1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.621 2010/06/05 11:13:29 pdp Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
8 PP/01 Date: & Message-Id: revert to normally being appended to a message,
9 only prepend for the Resent-* case. Fixes regression introduced in
10 Exim 4.70 by NM/22 for Bugzilla 607.
12 PP/02 Include check_rfc2047_length in configure.default because we're seeing
13 increasing numbers of administrators be bitten by this.
15 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
17 PP/03 Bugzilla 994: added openssl_options main configuration option.
19 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
21 PP/05 Bugzilla 834: provide a permit_codedump option for pipe transports.
23 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
25 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
26 without a peer certificate (I believe), leading to a segfault because of
27 an assumption that peers always have certificates. Be a little more
28 paranoid. Problem reported by Martin Tscholak.
30 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
31 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
32 NB: ClamAV planning to remove STREAM in "middle of 2010".
38 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
39 $data_path, and $header_path variables; fixed documentation bugs and
42 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
43 exipick to access non-standard spools, including the "frozen" queue
46 NM/01 Bugzilla 965: Support mysql stored procedures.
47 Patch from Alain Williams
49 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
51 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
52 Patch from Andreas Metzler
54 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
55 Patch from Kirill Miazine
57 NM/05 Bugzilla 671: Added umask to procmail example.
59 JJ/03 installed exipick 20100323.0, fixing doc bug
61 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
62 directory. Notification and patch from Dan Rosenberg.
64 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
66 TK/02 Improve log output when DKIM signing operation fails.
68 MH/01 Treat the transport option dkim_domain as a colon separated
69 list, not as a single string, and sign the message with each element,
70 omitting multiple occurences of the same signer.
72 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
73 Bugzilla 985, 986. Patch by Simon Arlott
75 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
78 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
79 MBX locking. Notification from Dan Rosenberg.
85 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body
87 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
89 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
91 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
93 NM/04 Bugzilla 915: Flag broken perl installation during build.
99 TK/01 Added patch by Johannes Berg that expands the main option
100 "spamd_address" if it starts with a dollar sign.
102 TK/02 Write list of recipients to X-Envelope-Sender header when building
103 the mbox-format spool file for content scanning (suggested by Jakob
106 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
107 (http://www.dcc-servers.net/) support via dccifd. Activated by
108 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
110 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
111 by Mark Daniel Reidel <mr@df.eu>.
113 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
114 When building exim an external PCRE library is now needed -
115 PCRE is a system library on the majority of modern systems.
116 See entry on PCRE_LIBS in EDITME file.
118 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
119 conversation. Added nologin parameter to request.
120 Patch contributed by Kirill Miazine
122 TF/01 Do not log submission mode rewrites if they do not change the address.
124 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
126 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
127 log files in place. Contributed by Roberto Lima
129 NM/04 Bugzilla 667: close socket used by dovecot authenticator
131 TF/03 Bugzilla 615: When checking the local_parts router precondition
132 after a local_part_suffix or local_part_prefix option, Exim now
133 does not use the address's named list lookup cache, since this
134 contains cached lookups for the whole local part.
136 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
137 Robert Millan. Documentation is in experimental-spec.txt
139 TF/04 Bugzilla 668: Fix parallel build (make -j).
141 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000
143 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
144 Patch provided by Jan Srzednicki
146 TF/05 Leading white space used to be stripped from $spam_report which
147 wrecked the formatting. Now it is preserved.
149 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
150 that they are available at delivery time.
152 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
154 TF/08 TLS error reporting now respects the incoming_interface and
155 incoming_port log selectors.
157 TF/09 Produce a more useful error message if an SMTP transport's hosts
158 setting expands to an empty string.
160 NM/06 Bugzilla 744: EXPN did not work under TLS.
161 Patch provided by Phil Pennock
163 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
164 Patch provided by Richard Godbee
166 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
167 acl_smtp_notquit, added index entry.
169 NM/09 Bugzilla 787: Potential buffer overflow in string_format
170 Patch provided by Eugene Bujak
172 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to accept()
173 Patch provided by Maxim Dounin
175 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
176 Patch provided by Phil Pennock
178 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
180 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
181 Patch provided by Brad "anomie" Jorsch
183 NM/14 Bugzilla 622: Added timeout setting to transport filter.
184 Patch provided by Dean Brooks
186 TK/05 Add native DKIM support (does not depend on external libraries).
188 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
189 Patch provided by Graeme Fowler
191 NM/16 Bugzilla 851: Documentation example syntax fix.
193 NM/17 Changed NOTICE file to remove references to embedded PCRE.
195 NM/18 Bugzilla 894: Fix issue with very long lines including comments in lsearch
197 NM/19 Bugzilla 745: TLS version reporting
198 Patch provided by Phil Pennock
200 NM/20 Bugzilla 167: bool: condition support
201 Patch provided by Phil Pennock
203 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken clients
204 Patch provided by Phil Pennock
206 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date
207 Patch provided by Brad "anomie" Jorsch
209 NM/23 Bugzilla 687: Fix misparses in eximstats
210 Patch provided by Heiko Schlittermann
212 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid
213 Patch provided by Heiko Schlittermann
215 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file
216 plus update to original patch.
218 NM/26 Bugzilla 799: Documentation correction for ratelimit
220 NM/27 Bugzilla 802: Improvements to local interface IP addr detection
221 Patch provided by David Brownlee
223 NM/28 Bugzilla 807: Improvements to LMTP delivery logging
225 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes
227 NM/30 Bugzilla 888: TLS documentation bugfixes
229 NM/31 Bugzilla 896: Dovecot buffer overrun fix
231 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
232 Unlike the original bugzilla I have changed all shell scripts in src tree
234 NM/33 Bugzilla 898: Transport filter timeout fix
235 Patch by Todd Rinaldo
237 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches
238 Patch by Serge Demonchaux
240 NM/35 Bugzilla 39: Base64 decode bug fixes
241 Patch by Jakob Hirsch
243 NM/36 Bugzilla 909: Correct connect() call in dcc code
245 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling
247 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed
249 NM/39 Bugzilla 911: Fixed MakeLinks build script
255 TK/01 Add preliminary DKIM support. Currently requires a forked version of
256 ALT-N's libdkim that I have put here:
257 http://duncanthrax.net/exim-experimental/
259 Note to Michael Haardt: I had to rename some vars in sieve.c. They
260 were called 'true' and it seems that C99 defines that as a reserved
261 keyword to be used with 'bool' variable types. That means you could
262 not include C99-style headers which use bools without triggering
263 build errors in sieve.c.
265 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
266 as mailq or other aliases. Changed the --help handling significantly
267 to do whats expected. exim_usage() emits usage/help information.
269 SC/01 Added the -bylocaldomain option to eximstats.
271 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr
273 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp
275 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall)
281 PH/01 Another patch from the Sieve maintainer.
283 PH/02 When an IPv6 address is converted to a string for single-key lookup
284 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
285 dots are used instead of colons so that keys in lsearch files need not
286 contain colons. This was done some time before quoting was made available
287 in lsearch files. However, iplsearch files do require colons in IPv6 keys
288 (notated using the quote facility) so as to distinguish them from IPv4
289 keys. This meant that lookups for IP addresses in host lists did not work
290 for iplsearch lookups.
292 This has been fixed by arranging for IPv6 addresses to be expressed with
293 colons if the lookup type is iplsearch. This is not incompatible, because
294 previously such lookups could never work.
296 The situation is now rather anomolous, since one *can* have colons in
297 ordinary lsearch keys. However, making the change in all cases is
298 incompatible and would probably break a number of configurations.
300 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
303 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
304 conversion specification without a maximum field width, thereby enabling
305 a rogue spamd server to cause a buffer overflow. While nobody in their
306 right mind would setup Exim to query an untrusted spamd server, an
307 attacker that gains access to a server running spamd could potentially
308 exploit this vulnerability to run arbitrary code as the Exim user.
310 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
311 $primary_hostname instead of what libspf2 thinks the hosts name is.
313 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
314 a directory entry by the name of the lookup key. Previously, if a
315 symlink pointed to a non-existing file or a file in a directory that
316 Exim lacked permissions to read, a lookup for a key matching that
317 symlink would fail. Now it is enough that a matching directory entry
318 exists, symlink or not. (Bugzilla 503.)
320 PH/03 The body_linecount and body_zerocount variables are now exported in the
323 PH/04 Added the $dnslist_matched variable.
325 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
326 This means they are set thereafter only if the connection becomes
329 PH/06 Added the client_condition to authenticators so that some can be skipped
330 by clients under certain conditions.
332 PH/07 The error message for a badly-placed control=no_multiline_responses left
333 "_responses" off the end of the name.
335 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
337 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
338 (without spaces) instead of just copying the configuration text.
340 PH/10 Added the /noupdate option to the ratelimit ACL condition.
342 PH/11 Added $max_received_linelength.
344 PH/12 Added +ignore_defer and +include_defer to host lists.
346 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
347 way in which PCRE > 7.0 is built.
349 PH/14 Implemented queue_only_load_latch.
351 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
352 MAIL command. The effect was to mangle the value on 64-bit systems.
354 PH/16 Another patch from the Sieve maintainer.
356 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
358 PH/18 If a system quota error occurred while trying to create the file for
359 a maildir delivery, the message "Mailbox is full" was not appended to the
360 bounce if the delivery eventually timed out. Change 4.67/27 below applied
361 only to a quota excession during the actual writing of the file.
363 PH/19 It seems that peer DN values may contain newlines (and other non-printing
364 characters?) which causes problems in log lines. The DN values are now
365 passed through string_printing() before being added to log lines.
367 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
368 and InterBase are left for another time.)
370 PH/21 Added message_body_newlines option.
372 PH/22 Guard against possible overflow in moan_check_errorcopy().
374 PH/23 POSIX allows open() to be a macro; guard against that.
376 PH/24 If the recipient of an error message contained an @ in the local part
377 (suitably quoted, of course), incorrect values were put in $domain and
378 $local_part during the evaluation of errors_copy.
384 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
385 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
388 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
389 issue a MAIL command.
391 PH/02 In an ACL statement such as
393 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
395 if a client was not listed at all, or was listed with a value other than
396 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
397 the condition was not true (as it should be), so access was not denied.
398 The bug was that the ! inversion was incorrectly passed on to the second
399 item. This has been fixed.
401 PH/03 Added additional dnslists conditions == and =& which are different from
402 = and & when the dns lookup returns more than one IP address.
404 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
405 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
407 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
408 FSYNC, which compiles an option called disable_fsync that allows for
409 bypassing fsync(). The documentation is heavily laced with warnings.
411 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
413 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
414 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
415 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
416 including adding "make clean"; (3) Added -fPIC when compiling the test
417 dynamically loaded module, to get rid of a warning.
419 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
420 message fails, move_frozen_messages = true and ignore_bounce_errors_after
421 = 0s. The bug is otherwise harmless.
423 PH/07 There was a bug in the dovecot authenticator such that the value of
424 $auth1 could be overwritten, and so not correctly preserved, after a
425 successful authentication. This usually meant that the value preserved by
426 the server_setid option was incorrect.
428 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
430 PH/09 Installed PCRE release 7.0.
432 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
433 run for batched SMTP input. It is now run at the start of every message
434 in the batch. While fixing this I discovered that the process information
435 (output by running exiwhat) was not always getting set for -bs and -bS
436 input. This is fixed, and it now also says "batched" for BSMTP.
438 PH/11 Added control=no_pipelining.
440 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
441 patch, slightly modified), and move the expansion of helo_data till after
442 the connection is made in the smtp transport (so it can use these
445 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
447 PH/14 Added log_selector = +pid.
449 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
451 PH/16 Add ${if forany and ${if forall.
453 PH/17 Added dsn_from option to vary the From: line in DSNs.
455 PH/18 Flush SMTP output before performing a callout, unless control =
456 no_callout_flush is set.
458 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
459 was true (the default) a successful delivery failed to delete the retry
460 item, thus causing premature timeout of the address. The bug is now
463 PH/20 Added hosts_avoid_pipelining to the smtp transport.
465 PH/21 Long custom messages for fakedefer and fakereject are now split up
466 into multiline reponses in the same way that messages for "deny" and
467 other ACL rejections are.
469 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
470 with slight modification.
472 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
473 draft, changing the syntax and factoring some duplicate code".
475 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
476 for deliveries of the second and subsequent messages over the same SMTP
479 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
480 ${reduce, with only minor "tidies".
482 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
484 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
485 expansion side effects.
487 PH/27 When a message times out after an over-quota error from an Exim-imposed
488 quota, the bounce message says "mailbox is full". This message was not
489 being given when it was a system quota that was exceeded. It now should
492 MH/03 Made $recipients available in local_scan(). local_scan() already has
493 better access to the recipient list through recipients_list[], but
494 $recipients can be useful in postmaster-provided expansion strings.
496 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
497 in the case of a MAIL command with additional options following the
498 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
499 were accidentally chopped off.
501 PH/29 SMTP synchronization checks are implemented when a command is read -
502 there is a check that no more input is waiting when there shouldn't be
503 any. However, for some commands, a delay in an ACL can mean that it is
504 some time before the response is written. In this time, more input might
505 arrive, invalidly. So now there are extra checks after an ACL has run for
506 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
507 pipelining has not been advertised.
509 PH/30 MH's patch to allow iscntrl() characters to be list separators.
511 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
512 returned in the SMTP response when smtp_return_error_details was false.
515 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
516 instead of the C I/O that was originally supplied, because problems were
519 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
520 Exim which did not show up earlier: it was assuming that a call to
521 SSL_CTX_set_info_callback() might give an error value. In fact, there is
522 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
523 was a macro that became an assignment, so it seemed to work. This has
524 changed to a proper function call with a void return, hence the compile
525 error. Exim's code has been fixed.
527 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
530 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
532 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
534 PH/37 If a message is not accepted after it has had an id assigned (e.g.
535 because it turns out to be too big or there is a timeout) there is no
536 "Completed" line in the log. When some messages of this type were
537 selected by exigrep, they were listed as "not completed". Others were
538 picked up by some special patterns. I have improved the selection
539 criteria to be more general.
541 PH/38 The host_find_failed option in the manualroute router can now be set
542 to "ignore", to completely ignore a host whose IP address cannot be
543 found. If all hosts are ignored, the behaviour is controlled by the new
544 host_all_ignored option.
546 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
547 homing or because of multiple MX records with /mx) generated more than
548 one IP address, and the following item turned out to be the local host,
549 all the secondary addresses of the first item were incorrectly removed
550 from the list, along with the local host and any following hosts (which
551 is what is supposed to happen).
553 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
554 whoever called Exim into the -H file. In the case of the daemon it was
555 behaving confusingly. When first started, it used values for whoever
556 started the daemon, but after a SIGHUP it used the Exim user (because it
557 calls itself on a restart). I have changed the code so that it now always
560 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
561 message are rejected with the same error (e.g. no authentication or bad
562 sender address), and a DATA command is nevertheless sent (as can happen
563 with PIPELINING or a stupid MUA), the error message that was given to the
564 RCPT commands is included in the rejection of the DATA command. This is
565 intended to be helpful for MUAs that show only the final error to their
568 PH/42 Another patch from the Sieve maintainer.
570 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
571 Eximstats - Fixed some broken HTML links and added missing column headers
573 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
574 columns for Rejects, Temp Rejects, Ham, and Spam rows.
576 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
578 PH/43 Yet another patch from the Sieve maintainer.
580 PH/44 I found a way to check for a TCP/IP connection going away before sending
581 the response to the final '.' that terminates a message, but only in the
582 case where the client has not sent further data following the '.'
583 (unfortunately, this is allowed). However, in many cases there won't be
584 any further data because there won't be any more messages to send. A call
585 to select() can be used: if it shows that the input is "ready", there is
586 either input waiting, or the socket has been closed. An attempt to read
587 the next input character can distinguish the two cases. Previously, Exim
588 would have sent an OK response which the client would never have see.
589 This could lead to message repetition. This fix should cure that, at
590 least in a lot of common cases.
592 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
593 advertised in response to EHLO.
599 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
600 fixed by 4.65/MH/01 (is this a record?) are fixed:
602 (i) An empty string was always treated as zero by the numeric comparison
603 operators. This behaviour has been restored.
605 (ii) It is documented that the numeric comparison operators always treat
606 their arguments as decimal numbers. This was broken in that numbers
607 starting with 0 were being interpreted as octal.
609 While fixing these problems I realized that there was another issue that
610 hadn't been noticed. Values of message_size_limit (both the global option
611 and the transport option) were treated as octal if they started with 0.
612 The documentation was vague. These values are now always treated as
613 decimal, and I will make that clear in the documentation.
619 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
620 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
623 MH/01 Don't check that the operands of numeric comparison operators are
624 integers when their expansion is in "skipping" mode (fixes bug
625 introduced by 4.64-PH/07).
627 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
628 child addresses, Exim now panics and dies. Previously, because the count
629 is held in a short int, deliveries were likely to be lost. As such a
630 large number of recipients for a single message is ridiculous
631 (performance will be very, very poor), I have chosen to impose a limit
632 rather than extend the field.
638 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
639 leftover -K file (the existence of which was triggered by #402).
640 While we were at it, introduced process PID as part of the -K
641 filename. This should rule out race conditions when creating
644 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
645 processing considerably. Previous code took too long for large mails,
646 triggering a timeout which in turn triggers #401.
648 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
649 in the DK code in transports.c. sendfile() is not really portable,
650 hence the _LINUX specificness.
652 TF/01 In the add_headers option to the mail command in an Exim filter,
653 there was a bug that Exim would claim a syntax error in any
654 header after the first one which had an odd number of characters
657 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
658 callout verification, Exim cached a "reject" for the entire domain. This
659 is correct for most verifications, but it is not correct for a recipient
660 verification with use_sender or use_postmaster set, because in that case
661 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
662 case of MAIL FROM:<> rejection from other early rejections (e.g.
663 rejection of HELO). When verifying a recipient using a non-null MAIL
664 address, the cache is ignored if it shows MAIL FROM:<> rejection.
665 Whatever the result of the callout, the value of the domain cache is
666 left unchanged (for any other kind of callout, getting as far as trying
667 RCPT means that the domain itself is ok).
669 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
672 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
673 manifest itself as EPIPE rather than ECONNECT. When tidying away a
674 session, the daemon ignores ECONNECT errors and logs others; it now
675 ignores EPIPE as well.
677 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
678 (quoted-printable decoding).
680 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
681 later the small subsequent patch to fix an introduced bug.
683 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
685 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
687 PH/08 An error is now given if message_size_limit is specified negative.
689 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
690 to be given (somewhat) arbitrary names.
692 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
695 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
696 miscellaneous code fixes
698 PH/10 Added the log_reject_target ACL modifier to specify where to log
701 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
702 hostname. This is wrong, because it relates to the incoming message (and
703 probably the interface on which it is arriving) and not to the outgoing
704 callout (which could be using a different interface). This has been
705 changed to use the value of the helo_data option from the smtp transport
706 instead - this is what is used when a message is actually being sent. If
707 there is no remote transport (possible with a router that sets up host
708 addresses), $smtp_active_hostname is used.
710 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
711 tweaks were necessary in order to get it to work (see also 21 below):
712 (a) The code assumed that strncpy() returns a negative number on buffer
713 overflow, which isn't the case. Replaced with Exim's string_format()
715 (b) There were several signed/unsigned issues. I just did the minimum
716 hacking in of casts. There is scope for a larger refactoring.
717 (c) The code used strcasecmp() which is not a standard C function.
718 Replaced with Exim's strcmpic() function.
719 (d) The code set only $1; it now sets $auth1 as well.
720 (e) A simple test gave the error "authentication client didn't specify
721 service in request". It would seem that Dovecot has changed its
722 interface. Fortunately there's a specification; I followed it and
723 changed what the client sends and it appears to be working now.
725 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
728 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
729 address A is aliased to B and C, where B exists and C does not. Without
730 -v the output is "A verified" because verification stops after a
731 successful redirection if more than one address is generated. However,
732 with -v the child addresses are also verified. Exim was outputting "A
733 failed to verify" and then showing the successful verification for C,
734 with its parentage. It now outputs "B failed to verify", showing B's
735 parentage before showing the successful verification of C.
737 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
738 look up a TXT record in a specific list after matching in a combined
741 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
742 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
743 they consult the DNS. I had assumed they would set it the way they
744 wanted; and indeed my experiments on Linux seem to show that in some
745 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
746 To be on the safe side, however, I have now made the interface to
747 host_find_byname() similar to host_find_bydns(), with an argument
748 containing the DNS resolver options. The host_find_byname() function now
749 sets these options at its start, just as host_find_bydns() does. The smtp
750 transport options dns_qualify_single and dns_search_parents are passed to
751 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
752 of host_find_byname() use the default settings of RES_DEFNAMES
753 (qualify_single) but not RES_DNSRCH (search_parents).
755 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
756 spool_read_header() do less string testing, by means of a preliminary
757 switch on the second character of optional "-foo" lines. (This is
758 overdue, caused by the large number of possibilities that now exist.
759 Originally there were few.) While I was there, I also converted the
760 str(n)cmp tests so they don't re-test the leading "-" and the first
761 character, in the hope this might squeeze out yet more improvement.
763 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
764 flag allowing group syntax was set by the header_syntax check but not
765 turned off, possible causing trouble later; (2) The flag was not being
766 set at all for the header_verify test, causing "group"-style headers to
767 be rejected. I have now set it in this case, and also caused header_
768 verify to ignore an empty address taken from a group. While doing this, I
769 came across some other cases where the code for allowing group syntax
770 while scanning a header line wasn't quite right (mostly, not resetting
771 the flag correctly in the right place). These bugs could have caused
772 trouble for malformed header lines. I hope it is now all correct.
774 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
775 with the "reply" argument non-NULL. The code, however (which originally
776 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
777 but it didn't always do it. This confused somebody who was copying the
778 code for some other use. I have removed all the tests.
780 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
781 feature that was used to support insecure browsers during the U.S. crypto
782 embargo. It requires special client support, and Exim is probably the
783 only MTA that supported it -- and would never use it because real RSA is
784 always available. This code has been removed, because it had the bad
785 effect of slowing Exim down by computing (never used) parameters for the
786 RSA_EXPORT functionality.
788 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
789 authenticator to fail if there's a tab character in the incoming data
790 (there should never be unless someone is messing about, as it's supposed
791 to be base64-encoded). Also added, on Timo's advice, the "secured" option
792 if the connection is using TLS or if the remote IP is the same as the
793 local IP, and the "valid-client-cert option" if a client certificate has
796 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
797 authenticators. This can be used for authorization after authentication
798 succeeds. (In the case of plaintext, it servers for both authentication
801 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
802 if any retry times were supplied.
804 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
805 connection, where there is no client IP address to check. In this
806 situation, the verify now always succeeds.
808 PH/25 Applied John Jetmore's -Mset patch.
810 PH/26 Added -bem to be like -Mset, but loading a message from a file.
812 PH/27 In a string expansion for a processed (not raw) header when multiple
813 headers of the same name were present, leading whitespace was being
814 removed from all of them, but trailing whitespace was being removed only
815 from the last one. Now trailing whitespace is removed from each header
816 before concatenation. Completely empty headers in a concatenation (as
819 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
820 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
822 PH/29 [Removed. This was a change that I later backed out, and forgot to
823 correct the ChangeLog entry (that I had efficiently created) before
824 committing the later change.]
826 PH/30 Exim was sometimes attempting to deliver messages that had suffered
827 address errors (4xx response to RCPT) over the same connection as other
828 messages routed to the same hosts. Such deliveries are always "forced",
829 so retry times are not inspected. This resulted in far too many retries
830 for the affected addresses. The effect occurred only when there were more
831 hosts than the hosts_max_try setting in the smtp transport when it had
832 the 4xx errors. Those hosts that it had tried were not added to the list
833 of hosts for which the message was waiting, so if all were tried, there
834 was no problem. Two fixes have been applied:
836 (i) If there are any address or message errors in an SMTP delivery, none
837 of the hosts (tried or untried) are now added to the list of hosts
838 for which the message is waiting, so the message should not be a
839 candidate for sending over the same connection that was used for a
840 successful delivery of some other message. This seems entirely
841 reasonable: after all the message is NOT "waiting for some host".
842 This is so "obvious" that I'm not sure why it wasn't done
843 previously. Hope I haven't missed anything, but it can't do any
844 harm, as the worst effect is to miss an optimization.
846 (ii) If, despite (i), such a delivery is accidentally attempted, the
847 routing retry time is respected, so at least it doesn't keep
848 hammering the server.
850 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
851 in ${readsocket because some servers need this prod.
853 PH/32 Added some extra debug output when updating a wait-xxx database.
855 PH/33 The hint "could be header name not terminated by colon", which has been
856 given for certain expansion errors for a long time, was not being given
857 for the ${if def:h_colon_omitted{... case.
859 PH/34 The spec says: "With one important exception, whenever a domain list is
860 being scanned, $domain contains the subject domain." There was at least
861 one case where this was not true.
863 PH/35 The error "getsockname() failed: connection reset by peer" was being
864 written to the panic log as well as the main log, but it isn't really
865 panic-worthy as it just means the connection died rather early on. I have
866 removed the panic log writing for the ECONNRESET error when getsockname()
869 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
870 runs only) independently of the message's sender address. This meant
871 that, if the 4xx error was in fact related to the sender, a different
872 message to the same recipient with a different sender could confuse
873 things. In particualar, this can happen when sending to a greylisting
874 server, but other circumstances could also provoke similar problems.
875 I have changed the default so that the retry time for these errors is now
876 based a combination of the sender and recipient addresses. This change
877 can be overridden by setting address_retry_include_sender=false in the
880 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
881 remote server are returned as part of bounce messages. This was not
882 happening for LMTP over a pipe (the lmtp transport), but now it is the
883 same for both kinds of LMTP.
885 PH/38 Despite being documented as not happening, Exim was rewriting addresses
886 in header lines that were in fact CNAMEs. This is no longer the case.
888 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
889 and queue runs started by the daemon processed all messages. This has
890 been fixed so that -R and -S can now usefully be given with -q<time>.
892 PH/40 Import PCRE release 6.7 (fixes some bugs).
894 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
896 PH/42 Give an error if -q is specified more than once.
898 PH/43 Renamed the variables $interface_address and $interface_port as
899 $received_ip_address and $received_port, to make it clear that these
900 values apply to message reception, and not to the outgoing interface when
901 a message is delivered. (The old names remain recognized, of course.)
903 PH/44 There was no timeout on the connect() call when using a Unix domain
904 socket in the ${readsocket expansion. There now is.
906 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
907 be meaningful with "accept".
909 SC/01 Eximstats V1.43
910 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
912 SC/02 Eximstats V1.44
913 Use a glob alias rather than an array ref in the generated
914 parser. This improves both readability and performance.
916 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
917 Collect SpamAssassin and rejection statistics.
918 Don't display local sender or destination tables unless
919 there is data to show.
920 Added average volumes into the top table text output.
922 SC/04 Eximstats V1.46
923 Collect data on the number of addresses (recipients)
924 as well as the number of messages.
926 SC/05 Eximstats V1.47
927 Added 'Message too big' to the list of mail rejection
928 reasons (thanks to Marco Gaiarin).
930 SC/06 Eximstats V1.48
931 Mainlog lines which have GMT offsets and are too short to
932 have a flag are now skipped.
934 SC/07 Eximstats V1.49 (Alain Williams)
935 Added the -emptyok flag.
937 SC/08 Eximstats V1.50
938 Fixes for obtaining the IP address from reject messages.
940 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
941 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
942 whitesspace changes from 4.64-PH/27
944 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
947 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
948 are found, allow negative numbers in numeric criteria)
950 JJ/06 exipick.20061117.2, added new $message_body_missing variable
952 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
953 to match changes made in 4.64-PH/43
955 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
957 PH/47 Put in an explicit test for a DNS lookup of an address record where the
958 "domain" is actually an IP address, and force a failure. This locks out
959 those revolvers/nameservers that support "A-for-A" lookups, in
960 contravention of the specifications.
962 PH/48 When a host name was looked up from an IP address, and the subsequent
963 forward lookup of the name timed out, the host name was left in
964 $sender_host_name, contrary to the specification.
966 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
967 restricted to single-key lookups, Exim was not diagnosing an error if
968 * or *@ was used with a query-style lookup.
970 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
972 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
973 long ago, but noone interested enough thought of it. Let's just say that
974 the "1.1" means that there are some new functions that weren't there at
975 some point in the past.
977 PH/51 Error processing for expansion failure of helo_data from an smtp
978 transport during callout processing was broken.
980 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
981 tested/used via the -bh/-bhc/-bs options.
983 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
984 bug, fixed in subsequent PCRE releases).
986 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
987 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
989 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
995 SC/01 Use a glob alias rather than an array ref in eximstats generated
996 parser. This improves both readability and performance.
998 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
999 Don't display local sender or destination tables in eximstats unless
1000 there is data to show.
1001 Added average volumes into the eximstats top table text output.
1003 SC/03 Collect data on the number of addresses (recipients) as well
1004 as the number of messages in eximstats.
1006 TF/01 Correct an error in the documentation for the redirect router. Exim
1007 does (usually) call initgroups() when daemonizing.
1009 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1010 with consistent privilege compared to when running as a daemon.
1012 TF/03 Note in the spec that $authenticated_id is not set for local
1013 submissions from trusted users.
1015 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1016 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1018 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1019 by adding some example configuration directives to the default
1020 configuration file. A little bit of work is required to uncomment the
1021 directives and define how usernames and passwords are checked, but
1022 there is now a framework to start from.
1024 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1025 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1026 without this. I don't know how relevant this is to other LDAP libraries.
1028 PH/02 Add the verb name to the "unknown ACL verb" error.
1030 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1032 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1034 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1035 directory not expanded when it should be if an expanded home directory
1036 was set for the address (which is overridden by the transport).
1038 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1041 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1042 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1043 because it is too late at that time, and has no effect.
1045 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1046 security issue with \' (bugzilla #107). I could not use the
1047 PQescapeStringConn() function, because it needs a PGconn value as one of
1050 PH/08 When testing addresses using -bt, indicate those final addresses that
1051 are duplicates that would not cause an additional delivery. At least one
1052 person was confused, thinking that -bt output corresponded to deliveries.
1053 (Suppressing duplicates isn't a good idea as you lose the information
1054 about possibly different redirections that led to the duplicates.)
1056 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1057 systems where poll() doesn't work, in particular OS X.
1059 PH/10 Added more information to debugging output for retry time not reached.
1061 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1062 operations in malware.c.
1064 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1067 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1068 syslog_duplication set false, log lines that would normally be written
1069 both the the main log and to the reject log were not written to syslog at
1072 PH/14 In the default configuration, change the use of "message" in ACL warn
1073 statements to "add_header".
1075 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1076 not followed by a command (e.g. "seen endif").
1078 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1079 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1082 PH/17 Added extra conditions to the default value of delay_warning_condition
1086 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1087 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1088 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1091 The Auto-Submitted: and various List- headers are standardised, whereas I
1092 don't think Precedence: ever was.
1094 PH/18 Refactored debugging code in route_finduser() to show more information,
1095 in particular, the error code if getpwnam() issues one.
1097 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1098 This is apparently needed in addition to the PH/07 change above to avoid
1099 any possible encoding problems.
1101 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1102 but not after initializing Perl.
1104 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1105 output them only if debugging. By default they are written stderr,
1106 apparently, which is not desirable.
1108 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1111 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1114 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1116 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1117 authenticated or an ident call has been made. Suppress the default
1118 values for $authenticated_id and $authenticated_sender (but permit -oMai
1119 and -oMas) when testing with -bh.
1121 PH/24 Re-jigged the order of the tests in the default configuration so that the
1122 tests for valid domains and recipients precede the DNS black list and CSA
1123 tests, on the grounds that those ones are more expensive.
1125 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1126 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1127 This bug exists in every version of Exim that I still have, right back to
1130 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1131 However, an attempt to turn on case-sensitivity in a regex key by
1132 including (?-i) didn't work because the subject string was already
1133 lowercased, and the effects were non-intuitive. It turns out that a
1134 one-line patch can be used to allow (?-i) to work as expected.
1140 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1141 other effects) broke the use of negated acl sub-conditions.
1143 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1146 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1147 "Deny" causes Exim to reject the incoming connection with a 554 error.
1148 Unfortunately, if there is a major crisis, such as a disk failure,
1149 tcp-wrappers gives "deny", whereas what one would like would be some
1150 kind of temporary error. A kludge has been added to help with this.
1151 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1152 554 error is used if errno is still zero or contains ENOENT (which occurs
1153 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1156 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1158 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1159 errors. Otherwise a message that provokes a temporary error (when other
1160 messages do not) can cause a whole host to time out.
1162 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1163 addresses were routed directly to files or pipes from a redirect router.
1164 File deliveries just didn't batch; pipe deliveries might have suffered
1167 PH/06 A failure to get a lock for a hints database would erroneously always say
1168 "Failed to get write lock", even when it was really a read lock.
1170 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1171 of 0600. This has been changed to use the value of the lockfile_mode
1172 option (which defaults to 0600).
1174 PH/08 Applied small patch from the Sieve maintainer.
1176 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1177 folder from quota calculations, a direct delivery into this folder messed
1178 up the contents of the maildirsize file. This was because the regex was
1179 used only to exclude .Trash (or whatever) when the size of the mailbox
1180 was calculated. There was no check that a delivery was happening into an
1181 excluded directory. This bug has been fixed by ignoring all quota
1182 processing for deliveries into excluded directories.
1184 PH/10 Added the maildirfolder_create_regex option to appendfile.
1190 PH/01 The code for finding all the local interface addresses on a FreeBSD
1191 system running IPv6 was broken. This may well have applied to all BSD
1192 systems, as well as to others that have similar system calls. The broken
1193 code found IPv4 interfaces correctly, but gave incorrect values for the
1194 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1195 that it would not match correctly against @[] and not recognize the IPv6
1198 PH/02 The ipliteral router was not recognizing addresses of the form user@
1199 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1201 PH/03 Added disable_ipv6.
1203 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1204 decoded form, because it is most often used to construct To: headers
1205 lines in autoreplies, and the decoded form may well be syntactically
1206 invalid. However, $reply_address has leading white space removed, and all
1207 newlines turned into spaces so that the autoreply transport does not
1210 PH/05 If group was specified without a user on a router, and no group or user
1211 was specified on a transport, the group from the router was ignored.
1213 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1214 for visible compile-time settings that can be used to change these
1215 numbers, for those that want even more. Backwards compatibility with old
1216 spool files has been maintained. However, going back to a previous Exim
1217 release will lost any variables that are in spool files.
1219 PH/07 Two small changes when running in the test harness: increase delay when
1220 passing a TCP/IP connection to a new process, in case the original
1221 process has to generate a bounce, and remove special handling of
1222 127.0.0.2 (sic), which is no longer necessary.
1224 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1225 be the same on different OS.
1227 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1230 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1231 whether --show-vars was specified or not
1233 JJ/02 exipick: Added support for new ACL variable spool format introduced
1236 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1237 syntactically invalid From: or Reply-to: line, and a filter used this to
1238 generate an autoreply, and therefore failed to obtain an address for the
1239 autoreply, Exim could try to deliver to a non-existent relative file
1240 name, causing unrelated and misleading errors. What now happens is that
1241 it logs this as a hard delivery error, but does not attempt to create a
1244 PH/11 The exinext utility has a -C option for testing purposes, but although
1245 the given file was scanned by exinext itself; it wasn't being passed on
1246 when Exim was called.
1248 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1249 an end-of-file indication when reading a command response.
1251 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1252 compiled. In many other places in Exim, IPv6 addresses are always
1253 recognized, so I have changed this. It also means that IPv4 domain
1254 literals of the form [IPV4:n.n.n.n] are now always recognized.
1256 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1257 used if the router is not running as root, for example, when verifying at
1258 ACL time, or when using -bh. The debugging output from this situation was
1259 non-existent - all you got was a failure to exec. I have made two
1262 (a) Failures to set uid/gid, the current directory, or a process leader
1263 in a subprocess such as that created by queryprogram now generate
1264 suitable debugging ouput when -d is set.
1266 (b) The queryprogram router detects when it is not running as root,
1267 outputs suitable debugging information if -d is set, and then runs
1268 the subprocess without attempting to change uid/gid.
1270 PH/15 Minor change to Makefile for building test_host (undocumented testing
1273 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1274 additional section of a DNS packet that returns MX or SRV records.
1275 Instead, it always explicitly searches for A/AAAA records. This avoids
1276 major problems that occur when a DNS server includes only records of one
1277 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1278 fixed another bug: if SRV records were looked up and the corresponding
1279 address records were *not* found in the additional section, the port
1280 values from the SRV records were lost.
1282 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1283 using the correct key (the original address) when searching the retry
1284 rules in order to find which one to use for generating the retry hint.
1286 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1287 adding the default one. Similarly, if it contains a Reply-To: header, the
1288 errors_reply_to option, if set, is not used.
1290 PH/19 When calculating a retry time, Exim used to measure the "time since
1291 failure" by looking at the "first failed" field in the retry record. Now
1292 it does not use this if it is later than than the arrival time of the
1293 message. Instead it uses the arrival time. This makes for better
1294 behaviour in cases where some deliveries succeed, thus re-setting the
1295 "first failed" field. An example is a quota failure for a huge message
1296 when small messages continue to be delivered. Without this change, the
1297 "time since failure" will always be short, possible causing more frequent
1298 delivery attempts for the huge message than are intended.
1299 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1301 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1302 $1, $2, $3) because the numerical variables can be reset during some
1303 expansion items (e.g. "match"), thereby losing the authentication data.
1305 PH/21 Make -bV show the size of off_t variables so that the test suite can
1306 decide whether to run tests for quotas > 2G.
1308 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1309 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1310 filecount value is greater than 2G or if a quota value is greater than 2G
1311 on a system where the size of off_t is not greater than 4, a panic error
1314 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1315 never match. The debug and -bh output now contains an explicit error
1316 message indicating a malformed IPv4 address or mask.
1318 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1319 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1320 PH/23 above applies.
1322 PH/25 Do not write to syslog when running in the test harness. The only
1323 occasion when this arises is a failure to open the main or panic logs
1324 (for which there is an explicit test).
1326 PH/26 Added the /no_tell option to "control=freeze".
1328 PH/27 If a host name lookup failed very early in a connection, for example, if
1329 the IP address matched host_lookup and the reverse lookup yielded a name
1330 that did not have a forward lookup, an error message of the form "no IP
1331 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1332 could be logged. Now it outputs the IP address instead of "NULL".
1334 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1335 allows the sender and the authenticated sender to be set when
1336 submitting a message from within Exim. Since child_open_exim() is
1337 documented for local_scan(), the new function should be too.
1339 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1340 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1341 results in an empty string is now treated as unset.
1343 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1345 PH/31 Added sender_verify_fail logging option.
1347 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1348 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1349 tidied the source and removed it altogether.
1351 PH/33 When a queue run was abandoned because the load average was too high, a
1352 log line was always written; now it is written only if the queue_run log
1353 selector is set. In addition, the log line for abandonment now contains
1354 information about the queue run such as the pid. This is always present
1355 in "start" and "stop" lines but was omitted from the "abandon" line.
1357 PH/34 Omit spaces between a header name and the colon in the error message that
1358 is given when verify = headers_syntax fails (if there are lots of them,
1359 the message gets confusing).
1361 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1362 names, as there are now some PTR records that contain slashes. This check
1363 is only to protect against broken name servers that fall over on strange
1364 characters, so the fact that it applies to all lookups doesn't matter.
1366 PH/36 Now that the new test suite is complete, we can remove some of the
1367 special code in Exim that was needed for the old test suite. For example,
1368 sorting DNS records because real resolvers return them in an arbitrary
1369 order. The new test suite's fake resolver always returns records in the
1372 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1373 bounces) except when queue_only is set, to avoid logging races between
1374 the different processes.
1376 PH/38 Panic-die if .include specifies a non-absolute path.
1378 PH/39 A tweak to the "H" retry rule from its user.
1380 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1381 a label. They prevented compilation on older perls.
1383 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1384 a warning to be raised on newish perls.
1386 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1387 on queue. Changes to match documented behaviour of showing count of
1388 messages matching specified criteria.
1390 PH/40 Changed the default ident timeout from 30s to 5s.
1392 PH/41 Added support for the use of login_cap features, on those BSD systems
1393 that have them, for controlling the resources used by pipe deliveries.
1395 PH/42 The content-scanning code uses fopen() to create files in which to put
1396 message data. Previously it was not paying any attention to the mode of
1397 the files. Exim runs with umask(0) because the rest of the code creates
1398 files with open(), and sets the required mode explicitly. Thus, these
1399 files were ending up world-writeable. This was not a big issue, because,
1400 being within the spool directory, they were not world-accessible. I have
1401 created a function called modefopen, which takes an additional mode
1402 argument. It sets umask(777), creates the file, chmods it to the required
1403 mode, then resets the umask. All the relevant calls to fopen() in the
1404 content scanning code have been changed to use this function.
1406 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1407 to 24 hours. This avoids potential overflow problems when processing G
1408 and H retry rules. I suspect nobody ever tinkers with this value.
1410 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1412 PH/45 When the plaintext authenticator is running as a client, the server's
1413 challenges are checked to ensure they are valid base64 strings. By
1414 default, the authentication attempt is cancelled if an invalid string is
1415 received. Setting client_ignore_invalid_base64 true ignores these errors.
1416 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1417 they are received. Thus, the responses can be made to depend on the
1418 challenges. If an invalid string is ignored, an empty string is placed in
1421 PH/46 Messages that are created by the autoreply transport now contains a
1422 References: header, in accordance with RFCs 2822 and 3834.
1424 PH/47 Added authenticated_sender_force to the smtp transport.
1426 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1428 PH/49 Installed latest patch from the Sieve maintainer.
1430 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1431 was also set, the appendfile transport was unnecessarily scanning a
1432 directory of message files (e.g. for maildir delivery) to find the count
1433 of files (along with the size), even though it did not need this
1434 information. It now does the scan only if it needs to find either the
1435 size of the count of files.
1437 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1439 PH/52 Two bugs concerned with error handling when the smtp transport is
1442 (i) Exim was not creating retry information for temporary errors given
1443 for individual recipients after the DATA command when the smtp transport
1444 was used in LMTP mode. This meant that they could be retried too
1445 frequently, and not timed out correctly.
1447 (ii) Exim was setting the flag that allows error details to be returned
1448 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1449 recipients that were returned after the DATA command.
1451 PH/53 This is related to PH/52, but is more general: for any failing address,
1452 when detailed error information was permitted to be returned to the
1453 sender, but the error was temporary, then after the final timeout, only
1454 "retry timeout exceeded" was returned. Now it returns the full error as
1455 well as "retry timeout exceeded".
1457 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1458 do this, and (what is worse) MTAs that accept it.
1460 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1461 will now be deprecated.
1463 PH/56 New os.c-cygwin from the Cygwin maintainer.
1465 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1466 formats (previously only available in exim formats via -bpr, -bpru,
1467 and -bpra. Now also available in native and exiqgrep formats)
1469 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1470 with very large, slow to parse queues
1472 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1474 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1476 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1477 responses to authentication challenges, though it was showing the
1478 challenges; (ii) I've removed the CR characters from the debug output for
1481 PH/58 Allow for the insertion of a newline as well as a space when a string
1482 is turned into more than one encoded-word during RFC 2047 encoding. The
1483 Sieve code now uses this.
1485 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1486 data_4xx, lost_connection, tls_required.
1488 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1489 message was being sent as an SMTP response.
1491 PH/61 Add -l and -k options to exicyclog.
1493 PH/62 When verifying, if an address was redirected to one new address, so that
1494 verification continued, and the new address failed or deferred after
1495 having set something in $address_data, the value of $address_data was not
1496 passed back to the ACL. This was different to the case when no
1497 redirection occurred. The value is now passed back in both cases.
1499 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1500 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1501 use login_cap.h, so on its own it isn't the distinguishing feature. The
1502 new name refers directly to the setclassresources() function.
1504 PH/65 Added configuration files for NetBSD3.
1506 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1508 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1509 is preferred over IPv4.
1511 PH/68 The bounce_return_message and bounce_return_body options were not being
1512 honoured for bounces generated during the reception of non-SMTP messages.
1513 In particular, this applied to messages rejected by the ACL. This bug has
1514 been fixed. However, if bounce_return_message is true and bounce_return_
1515 body is false, the headers that are returned for a non-SMTP message
1516 include only those that have been read before the error was detected.
1517 (In the case of an ACL rejection, they have all been read.)
1519 PH/69 The HTML version of the specification is now built in a directory called
1520 spec_html instead of spec.html, because the latter looks like a path with
1521 a MIME-type, and this confuses some software.
1523 PH/70 Catch two compiler warnings in sieve.c.
1525 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1526 function verify_get_ident() calls ip_connect() to connect a socket, but
1527 if the "connect()" function timed out, ip_connect() used to close the
1528 socket. However, verify_get_ident() also closes the socket later, and in
1529 between Exim writes to the log, which may get opened at this point. When
1530 the socket was closed in ip_connect(), the log could get the same file
1531 descriptor number as the socket. This naturally causes chaos. The fix is
1532 not to close the socket in ip_connect(); the socket should be closed by
1533 the function that creates it. There was only one place in the code where
1534 this was missing, in the iplookup router, which I don't think anybody now
1535 uses, but I've fixed it anyway.
1537 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1538 well as to direct DNS lookups. Otherwise the handling of names in host
1539 lists is inconsistent and therefore confusing.
1545 PH/01 Two changes to the default runtime configuration:
1547 (1) Move the checks for relay_from_hosts and authenticated clients from
1548 after to before the (commented out) DNS black list checks.
1550 (2) Add control=submission to the relay_from_hosts and authenticated
1551 clients checks, on the grounds that messages accepted by these
1552 statements are most likely to be submissions.
1554 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1556 (1) Generate an error if the third argument for the ${prvs expansion is
1559 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1562 (3) Reset the variables that are obtained from the first argument of
1563 ${prvscheck and used in the second argument before leaving the code,
1564 because their memory is reclaimed, so using them afterwards may do
1567 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1568 one (it's much easier than Tom thought :-).
1570 (5) Because of (4), we can now allow for the use of $prvscheck_result
1571 inside the third argument.
1573 PH/03 For some reason, the default setting of PATH when running a command from
1574 a pipe transport was just "/usr/bin". I have changed it to
1577 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1578 anything to be listed in the output from -bV.
1580 PH/05 When a filter generated an autoreply, the entire To: header line was
1581 quoted in the delivery log line, like this:
1583 => >A.N.Other <ano@some.domain> <original@ddress> ...
1585 This has been changed so that it extracts the operative address. There
1586 may be more than one such address. If so, they are comma-separated, like
1589 => >ano@some.domain,ona@other.domain <original@ddress> ...
1591 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1592 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1593 not being looked up in the rDNS to get a host name, Exim was showing the
1594 IP address twice in Received: lines, even though the IP addresses were
1595 identical. For example:
1597 Received: from [1.2.3.4] (helo=[1.2.3.4])
1599 However, if the real host name was known, it was omitting the HELO data
1600 if it matched the actual IP address. This has been tidied up so that it
1601 doesn't show the same IP address twice.
1603 PH/07 When both +timestamp and +memory debugging was on, the value given by
1604 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1605 called by the debug printing, thereby overwriting the timestamp buffer.
1606 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1608 PH/08 When the original message was included in an autoreply transport, it
1609 always said "this is a copy of the message, including all the headers",
1610 even if body_only or headers_only was set. It now gives an appropriate
1613 PH/09 Applied a patch from the Sieve maintainer which:
1615 o fixes some comments
1616 o adds the (disabled) notify extension core
1617 o adds some debug output for the result of if/elsif tests
1618 o points to the current vacation draft in the documentation
1619 and documents the missing references header update
1623 o fixes a bug in processing the envelope test (when testing
1624 multiple envelope elements, the last element determinted the
1627 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1628 Electronic Mail") by including:
1630 Auto-submitted: auto-generated
1632 in the messages that it generates (bounce messages and others, such as
1633 warnings). In the case of bounce messages for non-SMTP mesages, there was
1634 also a typo: it was using "Auto_submitted" (underscore instead of
1635 hyphen). Since every message generated by Exim is necessarily in response
1636 to another message, thes have all been changed to:
1638 Auto-Submitted: auto-replied
1640 in accordance with these statements in the RFC:
1642 The auto-replied keyword:
1644 - SHOULD be used on messages sent in direct response to another
1645 message by an automatic process,
1647 - MUST NOT be used on manually-generated messages,
1649 - MAY be used on Delivery Status Notifications (DSNs) and Message
1650 Disposition Notifications (MDNs),
1652 - MUST NOT be used on messages generated by automatic or periodic
1653 processes, except for messages which are automatic responses to
1656 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1657 to the default Received: header definition.
1659 PH/12 Added log selector acl_warn_skipped (default on).
1661 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1662 variables because (a) they are in the wrong storage pool and (b) even if
1663 they were copied, it wouldn't work properly because of the caching.
1665 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1666 checking when decoding. Apparently there are clients that generate
1667 overlong encoded strings. Why am I not surprised?
1669 PH/15 If the first argument of "${if match_address" was not empty, but did not
1670 contain an "@" character, Exim crashed. Now it writes a panic log message
1671 and treats the condition as false.
1673 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1675 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1676 extension "envelope-auth". The code is finished and in agreement with
1677 other implementations, but there is no documentation so far and in fact,
1678 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1679 not changing the active code.
1681 Print executed "if" and "elsif" statements when debugging is used. This
1682 helps a great deal to understand what a filter does.
1684 Document more things not specified clearly in RFC3028. I had all this
1685 sorted out, when out of a sudden new issues came to my mind. Oops."
1687 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1690 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1691 canonical form (as documented). However, after a host name lookup from
1692 the IP address, check_host() was doing a simple string comparison with
1693 addresses acquired from the DNS when checking that the found name did
1694 have the original IP as one of its addresses. Since any found IPv6
1695 addresses are likely to be in abbreviated form, the comparison could
1696 fail. Luckily, there already exists a function for doing the comparison
1697 by converting both addresses to binary, so now that is used instead of
1698 the text comparison.
1700 PH/20 There was another similar case to PH/19, when a complete host name was
1701 given in a host list; looking up its IP address could give an abbreviated
1702 form, whereas the current host's name might or might not be abbreviated.
1703 The same fix has been applied.
1709 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1710 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1713 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1714 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1716 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1718 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1719 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1720 Also, for "Auto-Submitted", treat anything other than "no" as
1721 non-personal, in accordance with RFC 3834. (Previously it treated
1722 anything starting "auto-" as non-personal.)
1724 TF/01 The control=submission/name=... option had a problem with syntax
1725 errors if the name included a slash character. The /name= option
1726 now slurps the rest of the string, so it can include any characters
1727 but it must come last in the list of options (after /sender_retain
1730 PH/05 Some modifications to the interface to the fake nameserver for the new
1738 TK/01 Added the "success_on_redirect" address verification option. See
1739 NewStuff for rationale and an example.
1741 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1743 PH/02 Patch to exigrep to allow it to work on syslog lines.
1745 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1746 fread() to skip over the body file's header line, because in Cygwin the
1747 header line is locked and is inaccessible.
1749 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1750 co-exist for some time) to make it clear that it is the Exim ID that is
1751 referenced, not the Message-ID: header line.
1753 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1754 string_format() function, because snprintf() does not exist on all
1757 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1758 ${stat: expansion item.
1760 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1761 protocol synchronization error", to keep the pedants happy.
1763 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1764 well as for IRIX systems, when gcc is being used. See the host.c source
1767 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1769 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1772 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1773 transport and to the smtp transport in LMTP mode.
1775 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1777 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1778 run a filter in a subprocess. This could lead to confusion in subsequent
1779 lookups in the parent process. There should also be a search_tidyup() at
1780 the end of the subprocess.
1782 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1783 only if the host matched helo_try_verify_hosts, which caused the
1784 verification to occur when the EHLO/HELO command was issued. The ACL just
1785 tested the remembered result. Now, if a previous verification attempt has
1786 not happened, "verify = helo" does it there and then.
1788 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1790 TK/03 Fix log output including CR from clamd.
1792 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1793 exist provoked a memory error which could cause a segfault.
1795 PH/15 Installed PCRE 6.2
1797 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1799 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1800 of the problem. Specifically, suggested +O2 rather than +O1 for the
1803 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1805 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1806 record, the port was not correctly logged when the outgoing_port log
1807 selector was set (it logged the transort's default port).
1809 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1810 fallback_hosts, and "hosts" in the smtp transport.
1812 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1813 host errors such as "Connection refused".
1815 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1816 authentication with radiusclient 0.4.9:
1818 - Error returned from rc_read_config was caught wrongly
1819 - Username/password not passed on to radius server due to wrong length.
1821 The presumption is that some radiusclient API changes for 4.51/PH/17
1822 were not taken care of correctly. The code is still untested by me (my
1823 Linux distribution still has 0.3.2 of radiusclient), but it was
1824 contributed by a Radius user.
1826 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1827 expanding the "port" option of the smtp transport.
1829 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1830 while reading a MIME header. Thanks to Tom Hughes for a patch.
1832 PH/24 Include config.h inside local_scan.h so that configuration settings are
1835 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1836 that in an ACL for RCPT (for example), you can examine exactly what was
1839 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1840 commands, but it was not correctly comparing the address with the actual
1841 client host address. Thus, it would show the EHLO address in Received:
1842 header lines when this was not necessary.
1844 PH/27 Added the % operator to ${eval:}.
1846 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1847 it should be ignoring failures (because with -C, for example, it has lost
1848 privilege). It wasn't ignoring creation failures other than "already
1851 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1854 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1855 because an input line was too long, either on its own, or by virtue of
1856 too many continuations, the temporary file was not being removed, and the
1857 return code was incorrect.
1859 PH/31 Missing "BOOL" in function definition in filtertest.c.
1861 PH/32 Applied Sieve patches from the maintainer.
1863 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1865 PH/33 Added "verify = not_blind".
1867 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1868 Local/Makefile (with some defaults set). These are used in built scripts
1869 such as exicyclog, but they have never been used in the exim_install
1870 script (though there are many overriding facilities there). I have
1871 arranged that the exim_install script now takes note of these two
1874 PH/35 Installed configuration files for Dragonfly.
1876 PH/36 When a locally submitted message by a trusted user did not contain a
1877 From: header, and the sender address was obtained from -f or from an SMTP
1878 MAIL command, and the trusted user did not use -F to supply a sender
1879 name, $originator_name was incorrectly used when constructing a From:
1880 header. Furthermore, $originator_name was used for submission mode
1881 messages from external hosts without From: headers in a similar way,
1882 which is clearly wrong.
1884 PH/37 Added control=suppress_local_fixups.
1886 PH/38 When log_selector = +received_sender was set, and the addition of the
1887 sender made the log line's construction buffer exactly full, or one byte
1888 less than full, an overflow happened when the terminating "\n" was
1891 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1892 when the result of a list match is failure because a DNS lookup failed.
1894 PH/40 RM_COMMAND is now used in the building process.
1896 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1897 the "build-* directories that it finds.
1899 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1900 address in a domain literal was a prefix of an interface address.
1902 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1903 when verifying a sender address, unless rewrite_headers is false.
1905 PH/44 Wrote a long comment about why errors_to addresses are verified as
1906 recipients, not senders.
1908 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1909 the ratelimit ACL was added.
1911 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1913 PH/46 Added extra information about PostgreSQL errors to the error string.
1915 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1916 avoiding the need to install special zones in a real server. This is
1917 backwards compatible; if it can't find the fake resolver, it drops back.
1918 Thus, both old and new test suites can be run.
1920 TF/02 Added util/ratelimit.pl
1922 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1923 clock is set back in time.
1925 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1926 Candler <B.Candler@pobox.com>.
1928 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1929 OK for addresses that are the result of redirections.
1931 PH/48 A number of further additions for the benefit of the new test suite,
1932 including a fake gethostbyname() that interfaces to the fake DNS resolver
1935 TF/06 The fix for widen_domains has also been applied to qualify_single and
1936 search_parents which are the other dnslookup options that can cause
1939 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1942 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1944 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1945 service" if it encountered an MX record with an empty target hostname.
1946 The message is now "an MX or SRV record indicated no SMTP service".
1948 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1949 if the DNS of the sending site is misconfigured. This is quite a
1950 common situation. This change restores the behaviour of treating a
1951 helo verification defer as a failure.
1953 PH/51 If self=fail was set on a router, the bounce message did not include the
1954 actual error message.
1960 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1962 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1963 command itself ended in error, the underlying message about the transport
1964 filter timeout was being overwritten with the pipe command error. Now the
1965 underlying error message should be appended to the second error message.
1967 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1969 PH/02 Reduce the amount of output that "make" produces by default. Full output
1970 can still be requested.
1972 PH/03 The warning log line about a condition test deferring for a "warn" verb
1973 was being output only once per connection, rather than after each
1974 occurrence (because it was using the same function as for successful
1975 "warn" verbs). This seems wrong, so I have changed it.
1977 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1978 it should not have, which might have caused a crash in the right
1979 circumstances, but probably never did.
1981 PH/04 Installed a modified version of Tony Finch's patch to make submission
1982 mode fix the return path as well as the Sender: header line, and to
1983 add a /name= option so that you can make the user's friendly name appear
1986 TF/03 Added the control = fakedefer ACL modifier.
1988 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1989 Mark Lowes for thorough testing.
1991 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1993 TK/03 Merged latest SRS patch from Miles Wilton.
1995 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1996 with the definition in sysexits.h (which is #included earlier).
1997 Fortunately, Exim does not actually use EX_OK. The code used to try to
1998 preserve the sysexits.h value, by assumimg that macro definitions were
1999 scanned for macro replacements. I have been disabused of this notion,
2000 so now the code just undefines EX_OK before #including unistd.h.
2002 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2003 in the smtp transport. When a block could not be written in a single
2004 write() function, the timeout was being re-applied to each part-write.
2005 This seems wrong - if the receiver was accepting one byte at a time it
2006 would take for ever. The timeout is now adjusted when this happens. It
2007 doesn't have to be particularly precise.
2009 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2010 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2012 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2013 without a domain if the check to <postmaster@domain> fails.
2015 SC/01 Eximstats: added -xls and the ability to specify output files
2016 (patch written by Frank Heydlauf).
2018 SC/02 Eximstats: use FileHandles for outputing results.
2020 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2022 SC/04 Eximstats: fixed display of large numbers with -nvr option
2024 SC/05 Eximstats: fixed merging of reports with empty tables.
2026 SC/06 Eximstats: added the -include_original_destination flag
2028 SC/07 Eximstats: removed tabs and trailing whitespace.
2030 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2032 TK/06 MBOX spool code: Add real "From " MBOX separator line
2033 so the .eml file is really in mbox format (even though
2034 most programs do not really care). Patch from Alex Miller.
2036 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2037 The latter is generated from $received_to and is only set if the
2038 message has one envelope recipient. SA can use these headers,
2039 obviously out-of-the-box. Patch from Alex Miller.
2041 PH/08 The ${def test on a variable was returning false if the variable's
2042 value was "0", contrary to what the specification has always said!
2043 The result should be true unless the variable is empty.
2045 PH/09 The syntax error of a character other than { following "${if
2046 def:variable_name" (after optional whitespace) was not being diagnosed.
2047 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2048 accidental colon was present, for example, could give incorrect results.
2050 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2051 result is used (not including appendfile, where other changes are about
2054 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2055 This involved changing a lot of size variables from int to off_t. It
2056 should work with maildirs and everything.
2058 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2059 spamd dying while we are connected to it.
2061 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2064 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2065 function for BDB 4.3.
2067 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2069 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2070 the sources for PCRE have been re-arranged and more files are now
2073 PH/15 The code I had for printing potentially long long variables in PH/11
2074 above was not the best (it lost precision). The length of off_t variables
2075 is now inspected at build time, and an appropriate printing format (%ld
2076 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2077 to be "long long int" or "long int". This is needed for the internal
2078 formatting function string_vformat().
2080 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2081 the configuration file to be ":syslog", then the script "guesses" where
2082 the logs files are, rather than using the compiled in default. In our
2083 case the guess is not the same as the compiled default, so the script
2084 suddenly stopped working when I started to use syslog. The patch checks
2085 to see if log_file_path is "". If so, it attempts to read it from exim
2086 with no configuration file to get the compiled in version, before it
2087 falls back to the previous guessing code."
2089 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2090 implementing BATV in an Exim configuration. See NewStuff for the gory
2093 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2094 Makefile that are specific to HP-UX.
2096 PH/18 If the "use_postmaster" option was set for a recipient callout together
2097 with the "random" option, the postmaster address was used as the MAIL
2098 FROM address for the random test, but not for the subsequent recipient
2099 test. It is now used for both.
2101 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2102 patch removes a few documentation additions to RFC 3028, because the
2103 latest draft now contains them. It adds the new en;ascii-case comparator
2104 and a new error check for 8bit text in MIME parts. Comparator and
2105 require names are now matched exactly. I enabled the subaddress
2106 extension, but it is not well tested yet (read: it works for me)."
2108 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2109 rework some of the code of TK/09 above to avoid the hardwired use of
2110 "%lld" and "long long". Replaced the call to snprintf() with a call to
2113 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2114 records point to non-existent hosts", "retry timeout exceeded", and
2115 "retry time not reached for any host after a long failure period".
2117 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2118 experimental DomainKeys support:
2120 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2121 (2) On an error such as an illegally used "control", the wrong name for
2122 the control was given.
2124 These problems did NOT occur unless DomainKeys support was compiled.
2126 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2128 PH/24 Added ${if match_ip condition.
2130 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2131 databases so that it will be absolutely obvious if a crash occurs in the
2132 DB library. This is a regular occurrence (often caused by mis-matched
2135 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2136 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2137 picked up on a user's system that detects such things. There doesn't seem
2138 to be a gcc warning option for this - only an attribute that has to be
2139 put on the function's prototype. It seems that in Fedora Core 4 they have
2140 set this on a number of new functions. No doubt there will be more in due
2143 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2144 specify a transport. However, if an address that was verified by such a
2145 router was the subject of a callout, Exim crashed because it tried to
2146 read the rcpt_include_affixes from the non-existent transport. Now it
2147 just assumes that the setting of that option is false. This bug was
2148 introduced by 4.51/PH/31.
2150 PH/28 Changed -d+all to exclude +memory, because that information is very
2151 rarely of interest, but it makes the output a lot bigger. People tend to
2152 do -d+all out of habit.
2154 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2155 code in os-type was giving problems when libc.so lives in lib64, like on
2158 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2159 aren't the modern standard, and it seems that some systems' include files
2160 don't always have them. Exim was already checking for some of the newer
2161 ones like T_AAAA, and defining it itself. I've added checks for all the
2162 record types that Exim uses.
2164 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2165 not automatically generating a new one, as it is supposed to. This
2166 prevented TLS from working. If the file did exist, but contained invalid
2167 data, a new version was generated, as expected. It was only the case of a
2168 non-existent file that was broken.
2170 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2171 with a change in libdomainkeys > 0.64.
2173 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2174 from DNS. If the selector record carries the flag, it now has
2175 precedence over the domain-wide flag.
2177 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2179 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2180 the use of a port name that isn't defined in /etc/services) occurred, the
2181 message was deferred as in a normal delivery, and thus remained on the
2182 spool, instead of being failed because of the mua_wrapper setting. This
2183 is now fixed, and I tidied up some of the mua_wrapper messages at the
2186 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2187 the messages in a hash of arrays rather than using individual hashes.
2188 This is a bit cleaner and results in dramatic memory savings, albeit
2189 at a slight CPU cost.
2191 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2192 as requested by Marc Sherman.
2194 SC/10 Eximstats: added histograms for user specified patterns as requested
2197 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2199 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2200 fopen() in the content-scanning modules that did not already have it.
2206 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2207 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2209 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2211 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2213 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2214 to test Sieve filters that use "vacation".
2216 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2217 that changes the way the GnuTLS parameters are stored in the cache file.
2218 The new format can be generated externally. For backward compatibility,
2219 if the data in the cache doesn't make sense, Exim assumes it has read an
2220 old-format file, and it generates new data and writes a new file. This
2221 means that you can't go back to an older release without removing the
2224 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2225 work if there are any delivery delays because "one_time" forces the
2226 parent to be marked "delivered", so its unseen clone is never tried
2227 again. For this reason, Exim now forbids the simultaneous setting of
2230 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2231 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2232 entry if you want to know the details. The fix, however, neglected to
2233 consider the case when local delivery batching is involved. The test for
2234 "previously delivered" was not happening when checking to see if an
2235 address could be batched with a previous (undelivered) one; under
2236 certain circumstances this could lead to multiple deliveries to the same
2239 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2240 in its include files, and this causes problems building Exim.
2242 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2243 header_syntax) but Exim was just ignoring anything given after a slash.
2244 In particular, this caused confusion with an attempt to use "verify =
2245 reverse_host_lookup/defer_ok". An error is now given when options are
2246 supplied for verify items that do not have them. (Maybe reverse_host_
2247 lookup should have a defer_ok option, but that's a different point.)
2249 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2250 defined by RFC 821) to 2048, because there were problems with some AUTH
2251 commands, and RFC 1869 says the size should be increased for extended
2252 SMTP commands that take arguments.
2254 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2257 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2258 "unknown" error; now it says that the functionality isn't in the binary.
2260 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2261 an address' error message when a string expansion fails (syntax or
2262 whatever). Otherwise the password may appear in the log. Following change
2263 PH/42 below, there is no longer a chance of it appearing in a bounce
2266 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2268 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2269 of its addresses was ever tried. (Bugzilla bug #2.)
2271 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2272 the result incorrectly in the debug output. (It correctly added a newline
2273 to what was transported.)
2275 TF/01 Added $received_time.
2277 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2278 commented out examples of how to interface to a virus scanner and to
2279 SpamAssassin. Also added commented examples of av_scanner and
2280 spamd_address settings.
2282 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2283 and controls are allowed in which ACLs. There were a couple of minor
2284 errors. Some of the entries in the conditions table (which is a table of
2285 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2286 negation of where the condition IS allowed.
2288 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2290 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2291 header file does not have a version number, so I've had to invent a new
2292 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2293 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2294 radiusclient), but it was contributed by a Radius user.
2296 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2297 files or directories, for OpenSSL.
2299 PH/19 When an Exim process that is running as root has to create an Exim log
2300 file, it does so in a subprocess that runs as exim:exim so as to get the
2301 ownership right at creation (otherwise, other Exim processes might see
2302 the file with the wrong ownership). There was no test for failure of this
2303 fork() call, which would lead to the process getting stuck as it waited
2304 for a non-existent subprocess. Forks do occasionally fail when resources
2305 run out. I reviewed all the other calls to fork(); they all seem to check
2308 PH/20 When checking for unexpected SMTP input at connect time (before writing
2309 the banner), Exim was not dealing correctly with a non-positive return
2310 from the read() function. If the client had disconnected by this time,
2311 the result was a log entry for a synchronization error with an empty
2312 string after "input=" when read() returned zero. If read() returned -1
2313 (an event I could not check), uninitialized data bytes were printed.
2314 There were reports of junk text (parts of files, etc) appearing after
2317 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2319 PH/22 Added support for macro redefinition, and (re)definition in between
2320 driver and ACL definitions.
2322 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2323 forgetting to use the resulting value; it was using the unexpanded value.
2325 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2326 hadn't been configured. The fix is from Juergen Kreileder, who
2327 understands it better than I do:
2329 "Here's what I see happening with three configured cyrus_sasl
2330 authenticators configured (plain, login, cram-md5):
2332 On startup auth_cyrus_sasl_init() gets called for each of these.
2333 This means three calls to sasl_listmech() without a specified mech_list.
2334 => SASL tests which mechs of all available mechs actually work
2335 => three warnings about OTP not working
2336 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2338 With the patch, sasl_listmech() also gets called three times. But now
2339 SASL's mech_list option is set to the server_mech specified in the the
2340 authenticator. Or in other words, the answer from sasl_listmech()
2341 gets limited to just the mech you're testing for (which is different
2343 => the return list contains just 'plain' or 'login', 'cram-md5' or
2344 nothing depending on the value of ob->server_mech.
2346 I've just tested the patch: Authentication still works fine,
2347 unavailable mechs specified in the exim configuration are still
2348 caught, and the auth.log warnings about OTP are gone."
2350 PH/25 When debugging is enabled, the contents of the command line are added
2351 to the debugging output, even when log_selector=+arguments is not
2354 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2355 answer is "GNU", and only if the return is "GNU/something" is the answer
2358 PH/27 $acl_verify_message is now set immediately after the failure of a
2359 verification in an ACL, and so is available in subsequent modifiers. In
2360 particular, the message can be preserved by coding like this:
2362 warn !verify = sender
2363 set acl_m0 = $acl_verify_message
2365 Previously, $acl_verify_message was set only while expanding "message"
2366 and "log_message" when a very denied access.
2368 PH/28 Modified OS/os.c-Linux with
2370 -#ifndef OS_LOAD_AVERAGE
2371 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2373 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2374 nomenclature these days.)
2376 PH/29 Installed patch from the Sieve maintainer that adds the options
2377 sieve_useraddress and sieve_subaddress to the redirect router.
2379 PH/30 In these circumstances:
2380 . Two addresses routed to the same list of hosts;
2381 . First host does not offer TLS;
2382 . First host accepts first address;
2383 . First host gives temporary error to second address;
2384 . Second host offers TLS and a TLS session is established;
2385 . Second host accepts second address.
2386 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2387 and peerdn, if requested) that were in fact used only for the second
2390 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2391 attention to any local part prefix or suffix that was matched by the
2392 router that accepted the address. It now behaves in the same way as it
2393 does for delivery: the affixes are removed from the local part unless
2394 rcpt_include_affixes is set on the transport.
2396 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2397 timeout during the DATA phase of an incoming message.
2399 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2400 applied a patch sanctioned by the Sieve maintainer.
2402 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2403 the uid or gid is negative. A case of a negative gid caused this to be
2404 noticed. The fix allows for either to be negative.
2406 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2407 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2410 PH/36 The change PH/12 above was broken. Fixed it.
2412 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2413 the grounds that routing the same address twice would always produce the
2414 same answer. This might have been true once, but it is certainly no
2415 longer true now. Routing a child address may depend on the previous
2416 routing that produced that child. Some complicated redirection strategies
2417 went wrong when messages had multiple recipients, and made Exim's
2418 behaviour dependent on the order in which the addresses were given.
2420 I have moved the duplicate checking until after the routing is complete.
2421 Exim scans the addresses that are assigned to local and remote
2422 transports, and removes any duplicates. This means that more work will be
2423 done, as duplicates will always all be routed, but duplicates are
2424 presumably rare, so I don't expect this is of any significance.
2426 For deliveries to pipes, files, and autoreplies, the duplicate checking
2427 still happens during the routing process, since they are not going to be
2430 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2431 It corrects a timeout issue with spamd. This is Ian's comment: "The
2432 background is that sometimes spamd either never reads data from a
2433 connection it has accepted, or it never writes response data. The exiscan
2434 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2435 blindly assumes that writes won't block so it may never time out."
2437 PH/39 Allow G after quota size as well as K and M.
2439 PH/40 The value set for $authenticated_id in an authenticator may not contain
2440 binary zeroes or newlines because the value is written to log lines and
2441 to spool files. There was no check on this. Now the value is run through
2442 the string_printing() function so that such characters are converted to
2443 printable escape sequences.
2445 PH/41 $message_linecount is a new variable that contains the total number of
2446 lines in the message. Compare $body_linecount, which is the count for the
2449 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2450 bounce and delay warning messages, except in certain special cases, which
2453 (a) An SMTP error message from a remote host;
2454 (b) A message specified in a :fail: redirection;
2455 (c) A message specified in a "fail" command in a system filter;
2456 (d) A message specified in a FAIL return from the queryprogram router;
2457 (e) A message specified by the cannot_route_message router option.
2459 In these cases only, Exim does include the error details in bounce and
2460 warning messages. There are also a few cases where bland messages such
2461 as "unrouteable address" or "local delivery error" are given.
2463 PH/43 $value is now also set for the "else" part of a ${run expansion.
2465 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2466 being worked on, but at least Exim now implements the latest version to
2469 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2470 process to complete was treated as a delivery failure, a timeout while
2471 writing the message to the pipe was logged, but erroneously treated as a
2472 successful delivery. Such timeouts include transport filter timeouts. For
2473 consistency with the overall process timeout, these timeouts are now
2474 treated as errors, giving rise to delivery failures by default. However,
2475 there is now a new Boolean option for the pipe transport called
2476 timeout_defer, which, if set TRUE, converts the failures into defers for
2477 both kinds of timeout. A transport filter timeout is now identified in
2480 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2481 systems where "make" and "gmake" are different, calling "gmake" at top
2482 level broke things. I've arranged for the value of $(MAKE) to be passed
2483 from the Makefile to this script so that it can call the same version of
2487 A note about Exim versions 4.44 and 4.50
2488 ----------------------------------------
2490 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2491 changes of various kinds. As a consequence, a big documentation update was
2492 needed. This delayed the release for rather longer than seemed good, especially
2493 in the light of a couple of (minor) security issues. Therefore, the changes
2494 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2495 release. So 4.44 and 4.50 are in effect two different branches that both start
2498 I have left the 4.50 change log unchanged; it contains all the changes since
2499 4.43. The change log for 4.44 is below; many of its items are identical to
2500 those for 4.50. This seems to be the most sensible way to preserve the
2501 historical information.
2507 1. Minor wording change to the doc/README.SIEVE file.
2509 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2510 computation of the current number of files was incorrect.
2512 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2513 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2516 4. Give more explanation in the error message when the command for a transport
2517 filter fails to execute.
2519 5. There are several places where Exim runs a non-Exim command in a
2520 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2521 was being done only for the command run by the queryprogram router. It is
2522 now done for all such subprocesses. The other cases are: ${run, transport
2523 filters, and the commands run by the lmtp and pipe transports.
2525 6. Added CONFIGURE_GROUP build-time option.
2527 7. Some older OS have a limit of 256 on the maximum number of file
2528 descriptors. Exim was using setrlimit() to set 1000 as a large value
2529 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2530 systems. I've change it so that if it can't get 1000, it tries for 256.
2532 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2533 was an oversight, and furthermore, ever since the addition of extra
2534 controls (e.g. 4.43/32), the checks on when to allow different forms of
2535 "control" were broken. There should now be diagnostics for all cases when a
2536 control that does not make sense is encountered.
2538 9. Added the /retain_sender option to "control=submission".
2540 10. $recipients is now available in the predata ACL (oversight).
2542 11. Tidy the search cache before the fork to do a delivery from a message
2543 received from the command line. Otherwise the child will trigger a lookup
2544 failure and thereby defer the delivery if it tries to use (for example) a
2545 cached ldap connection that the parent has called unbind on.
2547 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2548 of $address_data from the recipient verification was clobbered by the
2549 sender verification.
2551 13. The value of address_data from a sender verification is now available in
2552 $sender_address_data in subsequent conditions in the ACL statement.
2554 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2556 15. Added a new option "connect=<time>" to callout options, to set a different
2559 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2560 was its contents. (It was OK if the option was not defined at all.)
2562 17. A "Completed" log line is now written for messages that are removed from
2563 the spool by the -Mrm option.
2565 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2566 information about exactly what failed.
2568 19. Added -dd to debug only the daemon process.
2570 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2571 handles timeouts, both on the server side and network timeouts. Renamed the
2572 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2574 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2575 It is now set to "smtps".
2577 22. $host_address is now set to the target address during the checking of
2578 ignore_target_hosts.
2580 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2581 being passed; this would have caused $sender_host_name to have been used if
2582 matching the list had actually called for a host name (not very likely,
2583 since this list is usually IP addresses). A host name is now passed as
2586 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2587 code that specifies a non-restarting handler (typically sigaction() in
2588 modern systems) in an attempt to fix a rare and obscure crash bug.
2590 25. Narrowed the window for a race in the daemon that could cause it to ignore
2591 SIGCHLD signals. This is not a major problem, because they are used only to
2592 wake it up if nothing else does.
2594 26. A malformed maildirsize file could cause Exim to calculate negative values
2595 for the mailbox size or file count. Odd effects could occur as a result.
2596 The maildirsize information is now recalculated if the size or filecount
2599 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2600 support for a long time. Removed HAVE_SYS_VFS_H.
2602 28. Installed the latest version of exipick from John Jetmore.
2604 29. In an address list, if the pattern was not a regular expression, an empty
2605 subject address (from a bounce message) matched only if the pattern was an
2606 empty string. Non-empty patterns were not even tested. This was the wrong
2607 because it is perfectly reasonable to use an empty address as part of a
2608 database query. An empty address is now tested by patterns that are
2609 lookups. However, all the other forms of pattern expect the subject to
2610 contain a local part and a domain, and therefore, for them, an empty
2611 address still always fails if the pattern is not itself empty.
2613 30. Exim went into a mad DNS loop when attempting to do a callout where the
2614 host was specified on an smtp transport, and looking it up yielded more
2615 than one IP address.
2617 31. Re-factored the code for checking spool and log partition space into a
2618 function that finds that data and another that does the check. The former
2619 is then used to implement four new variables: $spool_space, $log_space,
2620 $spool_inodes, and $log_inodes.
2622 32. The RFC2047 encoding function was originally intended for short strings
2623 such as real names; it was not keeping to the 75-character limit for
2624 encoded words that the RFC imposes. It now respects the limit, and
2625 generates multiple encoded words if necessary. To be on the safe side, I
2626 have increased the buffer size for the ${rfc2047: expansion operator from
2629 33. It is now permitted to omit both strings after an "if" condition; if the
2630 condition is true, the result is "true". As before, when the second string
2631 is omitted, a false condition yields an empty string. This makes it less
2632 cumbersome to write custom ACL and router conditions.
2634 34. Failure to deliver a bounce message always caused it to be frozen, even if
2635 there was an errors_to setting on the router. The errors_to setting is now
2638 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2639 canonical form (fully expanded) before being placed in
2640 $sender_host_address.
2642 36. The table in the code that translates DNS record types into text (T_A to
2643 "A" for instance) was missing entries for NS and CNAME. It is just possible
2644 that this could have caused confusion if both these types were looked up
2645 for the same domain, because the text type is used as part of Exim's
2646 per-process caching. But the chance of anyone hitting this buglet seems
2649 37. The dnsdb lookup has been extended in a number of ways.
2651 (1) There is a new type, "zns", which walks up the domain tree until it
2652 finds some nameserver records. It should be used with care.
2654 (2) There is a new type, "mxh", which is like "mx" except that it returns
2655 just the host names, not the priorities.
2657 (3) It is now possible to give a list of domains (or IP addresses) to be
2658 looked up. The behaviour when one of the lookups defers can be
2659 controlled by a keyword.
2661 (4) It is now possible to specify the separator character for use when
2662 multiple records are returned.
2664 38. The dnslists ACL condition has been extended: it is now possible to supply
2665 a list of IP addresses and/or domains to be looked up in a particular DNS
2668 39. Added log_selector=+queue_time_overall.
2670 40. When running the queue in the test harness, wait just a tad after forking a
2671 delivery process, to get repeatability of debugging output.
2673 41. Include certificate and key file names in error message when GnuTLS fails
2674 to set them up, because the GnuTLS error message doesn't include the name
2675 of the failing file when there is a problem reading it.
2677 42. Allow both -bf and -bF in the same test run.
2679 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2681 44. The "Exiscan patch" is now merged into the mainline Exim source.
2683 45. Sometimes the final signoff response after QUIT could fail to get
2684 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2685 < 0 before doing a fflush(). This bug looks as though it goes back to the
2686 introduction of TLS in release 3.20, but "sometimes" must have been rare
2687 because the tests only now provoked it.
2689 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2690 (this can affect the format of dates).
2692 47. exim_tidydb, when checking for the continued existence of a message for
2693 which it has found a message-specific retry record, was not finding
2694 messages that were in split spool directories. Consequently, it was
2695 deleting retry records that should have stayed in existence.
2697 48. Steve fixed some bugs in eximstats.
2699 49. The SPA authentication driver was not abandoning authentication and moving
2700 on to the next authenticator when an expansion was forced to fail,
2701 contradicting the general specification for all authenticators. Instead it
2702 was generating a temporary error. It now behaves as specified.
2704 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2705 (the order specifies the preference for clients). The order is now AES256,
2706 AES128, 3DES, ARCFOUR128.
2708 51. Small patch to Sieve code - explicitly set From: when generating an
2711 52. Exim crashed if a remote delivery caused a very long error message to be
2712 recorded - for instance if somebody sent an entire SpamAssassin report back
2713 as a large number of 550 error lines. This bug was coincidentally fixed by
2714 increasing the size of one of Exim's internal buffers (big_buffer) that
2715 happened as part of the Exiscan merge. However, to be on the safe side, I
2716 have made the code more robust (and fixed the comments that describe what
2719 53. Now that there can be additional text after "Completed" in log lines (if
2720 the queue_time_overall log selector is set), a one-byte patch to exigrep
2721 was needed to allow it to recognize "Completed" as not the last thing in
2724 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2725 patch that reportedly fixes this has been added. I am not expert enough to
2726 create a test for it. This is what the patch creator wrote:
2728 "I found a little strange behaviour of ldap code when working with
2729 Windows 2003 AD Domain, where users was placed in more than one
2730 Organization Units. When I tried to give exim partial DN, the exit code
2731 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2732 But simultaneously result of request was absolutely normal ldap result,
2733 so I produce this patch..."
2735 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2736 REFERENCE, so I have modified the code to exclude the patch when that macro
2739 55. Some experimental protocols are using DNS PTR records for new purposes. The
2740 keys for these records are domain names, not reversed IP addresses. The
2741 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2742 leaves it alone. Component reversal etc. now happens only for IP addresses.
2745 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2747 57. Double the size of the debug message buffer (to 2048) so that more of very
2748 long debug lines gets shown.
2750 58. The exicyclog utility now does better if the number of log files to keep
2751 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2753 59. Two changes related to the smtp_active_hostname option:
2755 (1) $smtp_active_hostname is now available as a variable.
2756 (2) The default for smtp_banner uses $smtp_active_hostname instead
2757 of $primary_hostname.
2759 60. The host_aton() function is supposed to be passed a string that is known
2760 to be a valid IP address. However, in the case of IPv6 addresses, it was
2761 not checking this. This is a hostage to fortune. Exim now panics and dies
2762 if the condition is not met. A case was found where this could be provoked
2763 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2764 components; fortuitously, this particular loophole had already been fixed
2765 by change 4.50/55 above.
2767 If there are any other similar loopholes, the new check in host_aton()
2768 itself should stop them being exploited. The report I received stated that
2769 data on the command line could provoke the exploit when Exim was running as
2770 exim, but did not say which command line option was involved. All I could
2771 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2772 running as the user.
2775 61. There was a buffer overflow vulnerability in the SPA authentication code
2776 (which came originally from the Samba project). I have added a test to the
2777 spa_base64_to_bits() function which I hope fixes it.
2780 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2781 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2783 63. The daemon start-up calls getloadavg() while still root for those OS that
2784 need the first call to be done as root, but it missed one case: when
2785 deliver_queue_load_max is set with deliver_drop_privilege. This is
2786 necessary for the benefit of the queue runner, because there is no re-exec
2787 when deliver_drop_privilege is set.
2789 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2790 This has been fixed.
2792 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2793 was in use, was not putting the data itself into the right store pool;
2794 consequently, it could be overwritten for a subsequent message in the same
2795 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2798 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2800 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2801 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2802 the function were treating the return as a boolean value, which happened to
2803 work because 0=false and not-0=true, but is not correct code.
2805 68. The host_aton() function was not handling scoped IPv6 addresses (those
2806 with, for example, "%eth0" on the end) correctly.
2808 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2809 negated items (that is, ~something) in unsigned ints. Some compilers
2810 apparently mutter when there is no cast.
2812 70. If an address verification called from an ACL failed, and did not produce a
2813 user-specific message (i.e. there was only a "system" message), nothing was
2814 put in $acl_verify_message. In this situation, it now puts the system
2817 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2818 session; change 4.31/43 added the unwanted input to the log line - except
2819 that it did not do this in the start of session case. It now does.
2821 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2822 This is wrong and can cause the other end to generate a synchronization
2823 error if it is another Exim or anything else that does the synchronization
2824 check. A QUIT command is no longer sent after a timeout.
2826 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2827 during host lookups.
2829 74. The defer_ok option of callout verification was not working if it was used
2830 when verifying addresses in header lines, that is, for this case:
2832 verify = header_sender/callout=defer_ok
2834 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2835 those file descriptors could be used for SMTP connections. If anything
2836 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2837 could be sent to the SMTP client, causing chaos. The daemon now opens
2838 stdin, stdout, and stderr to /dev/null when it puts itself into the
2841 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2842 log by default. The user can override this with suitable Perl magic.
2844 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2845 the log message when discard triggers, was not working for the DATA ACL or
2846 for the non-SMTP ACL.
2848 78. Error message wording change in sieve.c.
2850 79. If smtp_accept_max_per_host was set, the number of connections could be
2851 restricted to fewer than expected, because the daemon was trying to set up
2852 a new connection before checking whether the processes handling previous
2853 connections had finished. The check for completed processes is now done
2854 earlier. On busy systems, this bug wouldn't be noticed because something
2855 else would have woken the daemon, and it would have reaped the completed
2858 80. If a message was submitted locally by a user whose login name contained one
2859 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2860 It caused a spool format error. I have fixed the spool reading code. A
2861 related problem was that the "from" clause in the Received: line became
2862 illegal because of the space(s). It is now covered by ${quote_local_part.
2864 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2867 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2869 83. Patch from Sieve maintainer to fix unterminated string problem in
2870 "vacation" handling.
2872 84. Some minor changes to the Linux configuration files to help with other
2873 OS variants using glibc.
2875 85. One more patch for Sieve to update vacation handling to latest spec.
2878 ----------------------------------------------------
2879 See the note above about the 4.44 and 4.50 releases.
2880 ----------------------------------------------------
2886 1. Change 4.43/35 introduced a bug that caused file counts to be
2887 incorrectly computed when quota_filecount was set in an appendfile
2890 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2891 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2894 3. Give more explanation in the error message when the command for a transport
2895 filter fails to execute.
2897 4. There are several places where Exim runs a non-Exim command in a
2898 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2899 was being done only for the command run by the queryprogram router. It is
2900 now done for all such subprocesses. The other cases are: ${run, transport
2901 filters, and the commands run by the lmtp and pipe transports.
2903 5. Some older OS have a limit of 256 on the maximum number of file
2904 descriptors. Exim was using setrlimit() to set 1000 as a large value
2905 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2906 systems. I've change it so that if it can't get 1000, it tries for 256.
2908 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2909 was an oversight, and furthermore, ever since the addition of extra
2910 controls (e.g. 4.43/32), the checks on when to allow different forms of
2911 "control" were broken. There should now be diagnostics for all cases when a
2912 control that does not make sense is encountered.
2914 7. $recipients is now available in the predata ACL (oversight).
2916 8. Tidy the search cache before the fork to do a delivery from a message
2917 received from the command line. Otherwise the child will trigger a lookup
2918 failure and thereby defer the delivery if it tries to use (for example) a
2919 cached ldap connection that the parent has called unbind on.
2921 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2922 of $address_data from the recipient verification was clobbered by the
2923 sender verification.
2925 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2926 was its contents. (It was OK if the option was not defined at all.)
2928 11. A "Completed" log line is now written for messages that are removed from
2929 the spool by the -Mrm option.
2931 12. $host_address is now set to the target address during the checking of
2932 ignore_target_hosts.
2934 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2935 being passed; this would have caused $sender_host_name to have been used if
2936 matching the list had actually called for a host name (not very likely,
2937 since this list is usually IP addresses). A host name is now passed as
2940 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2941 code that specifies a non-restarting handler (typically sigaction() in
2942 modern systems) in an attempt to fix a rare and obscure crash bug.
2944 15. Narrowed the window for a race in the daemon that could cause it to ignore
2945 SIGCHLD signals. This is not a major problem, because they are used only to
2946 wake it up if nothing else does.
2948 16. A malformed maildirsize file could cause Exim to calculate negative values
2949 for the mailbox size or file count. Odd effects could occur as a result.
2950 The maildirsize information is now recalculated if the size or filecount
2953 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2954 support for a long time. Removed HAVE_SYS_VFS_H.
2956 18. Updated exipick to current release from John Jetmore.
2958 19. Allow an empty sender to be matched against a lookup in an address list.
2959 Previously the only cases considered were a regular expression, or an
2962 20. Exim went into a mad DNS lookup loop when doing a callout where the
2963 host was specified on the transport, if the DNS lookup yielded more than
2966 21. The RFC2047 encoding function was originally intended for short strings
2967 such as real names; it was not keeping to the 75-character limit for
2968 encoded words that the RFC imposes. It now respects the limit, and
2969 generates multiple encoded words if necessary. To be on the safe side, I
2970 have increased the buffer size for the ${rfc2047: expansion operator from
2973 22. Failure to deliver a bounce message always caused it to be frozen, even if
2974 there was an errors_to setting on the router. The errors_to setting is now
2977 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2978 canonical form (fully expanded) before being placed in
2979 $sender_host_address.
2981 24. Updated eximstats to version 1.33
2983 25. Include certificate and key file names in error message when GnuTLS fails
2984 to set them up, because the GnuTLS error message doesn't include the name
2985 of the failing file when there is a problem reading it.
2987 26. Expand error message when OpenSSL has problems setting up cert/key files.
2990 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2991 (this can affect the format of dates).
2993 28. exim_tidydb, when checking for the continued existence of a message for
2994 which it has found a message-specific retry record, was not finding
2995 messages that were in split spool directories. Consequently, it was
2996 deleting retry records that should have stayed in existence.
2998 29. eximstats updated to version 1.35
2999 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3000 1.35 - bugfix such that pie charts by volume are generated correctly
3002 30. The SPA authentication driver was not abandoning authentication and moving
3003 on to the next authenticator when an expansion was forced to fail,
3004 contradicting the general specification for all authenticators. Instead it
3005 was generating a temporary error. It now behaves as specified.
3007 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3008 (the order specifies the preference for clients). The order is now AES256,
3009 AES128, 3DES, ARCFOUR128.
3011 31. Small patch to Sieve code - explicitly set From: when generating an
3014 32. Exim crashed if a remote delivery caused a very long error message to be
3015 recorded - for instance if somebody sent an entire SpamAssassin report back
3016 as a large number of 550 error lines. This bug was coincidentally fixed by
3017 increasing the size of one of Exim's internal buffers (big_buffer) that
3018 happened as part of the Exiscan merge. However, to be on the safe side, I
3019 have made the code more robust (and fixed the comments that describe what
3022 33. Some experimental protocols are using DNS PTR records for new purposes. The
3023 keys for these records are domain names, not reversed IP addresses. The
3024 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3025 leaves it alone. Component reversal etc. now happens only for IP addresses.
3028 34. The host_aton() function is supposed to be passed a string that is known
3029 to be a valid IP address. However, in the case of IPv6 addresses, it was
3030 not checking this. This is a hostage to fortune. Exim now panics and dies
3031 if the condition is not met. A case was found where this could be provoked
3032 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3033 components; fortuitously, this particular loophole had already been fixed
3034 by change 4.50/55 or 4.44/33 above.
3036 If there are any other similar loopholes, the new check in host_aton()
3037 itself should stop them being exploited. The report I received stated that
3038 data on the command line could provoke the exploit when Exim was running as
3039 exim, but did not say which command line option was involved. All I could
3040 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3041 running as the user.
3044 35. There was a buffer overflow vulnerability in the SPA authentication code
3045 (which came originally from the Samba project). I have added a test to the
3046 spa_base64_to_bits() function which I hope fixes it.
3049 36. The daemon start-up calls getloadavg() while still root for those OS that
3050 need the first call to be done as root, but it missed one case: when
3051 deliver_queue_load_max is set with deliver_drop_privilege. This is
3052 necessary for the benefit of the queue runner, because there is no re-exec
3053 when deliver_drop_privilege is set.
3055 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3056 was in use, was not putting the data itself into the right store pool;
3057 consequently, it could be overwritten for a subsequent message in the same
3058 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3061 38. Sometimes the final signoff response after QUIT could fail to get
3062 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3063 < 0 before doing a fflush(). This bug looks as though it goes back to the
3064 introduction of TLS in release 3.20, but "sometimes" must have been rare
3065 because the tests only now provoked it.
3071 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3072 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3073 Now it returns an integer. A number of calls were still expecting a T/F
3074 return. Fortuitously, in all cases, the tests worked in OK situations,
3075 which is the norm. However, things would have gone wrong on any write
3076 failures on the smtp file descriptor. This function is used when sending
3077 messages over SMTP and also when doing verify callouts.
3079 2. When Exim is called to do synchronous delivery of a locally submitted
3080 message (the -odf or -odi options), it no longer closes stderr before doing
3083 3. Implemented the mua_wrapper option.
3085 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3087 5. Implemented the functions header_remove(), header_testname(),
3088 header_add_at_position(), and receive_remove_recipient(), and exported them
3091 6. If an ACL "warn" statement specified the addition of headers, Exim already
3092 inserted X-ACL-Warn: at the start if there was no header name. However, it
3093 was not making this test for the second and subsequent header lines if
3094 there were newlines in the string. This meant that an invalid header could
3095 be inserted if Exim was badly configured.
3097 7. Allow an ACL "warn" statement to add header lines at the start or after all
3098 the Received: headers, as well as at the end.
3100 8. Added the rcpt_4xx retry error code.
3102 9. Added postmaster_mailfrom=xxx to callout verification option.
3104 10. Added mailfrom=xxxx to the callout verification option, for verify=
3107 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3108 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3110 12. Inserted some casts to stop certain compilers warning when using pointer
3111 differences as field lengths or precisions in printf-type calls (mostly
3112 affecting debugging statements).
3114 13. Added optional readline() support for -be (dynamically loaded).
3116 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3117 same clock tick as a message's arrival, so that its received time was the
3118 same as the "first fail" time on the retry record, and that message
3119 remained on the queue past the ultimate address timeout, every queue runner
3120 would try a delivery (because it was past the ultimate address timeout) but
3121 after another failure, the ultimate address timeout, which should have then
3122 bounced the address, did not kick in. This was a "< instead of <=" error;
3123 in most cases the first failure would have been in the next clock tick
3124 after the received time, and all would be well.
3126 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3127 being recognized when the domain list was tested by the match_domain
3128 condition in an expansion string.
3130 16. Added the ${str2b64: operator.
3132 17. Exim was always calling setrlimit() to set a large limit for the number of
3133 processes, without checking whether the existing limit was already
3134 adequate. (It did check for the limit on file descriptors.) Furthermore,
3135 errors from getrlimit() and setrlimit() were being ignored. Now they are
3136 logged to the main and panic logs, but Exim does carry on, to try to do its
3137 job under whatever limits there are.
3139 18. Imported PCRE 5.0.
3141 19. Trivial typo in log message " temporarily refused connection" (the leading
3144 20. If the log selector return_path_on_delivery was set and an address was
3145 redirected to /dev/null, the delivery process crashed because it assumed
3146 that a return path would always be set for a "successful" delivery. In this
3147 case, the whole delivery is bypassed as an optimization, and therefore no
3150 21. Internal re-arrangement: the function for sending a challenge and reading
3151 a response while authentication was assuming a zero-terminated challenge
3152 string. It's now changed to take a pointer and a length, to allow for
3153 binary data in such strings.
3155 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3157 23. Exim was not respecting finduser_retries when seeking the login of the
3158 uid under which it was called; it was always trying 10 times. (The default
3159 setting of finduser_retries is zero.) Also, it was sleeping after the final
3160 failure, which is pointless.
3162 24. Implemented tls_on_connect_ports.
3164 25. Implemented acl_smtp_predata.
3166 26. If the domain in control=submission is set empty, Exim assumes that the
3167 authenticated id is a complete email address when it generates From: or
3168 Sender: header lines.
3170 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3171 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3172 chown and chgrp in /bin and hostname in /usr/bin.
3174 28. Exim was keeping the "process log" file open after each use, just as it
3175 does for the main log. This opens the possibility of it remaining open for
3176 long periods when the USR1 signal hits a daemon. Occasional processlog
3177 errors were reported, that could have been caused by this. Anyway, it seems
3178 much more sensible not to leave this file open at all, so that is what now
3181 29. The long-running daemon process does not normally write to the log once it
3182 has entered its main loop, and it closes the log before doing so. This is
3183 so that log files can straightforwardly be renamed and moved. However,
3184 there are a couple of unusual error situations where the daemon does write
3185 log entries, and I had neglected to close the log afterwards.
3187 30. The text of an SMTP error response that was received during a remote
3188 delivery was being truncated at 512 bytes. This is too short for some of
3189 the long messages that one sometimes sees. I've increased the limit to
3192 31. It is now possible to make retry rules that apply only when a message has a
3193 specific sender, in particular, an empty sender.
3195 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3196 it possible to be selective about when SMTP synchronization is enforced.
3198 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3200 32. Implemented hosts_connection_nolog.
3202 33. Added an ACL for QUIT.
3204 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3207 35. Added mailbox_size and mailbox_filecount to appendfile.
3209 36. Added control = no_multiline_responses to ACLs.
3211 37. There was a bug in the logic of the code that waits for the clock to tick
3212 in the case where the clock went backwards by a substantial amount such
3213 that the microsecond fraction of "now" was more than the microsecond
3214 fraction of "then" (but the whole seconds number was less).
3216 38. Added support for the libradius Radius client library this is found on
3217 FreeBSD (previously only the radiusclient library was supported).
3223 1. When certain lookups returned multiple values in the form name=value, the
3224 quoting of the values was not always being done properly. Specifically:
3225 (a) If the value started with a double quote, but contained no whitespace,
3227 (b) If the value contained whitespace other than a space character (i.e.
3228 tabs or newlines or carriage returns) it was not quoted.
3229 This fix has been applied to the mysql and pgsql lookups by writing a
3230 separate quoting function and calling it from the lookup code. The fix
3231 should probably also be applied to nisplus, ibase and oracle lookups, but
3232 since I cannot test any of those, I have not disturbed their existing code.
3234 2. A hit in the callout cache for a specific address caused a log line with no
3235 reason for rejecting RCPT. Now it says "Previous (cached) callout
3236 verification failure".
3238 3. There was an off-by-one bug in the queryprogram router. An over-long
3239 return line was truncated at 256 instead of 255 characters, thereby
3240 overflowing its buffer with the terminating zero. As well as fixing this, I
3241 have increased the buffer size to 1024 (and made a note to document this).
3243 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3244 when Exim is waiting for an SMTP response from a remote server, Exim
3245 restarts its select() call on the socket, thereby resetting its timeout.
3246 This is not a problem when such interrupts are rare. Somebody set up a cron
3247 job to run exiwhat every 2 minutes, which is less than the normal select()
3248 timeout (5 or 10 minutes). This meant that the select() timeout never
3249 kicked in because it was always reset. I have fixed this by comparing the
3250 time when an interrupt arrives with the time at the start of the first call
3251 to select(). If more time than the timeout has elapsed, the interrupt is
3252 treated as a timeout.
3254 5. Some internal re-factoring in preparation for the addition of Sieve
3255 extensions (by MH). In particular, the "personal" test is moved to a
3256 separate function, and given an option for scanning Cc: and Bcc: (which is
3257 not set for Exim filters).
3259 6. When Exim created an email address using the login of the caller as the
3260 local part (e.g. when creating a From: or Sender: header line), it was not
3261 quoting the local part when it contained special characters such as @.
3263 7. Installed new OpenBSD configuration files.
3265 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3266 try to make them clearer.
3268 9. Callout options, other than the timeout value, were being ignored when
3269 verifying sender addresses in header lines. For example, when using
3271 verify = header_sender/callout=no_cache
3273 the cache was (incorrectly) being used.
3275 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3276 only the Cygwin environment.
3278 11. When return_path_on_delivery was set as a log selector, if different remote
3279 addresses in the same message used different return paths and parallel
3280 remote delivery occurred, the wrong values would sometimes be logged.
3281 (Whenever a remote delivery process finished, the return path value from
3282 the most recently started remote delivery process was logged.)
3284 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3285 lines when AUTH and/or TLS are in use. This is the "received protocol"
3286 field. Exim used to use "asmtp" for authenticated SMTP, without any
3287 indication (in the protocol name) for TLS use. Now it follows the RFC and
3288 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3289 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3290 names appear in log lines as well as in Received: header lines.
3292 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3293 extensions, and comparison tests, and to fix some bugs.
3295 14. Changes to the "personal" filter test:
3297 (1) The test was buggy in that it was just doing the equivalent of
3298 "contains" tests on header lines. For example, if a user's address was
3299 anne@some.where, the "personal" test would incorrectly be true for
3301 To: susanne@some.where
3303 This test is now done by extracting each address from the header in turn,
3304 and checking the entire address. Other tests that are part of "personal"
3305 are now done using regular expressions (for example, to check local parts
3306 of addresses in From: header lines).
3308 (2) The list of non-personal local parts in From: addresses has been
3309 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3310 taken from the Sieve specification recommendations.
3312 (3) If the message contains any header line starting with "List-" it is
3313 treated as non-personal.
3315 (4) The test for "circular" in the Subject: header line has been removed
3316 because it now seems ill-conceived.
3318 15. Minor typos in src/EDITME comments corrected.
3320 16. Installed latest exipick from John Jetmore.
3322 17. If headers_add on a router specified a text string that was too long for
3323 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3324 of string_sprintf() is now avoided.
3326 18. $message_body_size was not set (it was always zero) when running the DATA
3327 ACL and the local_scan() function.
3329 19. For the "mail" command in an Exim filter, no default was being set for
3330 the once_repeat time, causing a random time value to be used if "once" was
3331 specified. (If the value happened to be <= 0, no repeat happened.) The
3332 default is now 0s, meaning "never repeat". The "vacation" command was OK
3333 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3334 (I found it when inspecting the code).
3336 20. There is now an overall timeout for performing a callout verification. It
3337 defaults to 4 times the callout timeout, which applies to individual SMTP
3338 commands during the callout. The overall timeout applies when there is more
3339 than one host that can be tried. The timeout is checked before trying the
3340 next host. This prevents very long delays if there are a large number of
3341 hosts and all are timing out (e.g. when the network connections are timing
3342 out). The value of the overall timeout can be changed by specifying an
3343 additional sub-option for "callout", called "maxwait". For example:
3345 verify = sender/callout=5s,maxwait=20s
3347 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3348 to the end before writing, but this should make it even safer).
3350 22. Exim was forgetting that it had advertised PIPELINING for the second and
3351 subsequent messages on an SMTP connection. It was also not resetting its
3352 memory on STARTTLS and an internal HELO.
3354 23. When Exim logs an SMTP synchronization error within a session, it now
3355 records whether PIPELINING has been advertised or not.
3357 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3358 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3359 rather than long int.
3361 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3363 26. Added the never_mail option to autoreply.
3369 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3370 crash if the getsockname() call failed; this can happen if a connection is
3371 closed very soon after it is established. The problem was simply in the
3372 order in which certain operations were done, causing Exim to try to write
3373 to the SMTP stream before it had set up the file descriptor. The bug has
3374 been fixed by making things happen in the correct order.
3380 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3381 before the connection was closed, thus losing the rejection response.
3383 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3384 some early Solaris releases, but causes trouble in current releases where
3385 socklen_t is defined.
3387 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3390 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3393 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3394 that releasing the top of it at the end releases what was used for sub-
3395 expansions (unless the block got too big). However, discard this block if
3396 the first thing is a variable or header, so that we can use its block when
3397 it is dynamic (useful for very large $message_headers, for example).
3399 6. Lookups now cache *every* query, not just the most recent. A new, separate
3400 store pool is used for this. It can be recovered when all lookup caches are
3401 flushed. Lookups now release memory at the end of their result strings.
3402 This has involved some general refactoring of the lookup sources.
3404 7. Some code has been added to the store_xxx() functions to reduce the amount
3405 of flapping under certain conditions.
3407 8. log_incoming_interface used to affect only the <= reception log lines. Now
3408 it causes the local interface and port to be added to several more SMTP log
3409 lines, for example "SMTP connection from", and rejection lines.
3411 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3413 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3415 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3416 could be overwritten at the end of the current message (or the start of a
3417 new message if it was set in a HELO ACL). The value is now preserved for
3418 the duration of the SMTP connection.
3420 12. If a transport had a headers_rewrite setting, and a matching header line
3421 contained an unqualified address, that address was qualified, even if it
3422 did not match any rewriting rules. The underlying bug was that the values
3423 of the flags that permit the existence of unqualified sender and recipient
3424 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3425 non-local messages, and by -bnq for local messages) were not being
3426 preserved with the message after it was received.
3428 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3429 "next input=" as part of the text comprising the host identity instead of
3430 the correct text. The code was using the same buffer for two different
3431 strings. However, depending on which order the printing function evaluated
3432 its arguments, the bug did not always show up. Under Linux, for example, my
3433 test suite worked just fine.
3435 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3436 doesn't work with some older versions of Perl. It has been changed to "my",
3437 which in any case is probably the better facility to use.
3439 15. A really picky compiler found some instances of statements for creating
3440 error messages that either had too many or two few arguments for the format
3443 16. The size of the buffer for calls to the DNS resolver has been increased
3444 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3445 for addresses that have a lot of PTR records. This alleviates a problem; it
3446 does not fully solve it.
3448 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3449 buffer now truncates the list and logs the incident, which is the same
3450 action as happens when Exim is looking up a host name and its aliases.
3451 Previously in this situation something unpredictable would happen;
3452 sometimes it was "internal error: store_reset failed".
3454 18. If a server dropped the connection unexpectedly when an Exim client was
3455 using GnuTLS and trying to read a response, the client delivery process
3456 crashed while trying to generate an error log message.
3458 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3459 string, for example:
3461 warn message = H1: something\nH2: something
3463 the text was added as a single header line from Exim's point of view
3464 though it ended up OK in the delivered message. However, searching for the
3465 second and subsequent header lines using $h_h2: did not work. This has been
3466 fixed. Similarly, if a system filter added multiple headers in this way,
3467 the routers could not see them.
3469 20. Expanded the error message when iplsearch is called with an invalid key to
3470 suggest using net-iplsearch in a host list.
3472 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3473 ACLs are no longer actually imposed (and a message to that effect is
3476 22. If a "gecos" field in a passwd entry contained escaped characters, in
3477 particular, if it contained a \" sequence, Exim got it wrong when building
3478 a From: or a Sender: header from that name. A second bug also caused
3479 incorrect handling when an unquoted " was present following a character
3480 that needed quoting.
3482 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3483 was not being matched caselessly.
3485 24. Arranged for all hyphens in the exim.8 source to be escaped with
3488 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3489 properly. Recipient callout cache records were still being keyed to include
3490 the sender, even when use_sender was set false. This led to far more
3491 callouts that were necessary. The sender is no longer included in the key
3492 when use_sender is false.
3494 26. Added "control = submission" modifier to ACLs.
3496 27. Added the ${base62d: operator to decode base 62 numbers.
3498 28. dnsdb lookups can now access SRV records.
3500 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3501 the configuration file.
3503 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3504 (-v) mode. This makes the output for a verbose queue run more intelligible.
3506 31. Added a use_postmaster feature to recipient callouts.
3508 32. Added the $body_zerocount variable, containing the number of binary zero
3509 bytes in the message body.
3511 33. The time of last modification of the "new" subdirectory is now used as the
3512 "mailbox time last read" when there is a quota error for a maildir
3515 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3517 35. Added +ignore_unknown as a special item in host lists.
3519 36. Code for decoding IPv6 addresses in host lists is now included, even if
3520 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3521 address was recognized as an IP address, but was then not correctly decoded
3522 into binary, causing unexpected and incorrect effects when compared with
3529 1. Very minor rewording of debugging text in manualroute to say "list of
3530 hosts" instead of "hostlist".
3532 2. If verify=header_syntax was set, and a header line with an unqualified
3533 address (no domain) and a large number of spaces between the end of the
3534 name and the colon was received, the reception process suffered a buffer
3535 overflow, and (when I tested it) crashed. This was caused by some obsolete
3536 code that should have been removed. The fix is to remove it!
3538 3. When running in the test harness, delay a bit after writing a bounce
3539 message to get a bit more predictability in the log output.
3541 4. Added a call to search_tidyup() just before forking a reception process. In
3542 theory, someone could use a lookup in the expansion of smtp_accept_max_
3543 per_host which, without the tidyup, could leave open a database connection.
3545 5. Added the variables $recipient_data and $sender_data which get set from a
3546 lookup success in an ACL "recipients" or "senders" condition, or a router
3547 "senders" option, similar to $domain_data and $local_part_data.
3549 6. Moved the writing of debug_print from before to after the "senders" test
3552 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3553 problems for message scanning, either using a data ACL, or using
3554 local_scan() because the Received: header was not generated till after they
3555 were called (in order to set the time as the time of reception completion).
3556 I have revised the way this works. The header is now generated after the
3557 body is received, but before the ACL or local_scan() are called. After they
3558 are run, the timestamp in the header is updated.
3564 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3565 before starting a queue runner without re-exec. This happened only when
3566 deliver_drop_privilege was set or when the Exim user was set to root. The
3567 effect of the bug was that timeouts during subsequent deliveries caused
3568 crashes instead of being properly handled. The handler is now left at its
3569 default (and expected) setting.
3571 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3572 message, again when deliver_drop_privilege is set or Exim is run as root.
3573 The bug described in (1) was not present in this case, but the tidying up
3574 of the other signals was missing. I have made the two cases consistent.
3576 3. The ignore_target_hosts setting on a manualroute router was being ignored
3577 for hosts that were looked up using the /MX notation.
3579 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3582 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3583 operated on the sender address. After changing the $sender_address to <>
3584 for the sender address verify, Exim was re-instated it as the original
3585 (before rewriting) address, but remembering that it had rewritten it, so it
3586 wasn't rewriting it again. This bug also had the effect of breaking the
3587 sender address verification caching when the sender address was rewritten.
3589 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3590 This has been changed so that if the ip literal address matches
3591 ignore_target_hosts, the router declines.
3593 7. Added expansion conditions match_domain, match_address, and match_local_
3594 part (NOT match_host).
3596 8. The placeholder for the Received: header didn't have a length field set.
3598 9. Added code to Exim itself and to exim_lock to test for a specific race
3599 condition that could lead to file corruption when using MBX delivery. The
3600 issue is with the lockfile that is created in /tmp. If this file is removed
3601 after a process has opened it but before that process has acquired a lock,
3602 there is the potential for a second process to recreate the file and also
3603 acquire a lock. This could lead to two Exim processes writing to the file
3604 at the same time. The added code performs the same test as UW imapd; it
3605 checks after acquiring the lock that its file descriptor still refers to
3606 the same named file.
3608 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3609 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3612 11. Added the smtp_active_hostname option. If used, this will typically be made
3613 to depend on the incoming interface address. Because $interface_address is
3614 not set up until the daemon has forked a reception process, error responses
3615 that can happen earlier (such as "too many connections") no longer contain
3618 12. If an expansion in a condition on a "warn" statement fails because a lookup
3619 defers, the "warn" statement is abandoned, and the next ACL statement is
3620 processed. Previously this caused the whole ACL to be aborted.
3622 13. Added the iplsearch lookup type.
3624 14. Added ident_timeout as a log selector.
3626 15. Added tls_certificate_verified as a log selector.
3628 16. Added a global option tls_require_ciphers (compare the smtp transport
3629 option of the same name). This controls incoming TLS connections.
3631 17. I finally figured out how to make tls_require_ciphers do a similar thing
3632 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3633 before starting the TLS session.
3635 18. Tabs are now shown as \t in -bP output.
3637 19. If the log selector return_path_on_delivery was set, Exim crashed when
3638 bouncing a message because it had too many Received: header lines.
3640 20. If two routers both had headers_remove settings, and the first one included
3641 a superfluous trailing colon, the final name in the first list and the
3642 first name in the second list were incorrectly joined into one item (with a
3643 colon in the middle).
3649 1. Added -C and -D options to the exinext utility, mainly to make it easier
3650 to include in the automated testing, but these could be helpful when
3651 multiple configurations are in use.
3653 2. The exinext utility was not formatting the output nicely when there was
3654 an alternate port involved in the retry record key, nor when there was a
3655 message id as well (for retries that were specific to a specific message
3656 and a specific host). It was also confused by IPv6 addresses, because of
3657 the additional colons they contain. I have fixed the IPv4 problem, and
3658 patched it up to do a reasonable job for IPv6.
3660 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3661 delivery, the log line now contains "pipelined" if PIPELINING was used.
3663 4. An SMTP transport process used to panic and die if the bind() call to set
3664 an explicit outgoing interface failed. This has been changed; it is now
3665 treated in the same way as a connect() failure.
3667 5. A reference to $sender_host_name in the part of a conditional expansion
3668 that was being skipped was still causing a DNS lookup. This no longer
3671 6. The def: expansion condition was not recognizing references to header lines
3672 that used bh_ and bheader_.
3674 7. Added the _cache feature to named lists.
3676 8. The code for checking quota_filecount in the appendfile transport was
3677 allowing one more file than it should have been.
3679 9. For compatibility with Sendmail, the command line option
3687 and sets the incoming protocol and host name (for trusted callers). The
3688 host name and its colon can be omitted when only the protocol is to be set.
3689 Note the Exim already has two private options, -pd and -ps, that refer to
3690 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3691 "s", but I don't think that's a major issue.
3693 10. A number of refactoring changes to the code, none of which should affect
3696 (a) The number of logging options was getting close to filling up the
3697 32-bit word that was used as a bit map. I have split them into two classes:
3698 those that are passed in the argument to log_write(), and those that are
3699 only ever tested independently outside of that function. These are now in
3700 separate 32-bit words, so there is plenty of room for expansion again.
3701 There is no change in the user interface or the logging behaviour.
3703 (b) When building, for example, log lines, the code previously used a
3704 macro that called string_cat() twice, in order to add two strings. This is
3705 not really sufficiently general. Furthermore, there was one instance where
3706 it was actually wrong because one of the argument was used twice, and in
3707 one call a function was used. (As it happened, calling the function twice
3708 did not affect the overall behaviour.) The macro has been replaced by a
3709 function that can join an arbitrary number of extra strings onto a growing
3712 (c) The code for expansion conditions now uses a table and a binary chop
3713 instead of a serial search (which was left over from when there were very
3714 few conditions). Also, it now recognizes conditions like "pam" even when
3715 the relevant support is not compiled in: a suitably worded error message is
3716 given if an attempt is made to use such a condition.
3718 11. Added ${time_interval:xxxxx}.
3720 12. A bug was causing one of the ddress fields not to be passed back correctly
3721 from remote delivery subprocesses. The field in question was not being
3722 subsequently used, so this caused to problems in practice.
3724 13. Added new log selectors queue_time and deliver_time.
3726 14. Might have fixed a bug in maildirsizefile handling that threw up
3727 "unexpected character" debug warnings, and recalculated the data
3728 unnecessarily. In any case, I expanded the warning message to give more
3731 15. Added the message "Restricted characters in address" to the statements in
3732 the default ACL that block characters like @ and % in local parts.
3734 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3735 Three changes have been made:
3737 (a) There was a serious bug; a negative response to MAIL caused the whole
3738 recipient domain to be cached as invalid, thereby blocking all messages
3739 to all local parts at the same domain, from all senders. This bug has
3740 been fixed. The domain is no longer cached after a negative response to
3741 MAIL if the sender used is not empty.
3743 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3746 (c) A new callout option, "use_sender" has been added for people who want
3747 the modified behaviour.
3753 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3756 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3757 indeed breaks things for older releases.
3759 3. Added additional logging to the case where there is a problem reading data
3760 from a filter that is running in a subprocess using a pipe, in order to
3761 try to track down a specific problem.
3763 4. Testing facility fudge: when running in the test harness and attempting
3764 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3765 getting "No route to host". Convert this to a timeout.
3767 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3770 6. Some OS don't have socklen_t but use size_t instead. This affects the
3771 fifth argument of getsockopt() amongst other things. This is now
3772 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3773 can be set for individual OS. I have set it for SunOS5, OSF1, and
3774 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3775 some earlier ones do not.
3777 7. Change 4.30/15 was not doing the test caselessly.
3779 8. The standard form for an IPv6 address literal was being rejected by address
3780 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3781 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3782 this, as well as the form without the "IPv6" on the front (but only when
3783 address literals are enabled, of course).
3785 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3787 10. Exim crashed if a message with an empty sender address specified by -f
3788 encountered a router with an errors_to setting. This could be provoked only
3789 by a command such as
3793 where an empty string was supplied; "<>" did not hit this bug.
3795 11. Installed PCRE release 4.5.
3797 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3798 remained set. It is now erased.
3800 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3801 times from message ids (which are base 36 rather than the normal 62).
3803 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3804 were being counted as actual protocol errors, and logged if the log
3805 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3806 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3807 and DATA following a set of rejected RCPTs do not count as protocol errors.
3808 In other words, Exim assumes they were pipelined, though this may not
3809 actually be the case. Of course, in all cases the client gets an
3810 appropriate error code.
3812 15. If a lookup fails in an ACL condition, a message about the failure may
3813 be available; it is used if testing the ACL cannot continue, because most
3814 such messages specify what the cause of the deferral is. However, some
3815 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3816 that caused an old message to be retained and used if a later statement
3817 caused a defer, replacing the real cause of the deferral.
3819 16. If an IP address had so many PTR records that the DNS lookup buffer
3820 was not large enough to hold them, Exim could crash while trying to process
3821 the truncated data. It now detects and logs this case.
3823 17. Further to 4.21/58, another change has been made: if (and only if) the
3824 first line of a message (the first header line) ends with CRLF, a bare LF
3825 in a subsequent header line has a space inserted after it, so as not to
3826 terminate the header.
3828 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3829 unnecessarily, used atoi() instead of strtol(), and didn't check the
3830 termination when getting file sizes from file names by regex.
3832 19. Completely re-implemented the support for maildirsize files, in the light
3833 of a number of problems with the previous contributed implementation
3834 (4.30/29). In particular:
3836 . If the quota is zero, the maildirsize file is maintained, but no quota is
3839 . If the maildir directory does not exist, it is created before any attempt
3840 to write a maildirsize file.
3842 . The quota value in the file is just a cache; if the quota is changed in
3843 the transport, the new value overrides.
3845 . A regular expression is available for excluding directories from the
3848 20. The autoreply transport checks the characters in options that define the
3849 message's headers; it allows continued headers, but it was checking with
3850 isspace() after an embedded newline instead of explicitly looking for a
3853 21. If all the "regular" hosts to which an address was routed had passed their
3854 expiry times, and had not reached their retry times, the address was
3855 bounced, even if fallback hosts were defined. Now Exim should go on to try
3858 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3859 equivalent code in the SMTP transport. Some hosts send humungous responses
3860 to HELO/EHLO, more than 1024 it seems.
3862 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3863 gives compiler warnings in some environments. I've now done it "properly",
3866 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3867 (because of problems with the built-in one) was declared to return uschar *
3868 instead of char *, causing compiler failure.
3870 25. Fixed a file descriptor leak when processing alias/forward files.
3872 26. Fixed a minor format string issue in dbfn.c.
3874 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3876 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3877 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3880 29. When a sender address is verified, it is cached, to save repeating the test
3881 when there is more than one recipient in a message. However, when the
3882 verification involves a callout, it is possible for different callout
3883 options to be set for different recipients. It is too complicated to keep
3884 track of this in the cache, so now Exim always runs a verification when a
3885 callout is required, relying on the callout cache for the optimization.
3886 The overhead is duplication of the address routing, but this should not be
3889 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3890 to be verified with callout=postmaster, and the main callout worked but the
3891 postmaster check failed, the verification correctly failed. However, if a
3892 subsequent RCPT command asked for sender verification *without* the
3893 postmaster check, incorrect caching caused this verification also to fail,
3896 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3897 it was not caching the DNS options (qualify_single, search_parents) that
3898 were used when the lookup failed. A subsequent lookup with different
3899 options therefore always gave the same answer, though there were cases
3900 where it should not have. (Example: a "domains = !$mx_any" option on a
3901 dnslookup router: the "domains" option is always processed without any
3902 widening, but the router might have qualify_single set.) Now Exim uses the
3903 cached value only when the same options are set.
3905 32. Added John Jetmore's "exipick" utility to the distribution.
3907 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3908 than a timeout (e.g. a certificate is required, and is not provided), an
3909 Exim server now closes the connection immediately. Previously it waited for
3910 the client to close - but if the client is SSL, it seems that they each
3911 wait for each other, leading to a delay before one of them times out.
3913 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3914 maintained 0.8.x compatibility because I don't think many are using it, and
3915 it is clearly obsolete.
3917 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3920 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3921 client certificate was expired. A simple patch fixes this, though I don't
3922 understand the full logic of why the verify callback is called multiple
3925 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3926 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3927 which causes problems with some clients (such as the Certicom SSL Plus
3928 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3929 disables the coutermeasure allowing Eudora to connect."
3931 38. Exim was not checking that a write() to a log file succeeded. This could
3932 lead to Bad Things if a log got too big, in particular if it hit a file
3933 size limit. Exim now panics and dies if it cannot write to a log file, just
3934 as it does if it cannot open a log file.
3936 39. Modified OS/Makefile-Linux so that it now contains
3938 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3940 The two -D definitions ensure that Exim is compiled with large file
3941 support, which makes it possible to handle log files that are bigger than
3944 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3945 instance) a domain was checked against a named list that involved a lookup,
3946 causing $domain_data to be set, then another domain was checked against the
3947 same list, then the first domain was re-checked, the value of $domain_data
3948 after the final check could be wrong. In particular, if the second check
3949 failed, it could be set empty. This bug probably also applied to
3952 41. The strip_trailing_dot option was not being applied to the address given
3953 with the -f command-line option.
3955 42. The code for reading a message's header from the spool was incrementing
3956 $received_count, but never initializing it. This meant that the value was
3957 incorrect (doubled) while delivering a message in the same process in which
3958 it was received. In the most common configuration of Exim, this never
3959 happens - a fresh exec is done - but it can happen when
3960 deliver_drop_privilege is set.
3962 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3963 it now includes up to 150 characters of the unexpected data in the log
3966 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3967 and building data strings. The size of both of these buffers was 10 000
3968 bytes - far larger than anybody would *ever* want, thought I. Needless to
3969 say, somebody hit the limit. I have increased the maximum line length to
3970 20 000 and the maximum data length of concatenated lines to 100 000. I have
3971 also fixed two bugs, because there was no checking on these buffers. Tsk,
3972 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3973 buffer is too small.
3975 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3976 lsearch lookups. Now it does.
3978 46. When parsing a route_list item in a manualroute router, a fixed-length
3979 buffer was used for the list of hosts. I made this 1024 bytes long,
3980 thinking that nobody would ever have a list of hosts that long. Wrong.
3981 Somebody had a whole pile of complicated expansion conditions, and the
3982 string was silently truncated, leading to an expansion error. It turns out
3983 that it is easier to change to an unlimited length (owing to other changes
3984 that have happened since this code was originally written) than to build
3985 structure for giving a limitation error. The length of the item that
3986 expands into the list of hosts is now unlimited.
3988 47. The lsearch lookup could not handle data where the length of text line was
3989 more than 4095 characters. Such lines were truncated, leading to shortened
3990 data being returned. It should now handle lines of any length.
3992 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3993 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3996 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3997 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3998 no longer happens in comment lines. A list of replacements is now placed
3999 at the head of all of the source files, except those whose only change is
4000 to replace PERL_COMMAND in the very first #! line.
4002 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4003 messages on the queue, with a bottom-up merge sort, using code contributed
4004 by Michael Haardt. This should make operations like -bp somewhat faster on
4005 large queues. It won't affect queue runners, except when queue_run_in_order
4008 51. Installed eximstats 1.31 in the distribution.
4010 52. Added support for SRV lookups to the dnslookup router.
4012 53. If an ACL referred to $message_body or $message_body_end, the value was not
4013 reset for any messages that followed in the same SMTP session.
4015 54. The store-handling optimization for building very long strings was not
4016 differentiating between the different store pools. I don't think this
4017 actually made any difference in practice, but I've tidied it.
4019 55. While running the routers to verify a sender address, $sender_address
4020 was still set to the sender address. This is wrong, because when routing to
4021 send a bounce to the sender, it would be empty. Therefore, I have changed
4022 it so that, while verifying a sender address, $sender_address is set to <>.
4023 (There is no change to what happens when verifying a recipient address.)
4025 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4026 target A or AAAA records (if not already returned) without resetting the
4027 qualify_single or search_parents options of the DNS resolver. These are
4028 inappropriate in this case because the targets of MX and SRV records must
4029 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4030 target that, when qualified, matched something in the local domain. These
4031 two options are now turned off when doing these lookups.
4033 57. It seems that at least some releases of Reiserfs (which does not have the
4034 concept of a fixed number of inodes) returns zero and not -1 for the
4035 number of available inodes. This interacted badly with check_spool_inodes,
4036 which assumed that -1 was the "no such thing" setting. What I have done is
4037 to check that the total number of inodes is greater than zero before doing
4038 the test of how many are available.
4040 58. When a "warn" ACL statement has a log_message modifier, the message is
4041 remembered, and not repeated. This is to avoid a lot of repetition when a
4042 message has many recipients that cause the same warning to be written.
4043 Howewer, Exim was preserving the list of already written lines for an
4044 entire SMTP session, which doesn't seem right. The memory is now reset if a
4045 new message is started.
4047 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4048 debugging output unless log_rewrite was also set.
4050 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4051 of (int)(handle) when we know that handle contains (void *)(-1).
4053 61. The Exim daemon panic-logs an error return when it closes the incoming
4054 connection. However "connection reset by peer" seems to be common, and
4055 isn't really an error worthy of noting specially, so that particular error
4058 62. When Exim is trying to find all the local interfaces, it used to panic and
4059 die if the ioctl to get the interface flags failed. However, it seems that
4060 on at least one OS (Solaris 9) it is possible to have an interface that is
4061 included in the list of interfaces, but for which you get a failure error
4062 for this call. This happens when the interface is not "plumbed" into a
4063 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4064 failure of the "get flags" call assumes that the interface is down.
4066 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4067 makes life easier for people who are doing arithmetic on fields extracted
4068 from dates, where you often get leading zeros that should not be
4069 interpreted as octal.
4071 64. Added qualify_domain to the redirect router, to override the global
4074 65. If a pathologically long header line contained very many addresses (the
4075 report of this problem mentioned 10 000) and each of them was rewritten,
4076 Exim could use up a very large amount of memory. (It kept on making new
4077 copies of the header line as it rewrote, and never released the old ones.)
4078 At the expense of a bit more processing, the header rewriting function has
4079 been changed so that it no longer eats memory in this way.
4081 66. The generation of the Received: header has been moved from the time that a
4082 message starts to be received, to the time that it finishes. The timestamp
4083 in the Received: header should now be very close to that of the <= log
4084 line. There are two side-effects of this change:
4086 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4087 logged header lines no longer include the local Received: line, because
4088 it has not yet been created. The same applies to a copy of the message
4089 that is returned to a non-SMTP sender when a message is rejected.
4091 (b) When a filter file is tested using -bf, no additional Received: header
4092 is added to the test message. After some thought, I decided that this
4095 This change does not affect the value of $received_for. It is still set
4096 after address rewriting, but before local_scan() is called.
4098 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4100 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4101 gave an unhelpful panic error message, and a defer error. I have managed to
4102 change this behaviour so that it now rejects any supplied certificate,
4103 which seems right, as the list of acceptable certificates is empty.
4105 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4106 gave an unhelpful defer error. I have not managed to make this reject any
4107 supplied certificates, but the error message it gives is "no certificate
4108 supplied", which is not helpful.
4110 70. exigrep's output now also includes lines that are not associated with any
4111 message, but which match the given pattern. Implemented by a patch from
4112 Martin Sluka, which also tidied up the Perl a bit.
4114 71. Recipient callout verification, like sender verification, was using <> in
4115 the MAIL FROM command. This isn't really the right thing, since the actual
4116 sender may affect whether the remote host accepts the recipient or not. I
4117 have changed it to use the actual sender in the callout; this means that
4118 the cache record is now keyed on a recipient/sender pair, not just the
4119 recipient address. There doesn't seem to be a real danger of callout loops,
4120 since a callout by the remote host to check the sender would use <>.
4121 [SEE ABOVE: changed after hitting problems.]
4123 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4124 temporary errors. However, in the case of such a code being given after
4125 the end of a data transmission (i.e. after ".") Exim was failing to write
4126 a retry record for the message. (Yes, there was some broken host that was
4127 actually sending 8xx at this point.)
4129 73. An unknown lookup type in a host list could cause Exim to panic-die when
4130 the list was checked. (An example that provoked this was putting <; in the
4131 middle of a list instead of at the start.) If this happened during a DATA
4132 ACL check, a -D file could be left lying around. This kind of configuration
4133 error no longer causes Exim to die; instead it causes a defer errror. The
4134 incident is still logged to the main and panic logs.
4136 74. Buglet left over from Exim 3 conversion. The message "too many messages
4137 in one connection" was written to the rejectlog but not the mainlog, except
4138 when address rewriting (yes!) was being logged.
4140 75. Added write_rejectlog option.
4142 76. When a system filter was run not as root (that is, when system_filter_user
4143 was set), the values of the $n variables were not being returned to the
4144 main process; thus, they were not subsequently available in the $sn
4147 77. Added +return_path_on_delivery log selector.
4149 78. A connection timeout was being treated differently from recipients deferred
4150 when testing hosts_max_try with a message that was older than the host's
4151 retry timeout. (The host should not be counted, thus allowing all hosts to
4152 be tried at least once before bouncing.) This may have been the cause of an
4153 occasionally reported bug whereby a message would remain on the queue
4154 longer than the retry timeout, but would be bounced if a delivery was
4155 forced. I say "may" because I never totally pinned down the problem;
4156 setting up timeout/retry tests is difficult. See also the next item.
4158 79. The ultimate address timeout was not being applied to errors that involved
4159 a combination of host plus message (for example, a timeout on a MAIL
4160 command). When an address resolved to a number of possible hosts, and they
4161 were not all tried for each delivery (e.g. because of hosts_max_try), a
4162 message could remain on the queue longer than the retry timeout.
4164 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4167 81. Fixed an obscure SMTP outgoing bug which required at least the following
4168 conditions: (a) there was another message waiting for the same server;
4169 (b) the server returned 5xx to all RCPT commands in the first message so
4170 that the message was not completed; (c) the server dropped the connection
4171 or gave a negative response to the RSET that Exim sends to abort the
4172 transaction. The observed case was a dropped connection after DATA that had
4173 been sent in pipelining mode. That is, the server had advertised PIPELINING
4174 but was not implementing it correctly. The effect of the bug was incorrect
4175 behaviour, such as trying another host, and this could lead to a crash.
4181 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4182 and daemon.c were passed as pointers to ints; they should have been
4183 pointers to socklen_t variables (which are typically unsigned ints).
4185 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4188 3. Fixed a really odd bug that affected only the testing scheme; patching a
4189 certain fixed string in the binary changed the value of another string that
4190 happened to be identical to the end of the original first string.
4192 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4193 name", it returns that address as the IP address. On some operating
4194 systems (e.g. Solaris), it also passes back the IP address string as the
4195 "host name". However, on others (e.g. Linux), it passes back an empty
4196 string. Exim wasn't checking for this, and was changing the host name to an
4197 empty string, assuming it had been canonicized.
4199 5. Although rare, it is permitted to have more than one PTR record for a given
4200 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4201 all the names associated with an address, because they do in Solaris.
4202 However, it seems that they do not in Linux for data that comes from the
4203 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4204 I found this out when I moved to a new Linux workstation and tried to run
4205 the Exim test suite.
4207 To get round this problem I have changed the code so that it now does its
4208 own call to the DNS to look up PTR records when searching for a host name.
4209 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4210 addresses that are only in /etc/hosts are still found.
4212 This behaviour is, however, controlled by an option called host_lookup_
4213 order, which defaults to "bydns:byaddr". If people want to use the other
4214 order, or indeed, just use one or the other means of lookup, they can
4215 specify it in this variable.
4217 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4218 some operating systems, this comes back from gethostbyaddr() as an empty
4219 string, and this is what Exim used to test for. However, it seems that in
4220 other systems, "." is yielded. Exim now tests for this case too.
4222 7. The values of check_spool_space and check_log_space are now held internally
4223 as a number of kilobytes instead of an absolute number of bytes. If a
4224 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4225 kilobyte. This means that much larger values can be stored.
4227 8. Exim monitor: an attempt to get the action menu when not actually pointing
4228 at a message produces an empty menu entitled "No message selected". This
4229 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4230 no entries in it ("Shell widget menu has zero width and/or height"). So I
4231 have added a single, blank menu entry in this case.
4233 9. Added ${quote_local_part.
4235 10. MIME decoding is now applied to the contents of Subject: header lines when
4238 11. Now that a reference to $sender_host_address automatically causes a reverse
4239 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4240 host lookup before query-style lookups in lists that might use this
4241 variable. This has therefore been abolished, and the "net-" prefix is no
4242 longer necessary for query-style lookups.
4244 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4245 have been a typo for LFLAGS, so it has been changed.
4247 13. The install script calls Exim with "-C /dev/null" in order to find the
4248 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4249 to be output. Howeve, since Exim outputs its version number before the
4250 error, it didn't break the script. It just looked ugly. I fixed this by
4251 always allowing "-C /dev/null" if the caller is root.
4253 14. Ignore overlarge ACL variable number when reading spool file - insurance
4254 against a later release with more variables having written the file.
4256 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4257 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4258 this, as well as the form without the "IPv6" on the front.
4260 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4261 OS/Makefile-Darwin file.
4263 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4264 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4266 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4267 variable. However, some OS use ints for inodes. I've added cast to long int
4268 to get rid of the compiler warning.
4270 19. I had forgotten to lock out "/../" in configuration file names when
4271 ALT_CONFIG_PREFIX was set.
4273 20. Routers used for verification do not need to specify transports. However,
4274 if such a router generated a host list, and callout was configured, Exim
4275 crashed, because it could not find a port number from the (non-existent)
4276 transport. It now assumes port 25 in this circumstance.
4278 21. Added the -t option to exigrep.
4280 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4281 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4282 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4283 them is set, LOOKUP_LSEARCH is forced.
4285 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4287 24. Added sender and host information to the "rejected by local_scan()" log
4288 line; previously there was no indication of these.
4290 25. Added .include_if_exists.
4292 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4293 Linux. It turns out that not all file systems support this. Apparently some
4294 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4295 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4296 not supported on the file descriptor, is now ignored when Exim is trying to
4297 sync a directory. This applies only to Linux.
4299 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4301 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4302 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4305 29. Added support for maildirsize files from supplied patch (modified a bit).
4307 30. The use of :fail: followed by an empty string could lead Exim to respond to
4308 sender verification failures with (e.g.):
4310 550 Verification failed for <xxx>
4311 550 Sender verify failed
4313 where the first response line was missing the '-' that indicates it is not
4314 the final line of the response.
4316 31. The loop for finding the name of the user that called Exim had a hardwired
4317 limit of 10; it now uses the value of finduser_retries, which is used for
4318 all other user lookups.
4320 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4323 33. Exim was neglecting to zero errno before one call of strtol() when
4324 expanding a string and expecting an integer value. On some systems this
4325 resulted in spurious "integer overflow" errors. Also, it was casting the
4326 result into an int without checking.
4328 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4329 did not work. The code looks as if it has *never* worked, though it appears
4330 to have been documented since at least releast 1.62. I have made it work.
4332 35. The "timeout_DNS" error in retry rules, also documented since at least
4333 1.62, also never worked. As it isn't clear exactly what this means, and
4334 clearly it isn't a major issue, I have abolished the feature by treating it
4335 as "timeout", and writing a warning to the main and panic logs.
4337 36. The display of retry rules for -brt wasn't always showing the error code
4340 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4341 timeout_connect_A, timeout_connect_MX.
4343 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4344 to the empty sender.
4346 39. The daemon was not analysing the content of -oX till after it had closed
4347 stderr and disconnected from the controlling terminal. This meant that any
4348 syntax errors were only noted on the panic log, and the return code from
4349 the command was 0. By re-arranging the code a little, I've made the
4350 decoding happen first, so such errors now appear on stderr, and the return
4351 code is 1. However, the actual setting up of the sockets still happens in
4352 the disconnected process, so errors there are still only recorded on the
4355 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4356 connections (as happens on some IP stacks) was logged at start up time as
4357 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4358 it from "IPv6 and IPv4", which means that two separate sockets are being
4361 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4362 says whether AF_INET or AF_INET6 was passed as an argument.
4364 42. Exiwhat output was messed up when time zones were included in log
4367 43. Exiwhat now gives more information about the daemon's listening ports,
4368 and whether -tls-on-connect was used.
4370 44. The "port" option of the smtp transport is now expanded.
4372 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4373 silently ignored. Now an error message is written to the main and panic
4376 46. There's a new ACL modifier called "logwrite" which writes to a log file
4377 as soon as it is encountered.
4379 47. Added $local_user_uid and $local_user_gid at routing time.
4381 48. Exim crashed when trying to verify a sender address that was being
4384 49. Exim was recognizing only a space character after ".include". It now also
4385 recognizes a tab character.
4387 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4388 extracting the relevant information from the specification. The man page no
4389 longer contains scrambled data for the -d option, and I've added a section
4390 at the front about calling Exim under different names.
4392 51. Added "extra_headers" argument to the "mail" command in filter files.
4394 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4397 53. Installed eximstats 1.29.
4399 54. Added transport_filter_timeout as a generic transport option.
4401 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4402 Cc: header lines. This was required by RFC 822, but it not required by RFC
4405 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4406 incoming messages that did not have them. Now it does so only if the
4407 message originates locally, that is, if there is no associated remote host
4408 address. When Resent- header lines are present, this applies to the Resent-
4409 lines rather than the non-Resent- lines.
4411 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4412 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4414 58. Messages for configuration errors now include the name of the main
4415 configuration file - useful now that there may be more than one file in a
4416 list (.included file names were always shown).
4418 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4419 for those rare installations that do not start the daemon as root or run it
4420 setuid root. I've cut out the call to initgroups() if the daemon is not
4423 60. The Exim user and group can now be bound into the binary as text strings
4424 that are looked up at the start of Exim's processing.
4426 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4428 62. Added $mailstore_basename variable.
4430 63. Installed patch to sieve.c from Michael Haardt.
4432 64. When Exim failed to open the panic log after failing to open the main log,
4433 the original message it was trying to log was written to stderr and debug
4434 output, but if they were not available (the usual case in production), it
4435 was lost. Now it is written to syslog before the two lines that record the
4436 failures to open the logs.
4438 65. Users' Exim filters run in subprocesses under the user's uid. It is
4439 possible for a "deliver" command or an alias in a "personal" command to
4440 provoke an address rewrite. If logging of address rewriting is configured,
4441 this fails because the process is not running as root or exim. There may be
4442 a better way of dealing with this, but for the moment (because 4.30 needs
4443 to be released), I have disabled address rewrite logging when running a
4444 filter in a non-root, non-exim process.
4450 1. The buildconfig auxiliary program wasn't quoting the value set for
4451 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4452 not defined. This bug was masked in 4.22 by the effect that was fixed in
4455 2. Some messages that were rejected after a message id was allocated were
4456 shown as "incomplete" by exigrep. It no longer does this for messages that
4457 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4459 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4460 have allow_domain_literals set, the ID did not get logged in the <= line.
4461 Domain literals are now always recognized in Message-ID: header lines.
4463 4. The first argument for a ${extract expansion item is the key name or field
4464 number. Leading and trailing spaces in this item were not being ignored,
4465 causing some misleading effects.
4467 5. When deliver_drop_privilege was set, single queue runner processes started
4468 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4469 same command in the process it spins off) were not dropping privilege.
4471 6. When the daemon running as "exim" started a queue runner, it always
4472 re-executed Exim in the spun-off process. This is a waste of effort when
4473 deliver_drop_privilege is set. The new process now just calls the
4474 queue-runner function directly.
4480 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4483 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4484 ignored. Though the use of -oP was forcing the writing of a pid file, it
4485 was always written to the default place.
4487 3. If the message "no IP address found for host xxxx" is generated during
4488 incoming verification, it is now followed by identification of the incoming
4489 connection (so you can more easily find what provoked it).
4491 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4493 5. Added some features to "harden" Exim a bit more against certain attacks:
4495 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4496 be put in Local/Makefile. This is like the never_users runtime option,
4497 but it cannot be overridden. The default setting is "root".
4499 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4500 prefix string with which any file named in a -C command line option
4503 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4504 is retained for -C and -D only if the caller of Exim is root. Without
4505 it, the exim user may also use -C and -D and retain privilege.
4507 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4508 command line option is disabled.
4510 6. Macro names set by the -D option must start with an upper case letter, just
4511 like macro names defined in the configuration file.
4513 7. Added "dereference=" facility to LDAP.
4515 8. Two instances of the typo "uknown" in the source files are fixed.
4517 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4518 the Configure-Makefile script screwed up while processing it.
4520 10. Incorporated PCRE 4.4.
4522 11. The SMTP synchronization check was not operating right at the start of an
4523 SMTP session. For example, it could not catch a HELO sent before the client
4524 waited for the greeting. There is now a check for outstanding input at the
4525 point when the greeting is written. Because of the duplex, asynchronous
4526 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4527 way, but not yet received, when the check is performed.
4529 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4530 on TCP/IP sockets, because this apparently causes some broken clients to
4533 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4534 unchanged) from the Cygwin maintainer.
4536 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4537 was supported instead of testing for mbx. Effectively a typo.
4539 15. The spa authenticator server code was not checking that the input it
4540 received was valid base64.
4542 16. The debug output line for the "set" modifier in ACLs was not showing the
4543 name of the variable that was being set.
4545 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4547 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4548 lookup on its own. Something else was needed to trigger the lookup. For
4549 example, a match in host_lookup or the need for a host name in a host list.
4550 Now, if $sender_host_name is referenced and the host name has not yet been
4551 looked up, a lookup is performed. If the lookup fails, the variable remains
4552 empty, and $host_lookup_failed is set to "1".
4554 19. Added "eqi" as a case-independent comparison operator.
4556 20. The saslauthd authentication condition could segfault if neither service
4557 nor realm was specified.
4559 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4560 error message that was logged was misleading, and incoming SMTP
4561 connections were dropped. The message is now more accurate, and temporary
4562 errors are given to SMTP connections.
4564 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4565 (except RSET) with a 421 error, until QUIT is received. However, it was
4566 failing to send a response to QUIT.
4568 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4569 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4570 ACL is now run after the helo_try_verify_hosts code.
4572 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4573 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4574 case-independent, so other case variants are also recognized). Apparently
4575 some systems use these upper case variants.
4577 25. If more than two messages were waiting for the same host, and a transport
4578 filter was specified for the transport, Exim sent two messages over the
4579 same TCP/IP connection, and then failed with "socket operation on non-
4580 socket" when it tried to send the third.
4582 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4584 27. The extern definition of crypt16() in expand.c was not being excluded when
4585 the OS had its own crypt16() function.
4587 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4588 as a preferred synonym for return_size_limit, both as an option and as an
4591 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4593 30. Changed the default configuration ACL to relax the local part checking rule
4594 for addresses that are not in any local domains. For these addresses,
4595 slashes and pipe symbols are allowed within local parts, but the sequence
4596 /../ is explicitly forbidden.
4598 31. SPA server authentication was not clearing the challenge buffer before
4601 32. log_message in a "warn" ACL statement was writing to the reject log as
4602 well as to the main log, which contradicts the documentation and doesn't
4603 seem right (because no rejection is happening). So I have stopped it.
4605 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4606 However, I am unable to do any testing of this.
4608 34. Fixed an infelicity in the appendfile transport. When checking directories
4609 for a mailbox, to see if any needed to be created, it was accidentally
4610 using path names with one or more superfluous leading slashes; tracing
4611 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4613 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4614 contents are added to the log line that is written for every discarded
4615 recipient. (Previously a log_message setting was ignored.)
4617 36. The ${quote: operator now quotes the string if it is empty.
4619 37. The install script runs exim in order to find its version number. If for
4620 some reason other than non-existence or emptiness, which it checks, it
4621 could not run './exim', it was installing it with an empty version number,
4622 i.e. as "exim-". This error state is now caught, and the installation is
4625 38. An argument was missing from the function that creates an error message
4626 when Exim fails to connect to the socket for saslauthd authentication.
4627 This could cause Exim to crash, or give a corrupted message.
4629 39. Added isip, isip4, and isip6 to ${if conditions.
4631 40. The ACL variables $acl_xx are now saved with the message, and can be
4632 accessed later in routers, transports, and filters.
4634 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4635 strings in the file are not string-expanded.
4637 42. If a MAIL command specified a SIZE value that was too large to fit into an
4638 int variable, the check against message_size_limit failed. Such values are
4639 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4640 day this will have to be increased, but I don't think I want to be around
4641 when emails are that large.
4648 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4649 iconv() is not standard in FreeBSD.
4651 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4652 IPv6 enabled. The observed symptom was a segmentation fault on return from
4653 the function os_common_find_running_interfaces() in src/os.c.
4655 3. In the check_special_case() function in daemon.c I had used "errno" as an
4656 argument name, which causes warnings on some systems. This was basically a
4657 typo, since it was named "eno" in the comments!
4659 4. The code that waits for the clock to tick (at a resolution of some fraction
4660 of a second) so as to ensure message-id uniqueness was always waiting for
4661 at least one whole tick, when it could have waited for less. [This is
4662 almost certainly not relevant at current processor speeds, where it is
4663 unlikely to ever wait at all. But we try to future-proof.]
4665 5. The function that sleeps for a time interval that includes fractions of a
4666 second contained a race. It did not block SIGALRM between setting the
4667 timer, and suspending (a couple of lines later). If the interval was short
4668 and the sigsuspend() was delayed until after it had expired, the suspension
4669 never ended. On busy systems this could lead to processes getting stuck for
4672 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4673 process, before it forks any delivery processes. The open lookup caching
4674 mechanism meant that the open file or database connection was passed into
4675 the delivery process. The problem was that delivery processes always tidy
4676 up cached lookup data. This could cause a problem for the next delivery
4677 process started by the queue runner, because the external queue runner
4678 process does not know about the closure. So the next delivery process
4679 still has data in the lookup cache. In the case of a file lookup, there was
4680 no problem because closing a file descriptor in a subprocess doesn't affect
4681 the parent. However, if the lookup was caching a connection to a database,
4682 the connection was closed, and the second delivery process was likely to
4683 see errors such as "PGSQL: query failed: server closed the connection
4684 unexpectedly". The problem has been fixed by closing all cached lookups
4685 in a queue runner before running a delivery process.
4687 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4688 seem to have the "const" qualifier which it has on other OS. I've
4691 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4692 *of the same type* (client or server) with the same public name that an
4693 error should be diagnosed.
4695 9. When Exim looked up a host name for an IP address, but failed to find the
4696 original IP address when looking up the host name (a safety check), it
4697 output the message "<ip address> does not match any IP for NULL", which was
4698 confusing, to say the least. The bug was that the host name should have
4699 appeared instead of "NULL".
4701 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4702 user that is built into the binary, and the -C or -D options is used, root
4703 privilege is dropped before the configuration file is read. In addition,
4704 logging is switched to stderr instead of the normal log files. If the
4705 configuration then re-defines the Exim user, the unprivileged environment
4706 is probably not what is expected, so Exim logs a panic warning message (but
4709 However, if deliver_drop_privilege is set, the unprivileged state may well
4710 be exactly what is intended, so the warning has been cut out in that case,
4711 and Exim is allowed to try to write to its normal log files.
4717 1. smtp_return_error_details was not giving details for temporary sender
4718 or receiver verification errors.
4720 2. Diagnose a configuration error if two authenticators have the same public
4723 3. Exim used not to create the message log file for a message until the first
4724 delivery attempt. This could be confusing when incoming messages were held
4725 for policy or load reasons. The message log file is now created at the time
4726 the message is received, and an initial "Received" line is written to it.
4728 4. The automatically generated man page for command line options had a minor
4729 bug that caused no ill effects; however, a more serious problem was that
4730 the procedure for building the man page automatically didn't always
4731 operate. Consequently, release 4.20 contains an out-of-date version. This
4732 shouldn't happen again.
4734 5. When building Exim with embedded Perl support, the script that builds the
4735 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4736 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4738 6. The freeze_tell option was not being used for messages that were frozen on
4739 arrival, either by an ACL or by local_scan().
4741 7. Added the smtp_incomplete_transaction log selector.
4743 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4744 was accepting AUTH without a new EHLO.
4746 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4747 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4750 10. Logging of TCP/IP connections (when configured) now happens in the main
4751 daemon process instead of the child process, so that the TCP/IP connection
4752 count is more accurate (but it can never be perfect).
4754 11. The use of "drop" in a nested ACL was not being handled correctly in the
4755 outer ACL. Now, if condition failure induced by the nested "drop" causes
4756 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4757 or "require"), the connection is dropped.
4759 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4760 that yield "discard" can now be used with an "accept" or a "discard" verb,
4761 but an error is generated for any others (because I can't see a useful way
4762 to define what should happen).
4764 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4765 are now processed in the same way as lines in the Exim configuration file.
4766 In particular, continuation lines are supported.
4768 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4770 15. Added -ti meaning -t -i.
4772 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4773 domains, and warn by logging if others are found.
4775 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4776 returned from GIFCONF when Exim is trying to find the list of interfaces on
4777 a host. The code in os.c has been modified to copy each ifreq to an aligned
4778 structure in all cases.
4780 Also, in some cases, the returned ifreq's were being copied to a 'struct
4781 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4782 means the last couple of bytes of an IPv6 address could be chopped if the
4783 ifreq contained only a normal sockaddr (14 bytes storage).
4785 18. Named domain lists were not supported in the hosts_treat_as_local option.
4786 An entry such as +xxxx was not recognized, and was treated as a literal
4789 19. Ensure that header lines added by a DATA ACL are included in the reject log
4790 if the ACL subsequently rejects the message.
4792 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4793 MD5 (which is deprecated).
4795 21. When testing a filter file using -bf, Exim was writing a message when it
4796 took the sender from a "From " line in the message, but it was not doing so
4797 when it took $return_path from a Return-Path: header line. It now does.
4799 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4800 with a valid header line field name (a series of printing characters
4801 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4803 23. Changed "disc" in the source to "disk" to conform to the documentation and
4804 the book and for uniformity.
4806 24. Ignore Sendmail's -Ooption=value command line item.
4808 25. When execve() failed while trying to run a command in a pipe transport,
4809 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4810 could be confused with a return value of 69 from the command itself. This
4811 has been changed to 127, the value the shell returns if it is asked to run
4812 a non-existent command. The wording for the related log line suggests a
4813 non-existent command as the problem.
4815 26. If received_header_text expands to an empty string, do not add a Received:
4816 header line to the message. (Well, it adds a token one on the spool, but
4817 marks it "old" so that it doesn't get used or transmitted.)
4819 27. Installed eximstats 1.28 (addition of -nt option).
4821 28. There was no check for failure on the call to getsockname() in the daemon
4822 code. This can fail if there is a shortage of resources on the system, with
4823 ENOMEM, for example. A temporary error is now given on failure.
4825 29. Contrary to the C standard, it seems that in some environments, the
4826 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4827 program. Exim now does this explicitly; it affects the formatting of
4828 timestamps using strftime().
4830 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4831 complaints. I've now initialized all the variables, to avoid this.
4833 32. Header lines added by a system filter were not being "seen" during
4834 transport-time rewrites.
4836 33. The info_callback() function passed to OpenSSL is set up with type void
4837 (*)(SSL *, int, int), as described somewhere. However, when calling the
4838 function (actually a macro) that sets it up, the type void(*)() is
4839 expected. I've put in a cast to prevent warnings from picky compilers.
4841 34. If a DNS black list lookup found a CNAME record, but there were no A
4842 records associated with the domain it pointed at, Exim crashed.
4844 35. If a DNS black list lookup returned more than one A record, Exim ignored
4845 all but the first. It now scans all returned addresses if a particular IP
4846 value is being sought. In this situation, the contents of the
4847 $dnslist_value variable are a list of all the addresses, separated by a
4850 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4851 to accept a host name and all its aliases if the forward lookup for any of
4852 them yielded the IP address of the incoming connection. Now it accepts only
4853 those names whose forward lookup yields the correct IP address. Any other
4854 names are discarded. This closes a loophole whereby a rogue DNS
4855 administrator could create reverse DNS records to break through a
4856 wildcarded host restriction in an ACL.
4858 37. If a user filter or a system filter that ran in a subprocess used any of
4859 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4860 the wrong values were passed to the pipe command ($thisaddress had the
4861 value of $0, $0 had the value of $1, etc). This bug was introduced by
4862 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4864 38. Improved the line breaking for long SMTP error messages from ACLs.
4865 Previously, if there was no break point between 40 and 75 characters, Exim
4866 left the rest of the message alone. Two changes have been made: (a) I've
4867 reduced the minimum length to 35 characters; (b) if it can't find a break
4868 point between 35 and 75 characters, it looks ahead and uses the first one
4869 that it finds. This may give the occasional overlong line, but at least the
4870 remaining text gets split now.
4872 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4873 file descriptors might be low, and that setting 1000 would always raise it.
4874 It turns out that in some environments, the limit is already over 1000 and
4875 that lowering it causes trouble. So now Exim takes care not to decrease it.
4877 40. When delivering a message, the value of $return_path is set to $sender_
4878 address at the start of routing (routers may change the value). By an
4879 oversight, this default was not being set up when an address was tested by
4880 -bt or -bv, which affected the outcome if any router or filter referred to
4883 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4884 log only when "message" or "log_message" are set. However, if one of the
4885 conditions was an address verification, or a call to a nested ACL, the
4886 messages generated by the underlying test were being passed through. This
4887 no longer happens. The underlying message is available in $acl_verify_
4888 message for both "message" and "log_message" expansions, so it can be
4889 passed through if needed.
4891 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4892 new expansion $bh_ to give the encoded byte string without charset
4893 translation. Translation happens only if iconv() is available; HAVE_ICONV
4894 indicates this at build time. HEADERS_CHARSET gives the charset to
4895 translate to; headers_charset can change it in the configuration, and
4896 "headers charset" can change it in an individual filter file.
4898 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4899 that creates RFC 2047 encoded "words" labels them as that charset instead
4900 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4901 expansion operator; (ii) when Exim creates a From: line for a local
4902 message; (iii) when a header line is rewritten to include a "phrase" part.
4904 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4905 buggy, causing it to skip the first lines of messages whose message ID
4906 ended in 'D'. This would not have bitten before Exim release 4.14, because
4907 message IDs were unlikely to end in 'D' before then. The effect was to have
4908 incorrect size information for certain domains.
4910 45. #include "config.h" was missing at the start of the crypt16.c module. This
4911 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4914 46. If there was a timeout during a "random" callout check, Exim treated it as
4915 a failure of the random address, and carried on sending RSET and the real
4916 address. If the delay was just some slowness somewhere, the response to the
4917 original RCPT would be taken as a response to RSET and so on, causing
4918 mayhem of various kinds.
4920 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4921 when I implemented it. It didn't allow for the fact that some option values
4922 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4923 the right test for positive values.
4925 48. Domain names in DNS records are case-independent. Exim always looks them up
4926 in lower case. Some resolvers return domain names in exactly the case they
4927 appear in the zone file, that is, they may contain uppercase letters. Not
4928 all resolvers do this - some return always lower case. Exim was treating a
4929 change of case by a resolver as a change of domain, similar to a widening
4930 of a domain abbreviation. This triggered its re-routing code and so it was
4931 trying to route what was effectively the same domain again. This normally
4932 caused routing to fail (because the router wouldn't handle the domain
4933 twice). Now Exim checks for this case specially, and just changes the
4934 casing of the domain that it ultimately uses when it transmits the message
4937 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4940 50. If a filter generated a file delivery with a non-absolute name (possible if
4941 no home directory exists for the router), the forbid_file option was not
4944 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4945 the existing equality matching.
4947 52. Exim was using ints instead of ino_t variables in some places where it was
4948 dealing with inode numbers.
4950 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4951 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4952 called TMPDIR, and if it finds it is different, it changes its value.
4954 54. The smtp_printf() function is now made available to local_scan() so
4955 additional output lines can be written before returning. There is also an
4956 smtp_fflush() function to enable the detection of a dropped connection.
4957 The variables smtp_input and smtp_batched_input are exported to
4960 55. Changed the default runtime configuration: the message "Unknown user"
4961 has been removed from the ACL, and instead placed on the localuser router,
4962 using the cannot_route_message feature. This means that any verification
4963 failures that generate their own messages won't get overridden. Similarly,
4964 the "Unrouteable address" message that was in the ACL for unverifiable
4965 relay addresses has also been removed.
4967 56. Added hosts_avoid_esmtp to the smtp transport.
4969 57. The exicyclog script was not checking for the esoteric option
4970 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4971 will work only if exicyclog is run under the appropriate euid.
4973 58. Following a discussion on the list, the rules by which Exim recognises line
4974 endings on incoming messages have been changed. The -dropcr and drop_cr
4975 options are now no-ops, retained only for backwards compatibility. The
4976 following line terminators are recognized: LF CRLF CR. However, special
4977 processing applies to CR:
4979 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4980 nor a local message in the state where . is a terminator.
4982 (ii) If a bare CR is encountered in a header line, an extra space is added
4983 after the line terminator so as not to end the header. The reasoning
4984 behind this is that bare CRs in header lines are most likely either
4985 to be mistakes, or people trying to play silly games.
4987 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4988 was being incorrectly given as 18 bytes larger than it should have been.
4989 This is a VOB (very old bug).
4991 60. This may never have affected anything current, but just in case it has:
4992 When the local host is found other than at the start of a list of hosts,
4993 the local host, those with the same MX, and any that follow, are discarded.
4994 When the list in question was part of a longer list of hosts, the following
4995 hosts (not currently being processed) were also being discarded. This no
4996 longer happens. I'm not sure if this situation could ever has previously
4999 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5002 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5003 header. This is something that is recommended in a new Internet Draft, and
5004 is something that is documented as being done by Sendmail. There are two
5005 possible values. For messages generated by the autoreply transport, Exim
5008 Auto-Submitted: auto-replied
5010 whereas for all other generated messages (e.g. bounces) it adds
5012 Auto-Submitted: auto-generated
5014 63. The "personal" condition in filters now includes a test for the
5015 Auto-Submitted: header. If it contains the string "auto-" the message it
5016 not considered personal.
5018 64. Added rcpt_include_affixes as a generic transport option.
5020 65. Added queue_only_override (default true).
5022 66. Added the syslog_duplication option.
5024 67. If what should have been the first header line of a message consisted of
5025 a space followed by a colon, Exim was mis-interpreting it as a header line.
5026 It isn't of course - it is syntactically invalid and should therefore be
5027 treated as the start of the message body. The misbehaviour could have
5028 caused a number of strange effects, including loss of data in subsequent
5029 header lines, and spool format errors.
5031 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5032 client host had authenticated. This control can now be exercised by an ACL
5033 for more flexibility.
5035 69. By default, callouts do not happen when testing with -bh. There is now a
5036 variant, -bhc, which does actually run the callout code, including
5037 consulting and updating the callout cache.
5039 70. Added support for saslauthd authentication, courtesy of Alexander
5042 71. If statvfs() failed on the spool or log directories while checking their
5043 size for availability, Exim confusingly gave the error "space shortage".
5044 Furthermore, in debugging mode it crashed with a floating point exception.
5045 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5046 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5047 serious problem, Exim now writes to the main and panic logs when this
5048 happens, with details of the failure. It then refuses to accept the
5049 incoming message, giving the message "spool directory problem" or "log
5050 directory problem" with a 421 code for SMTP messages.
5052 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5053 0, 1, and 2 exist, because some OS complain for execs without them (see
5054 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5055 descriptors. However, the code omitted to check that the open succeeded,
5056 causing mysterious errors if for some reason the permissions on /dev/null
5057 got screwed. Now Exim writes a message to the main and panic logs, and
5058 bombs out if it can't open /dev/null.
5060 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5061 interact so that it is all more flexible. It is supposed to remain
5062 backwards compatible. Also added extra_local_interfaces.
5064 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5065 to bomb out with an assertion failure - to the client this appears as a
5066 connection drop. This problem occurs in the part of the code that was taken
5067 from the Samba project. Fortunately, the assertion is in a very simple
5068 function, so I have fixed this by reproducing the function inline in the
5069 one place where it is called, and arranging for authentication to fail
5070 instead of killing the process with assert().
5072 75. The SPA client code was not working when the server requested OEM rather
5073 than Unicode encoding.
5075 76. Added code to make require_files with a specific uid setting more usable in
5076 the case where statting the file as root fails - usually a non-root-mounted
5077 NFS file system. When this happens and the failure is EACCES, Exim now
5078 forks a subprocess and does the per-uid checking as the relevant uid.
5080 77. Added process_log_path.
5082 78. If log_file_path was not explicitly set, a setting of check_log_space or
5083 check_log_inodes was ignored.
5085 79. If a space check for the spool or log partitions fails, the incident is now
5086 logged. Of course, in the latter case the data may get lost...
5088 80. Added the %p formatting code to string_format() so that it can be used to
5089 print addresses in debug_print(). Adjusted all the address printing in the
5090 debugging in store.c to use %p rather than %d.
5092 81. There was a concern that a line of code in smtp_in.c could overflow a
5093 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5094 initially expressed, the concern was not well-founded, because trailing
5095 spaces are removed early. However, if the trailing spaces were followed by
5096 a NULL, they did not get removed, so the overflow was possible. Two fixes
5099 (a) I re-wrote the offending code in a cleaner fashion.
5100 (b) If an incoming SMTP command contains a NULL character, it is rejected
5103 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5104 runs initgroups(), so that if the Exim user is in any additional groups,
5105 they will be used during message reception.
5111 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob