log_message in a "warn" ACL statement was writing to the reject log as
4616 well as to the main log, which contradicts the documentation and doesn't
4617 seem right (because no rejection is happening). So I have stopped it.
4619 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4620 However, I am unable to do any testing of this.
4622 34. Fixed an infelicity in the appendfile transport. When checking directories
4623 for a mailbox, to see if any needed to be created, it was accidentally
4624 using path names with one or more superfluous leading slashes; tracing
4625 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4627 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4628 contents are added to the log line that is written for every discarded
4629 recipient. (Previously a log_message setting was ignored.)
4631 36. The ${quote: operator now quotes the string if it is empty.
4633 37. The install script runs exim in order to find its version number. If for
4634 some reason other than non-existence or emptiness, which it checks, it
4635 could not run './exim', it was installing it with an empty version number,
4636 i.e. as "exim-". This error state is now caught, and the installation is
4639 38. An argument was missing from the function that creates an error message
4640 when Exim fails to connect to the socket for saslauthd authentication.
4641 This could cause Exim to crash, or give a corrupted message.
4643 39. Added isip, isip4, and isip6 to ${if conditions.
4645 40. The ACL variables $acl_xx are now saved with the message, and can be
4646 accessed later in routers, transports, and filters.
4648 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4649 strings in the file are not string-expanded.
4651 42. If a MAIL command specified a SIZE value that was too large to fit into an
4652 int variable, the check against message_size_limit failed. Such values are
4653 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4654 day this will have to be increased, but I don't think I want to be around
4655 when emails are that large.
4662 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4663 iconv() is not standard in FreeBSD.
4665 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4666 IPv6 enabled. The observed symptom was a segmentation fault on return from
4667 the function os_common_find_running_interfaces() in src/os.c.
4669 3. In the check_special_case() function in daemon.c I had used "errno" as an
4670 argument name, which causes warnings on some systems. This was basically a
4671 typo, since it was named "eno" in the comments!
4673 4. The code that waits for the clock to tick (at a resolution of some fraction
4674 of a second) so as to ensure message-id uniqueness was always waiting for
4675 at least one whole tick, when it could have waited for less. [This is
4676 almost certainly not relevant at current processor speeds, where it is
4677 unlikely to ever wait at all. But we try to future-proof.]
4679 5. The function that sleeps for a time interval that includes fractions of a
4680 second contained a race. It did not block SIGALRM between setting the
4681 timer, and suspending (a couple of lines later). If the interval was short
4682 and the sigsuspend() was delayed until after it had expired, the suspension
4683 never ended. On busy systems this could lead to processes getting stuck for
4686 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4687 process, before it forks any delivery processes. The open lookup caching
4688 mechanism meant that the open file or database connection was passed into
4689 the delivery process. The problem was that delivery processes always tidy
4690 up cached lookup data. This could cause a problem for the next delivery
4691 process started by the queue runner, because the external queue runner
4692 process does not know about the closure. So the next delivery process
4693 still has data in the lookup cache. In the case of a file lookup, there was
4694 no problem because closing a file descriptor in a subprocess doesn't affect
4695 the parent. However, if the lookup was caching a connection to a database,
4696 the connection was closed, and the second delivery process was likely to
4697 see errors such as "PGSQL: query failed: server closed the connection
4698 unexpectedly". The problem has been fixed by closing all cached lookups
4699 in a queue runner before running a delivery process.
4701 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4702 seem to have the "const" qualifier which it has on other OS. I've
4705 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4706 *of the same type* (client or server) with the same public name that an
4707 error should be diagnosed.
4709 9. When Exim looked up a host name for an IP address, but failed to find the
4710 original IP address when looking up the host name (a safety check), it
4711 output the message "<ip address> does not match any IP for NULL", which was
4712 confusing, to say the least. The bug was that the host name should have
4713 appeared instead of "NULL".
4715 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4716 user that is built into the binary, and the -C or -D options is used, root
4717 privilege is dropped before the configuration file is read. In addition,
4718 logging is switched to stderr instead of the normal log files. If the
4719 configuration then re-defines the Exim user, the unprivileged environment
4720 is probably not what is expected, so Exim logs a panic warning message (but
4723 However, if deliver_drop_privilege is set, the unprivileged state may well
4724 be exactly what is intended, so the warning has been cut out in that case,
4725 and Exim is allowed to try to write to its normal log files.
4731 1. smtp_return_error_details was not giving details for temporary sender
4732 or receiver verification errors.
4734 2. Diagnose a configuration error if two authenticators have the same public
4737 3. Exim used not to create the message log file for a message until the first
4738 delivery attempt. This could be confusing when incoming messages were held
4739 for policy or load reasons. The message log file is now created at the time
4740 the message is received, and an initial "Received" line is written to it.
4742 4. The automatically generated man page for command line options had a minor
4743 bug that caused no ill effects; however, a more serious problem was that
4744 the procedure for building the man page automatically didn't always
4745 operate. Consequently, release 4.20 contains an out-of-date version. This
4746 shouldn't happen again.
4748 5. When building Exim with embedded Perl support, the script that builds the
4749 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4750 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4752 6. The freeze_tell option was not being used for messages that were frozen on
4753 arrival, either by an ACL or by local_scan().
4755 7. Added the smtp_incomplete_transaction log selector.
4757 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4758 was accepting AUTH without a new EHLO.
4760 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4761 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4764 10. Logging of TCP/IP connections (when configured) now happens in the main
4765 daemon process instead of the child process, so that the TCP/IP connection
4766 count is more accurate (but it can never be perfect).
4768 11. The use of "drop" in a nested ACL was not being handled correctly in the
4769 outer ACL. Now, if condition failure induced by the nested "drop" causes
4770 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4771 or "require"), the connection is dropped.
4773 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4774 that yield "discard" can now be used with an "accept" or a "discard" verb,
4775 but an error is generated for any others (because I can't see a useful way
4776 to define what should happen).
4778 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4779 are now processed in the same way as lines in the Exim configuration file.
4780 In particular, continuation lines are supported.
4782 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4784 15. Added -ti meaning -t -i.
4786 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4787 domains, and warn by logging if others are found.
4789 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4790 returned from GIFCONF when Exim is trying to find the list of interfaces on
4791 a host. The code in os.c has been modified to copy each ifreq to an aligned
4792 structure in all cases.
4794 Also, in some cases, the returned ifreq's were being copied to a 'struct
4795 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4796 means the last couple of bytes of an IPv6 address could be chopped if the
4797 ifreq contained only a normal sockaddr (14 bytes storage).
4799 18. Named domain lists were not supported in the hosts_treat_as_local option.
4800 An entry such as +xxxx was not recognized, and was treated as a literal
4803 19. Ensure that header lines added by a DATA ACL are included in the reject log
4804 if the ACL subsequently rejects the message.
4806 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4807 MD5 (which is deprecated).
4809 21. When testing a filter file using -bf, Exim was writing a message when it
4810 took the sender from a "From " line in the message, but it was not doing so
4811 when it took $return_path from a Return-Path: header line. It now does.
4813 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4814 with a valid header line field name (a series of printing characters
4815 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4817 23. Changed "disc" in the source to "disk" to conform to the documentation and
4818 the book and for uniformity.
4820 24. Ignore Sendmail's -Ooption=value command line item.
4822 25. When execve() failed while trying to run a command in a pipe transport,
4823 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4824 could be confused with a return value of 69 from the command itself. This
4825 has been changed to 127, the value the shell returns if it is asked to run
4826 a non-existent command. The wording for the related log line suggests a
4827 non-existent command as the problem.
4829 26. If received_header_text expands to an empty string, do not add a Received:
4830 header line to the message. (Well, it adds a token one on the spool, but
4831 marks it "old" so that it doesn't get used or transmitted.)
4833 27. Installed eximstats 1.28 (addition of -nt option).
4835 28. There was no check for failure on the call to getsockname() in the daemon
4836 code. This can fail if there is a shortage of resources on the system, with
4837 ENOMEM, for example. A temporary error is now given on failure.
4839 29. Contrary to the C standard, it seems that in some environments, the
4840 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4841 program. Exim now does this explicitly; it affects the formatting of
4842 timestamps using strftime().
4844 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4845 complaints. I've now initialized all the variables, to avoid this.
4847 32. Header lines added by a system filter were not being "seen" during
4848 transport-time rewrites.
4850 33. The info_callback() function passed to OpenSSL is set up with type void
4851 (*)(SSL *, int, int), as described somewhere. However, when calling the
4852 function (actually a macro) that sets it up, the type void(*)() is
4853 expected. I've put in a cast to prevent warnings from picky compilers.
4855 34. If a DNS black list lookup found a CNAME record, but there were no A
4856 records associated with the domain it pointed at, Exim crashed.
4858 35. If a DNS black list lookup returned more than one A record, Exim ignored
4859 all but the first. It now scans all returned addresses if a particular IP
4860 value is being sought. In this situation, the contents of the
4861 $dnslist_value variable are a list of all the addresses, separated by a
4864 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4865 to accept a host name and all its aliases if the forward lookup for any of
4866 them yielded the IP address of the incoming connection. Now it accepts only
4867 those names whose forward lookup yields the correct IP address. Any other
4868 names are discarded. This closes a loophole whereby a rogue DNS
4869 administrator could create reverse DNS records to break through a
4870 wildcarded host restriction in an ACL.
4872 37. If a user filter or a system filter that ran in a subprocess used any of
4873 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4874 the wrong values were passed to the pipe command ($thisaddress had the
4875 value of $0, $0 had the value of $1, etc). This bug was introduced by
4876 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4878 38. Improved the line breaking for long SMTP error messages from ACLs.
4879 Previously, if there was no break point between 40 and 75 characters, Exim
4880 left the rest of the message alone. Two changes have been made: (a) I've
4881 reduced the minimum length to 35 characters; (b) if it can't find a break
4882 point between 35 and 75 characters, it looks ahead and uses the first one
4883 that it finds. This may give the occasional overlong line, but at least the
4884 remaining text gets split now.
4886 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4887 file descriptors might be low, and that setting 1000 would always raise it.
4888 It turns out that in some environments, the limit is already over 1000 and
4889 that lowering it causes trouble. So now Exim takes care not to decrease it.
4891 40. When delivering a message, the value of $return_path is set to $sender_
4892 address at the start of routing (routers may change the value). By an
4893 oversight, this default was not being set up when an address was tested by
4894 -bt or -bv, which affected the outcome if any router or filter referred to
4897 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4898 log only when "message" or "log_message" are set. However, if one of the
4899 conditions was an address verification, or a call to a nested ACL, the
4900 messages generated by the underlying test were being passed through. This
4901 no longer happens. The underlying message is available in $acl_verify_
4902 message for both "message" and "log_message" expansions, so it can be
4903 passed through if needed.
4905 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4906 new expansion $bh_ to give the encoded byte string without charset
4907 translation. Translation happens only if iconv() is available; HAVE_ICONV
4908 indicates this at build time. HEADERS_CHARSET gives the charset to
4909 translate to; headers_charset can change it in the configuration, and
4910 "headers charset" can change it in an individual filter file.
4912 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4913 that creates RFC 2047 encoded "words" labels them as that charset instead
4914 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4915 expansion operator; (ii) when Exim creates a From: line for a local
4916 message; (iii) when a header line is rewritten to include a "phrase" part.
4918 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4919 buggy, causing it to skip the first lines of messages whose message ID
4920 ended in 'D'. This would not have bitten before Exim release 4.14, because
4921 message IDs were unlikely to end in 'D' before then. The effect was to have
4922 incorrect size information for certain domains.
4924 45. #include "config.h" was missing at the start of the crypt16.c module. This
4925 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4928 46. If there was a timeout during a "random" callout check, Exim treated it as
4929 a failure of the random address, and carried on sending RSET and the real
4930 address. If the delay was just some slowness somewhere, the response to the
4931 original RCPT would be taken as a response to RSET and so on, causing
4932 mayhem of various kinds.
4934 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4935 when I implemented it. It didn't allow for the fact that some option values
4936 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4937 the right test for positive values.
4939 48. Domain names in DNS records are case-independent. Exim always looks them up
4940 in lower case. Some resolvers return domain names in exactly the case they
4941 appear in the zone file, that is, they may contain uppercase letters. Not
4942 all resolvers do this - some return always lower case. Exim was treating a
4943 change of case by a resolver as a change of domain, similar to a widening
4944 of a domain abbreviation. This triggered its re-routing code and so it was
4945 trying to route what was effectively the same domain again. This normally
4946 caused routing to fail (because the router wouldn't handle the domain
4947 twice). Now Exim checks for this case specially, and just changes the
4948 casing of the domain that it ultimately uses when it transmits the message
4951 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4954 50. If a filter generated a file delivery with a non-absolute name (possible if
4955 no home directory exists for the router), the forbid_file option was not
4958 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4959 the existing equality matching.
4961 52. Exim was using ints instead of ino_t variables in some places where it was
4962 dealing with inode numbers.
4964 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4965 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4966 called TMPDIR, and if it finds it is different, it changes its value.
4968 54. The smtp_printf() function is now made available to local_scan() so
4969 additional output lines can be written before returning. There is also an
4970 smtp_fflush() function to enable the detection of a dropped connection.
4971 The variables smtp_input and smtp_batched_input are exported to
4974 55. Changed the default runtime configuration: the message "Unknown user"
4975 has been removed from the ACL, and instead placed on the localuser router,
4976 using the cannot_route_message feature. This means that any verification
4977 failures that generate their own messages won't get overridden. Similarly,
4978 the "Unrouteable address" message that was in the ACL for unverifiable
4979 relay addresses has also been removed.
4981 56. Added hosts_avoid_esmtp to the smtp transport.
4983 57. The exicyclog script was not checking for the esoteric option
4984 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4985 will work only if exicyclog is run under the appropriate euid.
4987 58. Following a discussion on the list, the rules by which Exim recognises line
4988 endings on incoming messages have been changed. The -dropcr and drop_cr
4989 options are now no-ops, retained only for backwards compatibility. The
4990 following line terminators are recognized: LF CRLF CR. However, special
4991 processing applies to CR:
4993 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4994 nor a local message in the state where . is a terminator.
4996 (ii) If a bare CR is encountered in a header line, an extra space is added
4997 after the line terminator so as not to end the header. The reasoning
4998 behind this is that bare CRs in header lines are most likely either
4999 to be mistakes, or people trying to play silly games.
5001 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5002 was being incorrectly given as 18 bytes larger than it should have been.
5003 This is a VOB (very old bug).
5005 60. This may never have affected anything current, but just in case it has:
5006 When the local host is found other than at the start of a list of hosts,
5007 the local host, those with the same MX, and any that follow, are discarded.
5008 When the list in question was part of a longer list of hosts, the following
5009 hosts (not currently being processed) were also being discarded. This no
5010 longer happens. I'm not sure if this situation could ever has previously
5013 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5016 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5017 header. This is something that is recommended in a new Internet Draft, and
5018 is something that is documented as being done by Sendmail. There are two
5019 possible values. For messages generated by the autoreply transport, Exim
5022 Auto-Submitted: auto-replied
5024 whereas for all other generated messages (e.g. bounces) it adds
5026 Auto-Submitted: auto-generated
5028 63. The "personal" condition in filters now includes a test for the
5029 Auto-Submitted: header. If it contains the string "auto-" the message it
5030 not considered personal.
5032 64. Added rcpt_include_affixes as a generic transport option.
5034 65. Added queue_only_override (default true).
5036 66. Added the syslog_duplication option.
5038 67. If what should have been the first header line of a message consisted of
5039 a space followed by a colon, Exim was mis-interpreting it as a header line.
5040 It isn't of course - it is syntactically invalid and should therefore be
5041 treated as the start of the message body. The misbehaviour could have
5042 caused a number of strange effects, including loss of data in subsequent
5043 header lines, and spool format errors.
5045 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5046 client host had authenticated. This control can now be exercised by an ACL
5047 for more flexibility.
5049 69. By default, callouts do not happen when testing with -bh. There is now a
5050 variant, -bhc, which does actually run the callout code, including
5051 consulting and updating the callout cache.
5053 70. Added support for saslauthd authentication, courtesy of Alexander
5056 71. If statvfs() failed on the spool or log directories while checking their
5057 size for availability, Exim confusingly gave the error "space shortage".
5058 Furthermore, in debugging mode it crashed with a floating point exception.
5059 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5060 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5061 serious problem, Exim now writes to the main and panic logs when this
5062 happens, with details of the failure. It then refuses to accept the
5063 incoming message, giving the message "spool directory problem" or "log
5064 directory problem" with a 421 code for SMTP messages.
5066 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5067 0, 1, and 2 exist, because some OS complain for execs without them (see
5068 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5069 descriptors. However, the code omitted to check that the open succeeded,
5070 causing mysterious errors if for some reason the permissions on /dev/null
5071 got screwed. Now Exim writes a message to the main and panic logs, and
5072 bombs out if it can't open /dev/null.
5074 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5075 interact so that it is all more flexible. It is supposed to remain
5076 backwards compatible. Also added extra_local_interfaces.
5078 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5079 to bomb out with an assertion failure - to the client this appears as a
5080 connection drop. This problem occurs in the part of the code that was taken
5081 from the Samba project. Fortunately, the assertion is in a very simple
5082 function, so I have fixed this by reproducing the function inline in the
5083 one place where it is called, and arranging for authentication to fail
5084 instead of killing the process with assert().
5086 75. The SPA client code was not working when the server requested OEM rather
5087 than Unicode encoding.
5089 76. Added code to make require_files with a specific uid setting more usable in
5090 the case where statting the file as root fails - usually a non-root-mounted
5091 NFS file system. When this happens and the failure is EACCES, Exim now
5092 forks a subprocess and does the per-uid checking as the relevant uid.
5094 77. Added process_log_path.
5096 78. If log_file_path was not explicitly set, a setting of check_log_space or
5097 check_log_inodes was ignored.
5099 79. If a space check for the spool or log partitions fails, the incident is now
5100 logged. Of course, in the latter case the data may get lost...
5102 80. Added the %p formatting code to string_format() so that it can be used to
5103 print addresses in debug_print(). Adjusted all the address printing in the
5104 debugging in store.c to use %p rather than %d.
5106 81. There was a concern that a line of code in smtp_in.c could overflow a
5107 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5108 initially expressed, the concern was not well-founded, because trailing
5109 spaces are removed early. However, if the trailing spaces were followed by
5110 a NULL, they did not get removed, so the overflow was possible. Two fixes
5113 (a) I re-wrote the offending code in a cleaner fashion.
5114 (b) If an incoming SMTP command contains a NULL character, it is rejected
5117 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5118 runs initgroups(), so that if the Exim user is in any additional groups,
5119 they will be used during message reception.
5125 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob