1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.628 2010/06/07 07:09:10 pdp Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PP/01 Date: & Message-Id: revert to normally being appended to a message,
10 only prepend for the Resent-* case. Fixes regression introduced in
11 Exim 4.70 by NM/22 for Bugzilla 607.
13 PP/02 Include check_rfc2047_length in configure.default because we're seeing
14 increasing numbers of administrators be bitten by this.
16 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
18 PP/03 Bugzilla 994: added openssl_options main configuration option.
20 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
22 PP/05 Bugzilla 834: provide a permit_codedump option for pipe transports.
24 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
26 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
27 without a peer certificate (I believe), leading to a segfault because of
28 an assumption that peers always have certificates. Be a little more
29 paranoid. Problem reported by Martin Tscholak.
31 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
32 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
33 NB: ClamAV planning to remove STREAM in "middle of 2010".
35 PP/09 Implemented reverse_ip expansion operator.
37 PP/10 Bugzilla 937: provide a "debug" ACL control.
39 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
41 PP/12 Bugzilla 973: Implement --version.
43 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
45 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
47 PP/15 Bugzilla 816: support multiple condition rules on Routers.
53 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
54 $data_path, and $header_path variables; fixed documentation bugs and
57 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
58 exipick to access non-standard spools, including the "frozen" queue
61 NM/01 Bugzilla 965: Support mysql stored procedures.
62 Patch from Alain Williams
64 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
66 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
67 Patch from Andreas Metzler
69 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
70 Patch from Kirill Miazine
72 NM/05 Bugzilla 671: Added umask to procmail example.
74 JJ/03 installed exipick 20100323.0, fixing doc bug
76 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
77 directory. Notification and patch from Dan Rosenberg.
79 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
81 TK/02 Improve log output when DKIM signing operation fails.
83 MH/01 Treat the transport option dkim_domain as a colon separated
84 list, not as a single string, and sign the message with each element,
85 omitting multiple occurences of the same signer.
87 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
88 Bugzilla 985, 986. Patch by Simon Arlott
90 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
93 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
94 MBX locking. Notification from Dan Rosenberg.
100 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
102 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
104 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
106 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
108 NM/04 Bugzilla 915: Flag broken perl installation during build.
114 TK/01 Added patch by Johannes Berg that expands the main option
115 "spamd_address" if it starts with a dollar sign.
117 TK/02 Write list of recipients to X-Envelope-Sender header when building
118 the mbox-format spool file for content scanning (suggested by Jakob
121 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
122 (http://www.dcc-servers.net/) support via dccifd. Activated by
123 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
125 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
126 by Mark Daniel Reidel <mr@df.eu>.
128 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
129 When building exim an external PCRE library is now needed -
130 PCRE is a system library on the majority of modern systems.
131 See entry on PCRE_LIBS in EDITME file.
133 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
134 conversation. Added nologin parameter to request.
135 Patch contributed by Kirill Miazine.
137 TF/01 Do not log submission mode rewrites if they do not change the address.
139 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
141 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
142 log files in place. Contributed by Roberto Lima.
144 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
146 TF/03 Bugzilla 615: When checking the local_parts router precondition
147 after a local_part_suffix or local_part_prefix option, Exim now
148 does not use the address's named list lookup cache, since this
149 contains cached lookups for the whole local part.
151 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
152 Robert Millan. Documentation is in experimental-spec.txt.
154 TF/04 Bugzilla 668: Fix parallel build (make -j).
156 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
158 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
159 Patch provided by Jan Srzednicki.
161 TF/05 Leading white space used to be stripped from $spam_report which
162 wrecked the formatting. Now it is preserved.
164 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
165 that they are available at delivery time.
167 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
169 TF/08 TLS error reporting now respects the incoming_interface and
170 incoming_port log selectors.
172 TF/09 Produce a more useful error message if an SMTP transport's hosts
173 setting expands to an empty string.
175 NM/06 Bugzilla 744: EXPN did not work under TLS.
176 Patch provided by Phil Pennock.
178 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
179 Patch provided by Richard Godbee.
181 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
182 acl_smtp_notquit, added index entry.
184 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
185 Patch provided by Eugene Bujak.
187 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
188 accept(). Patch provided by Maxim Dounin.
190 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
191 Patch provided by Phil Pennock.
193 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
195 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
196 Patch provided by Brad "anomie" Jorsch.
198 NM/14 Bugzilla 622: Added timeout setting to transport filter.
199 Patch provided by Dean Brooks.
201 TK/05 Add native DKIM support (does not depend on external libraries).
203 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
204 Patch provided by Graeme Fowler.
206 NM/16 Bugzilla 851: Documentation example syntax fix.
208 NM/17 Changed NOTICE file to remove references to embedded PCRE.
210 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
213 NM/19 Bugzilla 745: TLS version reporting.
214 Patch provided by Phil Pennock.
216 NM/20 Bugzilla 167: bool: condition support.
217 Patch provided by Phil Pennock.
219 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
220 clients. Patch provided by Phil Pennock.
222 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
223 Patch provided by Brad "anomie" Jorsch.
225 NM/23 Bugzilla 687: Fix misparses in eximstats.
226 Patch provided by Heiko Schlittermann.
228 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
229 Patch provided by Heiko Schlittermann.
231 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
232 plus update to original patch.
234 NM/26 Bugzilla 799: Documentation correction for ratelimit.
236 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
237 Patch provided by David Brownlee.
239 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
241 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
243 NM/30 Bugzilla 888: TLS documentation bugfixes.
245 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
247 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
248 Unlike the original bugzilla I have changed all shell scripts in src tree.
250 NM/33 Bugzilla 898: Transport filter timeout fix.
251 Patch by Todd Rinaldo.
253 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
254 Patch by Serge Demonchaux.
256 NM/35 Bugzilla 39: Base64 decode bug fixes.
257 Patch by Jakob Hirsch.
259 NM/36 Bugzilla 909: Correct connect() call in dcc code.
261 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
263 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
265 NM/39 Bugzilla 911: Fixed MakeLinks build script.
271 TK/01 Add preliminary DKIM support. Currently requires a forked version of
272 ALT-N's libdkim that I have put here:
273 http://duncanthrax.net/exim-experimental/
275 Note to Michael Haardt: I had to rename some vars in sieve.c. They
276 were called 'true' and it seems that C99 defines that as a reserved
277 keyword to be used with 'bool' variable types. That means you could
278 not include C99-style headers which use bools without triggering
279 build errors in sieve.c.
281 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
282 as mailq or other aliases. Changed the --help handling significantly
283 to do whats expected. exim_usage() emits usage/help information.
285 SC/01 Added the -bylocaldomain option to eximstats.
287 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
289 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
291 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
297 PH/01 Another patch from the Sieve maintainer.
299 PH/02 When an IPv6 address is converted to a string for single-key lookup
300 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
301 dots are used instead of colons so that keys in lsearch files need not
302 contain colons. This was done some time before quoting was made available
303 in lsearch files. However, iplsearch files do require colons in IPv6 keys
304 (notated using the quote facility) so as to distinguish them from IPv4
305 keys. This meant that lookups for IP addresses in host lists did not work
306 for iplsearch lookups.
308 This has been fixed by arranging for IPv6 addresses to be expressed with
309 colons if the lookup type is iplsearch. This is not incompatible, because
310 previously such lookups could never work.
312 The situation is now rather anomolous, since one *can* have colons in
313 ordinary lsearch keys. However, making the change in all cases is
314 incompatible and would probably break a number of configurations.
316 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
319 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
320 conversion specification without a maximum field width, thereby enabling
321 a rogue spamd server to cause a buffer overflow. While nobody in their
322 right mind would setup Exim to query an untrusted spamd server, an
323 attacker that gains access to a server running spamd could potentially
324 exploit this vulnerability to run arbitrary code as the Exim user.
326 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
327 $primary_hostname instead of what libspf2 thinks the hosts name is.
329 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
330 a directory entry by the name of the lookup key. Previously, if a
331 symlink pointed to a non-existing file or a file in a directory that
332 Exim lacked permissions to read, a lookup for a key matching that
333 symlink would fail. Now it is enough that a matching directory entry
334 exists, symlink or not. (Bugzilla 503.)
336 PH/03 The body_linecount and body_zerocount variables are now exported in the
339 PH/04 Added the $dnslist_matched variable.
341 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
342 This means they are set thereafter only if the connection becomes
345 PH/06 Added the client_condition to authenticators so that some can be skipped
346 by clients under certain conditions.
348 PH/07 The error message for a badly-placed control=no_multiline_responses left
349 "_responses" off the end of the name.
351 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
353 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
354 (without spaces) instead of just copying the configuration text.
356 PH/10 Added the /noupdate option to the ratelimit ACL condition.
358 PH/11 Added $max_received_linelength.
360 PH/12 Added +ignore_defer and +include_defer to host lists.
362 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
363 way in which PCRE > 7.0 is built.
365 PH/14 Implemented queue_only_load_latch.
367 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
368 MAIL command. The effect was to mangle the value on 64-bit systems.
370 PH/16 Another patch from the Sieve maintainer.
372 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
374 PH/18 If a system quota error occurred while trying to create the file for
375 a maildir delivery, the message "Mailbox is full" was not appended to the
376 bounce if the delivery eventually timed out. Change 4.67/27 below applied
377 only to a quota excession during the actual writing of the file.
379 PH/19 It seems that peer DN values may contain newlines (and other non-printing
380 characters?) which causes problems in log lines. The DN values are now
381 passed through string_printing() before being added to log lines.
383 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
384 and InterBase are left for another time.)
386 PH/21 Added message_body_newlines option.
388 PH/22 Guard against possible overflow in moan_check_errorcopy().
390 PH/23 POSIX allows open() to be a macro; guard against that.
392 PH/24 If the recipient of an error message contained an @ in the local part
393 (suitably quoted, of course), incorrect values were put in $domain and
394 $local_part during the evaluation of errors_copy.
400 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
401 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
404 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
405 issue a MAIL command.
407 PH/02 In an ACL statement such as
409 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
411 if a client was not listed at all, or was listed with a value other than
412 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
413 the condition was not true (as it should be), so access was not denied.
414 The bug was that the ! inversion was incorrectly passed on to the second
415 item. This has been fixed.
417 PH/03 Added additional dnslists conditions == and =& which are different from
418 = and & when the dns lookup returns more than one IP address.
420 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
421 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
423 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
424 FSYNC, which compiles an option called disable_fsync that allows for
425 bypassing fsync(). The documentation is heavily laced with warnings.
427 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
429 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
430 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
431 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
432 including adding "make clean"; (3) Added -fPIC when compiling the test
433 dynamically loaded module, to get rid of a warning.
435 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
436 message fails, move_frozen_messages = true and ignore_bounce_errors_after
437 = 0s. The bug is otherwise harmless.
439 PH/07 There was a bug in the dovecot authenticator such that the value of
440 $auth1 could be overwritten, and so not correctly preserved, after a
441 successful authentication. This usually meant that the value preserved by
442 the server_setid option was incorrect.
444 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
446 PH/09 Installed PCRE release 7.0.
448 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
449 run for batched SMTP input. It is now run at the start of every message
450 in the batch. While fixing this I discovered that the process information
451 (output by running exiwhat) was not always getting set for -bs and -bS
452 input. This is fixed, and it now also says "batched" for BSMTP.
454 PH/11 Added control=no_pipelining.
456 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
457 patch, slightly modified), and move the expansion of helo_data till after
458 the connection is made in the smtp transport (so it can use these
461 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
463 PH/14 Added log_selector = +pid.
465 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
467 PH/16 Add ${if forany and ${if forall.
469 PH/17 Added dsn_from option to vary the From: line in DSNs.
471 PH/18 Flush SMTP output before performing a callout, unless control =
472 no_callout_flush is set.
474 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
475 was true (the default) a successful delivery failed to delete the retry
476 item, thus causing premature timeout of the address. The bug is now
479 PH/20 Added hosts_avoid_pipelining to the smtp transport.
481 PH/21 Long custom messages for fakedefer and fakereject are now split up
482 into multiline reponses in the same way that messages for "deny" and
483 other ACL rejections are.
485 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
486 with slight modification.
488 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
489 draft, changing the syntax and factoring some duplicate code".
491 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
492 for deliveries of the second and subsequent messages over the same SMTP
495 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
496 ${reduce, with only minor "tidies".
498 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
500 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
501 expansion side effects.
503 PH/27 When a message times out after an over-quota error from an Exim-imposed
504 quota, the bounce message says "mailbox is full". This message was not
505 being given when it was a system quota that was exceeded. It now should
508 MH/03 Made $recipients available in local_scan(). local_scan() already has
509 better access to the recipient list through recipients_list[], but
510 $recipients can be useful in postmaster-provided expansion strings.
512 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
513 in the case of a MAIL command with additional options following the
514 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
515 were accidentally chopped off.
517 PH/29 SMTP synchronization checks are implemented when a command is read -
518 there is a check that no more input is waiting when there shouldn't be
519 any. However, for some commands, a delay in an ACL can mean that it is
520 some time before the response is written. In this time, more input might
521 arrive, invalidly. So now there are extra checks after an ACL has run for
522 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
523 pipelining has not been advertised.
525 PH/30 MH's patch to allow iscntrl() characters to be list separators.
527 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
528 returned in the SMTP response when smtp_return_error_details was false.
531 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
532 instead of the C I/O that was originally supplied, because problems were
535 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
536 Exim which did not show up earlier: it was assuming that a call to
537 SSL_CTX_set_info_callback() might give an error value. In fact, there is
538 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
539 was a macro that became an assignment, so it seemed to work. This has
540 changed to a proper function call with a void return, hence the compile
541 error. Exim's code has been fixed.
543 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
546 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
548 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
550 PH/37 If a message is not accepted after it has had an id assigned (e.g.
551 because it turns out to be too big or there is a timeout) there is no
552 "Completed" line in the log. When some messages of this type were
553 selected by exigrep, they were listed as "not completed". Others were
554 picked up by some special patterns. I have improved the selection
555 criteria to be more general.
557 PH/38 The host_find_failed option in the manualroute router can now be set
558 to "ignore", to completely ignore a host whose IP address cannot be
559 found. If all hosts are ignored, the behaviour is controlled by the new
560 host_all_ignored option.
562 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
563 homing or because of multiple MX records with /mx) generated more than
564 one IP address, and the following item turned out to be the local host,
565 all the secondary addresses of the first item were incorrectly removed
566 from the list, along with the local host and any following hosts (which
567 is what is supposed to happen).
569 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
570 whoever called Exim into the -H file. In the case of the daemon it was
571 behaving confusingly. When first started, it used values for whoever
572 started the daemon, but after a SIGHUP it used the Exim user (because it
573 calls itself on a restart). I have changed the code so that it now always
576 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
577 message are rejected with the same error (e.g. no authentication or bad
578 sender address), and a DATA command is nevertheless sent (as can happen
579 with PIPELINING or a stupid MUA), the error message that was given to the
580 RCPT commands is included in the rejection of the DATA command. This is
581 intended to be helpful for MUAs that show only the final error to their
584 PH/42 Another patch from the Sieve maintainer.
586 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
587 Eximstats - Fixed some broken HTML links and added missing column headers
589 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
590 columns for Rejects, Temp Rejects, Ham, and Spam rows.
592 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
594 PH/43 Yet another patch from the Sieve maintainer.
596 PH/44 I found a way to check for a TCP/IP connection going away before sending
597 the response to the final '.' that terminates a message, but only in the
598 case where the client has not sent further data following the '.'
599 (unfortunately, this is allowed). However, in many cases there won't be
600 any further data because there won't be any more messages to send. A call
601 to select() can be used: if it shows that the input is "ready", there is
602 either input waiting, or the socket has been closed. An attempt to read
603 the next input character can distinguish the two cases. Previously, Exim
604 would have sent an OK response which the client would never have see.
605 This could lead to message repetition. This fix should cure that, at
606 least in a lot of common cases.
608 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
609 advertised in response to EHLO.
615 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
616 fixed by 4.65/MH/01 (is this a record?) are fixed:
618 (i) An empty string was always treated as zero by the numeric comparison
619 operators. This behaviour has been restored.
621 (ii) It is documented that the numeric comparison operators always treat
622 their arguments as decimal numbers. This was broken in that numbers
623 starting with 0 were being interpreted as octal.
625 While fixing these problems I realized that there was another issue that
626 hadn't been noticed. Values of message_size_limit (both the global option
627 and the transport option) were treated as octal if they started with 0.
628 The documentation was vague. These values are now always treated as
629 decimal, and I will make that clear in the documentation.
635 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
636 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
639 MH/01 Don't check that the operands of numeric comparison operators are
640 integers when their expansion is in "skipping" mode (fixes bug
641 introduced by 4.64-PH/07).
643 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
644 child addresses, Exim now panics and dies. Previously, because the count
645 is held in a short int, deliveries were likely to be lost. As such a
646 large number of recipients for a single message is ridiculous
647 (performance will be very, very poor), I have chosen to impose a limit
648 rather than extend the field.
654 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
655 leftover -K file (the existence of which was triggered by #402).
656 While we were at it, introduced process PID as part of the -K
657 filename. This should rule out race conditions when creating
660 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
661 processing considerably. Previous code took too long for large mails,
662 triggering a timeout which in turn triggers #401.
664 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
665 in the DK code in transports.c. sendfile() is not really portable,
666 hence the _LINUX specificness.
668 TF/01 In the add_headers option to the mail command in an Exim filter,
669 there was a bug that Exim would claim a syntax error in any
670 header after the first one which had an odd number of characters
673 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
674 callout verification, Exim cached a "reject" for the entire domain. This
675 is correct for most verifications, but it is not correct for a recipient
676 verification with use_sender or use_postmaster set, because in that case
677 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
678 case of MAIL FROM:<> rejection from other early rejections (e.g.
679 rejection of HELO). When verifying a recipient using a non-null MAIL
680 address, the cache is ignored if it shows MAIL FROM:<> rejection.
681 Whatever the result of the callout, the value of the domain cache is
682 left unchanged (for any other kind of callout, getting as far as trying
683 RCPT means that the domain itself is ok).
685 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
688 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
689 manifest itself as EPIPE rather than ECONNECT. When tidying away a
690 session, the daemon ignores ECONNECT errors and logs others; it now
691 ignores EPIPE as well.
693 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
694 (quoted-printable decoding).
696 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
697 later the small subsequent patch to fix an introduced bug.
699 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
701 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
703 PH/08 An error is now given if message_size_limit is specified negative.
705 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
706 to be given (somewhat) arbitrary names.
708 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
711 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
712 miscellaneous code fixes
714 PH/10 Added the log_reject_target ACL modifier to specify where to log
717 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
718 hostname. This is wrong, because it relates to the incoming message (and
719 probably the interface on which it is arriving) and not to the outgoing
720 callout (which could be using a different interface). This has been
721 changed to use the value of the helo_data option from the smtp transport
722 instead - this is what is used when a message is actually being sent. If
723 there is no remote transport (possible with a router that sets up host
724 addresses), $smtp_active_hostname is used.
726 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
727 tweaks were necessary in order to get it to work (see also 21 below):
728 (a) The code assumed that strncpy() returns a negative number on buffer
729 overflow, which isn't the case. Replaced with Exim's string_format()
731 (b) There were several signed/unsigned issues. I just did the minimum
732 hacking in of casts. There is scope for a larger refactoring.
733 (c) The code used strcasecmp() which is not a standard C function.
734 Replaced with Exim's strcmpic() function.
735 (d) The code set only $1; it now sets $auth1 as well.
736 (e) A simple test gave the error "authentication client didn't specify
737 service in request". It would seem that Dovecot has changed its
738 interface. Fortunately there's a specification; I followed it and
739 changed what the client sends and it appears to be working now.
741 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
744 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
745 address A is aliased to B and C, where B exists and C does not. Without
746 -v the output is "A verified" because verification stops after a
747 successful redirection if more than one address is generated. However,
748 with -v the child addresses are also verified. Exim was outputting "A
749 failed to verify" and then showing the successful verification for C,
750 with its parentage. It now outputs "B failed to verify", showing B's
751 parentage before showing the successful verification of C.
753 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
754 look up a TXT record in a specific list after matching in a combined
757 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
758 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
759 they consult the DNS. I had assumed they would set it the way they
760 wanted; and indeed my experiments on Linux seem to show that in some
761 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
762 To be on the safe side, however, I have now made the interface to
763 host_find_byname() similar to host_find_bydns(), with an argument
764 containing the DNS resolver options. The host_find_byname() function now
765 sets these options at its start, just as host_find_bydns() does. The smtp
766 transport options dns_qualify_single and dns_search_parents are passed to
767 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
768 of host_find_byname() use the default settings of RES_DEFNAMES
769 (qualify_single) but not RES_DNSRCH (search_parents).
771 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
772 spool_read_header() do less string testing, by means of a preliminary
773 switch on the second character of optional "-foo" lines. (This is
774 overdue, caused by the large number of possibilities that now exist.
775 Originally there were few.) While I was there, I also converted the
776 str(n)cmp tests so they don't re-test the leading "-" and the first
777 character, in the hope this might squeeze out yet more improvement.
779 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
780 flag allowing group syntax was set by the header_syntax check but not
781 turned off, possible causing trouble later; (2) The flag was not being
782 set at all for the header_verify test, causing "group"-style headers to
783 be rejected. I have now set it in this case, and also caused header_
784 verify to ignore an empty address taken from a group. While doing this, I
785 came across some other cases where the code for allowing group syntax
786 while scanning a header line wasn't quite right (mostly, not resetting
787 the flag correctly in the right place). These bugs could have caused
788 trouble for malformed header lines. I hope it is now all correct.
790 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
791 with the "reply" argument non-NULL. The code, however (which originally
792 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
793 but it didn't always do it. This confused somebody who was copying the
794 code for some other use. I have removed all the tests.
796 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
797 feature that was used to support insecure browsers during the U.S. crypto
798 embargo. It requires special client support, and Exim is probably the
799 only MTA that supported it -- and would never use it because real RSA is
800 always available. This code has been removed, because it had the bad
801 effect of slowing Exim down by computing (never used) parameters for the
802 RSA_EXPORT functionality.
804 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
805 authenticator to fail if there's a tab character in the incoming data
806 (there should never be unless someone is messing about, as it's supposed
807 to be base64-encoded). Also added, on Timo's advice, the "secured" option
808 if the connection is using TLS or if the remote IP is the same as the
809 local IP, and the "valid-client-cert option" if a client certificate has
812 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
813 authenticators. This can be used for authorization after authentication
814 succeeds. (In the case of plaintext, it servers for both authentication
817 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
818 if any retry times were supplied.
820 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
821 connection, where there is no client IP address to check. In this
822 situation, the verify now always succeeds.
824 PH/25 Applied John Jetmore's -Mset patch.
826 PH/26 Added -bem to be like -Mset, but loading a message from a file.
828 PH/27 In a string expansion for a processed (not raw) header when multiple
829 headers of the same name were present, leading whitespace was being
830 removed from all of them, but trailing whitespace was being removed only
831 from the last one. Now trailing whitespace is removed from each header
832 before concatenation. Completely empty headers in a concatenation (as
835 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
836 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
838 PH/29 [Removed. This was a change that I later backed out, and forgot to
839 correct the ChangeLog entry (that I had efficiently created) before
840 committing the later change.]
842 PH/30 Exim was sometimes attempting to deliver messages that had suffered
843 address errors (4xx response to RCPT) over the same connection as other
844 messages routed to the same hosts. Such deliveries are always "forced",
845 so retry times are not inspected. This resulted in far too many retries
846 for the affected addresses. The effect occurred only when there were more
847 hosts than the hosts_max_try setting in the smtp transport when it had
848 the 4xx errors. Those hosts that it had tried were not added to the list
849 of hosts for which the message was waiting, so if all were tried, there
850 was no problem. Two fixes have been applied:
852 (i) If there are any address or message errors in an SMTP delivery, none
853 of the hosts (tried or untried) are now added to the list of hosts
854 for which the message is waiting, so the message should not be a
855 candidate for sending over the same connection that was used for a
856 successful delivery of some other message. This seems entirely
857 reasonable: after all the message is NOT "waiting for some host".
858 This is so "obvious" that I'm not sure why it wasn't done
859 previously. Hope I haven't missed anything, but it can't do any
860 harm, as the worst effect is to miss an optimization.
862 (ii) If, despite (i), such a delivery is accidentally attempted, the
863 routing retry time is respected, so at least it doesn't keep
864 hammering the server.
866 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
867 in ${readsocket because some servers need this prod.
869 PH/32 Added some extra debug output when updating a wait-xxx database.
871 PH/33 The hint "could be header name not terminated by colon", which has been
872 given for certain expansion errors for a long time, was not being given
873 for the ${if def:h_colon_omitted{... case.
875 PH/34 The spec says: "With one important exception, whenever a domain list is
876 being scanned, $domain contains the subject domain." There was at least
877 one case where this was not true.
879 PH/35 The error "getsockname() failed: connection reset by peer" was being
880 written to the panic log as well as the main log, but it isn't really
881 panic-worthy as it just means the connection died rather early on. I have
882 removed the panic log writing for the ECONNRESET error when getsockname()
885 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
886 runs only) independently of the message's sender address. This meant
887 that, if the 4xx error was in fact related to the sender, a different
888 message to the same recipient with a different sender could confuse
889 things. In particualar, this can happen when sending to a greylisting
890 server, but other circumstances could also provoke similar problems.
891 I have changed the default so that the retry time for these errors is now
892 based a combination of the sender and recipient addresses. This change
893 can be overridden by setting address_retry_include_sender=false in the
896 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
897 remote server are returned as part of bounce messages. This was not
898 happening for LMTP over a pipe (the lmtp transport), but now it is the
899 same for both kinds of LMTP.
901 PH/38 Despite being documented as not happening, Exim was rewriting addresses
902 in header lines that were in fact CNAMEs. This is no longer the case.
904 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
905 and queue runs started by the daemon processed all messages. This has
906 been fixed so that -R and -S can now usefully be given with -q<time>.
908 PH/40 Import PCRE release 6.7 (fixes some bugs).
910 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
912 PH/42 Give an error if -q is specified more than once.
914 PH/43 Renamed the variables $interface_address and $interface_port as
915 $received_ip_address and $received_port, to make it clear that these
916 values apply to message reception, and not to the outgoing interface when
917 a message is delivered. (The old names remain recognized, of course.)
919 PH/44 There was no timeout on the connect() call when using a Unix domain
920 socket in the ${readsocket expansion. There now is.
922 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
923 be meaningful with "accept".
925 SC/01 Eximstats V1.43
926 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
928 SC/02 Eximstats V1.44
929 Use a glob alias rather than an array ref in the generated
930 parser. This improves both readability and performance.
932 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
933 Collect SpamAssassin and rejection statistics.
934 Don't display local sender or destination tables unless
935 there is data to show.
936 Added average volumes into the top table text output.
938 SC/04 Eximstats V1.46
939 Collect data on the number of addresses (recipients)
940 as well as the number of messages.
942 SC/05 Eximstats V1.47
943 Added 'Message too big' to the list of mail rejection
944 reasons (thanks to Marco Gaiarin).
946 SC/06 Eximstats V1.48
947 Mainlog lines which have GMT offsets and are too short to
948 have a flag are now skipped.
950 SC/07 Eximstats V1.49 (Alain Williams)
951 Added the -emptyok flag.
953 SC/08 Eximstats V1.50
954 Fixes for obtaining the IP address from reject messages.
956 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
957 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
958 whitesspace changes from 4.64-PH/27
960 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
963 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
964 are found, allow negative numbers in numeric criteria)
966 JJ/06 exipick.20061117.2, added new $message_body_missing variable
968 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
969 to match changes made in 4.64-PH/43
971 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
973 PH/47 Put in an explicit test for a DNS lookup of an address record where the
974 "domain" is actually an IP address, and force a failure. This locks out
975 those revolvers/nameservers that support "A-for-A" lookups, in
976 contravention of the specifications.
978 PH/48 When a host name was looked up from an IP address, and the subsequent
979 forward lookup of the name timed out, the host name was left in
980 $sender_host_name, contrary to the specification.
982 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
983 restricted to single-key lookups, Exim was not diagnosing an error if
984 * or *@ was used with a query-style lookup.
986 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
988 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
989 long ago, but noone interested enough thought of it. Let's just say that
990 the "1.1" means that there are some new functions that weren't there at
991 some point in the past.
993 PH/51 Error processing for expansion failure of helo_data from an smtp
994 transport during callout processing was broken.
996 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
997 tested/used via the -bh/-bhc/-bs options.
999 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1000 bug, fixed in subsequent PCRE releases).
1002 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1003 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1005 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1011 SC/01 Use a glob alias rather than an array ref in eximstats generated
1012 parser. This improves both readability and performance.
1014 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1015 Don't display local sender or destination tables in eximstats unless
1016 there is data to show.
1017 Added average volumes into the eximstats top table text output.
1019 SC/03 Collect data on the number of addresses (recipients) as well
1020 as the number of messages in eximstats.
1022 TF/01 Correct an error in the documentation for the redirect router. Exim
1023 does (usually) call initgroups() when daemonizing.
1025 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1026 with consistent privilege compared to when running as a daemon.
1028 TF/03 Note in the spec that $authenticated_id is not set for local
1029 submissions from trusted users.
1031 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1032 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1034 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1035 by adding some example configuration directives to the default
1036 configuration file. A little bit of work is required to uncomment the
1037 directives and define how usernames and passwords are checked, but
1038 there is now a framework to start from.
1040 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1041 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1042 without this. I don't know how relevant this is to other LDAP libraries.
1044 PH/02 Add the verb name to the "unknown ACL verb" error.
1046 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1048 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1050 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1051 directory not expanded when it should be if an expanded home directory
1052 was set for the address (which is overridden by the transport).
1054 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1057 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1058 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1059 because it is too late at that time, and has no effect.
1061 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1062 security issue with \' (bugzilla #107). I could not use the
1063 PQescapeStringConn() function, because it needs a PGconn value as one of
1066 PH/08 When testing addresses using -bt, indicate those final addresses that
1067 are duplicates that would not cause an additional delivery. At least one
1068 person was confused, thinking that -bt output corresponded to deliveries.
1069 (Suppressing duplicates isn't a good idea as you lose the information
1070 about possibly different redirections that led to the duplicates.)
1072 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1073 systems where poll() doesn't work, in particular OS X.
1075 PH/10 Added more information to debugging output for retry time not reached.
1077 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1078 operations in malware.c.
1080 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1083 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1084 syslog_duplication set false, log lines that would normally be written
1085 both the the main log and to the reject log were not written to syslog at
1088 PH/14 In the default configuration, change the use of "message" in ACL warn
1089 statements to "add_header".
1091 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1092 not followed by a command (e.g. "seen endif").
1094 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1095 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1098 PH/17 Added extra conditions to the default value of delay_warning_condition
1102 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1103 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1104 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1107 The Auto-Submitted: and various List- headers are standardised, whereas I
1108 don't think Precedence: ever was.
1110 PH/18 Refactored debugging code in route_finduser() to show more information,
1111 in particular, the error code if getpwnam() issues one.
1113 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1114 This is apparently needed in addition to the PH/07 change above to avoid
1115 any possible encoding problems.
1117 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1118 but not after initializing Perl.
1120 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1121 output them only if debugging. By default they are written stderr,
1122 apparently, which is not desirable.
1124 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1127 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1130 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1132 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1133 authenticated or an ident call has been made. Suppress the default
1134 values for $authenticated_id and $authenticated_sender (but permit -oMai
1135 and -oMas) when testing with -bh.
1137 PH/24 Re-jigged the order of the tests in the default configuration so that the
1138 tests for valid domains and recipients precede the DNS black list and CSA
1139 tests, on the grounds that those ones are more expensive.
1141 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1142 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1143 This bug exists in every version of Exim that I still have, right back to
1146 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1147 However, an attempt to turn on case-sensitivity in a regex key by
1148 including (?-i) didn't work because the subject string was already
1149 lowercased, and the effects were non-intuitive. It turns out that a
1150 one-line patch can be used to allow (?-i) to work as expected.
1156 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1157 other effects) broke the use of negated acl sub-conditions.
1159 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1162 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1163 "Deny" causes Exim to reject the incoming connection with a 554 error.
1164 Unfortunately, if there is a major crisis, such as a disk failure,
1165 tcp-wrappers gives "deny", whereas what one would like would be some
1166 kind of temporary error. A kludge has been added to help with this.
1167 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1168 554 error is used if errno is still zero or contains ENOENT (which occurs
1169 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1172 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1174 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1175 errors. Otherwise a message that provokes a temporary error (when other
1176 messages do not) can cause a whole host to time out.
1178 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1179 addresses were routed directly to files or pipes from a redirect router.
1180 File deliveries just didn't batch; pipe deliveries might have suffered
1183 PH/06 A failure to get a lock for a hints database would erroneously always say
1184 "Failed to get write lock", even when it was really a read lock.
1186 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1187 of 0600. This has been changed to use the value of the lockfile_mode
1188 option (which defaults to 0600).
1190 PH/08 Applied small patch from the Sieve maintainer.
1192 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1193 folder from quota calculations, a direct delivery into this folder messed
1194 up the contents of the maildirsize file. This was because the regex was
1195 used only to exclude .Trash (or whatever) when the size of the mailbox
1196 was calculated. There was no check that a delivery was happening into an
1197 excluded directory. This bug has been fixed by ignoring all quota
1198 processing for deliveries into excluded directories.
1200 PH/10 Added the maildirfolder_create_regex option to appendfile.
1206 PH/01 The code for finding all the local interface addresses on a FreeBSD
1207 system running IPv6 was broken. This may well have applied to all BSD
1208 systems, as well as to others that have similar system calls. The broken
1209 code found IPv4 interfaces correctly, but gave incorrect values for the
1210 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1211 that it would not match correctly against @[] and not recognize the IPv6
1214 PH/02 The ipliteral router was not recognizing addresses of the form user@
1215 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1217 PH/03 Added disable_ipv6.
1219 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1220 decoded form, because it is most often used to construct To: headers
1221 lines in autoreplies, and the decoded form may well be syntactically
1222 invalid. However, $reply_address has leading white space removed, and all
1223 newlines turned into spaces so that the autoreply transport does not
1226 PH/05 If group was specified without a user on a router, and no group or user
1227 was specified on a transport, the group from the router was ignored.
1229 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1230 for visible compile-time settings that can be used to change these
1231 numbers, for those that want even more. Backwards compatibility with old
1232 spool files has been maintained. However, going back to a previous Exim
1233 release will lost any variables that are in spool files.
1235 PH/07 Two small changes when running in the test harness: increase delay when
1236 passing a TCP/IP connection to a new process, in case the original
1237 process has to generate a bounce, and remove special handling of
1238 127.0.0.2 (sic), which is no longer necessary.
1240 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1241 be the same on different OS.
1243 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1246 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1247 whether --show-vars was specified or not
1249 JJ/02 exipick: Added support for new ACL variable spool format introduced
1252 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1253 syntactically invalid From: or Reply-to: line, and a filter used this to
1254 generate an autoreply, and therefore failed to obtain an address for the
1255 autoreply, Exim could try to deliver to a non-existent relative file
1256 name, causing unrelated and misleading errors. What now happens is that
1257 it logs this as a hard delivery error, but does not attempt to create a
1260 PH/11 The exinext utility has a -C option for testing purposes, but although
1261 the given file was scanned by exinext itself; it wasn't being passed on
1262 when Exim was called.
1264 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1265 an end-of-file indication when reading a command response.
1267 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1268 compiled. In many other places in Exim, IPv6 addresses are always
1269 recognized, so I have changed this. It also means that IPv4 domain
1270 literals of the form [IPV4:n.n.n.n] are now always recognized.
1272 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1273 used if the router is not running as root, for example, when verifying at
1274 ACL time, or when using -bh. The debugging output from this situation was
1275 non-existent - all you got was a failure to exec. I have made two
1278 (a) Failures to set uid/gid, the current directory, or a process leader
1279 in a subprocess such as that created by queryprogram now generate
1280 suitable debugging ouput when -d is set.
1282 (b) The queryprogram router detects when it is not running as root,
1283 outputs suitable debugging information if -d is set, and then runs
1284 the subprocess without attempting to change uid/gid.
1286 PH/15 Minor change to Makefile for building test_host (undocumented testing
1289 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1290 additional section of a DNS packet that returns MX or SRV records.
1291 Instead, it always explicitly searches for A/AAAA records. This avoids
1292 major problems that occur when a DNS server includes only records of one
1293 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1294 fixed another bug: if SRV records were looked up and the corresponding
1295 address records were *not* found in the additional section, the port
1296 values from the SRV records were lost.
1298 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1299 using the correct key (the original address) when searching the retry
1300 rules in order to find which one to use for generating the retry hint.
1302 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1303 adding the default one. Similarly, if it contains a Reply-To: header, the
1304 errors_reply_to option, if set, is not used.
1306 PH/19 When calculating a retry time, Exim used to measure the "time since
1307 failure" by looking at the "first failed" field in the retry record. Now
1308 it does not use this if it is later than than the arrival time of the
1309 message. Instead it uses the arrival time. This makes for better
1310 behaviour in cases where some deliveries succeed, thus re-setting the
1311 "first failed" field. An example is a quota failure for a huge message
1312 when small messages continue to be delivered. Without this change, the
1313 "time since failure" will always be short, possible causing more frequent
1314 delivery attempts for the huge message than are intended.
1315 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1317 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1318 $1, $2, $3) because the numerical variables can be reset during some
1319 expansion items (e.g. "match"), thereby losing the authentication data.
1321 PH/21 Make -bV show the size of off_t variables so that the test suite can
1322 decide whether to run tests for quotas > 2G.
1324 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1325 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1326 filecount value is greater than 2G or if a quota value is greater than 2G
1327 on a system where the size of off_t is not greater than 4, a panic error
1330 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1331 never match. The debug and -bh output now contains an explicit error
1332 message indicating a malformed IPv4 address or mask.
1334 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1335 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1336 PH/23 above applies.
1338 PH/25 Do not write to syslog when running in the test harness. The only
1339 occasion when this arises is a failure to open the main or panic logs
1340 (for which there is an explicit test).
1342 PH/26 Added the /no_tell option to "control=freeze".
1344 PH/27 If a host name lookup failed very early in a connection, for example, if
1345 the IP address matched host_lookup and the reverse lookup yielded a name
1346 that did not have a forward lookup, an error message of the form "no IP
1347 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1348 could be logged. Now it outputs the IP address instead of "NULL".
1350 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1351 allows the sender and the authenticated sender to be set when
1352 submitting a message from within Exim. Since child_open_exim() is
1353 documented for local_scan(), the new function should be too.
1355 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1356 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1357 results in an empty string is now treated as unset.
1359 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1361 PH/31 Added sender_verify_fail logging option.
1363 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1364 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1365 tidied the source and removed it altogether.
1367 PH/33 When a queue run was abandoned because the load average was too high, a
1368 log line was always written; now it is written only if the queue_run log
1369 selector is set. In addition, the log line for abandonment now contains
1370 information about the queue run such as the pid. This is always present
1371 in "start" and "stop" lines but was omitted from the "abandon" line.
1373 PH/34 Omit spaces between a header name and the colon in the error message that
1374 is given when verify = headers_syntax fails (if there are lots of them,
1375 the message gets confusing).
1377 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1378 names, as there are now some PTR records that contain slashes. This check
1379 is only to protect against broken name servers that fall over on strange
1380 characters, so the fact that it applies to all lookups doesn't matter.
1382 PH/36 Now that the new test suite is complete, we can remove some of the
1383 special code in Exim that was needed for the old test suite. For example,
1384 sorting DNS records because real resolvers return them in an arbitrary
1385 order. The new test suite's fake resolver always returns records in the
1388 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1389 bounces) except when queue_only is set, to avoid logging races between
1390 the different processes.
1392 PH/38 Panic-die if .include specifies a non-absolute path.
1394 PH/39 A tweak to the "H" retry rule from its user.
1396 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1397 a label. They prevented compilation on older perls.
1399 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1400 a warning to be raised on newish perls.
1402 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1403 on queue. Changes to match documented behaviour of showing count of
1404 messages matching specified criteria.
1406 PH/40 Changed the default ident timeout from 30s to 5s.
1408 PH/41 Added support for the use of login_cap features, on those BSD systems
1409 that have them, for controlling the resources used by pipe deliveries.
1411 PH/42 The content-scanning code uses fopen() to create files in which to put
1412 message data. Previously it was not paying any attention to the mode of
1413 the files. Exim runs with umask(0) because the rest of the code creates
1414 files with open(), and sets the required mode explicitly. Thus, these
1415 files were ending up world-writeable. This was not a big issue, because,
1416 being within the spool directory, they were not world-accessible. I have
1417 created a function called modefopen, which takes an additional mode
1418 argument. It sets umask(777), creates the file, chmods it to the required
1419 mode, then resets the umask. All the relevant calls to fopen() in the
1420 content scanning code have been changed to use this function.
1422 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1423 to 24 hours. This avoids potential overflow problems when processing G
1424 and H retry rules. I suspect nobody ever tinkers with this value.
1426 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1428 PH/45 When the plaintext authenticator is running as a client, the server's
1429 challenges are checked to ensure they are valid base64 strings. By
1430 default, the authentication attempt is cancelled if an invalid string is
1431 received. Setting client_ignore_invalid_base64 true ignores these errors.
1432 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1433 they are received. Thus, the responses can be made to depend on the
1434 challenges. If an invalid string is ignored, an empty string is placed in
1437 PH/46 Messages that are created by the autoreply transport now contains a
1438 References: header, in accordance with RFCs 2822 and 3834.
1440 PH/47 Added authenticated_sender_force to the smtp transport.
1442 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1444 PH/49 Installed latest patch from the Sieve maintainer.
1446 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1447 was also set, the appendfile transport was unnecessarily scanning a
1448 directory of message files (e.g. for maildir delivery) to find the count
1449 of files (along with the size), even though it did not need this
1450 information. It now does the scan only if it needs to find either the
1451 size of the count of files.
1453 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1455 PH/52 Two bugs concerned with error handling when the smtp transport is
1458 (i) Exim was not creating retry information for temporary errors given
1459 for individual recipients after the DATA command when the smtp transport
1460 was used in LMTP mode. This meant that they could be retried too
1461 frequently, and not timed out correctly.
1463 (ii) Exim was setting the flag that allows error details to be returned
1464 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1465 recipients that were returned after the DATA command.
1467 PH/53 This is related to PH/52, but is more general: for any failing address,
1468 when detailed error information was permitted to be returned to the
1469 sender, but the error was temporary, then after the final timeout, only
1470 "retry timeout exceeded" was returned. Now it returns the full error as
1471 well as "retry timeout exceeded".
1473 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1474 do this, and (what is worse) MTAs that accept it.
1476 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1477 will now be deprecated.
1479 PH/56 New os.c-cygwin from the Cygwin maintainer.
1481 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1482 formats (previously only available in exim formats via -bpr, -bpru,
1483 and -bpra. Now also available in native and exiqgrep formats)
1485 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1486 with very large, slow to parse queues
1488 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1490 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1492 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1493 responses to authentication challenges, though it was showing the
1494 challenges; (ii) I've removed the CR characters from the debug output for
1497 PH/58 Allow for the insertion of a newline as well as a space when a string
1498 is turned into more than one encoded-word during RFC 2047 encoding. The
1499 Sieve code now uses this.
1501 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1502 data_4xx, lost_connection, tls_required.
1504 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1505 message was being sent as an SMTP response.
1507 PH/61 Add -l and -k options to exicyclog.
1509 PH/62 When verifying, if an address was redirected to one new address, so that
1510 verification continued, and the new address failed or deferred after
1511 having set something in $address_data, the value of $address_data was not
1512 passed back to the ACL. This was different to the case when no
1513 redirection occurred. The value is now passed back in both cases.
1515 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1516 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1517 use login_cap.h, so on its own it isn't the distinguishing feature. The
1518 new name refers directly to the setclassresources() function.
1520 PH/65 Added configuration files for NetBSD3.
1522 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1524 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1525 is preferred over IPv4.
1527 PH/68 The bounce_return_message and bounce_return_body options were not being
1528 honoured for bounces generated during the reception of non-SMTP messages.
1529 In particular, this applied to messages rejected by the ACL. This bug has
1530 been fixed. However, if bounce_return_message is true and bounce_return_
1531 body is false, the headers that are returned for a non-SMTP message
1532 include only those that have been read before the error was detected.
1533 (In the case of an ACL rejection, they have all been read.)
1535 PH/69 The HTML version of the specification is now built in a directory called
1536 spec_html instead of spec.html, because the latter looks like a path with
1537 a MIME-type, and this confuses some software.
1539 PH/70 Catch two compiler warnings in sieve.c.
1541 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1542 function verify_get_ident() calls ip_connect() to connect a socket, but
1543 if the "connect()" function timed out, ip_connect() used to close the
1544 socket. However, verify_get_ident() also closes the socket later, and in
1545 between Exim writes to the log, which may get opened at this point. When
1546 the socket was closed in ip_connect(), the log could get the same file
1547 descriptor number as the socket. This naturally causes chaos. The fix is
1548 not to close the socket in ip_connect(); the socket should be closed by
1549 the function that creates it. There was only one place in the code where
1550 this was missing, in the iplookup router, which I don't think anybody now
1551 uses, but I've fixed it anyway.
1553 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1554 well as to direct DNS lookups. Otherwise the handling of names in host
1555 lists is inconsistent and therefore confusing.
1561 PH/01 Two changes to the default runtime configuration:
1563 (1) Move the checks for relay_from_hosts and authenticated clients from
1564 after to before the (commented out) DNS black list checks.
1566 (2) Add control=submission to the relay_from_hosts and authenticated
1567 clients checks, on the grounds that messages accepted by these
1568 statements are most likely to be submissions.
1570 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1572 (1) Generate an error if the third argument for the ${prvs expansion is
1575 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1578 (3) Reset the variables that are obtained from the first argument of
1579 ${prvscheck and used in the second argument before leaving the code,
1580 because their memory is reclaimed, so using them afterwards may do
1583 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1584 one (it's much easier than Tom thought :-).
1586 (5) Because of (4), we can now allow for the use of $prvscheck_result
1587 inside the third argument.
1589 PH/03 For some reason, the default setting of PATH when running a command from
1590 a pipe transport was just "/usr/bin". I have changed it to
1593 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1594 anything to be listed in the output from -bV.
1596 PH/05 When a filter generated an autoreply, the entire To: header line was
1597 quoted in the delivery log line, like this:
1599 => >A.N.Other <ano@some.domain> <original@ddress> ...
1601 This has been changed so that it extracts the operative address. There
1602 may be more than one such address. If so, they are comma-separated, like
1605 => >ano@some.domain,ona@other.domain <original@ddress> ...
1607 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1608 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1609 not being looked up in the rDNS to get a host name, Exim was showing the
1610 IP address twice in Received: lines, even though the IP addresses were
1611 identical. For example:
1613 Received: from [1.2.3.4] (helo=[1.2.3.4])
1615 However, if the real host name was known, it was omitting the HELO data
1616 if it matched the actual IP address. This has been tidied up so that it
1617 doesn't show the same IP address twice.
1619 PH/07 When both +timestamp and +memory debugging was on, the value given by
1620 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1621 called by the debug printing, thereby overwriting the timestamp buffer.
1622 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1624 PH/08 When the original message was included in an autoreply transport, it
1625 always said "this is a copy of the message, including all the headers",
1626 even if body_only or headers_only was set. It now gives an appropriate
1629 PH/09 Applied a patch from the Sieve maintainer which:
1631 o fixes some comments
1632 o adds the (disabled) notify extension core
1633 o adds some debug output for the result of if/elsif tests
1634 o points to the current vacation draft in the documentation
1635 and documents the missing references header update
1639 o fixes a bug in processing the envelope test (when testing
1640 multiple envelope elements, the last element determinted the
1643 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1644 Electronic Mail") by including:
1646 Auto-submitted: auto-generated
1648 in the messages that it generates (bounce messages and others, such as
1649 warnings). In the case of bounce messages for non-SMTP mesages, there was
1650 also a typo: it was using "Auto_submitted" (underscore instead of
1651 hyphen). Since every message generated by Exim is necessarily in response
1652 to another message, thes have all been changed to:
1654 Auto-Submitted: auto-replied
1656 in accordance with these statements in the RFC:
1658 The auto-replied keyword:
1660 - SHOULD be used on messages sent in direct response to another
1661 message by an automatic process,
1663 - MUST NOT be used on manually-generated messages,
1665 - MAY be used on Delivery Status Notifications (DSNs) and Message
1666 Disposition Notifications (MDNs),
1668 - MUST NOT be used on messages generated by automatic or periodic
1669 processes, except for messages which are automatic responses to
1672 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1673 to the default Received: header definition.
1675 PH/12 Added log selector acl_warn_skipped (default on).
1677 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1678 variables because (a) they are in the wrong storage pool and (b) even if
1679 they were copied, it wouldn't work properly because of the caching.
1681 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1682 checking when decoding. Apparently there are clients that generate
1683 overlong encoded strings. Why am I not surprised?
1685 PH/15 If the first argument of "${if match_address" was not empty, but did not
1686 contain an "@" character, Exim crashed. Now it writes a panic log message
1687 and treats the condition as false.
1689 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1691 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1692 extension "envelope-auth". The code is finished and in agreement with
1693 other implementations, but there is no documentation so far and in fact,
1694 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1695 not changing the active code.
1697 Print executed "if" and "elsif" statements when debugging is used. This
1698 helps a great deal to understand what a filter does.
1700 Document more things not specified clearly in RFC3028. I had all this
1701 sorted out, when out of a sudden new issues came to my mind. Oops."
1703 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1706 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1707 canonical form (as documented). However, after a host name lookup from
1708 the IP address, check_host() was doing a simple string comparison with
1709 addresses acquired from the DNS when checking that the found name did
1710 have the original IP as one of its addresses. Since any found IPv6
1711 addresses are likely to be in abbreviated form, the comparison could
1712 fail. Luckily, there already exists a function for doing the comparison
1713 by converting both addresses to binary, so now that is used instead of
1714 the text comparison.
1716 PH/20 There was another similar case to PH/19, when a complete host name was
1717 given in a host list; looking up its IP address could give an abbreviated
1718 form, whereas the current host's name might or might not be abbreviated.
1719 The same fix has been applied.
1725 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1726 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1729 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1730 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1732 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1734 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1735 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1736 Also, for "Auto-Submitted", treat anything other than "no" as
1737 non-personal, in accordance with RFC 3834. (Previously it treated
1738 anything starting "auto-" as non-personal.)
1740 TF/01 The control=submission/name=... option had a problem with syntax
1741 errors if the name included a slash character. The /name= option
1742 now slurps the rest of the string, so it can include any characters
1743 but it must come last in the list of options (after /sender_retain
1746 PH/05 Some modifications to the interface to the fake nameserver for the new
1754 TK/01 Added the "success_on_redirect" address verification option. See
1755 NewStuff for rationale and an example.
1757 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1759 PH/02 Patch to exigrep to allow it to work on syslog lines.
1761 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1762 fread() to skip over the body file's header line, because in Cygwin the
1763 header line is locked and is inaccessible.
1765 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1766 co-exist for some time) to make it clear that it is the Exim ID that is
1767 referenced, not the Message-ID: header line.
1769 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1770 string_format() function, because snprintf() does not exist on all
1773 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1774 ${stat: expansion item.
1776 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1777 protocol synchronization error", to keep the pedants happy.
1779 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1780 well as for IRIX systems, when gcc is being used. See the host.c source
1783 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1785 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1788 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1789 transport and to the smtp transport in LMTP mode.
1791 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1793 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1794 run a filter in a subprocess. This could lead to confusion in subsequent
1795 lookups in the parent process. There should also be a search_tidyup() at
1796 the end of the subprocess.
1798 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1799 only if the host matched helo_try_verify_hosts, which caused the
1800 verification to occur when the EHLO/HELO command was issued. The ACL just
1801 tested the remembered result. Now, if a previous verification attempt has
1802 not happened, "verify = helo" does it there and then.
1804 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1806 TK/03 Fix log output including CR from clamd.
1808 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1809 exist provoked a memory error which could cause a segfault.
1811 PH/15 Installed PCRE 6.2
1813 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1815 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1816 of the problem. Specifically, suggested +O2 rather than +O1 for the
1819 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1821 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1822 record, the port was not correctly logged when the outgoing_port log
1823 selector was set (it logged the transort's default port).
1825 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1826 fallback_hosts, and "hosts" in the smtp transport.
1828 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1829 host errors such as "Connection refused".
1831 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1832 authentication with radiusclient 0.4.9:
1834 - Error returned from rc_read_config was caught wrongly
1835 - Username/password not passed on to radius server due to wrong length.
1837 The presumption is that some radiusclient API changes for 4.51/PH/17
1838 were not taken care of correctly. The code is still untested by me (my
1839 Linux distribution still has 0.3.2 of radiusclient), but it was
1840 contributed by a Radius user.
1842 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1843 expanding the "port" option of the smtp transport.
1845 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1846 while reading a MIME header. Thanks to Tom Hughes for a patch.
1848 PH/24 Include config.h inside local_scan.h so that configuration settings are
1851 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1852 that in an ACL for RCPT (for example), you can examine exactly what was
1855 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1856 commands, but it was not correctly comparing the address with the actual
1857 client host address. Thus, it would show the EHLO address in Received:
1858 header lines when this was not necessary.
1860 PH/27 Added the % operator to ${eval:}.
1862 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1863 it should be ignoring failures (because with -C, for example, it has lost
1864 privilege). It wasn't ignoring creation failures other than "already
1867 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1870 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1871 because an input line was too long, either on its own, or by virtue of
1872 too many continuations, the temporary file was not being removed, and the
1873 return code was incorrect.
1875 PH/31 Missing "BOOL" in function definition in filtertest.c.
1877 PH/32 Applied Sieve patches from the maintainer.
1879 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1881 PH/33 Added "verify = not_blind".
1883 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1884 Local/Makefile (with some defaults set). These are used in built scripts
1885 such as exicyclog, but they have never been used in the exim_install
1886 script (though there are many overriding facilities there). I have
1887 arranged that the exim_install script now takes note of these two
1890 PH/35 Installed configuration files for Dragonfly.
1892 PH/36 When a locally submitted message by a trusted user did not contain a
1893 From: header, and the sender address was obtained from -f or from an SMTP
1894 MAIL command, and the trusted user did not use -F to supply a sender
1895 name, $originator_name was incorrectly used when constructing a From:
1896 header. Furthermore, $originator_name was used for submission mode
1897 messages from external hosts without From: headers in a similar way,
1898 which is clearly wrong.
1900 PH/37 Added control=suppress_local_fixups.
1902 PH/38 When log_selector = +received_sender was set, and the addition of the
1903 sender made the log line's construction buffer exactly full, or one byte
1904 less than full, an overflow happened when the terminating "\n" was
1907 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1908 when the result of a list match is failure because a DNS lookup failed.
1910 PH/40 RM_COMMAND is now used in the building process.
1912 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1913 the "build-* directories that it finds.
1915 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1916 address in a domain literal was a prefix of an interface address.
1918 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1919 when verifying a sender address, unless rewrite_headers is false.
1921 PH/44 Wrote a long comment about why errors_to addresses are verified as
1922 recipients, not senders.
1924 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1925 the ratelimit ACL was added.
1927 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1929 PH/46 Added extra information about PostgreSQL errors to the error string.
1931 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1932 avoiding the need to install special zones in a real server. This is
1933 backwards compatible; if it can't find the fake resolver, it drops back.
1934 Thus, both old and new test suites can be run.
1936 TF/02 Added util/ratelimit.pl
1938 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1939 clock is set back in time.
1941 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1942 Candler <B.Candler@pobox.com>.
1944 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1945 OK for addresses that are the result of redirections.
1947 PH/48 A number of further additions for the benefit of the new test suite,
1948 including a fake gethostbyname() that interfaces to the fake DNS resolver
1951 TF/06 The fix for widen_domains has also been applied to qualify_single and
1952 search_parents which are the other dnslookup options that can cause
1955 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1958 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1960 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1961 service" if it encountered an MX record with an empty target hostname.
1962 The message is now "an MX or SRV record indicated no SMTP service".
1964 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1965 if the DNS of the sending site is misconfigured. This is quite a
1966 common situation. This change restores the behaviour of treating a
1967 helo verification defer as a failure.
1969 PH/51 If self=fail was set on a router, the bounce message did not include the
1970 actual error message.
1976 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1978 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1979 command itself ended in error, the underlying message about the transport
1980 filter timeout was being overwritten with the pipe command error. Now the
1981 underlying error message should be appended to the second error message.
1983 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1985 PH/02 Reduce the amount of output that "make" produces by default. Full output
1986 can still be requested.
1988 PH/03 The warning log line about a condition test deferring for a "warn" verb
1989 was being output only once per connection, rather than after each
1990 occurrence (because it was using the same function as for successful
1991 "warn" verbs). This seems wrong, so I have changed it.
1993 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1994 it should not have, which might have caused a crash in the right
1995 circumstances, but probably never did.
1997 PH/04 Installed a modified version of Tony Finch's patch to make submission
1998 mode fix the return path as well as the Sender: header line, and to
1999 add a /name= option so that you can make the user's friendly name appear
2002 TF/03 Added the control = fakedefer ACL modifier.
2004 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2005 Mark Lowes for thorough testing.
2007 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2009 TK/03 Merged latest SRS patch from Miles Wilton.
2011 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2012 with the definition in sysexits.h (which is #included earlier).
2013 Fortunately, Exim does not actually use EX_OK. The code used to try to
2014 preserve the sysexits.h value, by assumimg that macro definitions were
2015 scanned for macro replacements. I have been disabused of this notion,
2016 so now the code just undefines EX_OK before #including unistd.h.
2018 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2019 in the smtp transport. When a block could not be written in a single
2020 write() function, the timeout was being re-applied to each part-write.
2021 This seems wrong - if the receiver was accepting one byte at a time it
2022 would take for ever. The timeout is now adjusted when this happens. It
2023 doesn't have to be particularly precise.
2025 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2026 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2028 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2029 without a domain if the check to <postmaster@domain> fails.
2031 SC/01 Eximstats: added -xls and the ability to specify output files
2032 (patch written by Frank Heydlauf).
2034 SC/02 Eximstats: use FileHandles for outputing results.
2036 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2038 SC/04 Eximstats: fixed display of large numbers with -nvr option
2040 SC/05 Eximstats: fixed merging of reports with empty tables.
2042 SC/06 Eximstats: added the -include_original_destination flag
2044 SC/07 Eximstats: removed tabs and trailing whitespace.
2046 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2048 TK/06 MBOX spool code: Add real "From " MBOX separator line
2049 so the .eml file is really in mbox format (even though
2050 most programs do not really care). Patch from Alex Miller.
2052 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2053 The latter is generated from $received_to and is only set if the
2054 message has one envelope recipient. SA can use these headers,
2055 obviously out-of-the-box. Patch from Alex Miller.
2057 PH/08 The ${def test on a variable was returning false if the variable's
2058 value was "0", contrary to what the specification has always said!
2059 The result should be true unless the variable is empty.
2061 PH/09 The syntax error of a character other than { following "${if
2062 def:variable_name" (after optional whitespace) was not being diagnosed.
2063 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2064 accidental colon was present, for example, could give incorrect results.
2066 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2067 result is used (not including appendfile, where other changes are about
2070 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2071 This involved changing a lot of size variables from int to off_t. It
2072 should work with maildirs and everything.
2074 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2075 spamd dying while we are connected to it.
2077 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2080 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2081 function for BDB 4.3.
2083 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2085 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2086 the sources for PCRE have been re-arranged and more files are now
2089 PH/15 The code I had for printing potentially long long variables in PH/11
2090 above was not the best (it lost precision). The length of off_t variables
2091 is now inspected at build time, and an appropriate printing format (%ld
2092 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2093 to be "long long int" or "long int". This is needed for the internal
2094 formatting function string_vformat().
2096 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2097 the configuration file to be ":syslog", then the script "guesses" where
2098 the logs files are, rather than using the compiled in default. In our
2099 case the guess is not the same as the compiled default, so the script
2100 suddenly stopped working when I started to use syslog. The patch checks
2101 to see if log_file_path is "". If so, it attempts to read it from exim
2102 with no configuration file to get the compiled in version, before it
2103 falls back to the previous guessing code."
2105 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2106 implementing BATV in an Exim configuration. See NewStuff for the gory
2109 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2110 Makefile that are specific to HP-UX.
2112 PH/18 If the "use_postmaster" option was set for a recipient callout together
2113 with the "random" option, the postmaster address was used as the MAIL
2114 FROM address for the random test, but not for the subsequent recipient
2115 test. It is now used for both.
2117 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2118 patch removes a few documentation additions to RFC 3028, because the
2119 latest draft now contains them. It adds the new en;ascii-case comparator
2120 and a new error check for 8bit text in MIME parts. Comparator and
2121 require names are now matched exactly. I enabled the subaddress
2122 extension, but it is not well tested yet (read: it works for me)."
2124 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2125 rework some of the code of TK/09 above to avoid the hardwired use of
2126 "%lld" and "long long". Replaced the call to snprintf() with a call to
2129 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2130 records point to non-existent hosts", "retry timeout exceeded", and
2131 "retry time not reached for any host after a long failure period".
2133 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2134 experimental DomainKeys support:
2136 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2137 (2) On an error such as an illegally used "control", the wrong name for
2138 the control was given.
2140 These problems did NOT occur unless DomainKeys support was compiled.
2142 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2144 PH/24 Added ${if match_ip condition.
2146 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2147 databases so that it will be absolutely obvious if a crash occurs in the
2148 DB library. This is a regular occurrence (often caused by mis-matched
2151 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2152 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2153 picked up on a user's system that detects such things. There doesn't seem
2154 to be a gcc warning option for this - only an attribute that has to be
2155 put on the function's prototype. It seems that in Fedora Core 4 they have
2156 set this on a number of new functions. No doubt there will be more in due
2159 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2160 specify a transport. However, if an address that was verified by such a
2161 router was the subject of a callout, Exim crashed because it tried to
2162 read the rcpt_include_affixes from the non-existent transport. Now it
2163 just assumes that the setting of that option is false. This bug was
2164 introduced by 4.51/PH/31.
2166 PH/28 Changed -d+all to exclude +memory, because that information is very
2167 rarely of interest, but it makes the output a lot bigger. People tend to
2168 do -d+all out of habit.
2170 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2171 code in os-type was giving problems when libc.so lives in lib64, like on
2174 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2175 aren't the modern standard, and it seems that some systems' include files
2176 don't always have them. Exim was already checking for some of the newer
2177 ones like T_AAAA, and defining it itself. I've added checks for all the
2178 record types that Exim uses.
2180 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2181 not automatically generating a new one, as it is supposed to. This
2182 prevented TLS from working. If the file did exist, but contained invalid
2183 data, a new version was generated, as expected. It was only the case of a
2184 non-existent file that was broken.
2186 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2187 with a change in libdomainkeys > 0.64.
2189 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2190 from DNS. If the selector record carries the flag, it now has
2191 precedence over the domain-wide flag.
2193 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2195 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2196 the use of a port name that isn't defined in /etc/services) occurred, the
2197 message was deferred as in a normal delivery, and thus remained on the
2198 spool, instead of being failed because of the mua_wrapper setting. This
2199 is now fixed, and I tidied up some of the mua_wrapper messages at the
2202 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2203 the messages in a hash of arrays rather than using individual hashes.
2204 This is a bit cleaner and results in dramatic memory savings, albeit
2205 at a slight CPU cost.
2207 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2208 as requested by Marc Sherman.
2210 SC/10 Eximstats: added histograms for user specified patterns as requested
2213 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2215 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2216 fopen() in the content-scanning modules that did not already have it.
2222 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2223 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2225 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2227 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2229 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2230 to test Sieve filters that use "vacation".
2232 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2233 that changes the way the GnuTLS parameters are stored in the cache file.
2234 The new format can be generated externally. For backward compatibility,
2235 if the data in the cache doesn't make sense, Exim assumes it has read an
2236 old-format file, and it generates new data and writes a new file. This
2237 means that you can't go back to an older release without removing the
2240 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2241 work if there are any delivery delays because "one_time" forces the
2242 parent to be marked "delivered", so its unseen clone is never tried
2243 again. For this reason, Exim now forbids the simultaneous setting of
2246 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2247 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2248 entry if you want to know the details. The fix, however, neglected to
2249 consider the case when local delivery batching is involved. The test for
2250 "previously delivered" was not happening when checking to see if an
2251 address could be batched with a previous (undelivered) one; under
2252 certain circumstances this could lead to multiple deliveries to the same
2255 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2256 in its include files, and this causes problems building Exim.
2258 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2259 header_syntax) but Exim was just ignoring anything given after a slash.
2260 In particular, this caused confusion with an attempt to use "verify =
2261 reverse_host_lookup/defer_ok". An error is now given when options are
2262 supplied for verify items that do not have them. (Maybe reverse_host_
2263 lookup should have a defer_ok option, but that's a different point.)
2265 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2266 defined by RFC 821) to 2048, because there were problems with some AUTH
2267 commands, and RFC 1869 says the size should be increased for extended
2268 SMTP commands that take arguments.
2270 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2273 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2274 "unknown" error; now it says that the functionality isn't in the binary.
2276 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2277 an address' error message when a string expansion fails (syntax or
2278 whatever). Otherwise the password may appear in the log. Following change
2279 PH/42 below, there is no longer a chance of it appearing in a bounce
2282 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2284 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2285 of its addresses was ever tried. (Bugzilla bug #2.)
2287 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2288 the result incorrectly in the debug output. (It correctly added a newline
2289 to what was transported.)
2291 TF/01 Added $received_time.
2293 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2294 commented out examples of how to interface to a virus scanner and to
2295 SpamAssassin. Also added commented examples of av_scanner and
2296 spamd_address settings.
2298 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2299 and controls are allowed in which ACLs. There were a couple of minor
2300 errors. Some of the entries in the conditions table (which is a table of
2301 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2302 negation of where the condition IS allowed.
2304 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2306 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2307 header file does not have a version number, so I've had to invent a new
2308 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2309 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2310 radiusclient), but it was contributed by a Radius user.
2312 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2313 files or directories, for OpenSSL.
2315 PH/19 When an Exim process that is running as root has to create an Exim log
2316 file, it does so in a subprocess that runs as exim:exim so as to get the
2317 ownership right at creation (otherwise, other Exim processes might see
2318 the file with the wrong ownership). There was no test for failure of this
2319 fork() call, which would lead to the process getting stuck as it waited
2320 for a non-existent subprocess. Forks do occasionally fail when resources
2321 run out. I reviewed all the other calls to fork(); they all seem to check
2324 PH/20 When checking for unexpected SMTP input at connect time (before writing
2325 the banner), Exim was not dealing correctly with a non-positive return
2326 from the read() function. If the client had disconnected by this time,
2327 the result was a log entry for a synchronization error with an empty
2328 string after "input=" when read() returned zero. If read() returned -1
2329 (an event I could not check), uninitialized data bytes were printed.
2330 There were reports of junk text (parts of files, etc) appearing after
2333 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2335 PH/22 Added support for macro redefinition, and (re)definition in between
2336 driver and ACL definitions.
2338 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2339 forgetting to use the resulting value; it was using the unexpanded value.
2341 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2342 hadn't been configured. The fix is from Juergen Kreileder, who
2343 understands it better than I do:
2345 "Here's what I see happening with three configured cyrus_sasl
2346 authenticators configured (plain, login, cram-md5):
2348 On startup auth_cyrus_sasl_init() gets called for each of these.
2349 This means three calls to sasl_listmech() without a specified mech_list.
2350 => SASL tests which mechs of all available mechs actually work
2351 => three warnings about OTP not working
2352 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2354 With the patch, sasl_listmech() also gets called three times. But now
2355 SASL's mech_list option is set to the server_mech specified in the the
2356 authenticator. Or in other words, the answer from sasl_listmech()
2357 gets limited to just the mech you're testing for (which is different
2359 => the return list contains just 'plain' or 'login', 'cram-md5' or
2360 nothing depending on the value of ob->server_mech.
2362 I've just tested the patch: Authentication still works fine,
2363 unavailable mechs specified in the exim configuration are still
2364 caught, and the auth.log warnings about OTP are gone."
2366 PH/25 When debugging is enabled, the contents of the command line are added
2367 to the debugging output, even when log_selector=+arguments is not
2370 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2371 answer is "GNU", and only if the return is "GNU/something" is the answer
2374 PH/27 $acl_verify_message is now set immediately after the failure of a
2375 verification in an ACL, and so is available in subsequent modifiers. In
2376 particular, the message can be preserved by coding like this:
2378 warn !verify = sender
2379 set acl_m0 = $acl_verify_message
2381 Previously, $acl_verify_message was set only while expanding "message"
2382 and "log_message" when a very denied access.
2384 PH/28 Modified OS/os.c-Linux with
2386 -#ifndef OS_LOAD_AVERAGE
2387 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2389 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2390 nomenclature these days.)
2392 PH/29 Installed patch from the Sieve maintainer that adds the options
2393 sieve_useraddress and sieve_subaddress to the redirect router.
2395 PH/30 In these circumstances:
2396 . Two addresses routed to the same list of hosts;
2397 . First host does not offer TLS;
2398 . First host accepts first address;
2399 . First host gives temporary error to second address;
2400 . Second host offers TLS and a TLS session is established;
2401 . Second host accepts second address.
2402 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2403 and peerdn, if requested) that were in fact used only for the second
2406 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2407 attention to any local part prefix or suffix that was matched by the
2408 router that accepted the address. It now behaves in the same way as it
2409 does for delivery: the affixes are removed from the local part unless
2410 rcpt_include_affixes is set on the transport.
2412 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2413 timeout during the DATA phase of an incoming message.
2415 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2416 applied a patch sanctioned by the Sieve maintainer.
2418 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2419 the uid or gid is negative. A case of a negative gid caused this to be
2420 noticed. The fix allows for either to be negative.
2422 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2423 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2426 PH/36 The change PH/12 above was broken. Fixed it.
2428 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2429 the grounds that routing the same address twice would always produce the
2430 same answer. This might have been true once, but it is certainly no
2431 longer true now. Routing a child address may depend on the previous
2432 routing that produced that child. Some complicated redirection strategies
2433 went wrong when messages had multiple recipients, and made Exim's
2434 behaviour dependent on the order in which the addresses were given.
2436 I have moved the duplicate checking until after the routing is complete.
2437 Exim scans the addresses that are assigned to local and remote
2438 transports, and removes any duplicates. This means that more work will be
2439 done, as duplicates will always all be routed, but duplicates are
2440 presumably rare, so I don't expect this is of any significance.
2442 For deliveries to pipes, files, and autoreplies, the duplicate checking
2443 still happens during the routing process, since they are not going to be
2446 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2447 It corrects a timeout issue with spamd. This is Ian's comment: "The
2448 background is that sometimes spamd either never reads data from a
2449 connection it has accepted, or it never writes response data. The exiscan
2450 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2451 blindly assumes that writes won't block so it may never time out."
2453 PH/39 Allow G after quota size as well as K and M.
2455 PH/40 The value set for $authenticated_id in an authenticator may not contain
2456 binary zeroes or newlines because the value is written to log lines and
2457 to spool files. There was no check on this. Now the value is run through
2458 the string_printing() function so that such characters are converted to
2459 printable escape sequences.
2461 PH/41 $message_linecount is a new variable that contains the total number of
2462 lines in the message. Compare $body_linecount, which is the count for the
2465 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2466 bounce and delay warning messages, except in certain special cases, which
2469 (a) An SMTP error message from a remote host;
2470 (b) A message specified in a :fail: redirection;
2471 (c) A message specified in a "fail" command in a system filter;
2472 (d) A message specified in a FAIL return from the queryprogram router;
2473 (e) A message specified by the cannot_route_message router option.
2475 In these cases only, Exim does include the error details in bounce and
2476 warning messages. There are also a few cases where bland messages such
2477 as "unrouteable address" or "local delivery error" are given.
2479 PH/43 $value is now also set for the "else" part of a ${run expansion.
2481 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2482 being worked on, but at least Exim now implements the latest version to
2485 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2486 process to complete was treated as a delivery failure, a timeout while
2487 writing the message to the pipe was logged, but erroneously treated as a
2488 successful delivery. Such timeouts include transport filter timeouts. For
2489 consistency with the overall process timeout, these timeouts are now
2490 treated as errors, giving rise to delivery failures by default. However,
2491 there is now a new Boolean option for the pipe transport called
2492 timeout_defer, which, if set TRUE, converts the failures into defers for
2493 both kinds of timeout. A transport filter timeout is now identified in
2496 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2497 systems where "make" and "gmake" are different, calling "gmake" at top
2498 level broke things. I've arranged for the value of $(MAKE) to be passed
2499 from the Makefile to this script so that it can call the same version of
2503 A note about Exim versions 4.44 and 4.50
2504 ----------------------------------------
2506 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2507 changes of various kinds. As a consequence, a big documentation update was
2508 needed. This delayed the release for rather longer than seemed good, especially
2509 in the light of a couple of (minor) security issues. Therefore, the changes
2510 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2511 release. So 4.44 and 4.50 are in effect two different branches that both start
2514 I have left the 4.50 change log unchanged; it contains all the changes since
2515 4.43. The change log for 4.44 is below; many of its items are identical to
2516 those for 4.50. This seems to be the most sensible way to preserve the
2517 historical information.
2523 1. Minor wording change to the doc/README.SIEVE file.
2525 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2526 computation of the current number of files was incorrect.
2528 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2529 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2532 4. Give more explanation in the error message when the command for a transport
2533 filter fails to execute.
2535 5. There are several places where Exim runs a non-Exim command in a
2536 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2537 was being done only for the command run by the queryprogram router. It is
2538 now done for all such subprocesses. The other cases are: ${run, transport
2539 filters, and the commands run by the lmtp and pipe transports.
2541 6. Added CONFIGURE_GROUP build-time option.
2543 7. Some older OS have a limit of 256 on the maximum number of file
2544 descriptors. Exim was using setrlimit() to set 1000 as a large value
2545 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2546 systems. I've change it so that if it can't get 1000, it tries for 256.
2548 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2549 was an oversight, and furthermore, ever since the addition of extra
2550 controls (e.g. 4.43/32), the checks on when to allow different forms of
2551 "control" were broken. There should now be diagnostics for all cases when a
2552 control that does not make sense is encountered.
2554 9. Added the /retain_sender option to "control=submission".
2556 10. $recipients is now available in the predata ACL (oversight).
2558 11. Tidy the search cache before the fork to do a delivery from a message
2559 received from the command line. Otherwise the child will trigger a lookup
2560 failure and thereby defer the delivery if it tries to use (for example) a
2561 cached ldap connection that the parent has called unbind on.
2563 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2564 of $address_data from the recipient verification was clobbered by the
2565 sender verification.
2567 13. The value of address_data from a sender verification is now available in
2568 $sender_address_data in subsequent conditions in the ACL statement.
2570 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2572 15. Added a new option "connect=<time>" to callout options, to set a different
2575 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2576 was its contents. (It was OK if the option was not defined at all.)
2578 17. A "Completed" log line is now written for messages that are removed from
2579 the spool by the -Mrm option.
2581 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2582 information about exactly what failed.
2584 19. Added -dd to debug only the daemon process.
2586 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2587 handles timeouts, both on the server side and network timeouts. Renamed the
2588 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2590 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2591 It is now set to "smtps".
2593 22. $host_address is now set to the target address during the checking of
2594 ignore_target_hosts.
2596 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2597 being passed; this would have caused $sender_host_name to have been used if
2598 matching the list had actually called for a host name (not very likely,
2599 since this list is usually IP addresses). A host name is now passed as
2602 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2603 code that specifies a non-restarting handler (typically sigaction() in
2604 modern systems) in an attempt to fix a rare and obscure crash bug.
2606 25. Narrowed the window for a race in the daemon that could cause it to ignore
2607 SIGCHLD signals. This is not a major problem, because they are used only to
2608 wake it up if nothing else does.
2610 26. A malformed maildirsize file could cause Exim to calculate negative values
2611 for the mailbox size or file count. Odd effects could occur as a result.
2612 The maildirsize information is now recalculated if the size or filecount
2615 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2616 support for a long time. Removed HAVE_SYS_VFS_H.
2618 28. Installed the latest version of exipick from John Jetmore.
2620 29. In an address list, if the pattern was not a regular expression, an empty
2621 subject address (from a bounce message) matched only if the pattern was an
2622 empty string. Non-empty patterns were not even tested. This was the wrong
2623 because it is perfectly reasonable to use an empty address as part of a
2624 database query. An empty address is now tested by patterns that are
2625 lookups. However, all the other forms of pattern expect the subject to
2626 contain a local part and a domain, and therefore, for them, an empty
2627 address still always fails if the pattern is not itself empty.
2629 30. Exim went into a mad DNS loop when attempting to do a callout where the
2630 host was specified on an smtp transport, and looking it up yielded more
2631 than one IP address.
2633 31. Re-factored the code for checking spool and log partition space into a
2634 function that finds that data and another that does the check. The former
2635 is then used to implement four new variables: $spool_space, $log_space,
2636 $spool_inodes, and $log_inodes.
2638 32. The RFC2047 encoding function was originally intended for short strings
2639 such as real names; it was not keeping to the 75-character limit for
2640 encoded words that the RFC imposes. It now respects the limit, and
2641 generates multiple encoded words if necessary. To be on the safe side, I
2642 have increased the buffer size for the ${rfc2047: expansion operator from
2645 33. It is now permitted to omit both strings after an "if" condition; if the
2646 condition is true, the result is "true". As before, when the second string
2647 is omitted, a false condition yields an empty string. This makes it less
2648 cumbersome to write custom ACL and router conditions.
2650 34. Failure to deliver a bounce message always caused it to be frozen, even if
2651 there was an errors_to setting on the router. The errors_to setting is now
2654 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2655 canonical form (fully expanded) before being placed in
2656 $sender_host_address.
2658 36. The table in the code that translates DNS record types into text (T_A to
2659 "A" for instance) was missing entries for NS and CNAME. It is just possible
2660 that this could have caused confusion if both these types were looked up
2661 for the same domain, because the text type is used as part of Exim's
2662 per-process caching. But the chance of anyone hitting this buglet seems
2665 37. The dnsdb lookup has been extended in a number of ways.
2667 (1) There is a new type, "zns", which walks up the domain tree until it
2668 finds some nameserver records. It should be used with care.
2670 (2) There is a new type, "mxh", which is like "mx" except that it returns
2671 just the host names, not the priorities.
2673 (3) It is now possible to give a list of domains (or IP addresses) to be
2674 looked up. The behaviour when one of the lookups defers can be
2675 controlled by a keyword.
2677 (4) It is now possible to specify the separator character for use when
2678 multiple records are returned.
2680 38. The dnslists ACL condition has been extended: it is now possible to supply
2681 a list of IP addresses and/or domains to be looked up in a particular DNS
2684 39. Added log_selector=+queue_time_overall.
2686 40. When running the queue in the test harness, wait just a tad after forking a
2687 delivery process, to get repeatability of debugging output.
2689 41. Include certificate and key file names in error message when GnuTLS fails
2690 to set them up, because the GnuTLS error message doesn't include the name
2691 of the failing file when there is a problem reading it.
2693 42. Allow both -bf and -bF in the same test run.
2695 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2697 44. The "Exiscan patch" is now merged into the mainline Exim source.
2699 45. Sometimes the final signoff response after QUIT could fail to get
2700 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2701 < 0 before doing a fflush(). This bug looks as though it goes back to the
2702 introduction of TLS in release 3.20, but "sometimes" must have been rare
2703 because the tests only now provoked it.
2705 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2706 (this can affect the format of dates).
2708 47. exim_tidydb, when checking for the continued existence of a message for
2709 which it has found a message-specific retry record, was not finding
2710 messages that were in split spool directories. Consequently, it was
2711 deleting retry records that should have stayed in existence.
2713 48. Steve fixed some bugs in eximstats.
2715 49. The SPA authentication driver was not abandoning authentication and moving
2716 on to the next authenticator when an expansion was forced to fail,
2717 contradicting the general specification for all authenticators. Instead it
2718 was generating a temporary error. It now behaves as specified.
2720 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2721 (the order specifies the preference for clients). The order is now AES256,
2722 AES128, 3DES, ARCFOUR128.
2724 51. Small patch to Sieve code - explicitly set From: when generating an
2727 52. Exim crashed if a remote delivery caused a very long error message to be
2728 recorded - for instance if somebody sent an entire SpamAssassin report back
2729 as a large number of 550 error lines. This bug was coincidentally fixed by
2730 increasing the size of one of Exim's internal buffers (big_buffer) that
2731 happened as part of the Exiscan merge. However, to be on the safe side, I
2732 have made the code more robust (and fixed the comments that describe what
2735 53. Now that there can be additional text after "Completed" in log lines (if
2736 the queue_time_overall log selector is set), a one-byte patch to exigrep
2737 was needed to allow it to recognize "Completed" as not the last thing in
2740 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2741 patch that reportedly fixes this has been added. I am not expert enough to
2742 create a test for it. This is what the patch creator wrote:
2744 "I found a little strange behaviour of ldap code when working with
2745 Windows 2003 AD Domain, where users was placed in more than one
2746 Organization Units. When I tried to give exim partial DN, the exit code
2747 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2748 But simultaneously result of request was absolutely normal ldap result,
2749 so I produce this patch..."
2751 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2752 REFERENCE, so I have modified the code to exclude the patch when that macro
2755 55. Some experimental protocols are using DNS PTR records for new purposes. The
2756 keys for these records are domain names, not reversed IP addresses. The
2757 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2758 leaves it alone. Component reversal etc. now happens only for IP addresses.
2761 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2763 57. Double the size of the debug message buffer (to 2048) so that more of very
2764 long debug lines gets shown.
2766 58. The exicyclog utility now does better if the number of log files to keep
2767 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2769 59. Two changes related to the smtp_active_hostname option:
2771 (1) $smtp_active_hostname is now available as a variable.
2772 (2) The default for smtp_banner uses $smtp_active_hostname instead
2773 of $primary_hostname.
2775 60. The host_aton() function is supposed to be passed a string that is known
2776 to be a valid IP address. However, in the case of IPv6 addresses, it was
2777 not checking this. This is a hostage to fortune. Exim now panics and dies
2778 if the condition is not met. A case was found where this could be provoked
2779 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2780 components; fortuitously, this particular loophole had already been fixed
2781 by change 4.50/55 above.
2783 If there are any other similar loopholes, the new check in host_aton()
2784 itself should stop them being exploited. The report I received stated that
2785 data on the command line could provoke the exploit when Exim was running as
2786 exim, but did not say which command line option was involved. All I could
2787 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2788 running as the user.
2791 61. There was a buffer overflow vulnerability in the SPA authentication code
2792 (which came originally from the Samba project). I have added a test to the
2793 spa_base64_to_bits() function which I hope fixes it.
2796 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2797 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2799 63. The daemon start-up calls getloadavg() while still root for those OS that
2800 need the first call to be done as root, but it missed one case: when
2801 deliver_queue_load_max is set with deliver_drop_privilege. This is
2802 necessary for the benefit of the queue runner, because there is no re-exec
2803 when deliver_drop_privilege is set.
2805 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2806 This has been fixed.
2808 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2809 was in use, was not putting the data itself into the right store pool;
2810 consequently, it could be overwritten for a subsequent message in the same
2811 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2814 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2816 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2817 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2818 the function were treating the return as a boolean value, which happened to
2819 work because 0=false and not-0=true, but is not correct code.
2821 68. The host_aton() function was not handling scoped IPv6 addresses (those
2822 with, for example, "%eth0" on the end) correctly.
2824 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2825 negated items (that is, ~something) in unsigned ints. Some compilers
2826 apparently mutter when there is no cast.
2828 70. If an address verification called from an ACL failed, and did not produce a
2829 user-specific message (i.e. there was only a "system" message), nothing was
2830 put in $acl_verify_message. In this situation, it now puts the system
2833 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2834 session; change 4.31/43 added the unwanted input to the log line - except
2835 that it did not do this in the start of session case. It now does.
2837 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2838 This is wrong and can cause the other end to generate a synchronization
2839 error if it is another Exim or anything else that does the synchronization
2840 check. A QUIT command is no longer sent after a timeout.
2842 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2843 during host lookups.
2845 74. The defer_ok option of callout verification was not working if it was used
2846 when verifying addresses in header lines, that is, for this case:
2848 verify = header_sender/callout=defer_ok
2850 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2851 those file descriptors could be used for SMTP connections. If anything
2852 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2853 could be sent to the SMTP client, causing chaos. The daemon now opens
2854 stdin, stdout, and stderr to /dev/null when it puts itself into the
2857 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2858 log by default. The user can override this with suitable Perl magic.
2860 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2861 the log message when discard triggers, was not working for the DATA ACL or
2862 for the non-SMTP ACL.
2864 78. Error message wording change in sieve.c.
2866 79. If smtp_accept_max_per_host was set, the number of connections could be
2867 restricted to fewer than expected, because the daemon was trying to set up
2868 a new connection before checking whether the processes handling previous
2869 connections had finished. The check for completed processes is now done
2870 earlier. On busy systems, this bug wouldn't be noticed because something
2871 else would have woken the daemon, and it would have reaped the completed
2874 80. If a message was submitted locally by a user whose login name contained one
2875 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2876 It caused a spool format error. I have fixed the spool reading code. A
2877 related problem was that the "from" clause in the Received: line became
2878 illegal because of the space(s). It is now covered by ${quote_local_part.
2880 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2883 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2885 83. Patch from Sieve maintainer to fix unterminated string problem in
2886 "vacation" handling.
2888 84. Some minor changes to the Linux configuration files to help with other
2889 OS variants using glibc.
2891 85. One more patch for Sieve to update vacation handling to latest spec.
2894 ----------------------------------------------------
2895 See the note above about the 4.44 and 4.50 releases.
2896 ----------------------------------------------------
2902 1. Change 4.43/35 introduced a bug that caused file counts to be
2903 incorrectly computed when quota_filecount was set in an appendfile
2906 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2907 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2910 3. Give more explanation in the error message when the command for a transport
2911 filter fails to execute.
2913 4. There are several places where Exim runs a non-Exim command in a
2914 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2915 was being done only for the command run by the queryprogram router. It is
2916 now done for all such subprocesses. The other cases are: ${run, transport
2917 filters, and the commands run by the lmtp and pipe transports.
2919 5. Some older OS have a limit of 256 on the maximum number of file
2920 descriptors. Exim was using setrlimit() to set 1000 as a large value
2921 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2922 systems. I've change it so that if it can't get 1000, it tries for 256.
2924 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2925 was an oversight, and furthermore, ever since the addition of extra
2926 controls (e.g. 4.43/32), the checks on when to allow different forms of
2927 "control" were broken. There should now be diagnostics for all cases when a
2928 control that does not make sense is encountered.
2930 7. $recipients is now available in the predata ACL (oversight).
2932 8. Tidy the search cache before the fork to do a delivery from a message
2933 received from the command line. Otherwise the child will trigger a lookup
2934 failure and thereby defer the delivery if it tries to use (for example) a
2935 cached ldap connection that the parent has called unbind on.
2937 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2938 of $address_data from the recipient verification was clobbered by the
2939 sender verification.
2941 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2942 was its contents. (It was OK if the option was not defined at all.)
2944 11. A "Completed" log line is now written for messages that are removed from
2945 the spool by the -Mrm option.
2947 12. $host_address is now set to the target address during the checking of
2948 ignore_target_hosts.
2950 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2951 being passed; this would have caused $sender_host_name to have been used if
2952 matching the list had actually called for a host name (not very likely,
2953 since this list is usually IP addresses). A host name is now passed as
2956 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2957 code that specifies a non-restarting handler (typically sigaction() in
2958 modern systems) in an attempt to fix a rare and obscure crash bug.
2960 15. Narrowed the window for a race in the daemon that could cause it to ignore
2961 SIGCHLD signals. This is not a major problem, because they are used only to
2962 wake it up if nothing else does.
2964 16. A malformed maildirsize file could cause Exim to calculate negative values
2965 for the mailbox size or file count. Odd effects could occur as a result.
2966 The maildirsize information is now recalculated if the size or filecount
2969 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2970 support for a long time. Removed HAVE_SYS_VFS_H.
2972 18. Updated exipick to current release from John Jetmore.
2974 19. Allow an empty sender to be matched against a lookup in an address list.
2975 Previously the only cases considered were a regular expression, or an
2978 20. Exim went into a mad DNS lookup loop when doing a callout where the
2979 host was specified on the transport, if the DNS lookup yielded more than
2982 21. The RFC2047 encoding function was originally intended for short strings
2983 such as real names; it was not keeping to the 75-character limit for
2984 encoded words that the RFC imposes. It now respects the limit, and
2985 generates multiple encoded words if necessary. To be on the safe side, I
2986 have increased the buffer size for the ${rfc2047: expansion operator from
2989 22. Failure to deliver a bounce message always caused it to be frozen, even if
2990 there was an errors_to setting on the router. The errors_to setting is now
2993 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2994 canonical form (fully expanded) before being placed in
2995 $sender_host_address.
2997 24. Updated eximstats to version 1.33
2999 25. Include certificate and key file names in error message when GnuTLS fails
3000 to set them up, because the GnuTLS error message doesn't include the name
3001 of the failing file when there is a problem reading it.
3003 26. Expand error message when OpenSSL has problems setting up cert/key files.
3006 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3007 (this can affect the format of dates).
3009 28. exim_tidydb, when checking for the continued existence of a message for
3010 which it has found a message-specific retry record, was not finding
3011 messages that were in split spool directories. Consequently, it was
3012 deleting retry records that should have stayed in existence.
3014 29. eximstats updated to version 1.35
3015 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3016 1.35 - bugfix such that pie charts by volume are generated correctly
3018 30. The SPA authentication driver was not abandoning authentication and moving
3019 on to the next authenticator when an expansion was forced to fail,
3020 contradicting the general specification for all authenticators. Instead it
3021 was generating a temporary error. It now behaves as specified.
3023 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3024 (the order specifies the preference for clients). The order is now AES256,
3025 AES128, 3DES, ARCFOUR128.
3027 31. Small patch to Sieve code - explicitly set From: when generating an
3030 32. Exim crashed if a remote delivery caused a very long error message to be
3031 recorded - for instance if somebody sent an entire SpamAssassin report back
3032 as a large number of 550 error lines. This bug was coincidentally fixed by
3033 increasing the size of one of Exim's internal buffers (big_buffer) that
3034 happened as part of the Exiscan merge. However, to be on the safe side, I
3035 have made the code more robust (and fixed the comments that describe what
3038 33. Some experimental protocols are using DNS PTR records for new purposes. The
3039 keys for these records are domain names, not reversed IP addresses. The
3040 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3041 leaves it alone. Component reversal etc. now happens only for IP addresses.
3044 34. The host_aton() function is supposed to be passed a string that is known
3045 to be a valid IP address. However, in the case of IPv6 addresses, it was
3046 not checking this. This is a hostage to fortune. Exim now panics and dies
3047 if the condition is not met. A case was found where this could be provoked
3048 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3049 components; fortuitously, this particular loophole had already been fixed
3050 by change 4.50/55 or 4.44/33 above.
3052 If there are any other similar loopholes, the new check in host_aton()
3053 itself should stop them being exploited. The report I received stated that
3054 data on the command line could provoke the exploit when Exim was running as
3055 exim, but did not say which command line option was involved. All I could
3056 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3057 running as the user.
3060 35. There was a buffer overflow vulnerability in the SPA authentication code
3061 (which came originally from the Samba project). I have added a test to the
3062 spa_base64_to_bits() function which I hope fixes it.
3065 36. The daemon start-up calls getloadavg() while still root for those OS that
3066 need the first call to be done as root, but it missed one case: when
3067 deliver_queue_load_max is set with deliver_drop_privilege. This is
3068 necessary for the benefit of the queue runner, because there is no re-exec
3069 when deliver_drop_privilege is set.
3071 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3072 was in use, was not putting the data itself into the right store pool;
3073 consequently, it could be overwritten for a subsequent message in the same
3074 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3077 38. Sometimes the final signoff response after QUIT could fail to get
3078 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3079 < 0 before doing a fflush(). This bug looks as though it goes back to the
3080 introduction of TLS in release 3.20, but "sometimes" must have been rare
3081 because the tests only now provoked it.
3087 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3088 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3089 Now it returns an integer. A number of calls were still expecting a T/F
3090 return. Fortuitously, in all cases, the tests worked in OK situations,
3091 which is the norm. However, things would have gone wrong on any write
3092 failures on the smtp file descriptor. This function is used when sending
3093 messages over SMTP and also when doing verify callouts.
3095 2. When Exim is called to do synchronous delivery of a locally submitted
3096 message (the -odf or -odi options), it no longer closes stderr before doing
3099 3. Implemented the mua_wrapper option.
3101 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3103 5. Implemented the functions header_remove(), header_testname(),
3104 header_add_at_position(), and receive_remove_recipient(), and exported them
3107 6. If an ACL "warn" statement specified the addition of headers, Exim already
3108 inserted X-ACL-Warn: at the start if there was no header name. However, it
3109 was not making this test for the second and subsequent header lines if
3110 there were newlines in the string. This meant that an invalid header could
3111 be inserted if Exim was badly configured.
3113 7. Allow an ACL "warn" statement to add header lines at the start or after all
3114 the Received: headers, as well as at the end.
3116 8. Added the rcpt_4xx retry error code.
3118 9. Added postmaster_mailfrom=xxx to callout verification option.
3120 10. Added mailfrom=xxxx to the callout verification option, for verify=
3123 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3124 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3126 12. Inserted some casts to stop certain compilers warning when using pointer
3127 differences as field lengths or precisions in printf-type calls (mostly
3128 affecting debugging statements).
3130 13. Added optional readline() support for -be (dynamically loaded).
3132 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3133 same clock tick as a message's arrival, so that its received time was the
3134 same as the "first fail" time on the retry record, and that message
3135 remained on the queue past the ultimate address timeout, every queue runner
3136 would try a delivery (because it was past the ultimate address timeout) but
3137 after another failure, the ultimate address timeout, which should have then
3138 bounced the address, did not kick in. This was a "< instead of <=" error;
3139 in most cases the first failure would have been in the next clock tick
3140 after the received time, and all would be well.
3142 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3143 being recognized when the domain list was tested by the match_domain
3144 condition in an expansion string.
3146 16. Added the ${str2b64: operator.
3148 17. Exim was always calling setrlimit() to set a large limit for the number of
3149 processes, without checking whether the existing limit was already
3150 adequate. (It did check for the limit on file descriptors.) Furthermore,
3151 errors from getrlimit() and setrlimit() were being ignored. Now they are
3152 logged to the main and panic logs, but Exim does carry on, to try to do its
3153 job under whatever limits there are.
3155 18. Imported PCRE 5.0.
3157 19. Trivial typo in log message " temporarily refused connection" (the leading
3160 20. If the log selector return_path_on_delivery was set and an address was
3161 redirected to /dev/null, the delivery process crashed because it assumed
3162 that a return path would always be set for a "successful" delivery. In this
3163 case, the whole delivery is bypassed as an optimization, and therefore no
3166 21. Internal re-arrangement: the function for sending a challenge and reading
3167 a response while authentication was assuming a zero-terminated challenge
3168 string. It's now changed to take a pointer and a length, to allow for
3169 binary data in such strings.
3171 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3173 23. Exim was not respecting finduser_retries when seeking the login of the
3174 uid under which it was called; it was always trying 10 times. (The default
3175 setting of finduser_retries is zero.) Also, it was sleeping after the final
3176 failure, which is pointless.
3178 24. Implemented tls_on_connect_ports.
3180 25. Implemented acl_smtp_predata.
3182 26. If the domain in control=submission is set empty, Exim assumes that the
3183 authenticated id is a complete email address when it generates From: or
3184 Sender: header lines.
3186 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3187 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3188 chown and chgrp in /bin and hostname in /usr/bin.
3190 28. Exim was keeping the "process log" file open after each use, just as it
3191 does for the main log. This opens the possibility of it remaining open for
3192 long periods when the USR1 signal hits a daemon. Occasional processlog
3193 errors were reported, that could have been caused by this. Anyway, it seems
3194 much more sensible not to leave this file open at all, so that is what now
3197 29. The long-running daemon process does not normally write to the log once it
3198 has entered its main loop, and it closes the log before doing so. This is
3199 so that log files can straightforwardly be renamed and moved. However,
3200 there are a couple of unusual error situations where the daemon does write
3201 log entries, and I had neglected to close the log afterwards.
3203 30. The text of an SMTP error response that was received during a remote
3204 delivery was being truncated at 512 bytes. This is too short for some of
3205 the long messages that one sometimes sees. I've increased the limit to
3208 31. It is now possible to make retry rules that apply only when a message has a
3209 specific sender, in particular, an empty sender.
3211 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3212 it possible to be selective about when SMTP synchronization is enforced.
3214 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3216 32. Implemented hosts_connection_nolog.
3218 33. Added an ACL for QUIT.
3220 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3223 35. Added mailbox_size and mailbox_filecount to appendfile.
3225 36. Added control = no_multiline_responses to ACLs.
3227 37. There was a bug in the logic of the code that waits for the clock to tick
3228 in the case where the clock went backwards by a substantial amount such
3229 that the microsecond fraction of "now" was more than the microsecond
3230 fraction of "then" (but the whole seconds number was less).
3232 38. Added support for the libradius Radius client library this is found on
3233 FreeBSD (previously only the radiusclient library was supported).
3239 1. When certain lookups returned multiple values in the form name=value, the
3240 quoting of the values was not always being done properly. Specifically:
3241 (a) If the value started with a double quote, but contained no whitespace,
3243 (b) If the value contained whitespace other than a space character (i.e.
3244 tabs or newlines or carriage returns) it was not quoted.
3245 This fix has been applied to the mysql and pgsql lookups by writing a
3246 separate quoting function and calling it from the lookup code. The fix
3247 should probably also be applied to nisplus, ibase and oracle lookups, but
3248 since I cannot test any of those, I have not disturbed their existing code.
3250 2. A hit in the callout cache for a specific address caused a log line with no
3251 reason for rejecting RCPT. Now it says "Previous (cached) callout
3252 verification failure".
3254 3. There was an off-by-one bug in the queryprogram router. An over-long
3255 return line was truncated at 256 instead of 255 characters, thereby
3256 overflowing its buffer with the terminating zero. As well as fixing this, I
3257 have increased the buffer size to 1024 (and made a note to document this).
3259 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3260 when Exim is waiting for an SMTP response from a remote server, Exim
3261 restarts its select() call on the socket, thereby resetting its timeout.
3262 This is not a problem when such interrupts are rare. Somebody set up a cron
3263 job to run exiwhat every 2 minutes, which is less than the normal select()
3264 timeout (5 or 10 minutes). This meant that the select() timeout never
3265 kicked in because it was always reset. I have fixed this by comparing the
3266 time when an interrupt arrives with the time at the start of the first call
3267 to select(). If more time than the timeout has elapsed, the interrupt is
3268 treated as a timeout.
3270 5. Some internal re-factoring in preparation for the addition of Sieve
3271 extensions (by MH). In particular, the "personal" test is moved to a
3272 separate function, and given an option for scanning Cc: and Bcc: (which is
3273 not set for Exim filters).
3275 6. When Exim created an email address using the login of the caller as the
3276 local part (e.g. when creating a From: or Sender: header line), it was not
3277 quoting the local part when it contained special characters such as @.
3279 7. Installed new OpenBSD configuration files.
3281 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3282 try to make them clearer.
3284 9. Callout options, other than the timeout value, were being ignored when
3285 verifying sender addresses in header lines. For example, when using
3287 verify = header_sender/callout=no_cache
3289 the cache was (incorrectly) being used.
3291 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3292 only the Cygwin environment.
3294 11. When return_path_on_delivery was set as a log selector, if different remote
3295 addresses in the same message used different return paths and parallel
3296 remote delivery occurred, the wrong values would sometimes be logged.
3297 (Whenever a remote delivery process finished, the return path value from
3298 the most recently started remote delivery process was logged.)
3300 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3301 lines when AUTH and/or TLS are in use. This is the "received protocol"
3302 field. Exim used to use "asmtp" for authenticated SMTP, without any
3303 indication (in the protocol name) for TLS use. Now it follows the RFC and
3304 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3305 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3306 names appear in log lines as well as in Received: header lines.
3308 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3309 extensions, and comparison tests, and to fix some bugs.
3311 14. Changes to the "personal" filter test:
3313 (1) The test was buggy in that it was just doing the equivalent of
3314 "contains" tests on header lines. For example, if a user's address was
3315 anne@some.where, the "personal" test would incorrectly be true for
3317 To: susanne@some.where
3319 This test is now done by extracting each address from the header in turn,
3320 and checking the entire address. Other tests that are part of "personal"
3321 are now done using regular expressions (for example, to check local parts
3322 of addresses in From: header lines).
3324 (2) The list of non-personal local parts in From: addresses has been
3325 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3326 taken from the Sieve specification recommendations.
3328 (3) If the message contains any header line starting with "List-" it is
3329 treated as non-personal.
3331 (4) The test for "circular" in the Subject: header line has been removed
3332 because it now seems ill-conceived.
3334 15. Minor typos in src/EDITME comments corrected.
3336 16. Installed latest exipick from John Jetmore.
3338 17. If headers_add on a router specified a text string that was too long for
3339 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3340 of string_sprintf() is now avoided.
3342 18. $message_body_size was not set (it was always zero) when running the DATA
3343 ACL and the local_scan() function.
3345 19. For the "mail" command in an Exim filter, no default was being set for
3346 the once_repeat time, causing a random time value to be used if "once" was
3347 specified. (If the value happened to be <= 0, no repeat happened.) The
3348 default is now 0s, meaning "never repeat". The "vacation" command was OK
3349 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3350 (I found it when inspecting the code).
3352 20. There is now an overall timeout for performing a callout verification. It
3353 defaults to 4 times the callout timeout, which applies to individual SMTP
3354 commands during the callout. The overall timeout applies when there is more
3355 than one host that can be tried. The timeout is checked before trying the
3356 next host. This prevents very long delays if there are a large number of
3357 hosts and all are timing out (e.g. when the network connections are timing
3358 out). The value of the overall timeout can be changed by specifying an
3359 additional sub-option for "callout", called "maxwait". For example:
3361 verify = sender/callout=5s,maxwait=20s
3363 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3364 to the end before writing, but this should make it even safer).
3366 22. Exim was forgetting that it had advertised PIPELINING for the second and
3367 subsequent messages on an SMTP connection. It was also not resetting its
3368 memory on STARTTLS and an internal HELO.
3370 23. When Exim logs an SMTP synchronization error within a session, it now
3371 records whether PIPELINING has been advertised or not.
3373 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3374 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3375 rather than long int.
3377 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3379 26. Added the never_mail option to autoreply.
3385 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3386 crash if the getsockname() call failed; this can happen if a connection is
3387 closed very soon after it is established. The problem was simply in the
3388 order in which certain operations were done, causing Exim to try to write
3389 to the SMTP stream before it had set up the file descriptor. The bug has
3390 been fixed by making things happen in the correct order.
3396 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3397 before the connection was closed, thus losing the rejection response.
3399 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3400 some early Solaris releases, but causes trouble in current releases where
3401 socklen_t is defined.
3403 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3406 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3409 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3410 that releasing the top of it at the end releases what was used for sub-
3411 expansions (unless the block got too big). However, discard this block if
3412 the first thing is a variable or header, so that we can use its block when
3413 it is dynamic (useful for very large $message_headers, for example).
3415 6. Lookups now cache *every* query, not just the most recent. A new, separate
3416 store pool is used for this. It can be recovered when all lookup caches are
3417 flushed. Lookups now release memory at the end of their result strings.
3418 This has involved some general refactoring of the lookup sources.
3420 7. Some code has been added to the store_xxx() functions to reduce the amount
3421 of flapping under certain conditions.
3423 8. log_incoming_interface used to affect only the <= reception log lines. Now
3424 it causes the local interface and port to be added to several more SMTP log
3425 lines, for example "SMTP connection from", and rejection lines.
3427 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3429 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3431 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3432 could be overwritten at the end of the current message (or the start of a
3433 new message if it was set in a HELO ACL). The value is now preserved for
3434 the duration of the SMTP connection.
3436 12. If a transport had a headers_rewrite setting, and a matching header line
3437 contained an unqualified address, that address was qualified, even if it
3438 did not match any rewriting rules. The underlying bug was that the values
3439 of the flags that permit the existence of unqualified sender and recipient
3440 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3441 non-local messages, and by -bnq for local messages) were not being
3442 preserved with the message after it was received.
3444 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3445 "next input=" as part of the text comprising the host identity instead of
3446 the correct text. The code was using the same buffer for two different
3447 strings. However, depending on which order the printing function evaluated
3448 its arguments, the bug did not always show up. Under Linux, for example, my
3449 test suite worked just fine.
3451 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3452 doesn't work with some older versions of Perl. It has been changed to "my",
3453 which in any case is probably the better facility to use.
3455 15. A really picky compiler found some instances of statements for creating
3456 error messages that either had too many or two few arguments for the format
3459 16. The size of the buffer for calls to the DNS resolver has been increased
3460 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3461 for addresses that have a lot of PTR records. This alleviates a problem; it
3462 does not fully solve it.
3464 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3465 buffer now truncates the list and logs the incident, which is the same
3466 action as happens when Exim is looking up a host name and its aliases.
3467 Previously in this situation something unpredictable would happen;
3468 sometimes it was "internal error: store_reset failed".
3470 18. If a server dropped the connection unexpectedly when an Exim client was
3471 using GnuTLS and trying to read a response, the client delivery process
3472 crashed while trying to generate an error log message.
3474 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3475 string, for example:
3477 warn message = H1: something\nH2: something
3479 the text was added as a single header line from Exim's point of view
3480 though it ended up OK in the delivered message. However, searching for the
3481 second and subsequent header lines using $h_h2: did not work. This has been
3482 fixed. Similarly, if a system filter added multiple headers in this way,
3483 the routers could not see them.
3485 20. Expanded the error message when iplsearch is called with an invalid key to
3486 suggest using net-iplsearch in a host list.
3488 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3489 ACLs are no longer actually imposed (and a message to that effect is
3492 22. If a "gecos" field in a passwd entry contained escaped characters, in
3493 particular, if it contained a \" sequence, Exim got it wrong when building
3494 a From: or a Sender: header from that name. A second bug also caused
3495 incorrect handling when an unquoted " was present following a character
3496 that needed quoting.
3498 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3499 was not being matched caselessly.
3501 24. Arranged for all hyphens in the exim.8 source to be escaped with
3504 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3505 properly. Recipient callout cache records were still being keyed to include
3506 the sender, even when use_sender was set false. This led to far more
3507 callouts that were necessary. The sender is no longer included in the key
3508 when use_sender is false.
3510 26. Added "control = submission" modifier to ACLs.
3512 27. Added the ${base62d: operator to decode base 62 numbers.
3514 28. dnsdb lookups can now access SRV records.
3516 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3517 the configuration file.
3519 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3520 (-v) mode. This makes the output for a verbose queue run more intelligible.
3522 31. Added a use_postmaster feature to recipient callouts.
3524 32. Added the $body_zerocount variable, containing the number of binary zero
3525 bytes in the message body.
3527 33. The time of last modification of the "new" subdirectory is now used as the
3528 "mailbox time last read" when there is a quota error for a maildir
3531 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3533 35. Added +ignore_unknown as a special item in host lists.
3535 36. Code for decoding IPv6 addresses in host lists is now included, even if
3536 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3537 address was recognized as an IP address, but was then not correctly decoded
3538 into binary, causing unexpected and incorrect effects when compared with
3545 1. Very minor rewording of debugging text in manualroute to say "list of
3546 hosts" instead of "hostlist".
3548 2. If verify=header_syntax was set, and a header line with an unqualified
3549 address (no domain) and a large number of spaces between the end of the
3550 name and the colon was received, the reception process suffered a buffer
3551 overflow, and (when I tested it) crashed. This was caused by some obsolete
3552 code that should have been removed. The fix is to remove it!
3554 3. When running in the test harness, delay a bit after writing a bounce
3555 message to get a bit more predictability in the log output.
3557 4. Added a call to search_tidyup() just before forking a reception process. In
3558 theory, someone could use a lookup in the expansion of smtp_accept_max_
3559 per_host which, without the tidyup, could leave open a database connection.
3561 5. Added the variables $recipient_data and $sender_data which get set from a
3562 lookup success in an ACL "recipients" or "senders" condition, or a router
3563 "senders" option, similar to $domain_data and $local_part_data.
3565 6. Moved the writing of debug_print from before to after the "senders" test
3568 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3569 problems for message scanning, either using a data ACL, or using
3570 local_scan() because the Received: header was not generated till after they
3571 were called (in order to set the time as the time of reception completion).
3572 I have revised the way this works. The header is now generated after the
3573 body is received, but before the ACL or local_scan() are called. After they
3574 are run, the timestamp in the header is updated.
3580 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3581 before starting a queue runner without re-exec. This happened only when
3582 deliver_drop_privilege was set or when the Exim user was set to root. The
3583 effect of the bug was that timeouts during subsequent deliveries caused
3584 crashes instead of being properly handled. The handler is now left at its
3585 default (and expected) setting.
3587 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3588 message, again when deliver_drop_privilege is set or Exim is run as root.
3589 The bug described in (1) was not present in this case, but the tidying up
3590 of the other signals was missing. I have made the two cases consistent.
3592 3. The ignore_target_hosts setting on a manualroute router was being ignored
3593 for hosts that were looked up using the /MX notation.
3595 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3598 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3599 operated on the sender address. After changing the $sender_address to <>
3600 for the sender address verify, Exim was re-instated it as the original
3601 (before rewriting) address, but remembering that it had rewritten it, so it
3602 wasn't rewriting it again. This bug also had the effect of breaking the
3603 sender address verification caching when the sender address was rewritten.
3605 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3606 This has been changed so that if the ip literal address matches
3607 ignore_target_hosts, the router declines.
3609 7. Added expansion conditions match_domain, match_address, and match_local_
3610 part (NOT match_host).
3612 8. The placeholder for the Received: header didn't have a length field set.
3614 9. Added code to Exim itself and to exim_lock to test for a specific race
3615 condition that could lead to file corruption when using MBX delivery. The
3616 issue is with the lockfile that is created in /tmp. If this file is removed
3617 after a process has opened it but before that process has acquired a lock,
3618 there is the potential for a second process to recreate the file and also
3619 acquire a lock. This could lead to two Exim processes writing to the file
3620 at the same time. The added code performs the same test as UW imapd; it
3621 checks after acquiring the lock that its file descriptor still refers to
3622 the same named file.
3624 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3625 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3628 11. Added the smtp_active_hostname option. If used, this will typically be made
3629 to depend on the incoming interface address. Because $interface_address is
3630 not set up until the daemon has forked a reception process, error responses
3631 that can happen earlier (such as "too many connections") no longer contain
3634 12. If an expansion in a condition on a "warn" statement fails because a lookup
3635 defers, the "warn" statement is abandoned, and the next ACL statement is
3636 processed. Previously this caused the whole ACL to be aborted.
3638 13. Added the iplsearch lookup type.
3640 14. Added ident_timeout as a log selector.
3642 15. Added tls_certificate_verified as a log selector.
3644 16. Added a global option tls_require_ciphers (compare the smtp transport
3645 option of the same name). This controls incoming TLS connections.
3647 17. I finally figured out how to make tls_require_ciphers do a similar thing
3648 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3649 before starting the TLS session.
3651 18. Tabs are now shown as \t in -bP output.
3653 19. If the log selector return_path_on_delivery was set, Exim crashed when
3654 bouncing a message because it had too many Received: header lines.
3656 20. If two routers both had headers_remove settings, and the first one included
3657 a superfluous trailing colon, the final name in the first list and the
3658 first name in the second list were incorrectly joined into one item (with a
3659 colon in the middle).
3665 1. Added -C and -D options to the exinext utility, mainly to make it easier
3666 to include in the automated testing, but these could be helpful when
3667 multiple configurations are in use.
3669 2. The exinext utility was not formatting the output nicely when there was
3670 an alternate port involved in the retry record key, nor when there was a
3671 message id as well (for retries that were specific to a specific message
3672 and a specific host). It was also confused by IPv6 addresses, because of
3673 the additional colons they contain. I have fixed the IPv4 problem, and
3674 patched it up to do a reasonable job for IPv6.
3676 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3677 delivery, the log line now contains "pipelined" if PIPELINING was used.
3679 4. An SMTP transport process used to panic and die if the bind() call to set
3680 an explicit outgoing interface failed. This has been changed; it is now
3681 treated in the same way as a connect() failure.
3683 5. A reference to $sender_host_name in the part of a conditional expansion
3684 that was being skipped was still causing a DNS lookup. This no longer
3687 6. The def: expansion condition was not recognizing references to header lines
3688 that used bh_ and bheader_.
3690 7. Added the _cache feature to named lists.
3692 8. The code for checking quota_filecount in the appendfile transport was
3693 allowing one more file than it should have been.
3695 9. For compatibility with Sendmail, the command line option
3703 and sets the incoming protocol and host name (for trusted callers). The
3704 host name and its colon can be omitted when only the protocol is to be set.
3705 Note the Exim already has two private options, -pd and -ps, that refer to
3706 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3707 "s", but I don't think that's a major issue.
3709 10. A number of refactoring changes to the code, none of which should affect
3712 (a) The number of logging options was getting close to filling up the
3713 32-bit word that was used as a bit map. I have split them into two classes:
3714 those that are passed in the argument to log_write(), and those that are
3715 only ever tested independently outside of that function. These are now in
3716 separate 32-bit words, so there is plenty of room for expansion again.
3717 There is no change in the user interface or the logging behaviour.
3719 (b) When building, for example, log lines, the code previously used a
3720 macro that called string_cat() twice, in order to add two strings. This is
3721 not really sufficiently general. Furthermore, there was one instance where
3722 it was actually wrong because one of the argument was used twice, and in
3723 one call a function was used. (As it happened, calling the function twice
3724 did not affect the overall behaviour.) The macro has been replaced by a
3725 function that can join an arbitrary number of extra strings onto a growing
3728 (c) The code for expansion conditions now uses a table and a binary chop
3729 instead of a serial search (which was left over from when there were very
3730 few conditions). Also, it now recognizes conditions like "pam" even when
3731 the relevant support is not compiled in: a suitably worded error message is
3732 given if an attempt is made to use such a condition.
3734 11. Added ${time_interval:xxxxx}.
3736 12. A bug was causing one of the ddress fields not to be passed back correctly
3737 from remote delivery subprocesses. The field in question was not being
3738 subsequently used, so this caused to problems in practice.
3740 13. Added new log selectors queue_time and deliver_time.
3742 14. Might have fixed a bug in maildirsizefile handling that threw up
3743 "unexpected character" debug warnings, and recalculated the data
3744 unnecessarily. In any case, I expanded the warning message to give more
3747 15. Added the message "Restricted characters in address" to the statements in
3748 the default ACL that block characters like @ and % in local parts.
3750 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3751 Three changes have been made:
3753 (a) There was a serious bug; a negative response to MAIL caused the whole
3754 recipient domain to be cached as invalid, thereby blocking all messages
3755 to all local parts at the same domain, from all senders. This bug has
3756 been fixed. The domain is no longer cached after a negative response to
3757 MAIL if the sender used is not empty.
3759 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3762 (c) A new callout option, "use_sender" has been added for people who want
3763 the modified behaviour.
3769 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3772 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3773 indeed breaks things for older releases.
3775 3. Added additional logging to the case where there is a problem reading data
3776 from a filter that is running in a subprocess using a pipe, in order to
3777 try to track down a specific problem.
3779 4. Testing facility fudge: when running in the test harness and attempting
3780 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3781 getting "No route to host". Convert this to a timeout.
3783 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3786 6. Some OS don't have socklen_t but use size_t instead. This affects the
3787 fifth argument of getsockopt() amongst other things. This is now
3788 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3789 can be set for individual OS. I have set it for SunOS5, OSF1, and
3790 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3791 some earlier ones do not.
3793 7. Change 4.30/15 was not doing the test caselessly.
3795 8. The standard form for an IPv6 address literal was being rejected by address
3796 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3797 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3798 this, as well as the form without the "IPv6" on the front (but only when
3799 address literals are enabled, of course).
3801 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3803 10. Exim crashed if a message with an empty sender address specified by -f
3804 encountered a router with an errors_to setting. This could be provoked only
3805 by a command such as
3809 where an empty string was supplied; "<>" did not hit this bug.
3811 11. Installed PCRE release 4.5.
3813 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3814 remained set. It is now erased.
3816 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3817 times from message ids (which are base 36 rather than the normal 62).
3819 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3820 were being counted as actual protocol errors, and logged if the log
3821 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3822 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3823 and DATA following a set of rejected RCPTs do not count as protocol errors.
3824 In other words, Exim assumes they were pipelined, though this may not
3825 actually be the case. Of course, in all cases the client gets an
3826 appropriate error code.
3828 15. If a lookup fails in an ACL condition, a message about the failure may
3829 be available; it is used if testing the ACL cannot continue, because most
3830 such messages specify what the cause of the deferral is. However, some
3831 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3832 that caused an old message to be retained and used if a later statement
3833 caused a defer, replacing the real cause of the deferral.
3835 16. If an IP address had so many PTR records that the DNS lookup buffer
3836 was not large enough to hold them, Exim could crash while trying to process
3837 the truncated data. It now detects and logs this case.
3839 17. Further to 4.21/58, another change has been made: if (and only if) the
3840 first line of a message (the first header line) ends with CRLF, a bare LF
3841 in a subsequent header line has a space inserted after it, so as not to
3842 terminate the header.
3844 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3845 unnecessarily, used atoi() instead of strtol(), and didn't check the
3846 termination when getting file sizes from file names by regex.
3848 19. Completely re-implemented the support for maildirsize files, in the light
3849 of a number of problems with the previous contributed implementation
3850 (4.30/29). In particular:
3852 . If the quota is zero, the maildirsize file is maintained, but no quota is
3855 . If the maildir directory does not exist, it is created before any attempt
3856 to write a maildirsize file.
3858 . The quota value in the file is just a cache; if the quota is changed in
3859 the transport, the new value overrides.
3861 . A regular expression is available for excluding directories from the
3864 20. The autoreply transport checks the characters in options that define the
3865 message's headers; it allows continued headers, but it was checking with
3866 isspace() after an embedded newline instead of explicitly looking for a
3869 21. If all the "regular" hosts to which an address was routed had passed their
3870 expiry times, and had not reached their retry times, the address was
3871 bounced, even if fallback hosts were defined. Now Exim should go on to try
3874 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3875 equivalent code in the SMTP transport. Some hosts send humungous responses
3876 to HELO/EHLO, more than 1024 it seems.
3878 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3879 gives compiler warnings in some environments. I've now done it "properly",
3882 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3883 (because of problems with the built-in one) was declared to return uschar *
3884 instead of char *, causing compiler failure.
3886 25. Fixed a file descriptor leak when processing alias/forward files.
3888 26. Fixed a minor format string issue in dbfn.c.
3890 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3892 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3893 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3896 29. When a sender address is verified, it is cached, to save repeating the test
3897 when there is more than one recipient in a message. However, when the
3898 verification involves a callout, it is possible for different callout
3899 options to be set for different recipients. It is too complicated to keep
3900 track of this in the cache, so now Exim always runs a verification when a
3901 callout is required, relying on the callout cache for the optimization.
3902 The overhead is duplication of the address routing, but this should not be
3905 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3906 to be verified with callout=postmaster, and the main callout worked but the
3907 postmaster check failed, the verification correctly failed. However, if a
3908 subsequent RCPT command asked for sender verification *without* the
3909 postmaster check, incorrect caching caused this verification also to fail,
3912 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3913 it was not caching the DNS options (qualify_single, search_parents) that
3914 were used when the lookup failed. A subsequent lookup with different
3915 options therefore always gave the same answer, though there were cases
3916 where it should not have. (Example: a "domains = !$mx_any" option on a
3917 dnslookup router: the "domains" option is always processed without any
3918 widening, but the router might have qualify_single set.) Now Exim uses the
3919 cached value only when the same options are set.
3921 32. Added John Jetmore's "exipick" utility to the distribution.
3923 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3924 than a timeout (e.g. a certificate is required, and is not provided), an
3925 Exim server now closes the connection immediately. Previously it waited for
3926 the client to close - but if the client is SSL, it seems that they each
3927 wait for each other, leading to a delay before one of them times out.
3929 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3930 maintained 0.8.x compatibility because I don't think many are using it, and
3931 it is clearly obsolete.
3933 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3936 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3937 client certificate was expired. A simple patch fixes this, though I don't
3938 understand the full logic of why the verify callback is called multiple
3941 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3942 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3943 which causes problems with some clients (such as the Certicom SSL Plus
3944 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3945 disables the coutermeasure allowing Eudora to connect."
3947 38. Exim was not checking that a write() to a log file succeeded. This could
3948 lead to Bad Things if a log got too big, in particular if it hit a file
3949 size limit. Exim now panics and dies if it cannot write to a log file, just
3950 as it does if it cannot open a log file.
3952 39. Modified OS/Makefile-Linux so that it now contains
3954 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3956 The two -D definitions ensure that Exim is compiled with large file
3957 support, which makes it possible to handle log files that are bigger than
3960 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3961 instance) a domain was checked against a named list that involved a lookup,
3962 causing $domain_data to be set, then another domain was checked against the
3963 same list, then the first domain was re-checked, the value of $domain_data
3964 after the final check could be wrong. In particular, if the second check
3965 failed, it could be set empty. This bug probably also applied to
3968 41. The strip_trailing_dot option was not being applied to the address given
3969 with the -f command-line option.
3971 42. The code for reading a message's header from the spool was incrementing
3972 $received_count, but never initializing it. This meant that the value was
3973 incorrect (doubled) while delivering a message in the same process in which
3974 it was received. In the most common configuration of Exim, this never
3975 happens - a fresh exec is done - but it can happen when
3976 deliver_drop_privilege is set.
3978 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3979 it now includes up to 150 characters of the unexpected data in the log
3982 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3983 and building data strings. The size of both of these buffers was 10 000
3984 bytes - far larger than anybody would *ever* want, thought I. Needless to
3985 say, somebody hit the limit. I have increased the maximum line length to
3986 20 000 and the maximum data length of concatenated lines to 100 000. I have
3987 also fixed two bugs, because there was no checking on these buffers. Tsk,
3988 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3989 buffer is too small.
3991 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3992 lsearch lookups. Now it does.
3994 46. When parsing a route_list item in a manualroute router, a fixed-length
3995 buffer was used for the list of hosts. I made this 1024 bytes long,
3996 thinking that nobody would ever have a list of hosts that long. Wrong.
3997 Somebody had a whole pile of complicated expansion conditions, and the
3998 string was silently truncated, leading to an expansion error. It turns out
3999 that it is easier to change to an unlimited length (owing to other changes
4000 that have happened since this code was originally written) than to build
4001 structure for giving a limitation error. The length of the item that
4002 expands into the list of hosts is now unlimited.
4004 47. The lsearch lookup could not handle data where the length of text line was
4005 more than 4095 characters. Such lines were truncated, leading to shortened
4006 data being returned. It should now handle lines of any length.
4008 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4009 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4012 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4013 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4014 no longer happens in comment lines. A list of replacements is now placed
4015 at the head of all of the source files, except those whose only change is
4016 to replace PERL_COMMAND in the very first #! line.
4018 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4019 messages on the queue, with a bottom-up merge sort, using code contributed
4020 by Michael Haardt. This should make operations like -bp somewhat faster on
4021 large queues. It won't affect queue runners, except when queue_run_in_order
4024 51. Installed eximstats 1.31 in the distribution.
4026 52. Added support for SRV lookups to the dnslookup router.
4028 53. If an ACL referred to $message_body or $message_body_end, the value was not
4029 reset for any messages that followed in the same SMTP session.
4031 54. The store-handling optimization for building very long strings was not
4032 differentiating between the different store pools. I don't think this
4033 actually made any difference in practice, but I've tidied it.
4035 55. While running the routers to verify a sender address, $sender_address
4036 was still set to the sender address. This is wrong, because when routing to
4037 send a bounce to the sender, it would be empty. Therefore, I have changed
4038 it so that, while verifying a sender address, $sender_address is set to <>.
4039 (There is no change to what happens when verifying a recipient address.)
4041 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4042 target A or AAAA records (if not already returned) without resetting the
4043 qualify_single or search_parents options of the DNS resolver. These are
4044 inappropriate in this case because the targets of MX and SRV records must
4045 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4046 target that, when qualified, matched something in the local domain. These
4047 two options are now turned off when doing these lookups.
4049 57. It seems that at least some releases of Reiserfs (which does not have the
4050 concept of a fixed number of inodes) returns zero and not -1 for the
4051 number of available inodes. This interacted badly with check_spool_inodes,
4052 which assumed that -1 was the "no such thing" setting. What I have done is
4053 to check that the total number of inodes is greater than zero before doing
4054 the test of how many are available.
4056 58. When a "warn" ACL statement has a log_message modifier, the message is
4057 remembered, and not repeated. This is to avoid a lot of repetition when a
4058 message has many recipients that cause the same warning to be written.
4059 Howewer, Exim was preserving the list of already written lines for an
4060 entire SMTP session, which doesn't seem right. The memory is now reset if a
4061 new message is started.
4063 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4064 debugging output unless log_rewrite was also set.
4066 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4067 of (int)(handle) when we know that handle contains (void *)(-1).
4069 61. The Exim daemon panic-logs an error return when it closes the incoming
4070 connection. However "connection reset by peer" seems to be common, and
4071 isn't really an error worthy of noting specially, so that particular error
4074 62. When Exim is trying to find all the local interfaces, it used to panic and
4075 die if the ioctl to get the interface flags failed. However, it seems that
4076 on at least one OS (Solaris 9) it is possible to have an interface that is
4077 included in the list of interfaces, but for which you get a failure error
4078 for this call. This happens when the interface is not "plumbed" into a
4079 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4080 failure of the "get flags" call assumes that the interface is down.
4082 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4083 makes life easier for people who are doing arithmetic on fields extracted
4084 from dates, where you often get leading zeros that should not be
4085 interpreted as octal.
4087 64. Added qualify_domain to the redirect router, to override the global
4090 65. If a pathologically long header line contained very many addresses (the
4091 report of this problem mentioned 10 000) and each of them was rewritten,
4092 Exim could use up a very large amount of memory. (It kept on making new
4093 copies of the header line as it rewrote, and never released the old ones.)
4094 At the expense of a bit more processing, the header rewriting function has
4095 been changed so that it no longer eats memory in this way.
4097 66. The generation of the Received: header has been moved from the time that a
4098 message starts to be received, to the time that it finishes. The timestamp
4099 in the Received: header should now be very close to that of the <= log
4100 line. There are two side-effects of this change:
4102 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4103 logged header lines no longer include the local Received: line, because
4104 it has not yet been created. The same applies to a copy of the message
4105 that is returned to a non-SMTP sender when a message is rejected.
4107 (b) When a filter file is tested using -bf, no additional Received: header
4108 is added to the test message. After some thought, I decided that this
4111 This change does not affect the value of $received_for. It is still set
4112 after address rewriting, but before local_scan() is called.
4114 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4116 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4117 gave an unhelpful panic error message, and a defer error. I have managed to
4118 change this behaviour so that it now rejects any supplied certificate,
4119 which seems right, as the list of acceptable certificates is empty.
4121 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4122 gave an unhelpful defer error. I have not managed to make this reject any
4123 supplied certificates, but the error message it gives is "no certificate
4124 supplied", which is not helpful.
4126 70. exigrep's output now also includes lines that are not associated with any
4127 message, but which match the given pattern. Implemented by a patch from
4128 Martin Sluka, which also tidied up the Perl a bit.
4130 71. Recipient callout verification, like sender verification, was using <> in
4131 the MAIL FROM command. This isn't really the right thing, since the actual
4132 sender may affect whether the remote host accepts the recipient or not. I
4133 have changed it to use the actual sender in the callout; this means that
4134 the cache record is now keyed on a recipient/sender pair, not just the
4135 recipient address. There doesn't seem to be a real danger of callout loops,
4136 since a callout by the remote host to check the sender would use <>.
4137 [SEE ABOVE: changed after hitting problems.]
4139 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4140 temporary errors. However, in the case of such a code being given after
4141 the end of a data transmission (i.e. after ".") Exim was failing to write
4142 a retry record for the message. (Yes, there was some broken host that was
4143 actually sending 8xx at this point.)
4145 73. An unknown lookup type in a host list could cause Exim to panic-die when
4146 the list was checked. (An example that provoked this was putting <; in the
4147 middle of a list instead of at the start.) If this happened during a DATA
4148 ACL check, a -D file could be left lying around. This kind of configuration
4149 error no longer causes Exim to die; instead it causes a defer errror. The
4150 incident is still logged to the main and panic logs.
4152 74. Buglet left over from Exim 3 conversion. The message "too many messages
4153 in one connection" was written to the rejectlog but not the mainlog, except
4154 when address rewriting (yes!) was being logged.
4156 75. Added write_rejectlog option.
4158 76. When a system filter was run not as root (that is, when system_filter_user
4159 was set), the values of the $n variables were not being returned to the
4160 main process; thus, they were not subsequently available in the $sn
4163 77. Added +return_path_on_delivery log selector.
4165 78. A connection timeout was being treated differently from recipients deferred
4166 when testing hosts_max_try with a message that was older than the host's
4167 retry timeout. (The host should not be counted, thus allowing all hosts to
4168 be tried at least once before bouncing.) This may have been the cause of an
4169 occasionally reported bug whereby a message would remain on the queue
4170 longer than the retry timeout, but would be bounced if a delivery was
4171 forced. I say "may" because I never totally pinned down the problem;
4172 setting up timeout/retry tests is difficult. See also the next item.
4174 79. The ultimate address timeout was not being applied to errors that involved
4175 a combination of host plus message (for example, a timeout on a MAIL
4176 command). When an address resolved to a number of possible hosts, and they
4177 were not all tried for each delivery (e.g. because of hosts_max_try), a
4178 message could remain on the queue longer than the retry timeout.
4180 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4183 81. Fixed an obscure SMTP outgoing bug which required at least the following
4184 conditions: (a) there was another message waiting for the same server;
4185 (b) the server returned 5xx to all RCPT commands in the first message so
4186 that the message was not completed; (c) the server dropped the connection
4187 or gave a negative response to the RSET that Exim sends to abort the
4188 transaction. The observed case was a dropped connection after DATA that had
4189 been sent in pipelining mode. That is, the server had advertised PIPELINING
4190 but was not implementing it correctly. The effect of the bug was incorrect
4191 behaviour, such as trying another host, and this could lead to a crash.
4197 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4198 and daemon.c were passed as pointers to ints; they should have been
4199 pointers to socklen_t variables (which are typically unsigned ints).
4201 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4204 3. Fixed a really odd bug that affected only the testing scheme; patching a
4205 certain fixed string in the binary changed the value of another string that
4206 happened to be identical to the end of the original first string.
4208 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4209 name", it returns that address as the IP address. On some operating
4210 systems (e.g. Solaris), it also passes back the IP address string as the
4211 "host name". However, on others (e.g. Linux), it passes back an empty
4212 string. Exim wasn't checking for this, and was changing the host name to an
4213 empty string, assuming it had been canonicized.
4215 5. Although rare, it is permitted to have more than one PTR record for a given
4216 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4217 all the names associated with an address, because they do in Solaris.
4218 However, it seems that they do not in Linux for data that comes from the
4219 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4220 I found this out when I moved to a new Linux workstation and tried to run
4221 the Exim test suite.
4223 To get round this problem I have changed the code so that it now does its
4224 own call to the DNS to look up PTR records when searching for a host name.
4225 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4226 addresses that are only in /etc/hosts are still found.
4228 This behaviour is, however, controlled by an option called host_lookup_
4229 order, which defaults to "bydns:byaddr". If people want to use the other
4230 order, or indeed, just use one or the other means of lookup, they can
4231 specify it in this variable.
4233 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4234 some operating systems, this comes back from gethostbyaddr() as an empty
4235 string, and this is what Exim used to test for. However, it seems that in
4236 other systems, "." is yielded. Exim now tests for this case too.
4238 7. The values of check_spool_space and check_log_space are now held internally
4239 as a number of kilobytes instead of an absolute number of bytes. If a
4240 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4241 kilobyte. This means that much larger values can be stored.
4243 8. Exim monitor: an attempt to get the action menu when not actually pointing
4244 at a message produces an empty menu entitled "No message selected". This
4245 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4246 no entries in it ("Shell widget menu has zero width and/or height"). So I
4247 have added a single, blank menu entry in this case.
4249 9. Added ${quote_local_part.
4251 10. MIME decoding is now applied to the contents of Subject: header lines when
4254 11. Now that a reference to $sender_host_address automatically causes a reverse
4255 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4256 host lookup before query-style lookups in lists that might use this
4257 variable. This has therefore been abolished, and the "net-" prefix is no
4258 longer necessary for query-style lookups.
4260 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4261 have been a typo for LFLAGS, so it has been changed.
4263 13. The install script calls Exim with "-C /dev/null" in order to find the
4264 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4265 to be output. Howeve, since Exim outputs its version number before the
4266 error, it didn't break the script. It just looked ugly. I fixed this by
4267 always allowing "-C /dev/null" if the caller is root.
4269 14. Ignore overlarge ACL variable number when reading spool file - insurance
4270 against a later release with more variables having written the file.
4272 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4273 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4274 this, as well as the form without the "IPv6" on the front.
4276 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4277 OS/Makefile-Darwin file.
4279 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4280 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4282 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4283 variable. However, some OS use ints for inodes. I've added cast to long int
4284 to get rid of the compiler warning.
4286 19. I had forgotten to lock out "/../" in configuration file names when
4287 ALT_CONFIG_PREFIX was set.
4289 20. Routers used for verification do not need to specify transports. However,
4290 if such a router generated a host list, and callout was configured, Exim
4291 crashed, because it could not find a port number from the (non-existent)
4292 transport. It now assumes port 25 in this circumstance.
4294 21. Added the -t option to exigrep.
4296 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4297 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4298 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4299 them is set, LOOKUP_LSEARCH is forced.
4301 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4303 24. Added sender and host information to the "rejected by local_scan()" log
4304 line; previously there was no indication of these.
4306 25. Added .include_if_exists.
4308 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4309 Linux. It turns out that not all file systems support this. Apparently some
4310 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4311 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4312 not supported on the file descriptor, is now ignored when Exim is trying to
4313 sync a directory. This applies only to Linux.
4315 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4317 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4318 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4321 29. Added support for maildirsize files from supplied patch (modified a bit).
4323 30. The use of :fail: followed by an empty string could lead Exim to respond to
4324 sender verification failures with (e.g.):
4326 550 Verification failed for <xxx>
4327 550 Sender verify failed
4329 where the first response line was missing the '-' that indicates it is not
4330 the final line of the response.
4332 31. The loop for finding the name of the user that called Exim had a hardwired
4333 limit of 10; it now uses the value of finduser_retries, which is used for
4334 all other user lookups.
4336 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4339 33. Exim was neglecting to zero errno before one call of strtol() when
4340 expanding a string and expecting an integer value. On some systems this
4341 resulted in spurious "integer overflow" errors. Also, it was casting the
4342 result into an int without checking.
4344 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4345 did not work. The code looks as if it has *never* worked, though it appears
4346 to have been documented since at least releast 1.62. I have made it work.
4348 35. The "timeout_DNS" error in retry rules, also documented since at least
4349 1.62, also never worked. As it isn't clear exactly what this means, and
4350 clearly it isn't a major issue, I have abolished the feature by treating it
4351 as "timeout", and writing a warning to the main and panic logs.
4353 36. The display of retry rules for -brt wasn't always showing the error code
4356 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4357 timeout_connect_A, timeout_connect_MX.
4359 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4360 to the empty sender.
4362 39. The daemon was not analysing the content of -oX till after it had closed
4363 stderr and disconnected from the controlling terminal. This meant that any
4364 syntax errors were only noted on the panic log, and the return code from
4365 the command was 0. By re-arranging the code a little, I've made the
4366 decoding happen first, so such errors now appear on stderr, and the return
4367 code is 1. However, the actual setting up of the sockets still happens in
4368 the disconnected process, so errors there are still only recorded on the
4371 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4372 connections (as happens on some IP stacks) was logged at start up time as
4373 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4374 it from "IPv6 and IPv4", which means that two separate sockets are being
4377 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4378 says whether AF_INET or AF_INET6 was passed as an argument.
4380 42. Exiwhat output was messed up when time zones were included in log
4383 43. Exiwhat now gives more information about the daemon's listening ports,
4384 and whether -tls-on-connect was used.
4386 44. The "port" option of the smtp transport is now expanded.
4388 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4389 silently ignored. Now an error message is written to the main and panic
4392 46. There's a new ACL modifier called "logwrite" which writes to a log file
4393 as soon as it is encountered.
4395 47. Added $local_user_uid and $local_user_gid at routing time.
4397 48. Exim crashed when trying to verify a sender address that was being
4400 49. Exim was recognizing only a space character after ".include". It now also
4401 recognizes a tab character.
4403 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4404 extracting the relevant information from the specification. The man page no
4405 longer contains scrambled data for the -d option, and I've added a section
4406 at the front about calling Exim under different names.
4408 51. Added "extra_headers" argument to the "mail" command in filter files.
4410 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4413 53. Installed eximstats 1.29.
4415 54. Added transport_filter_timeout as a generic transport option.
4417 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4418 Cc: header lines. This was required by RFC 822, but it not required by RFC
4421 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4422 incoming messages that did not have them. Now it does so only if the
4423 message originates locally, that is, if there is no associated remote host
4424 address. When Resent- header lines are present, this applies to the Resent-
4425 lines rather than the non-Resent- lines.
4427 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4428 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4430 58. Messages for configuration errors now include the name of the main
4431 configuration file - useful now that there may be more than one file in a
4432 list (.included file names were always shown).
4434 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4435 for those rare installations that do not start the daemon as root or run it
4436 setuid root. I've cut out the call to initgroups() if the daemon is not
4439 60. The Exim user and group can now be bound into the binary as text strings
4440 that are looked up at the start of Exim's processing.
4442 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4444 62. Added $mailstore_basename variable.
4446 63. Installed patch to sieve.c from Michael Haardt.
4448 64. When Exim failed to open the panic log after failing to open the main log,
4449 the original message it was trying to log was written to stderr and debug
4450 output, but if they were not available (the usual case in production), it
4451 was lost. Now it is written to syslog before the two lines that record the
4452 failures to open the logs.
4454 65. Users' Exim filters run in subprocesses under the user's uid. It is
4455 possible for a "deliver" command or an alias in a "personal" command to
4456 provoke an address rewrite. If logging of address rewriting is configured,
4457 this fails because the process is not running as root or exim. There may be
4458 a better way of dealing with this, but for the moment (because 4.30 needs
4459 to be released), I have disabled address rewrite logging when running a
4460 filter in a non-root, non-exim process.
4466 1. The buildconfig auxiliary program wasn't quoting the value set for
4467 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4468 not defined. This bug was masked in 4.22 by the effect that was fixed in
4471 2. Some messages that were rejected after a message id was allocated were
4472 shown as "incomplete" by exigrep. It no longer does this for messages that
4473 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4475 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4476 have allow_domain_literals set, the ID did not get logged in the <= line.
4477 Domain literals are now always recognized in Message-ID: header lines.
4479 4. The first argument for a ${extract expansion item is the key name or field
4480 number. Leading and trailing spaces in this item were not being ignored,
4481 causing some misleading effects.
4483 5. When deliver_drop_privilege was set, single queue runner processes started
4484 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4485 same command in the process it spins off) were not dropping privilege.
4487 6. When the daemon running as "exim" started a queue runner, it always
4488 re-executed Exim in the spun-off process. This is a waste of effort when
4489 deliver_drop_privilege is set. The new process now just calls the
4490 queue-runner function directly.
4496 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4499 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4500 ignored. Though the use of -oP was forcing the writing of a pid file, it
4501 was always written to the default place.
4503 3. If the message "no IP address found for host xxxx" is generated during
4504 incoming verification, it is now followed by identification of the incoming
4505 connection (so you can more easily find what provoked it).
4507 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4509 5. Added some features to "harden" Exim a bit more against certain attacks:
4511 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4512 be put in Local/Makefile. This is like the never_users runtime option,
4513 but it cannot be overridden. The default setting is "root".
4515 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4516 prefix string with which any file named in a -C command line option
4519 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4520 is retained for -C and -D only if the caller of Exim is root. Without
4521 it, the exim user may also use -C and -D and retain privilege.
4523 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4524 command line option is disabled.
4526 6. Macro names set by the -D option must start with an upper case letter, just
4527 like macro names defined in the configuration file.
4529 7. Added "dereference=" facility to LDAP.
4531 8. Two instances of the typo "uknown" in the source files are fixed.
4533 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4534 the Configure-Makefile script screwed up while processing it.
4536 10. Incorporated PCRE 4.4.
4538 11. The SMTP synchronization check was not operating right at the start of an
4539 SMTP session. For example, it could not catch a HELO sent before the client
4540 waited for the greeting. There is now a check for outstanding input at the
4541 point when the greeting is written. Because of the duplex, asynchronous
4542 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4543 way, but not yet received, when the check is performed.
4545 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4546 on TCP/IP sockets, because this apparently causes some broken clients to
4549 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4550 unchanged) from the Cygwin maintainer.
4552 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4553 was supported instead of testing for mbx. Effectively a typo.
4555 15. The spa authenticator server code was not checking that the input it
4556 received was valid base64.
4558 16. The debug output line for the "set" modifier in ACLs was not showing the
4559 name of the variable that was being set.
4561 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4563 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4564 lookup on its own. Something else was needed to trigger the lookup. For
4565 example, a match in host_lookup or the need for a host name in a host list.
4566 Now, if $sender_host_name is referenced and the host name has not yet been
4567 looked up, a lookup is performed. If the lookup fails, the variable remains
4568 empty, and $host_lookup_failed is set to "1".
4570 19. Added "eqi" as a case-independent comparison operator.
4572 20. The saslauthd authentication condition could segfault if neither service
4573 nor realm was specified.
4575 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4576 error message that was logged was misleading, and incoming SMTP
4577 connections were dropped. The message is now more accurate, and temporary
4578 errors are given to SMTP connections.
4580 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4581 (except RSET) with a 421 error, until QUIT is received. However, it was
4582 failing to send a response to QUIT.
4584 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4585 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4586 ACL is now run after the helo_try_verify_hosts code.
4588 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4589 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4590 case-independent, so other case variants are also recognized). Apparently
4591 some systems use these upper case variants.
4593 25. If more than two messages were waiting for the same host, and a transport
4594 filter was specified for the transport, Exim sent two messages over the
4595 same TCP/IP connection, and then failed with "socket operation on non-
4596 socket" when it tried to send the third.
4598 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4600 27. The extern definition of crypt16() in expand.c was not being excluded when
4601 the OS had its own crypt16() function.
4603 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4604 as a preferred synonym for return_size_limit, both as an option and as an
4607 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4609 30. Changed the default configuration ACL to relax the local part checking rule
4610 for addresses that are not in any local domains. For these addresses,
4611 slashes and pipe symbols are allowed within local parts, but the sequence
4612 /../ is explicitly forbidden.
4614 31. SPA server authentication was not clearing the challenge buffer before
4617 32. log_message in a "warn" ACL statement was writing to the reject log as
4618 well as to the main log, which contradicts the documentation and doesn't
4619 seem right (because no rejection is happening). So I have stopped it.
4621 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4622 However, I am unable to do any testing of this.
4624 34. Fixed an infelicity in the appendfile transport. When checking directories
4625 for a mailbox, to see if any needed to be created, it was accidentally
4626 using path names with one or more superfluous leading slashes; tracing
4627 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4629 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4630 contents are added to the log line that is written for every discarded
4631 recipient. (Previously a log_message setting was ignored.)
4633 36. The ${quote: operator now quotes the string if it is empty.
4635 37. The install script runs exim in order to find its version number. If for
4636 some reason other than non-existence or emptiness, which it checks, it
4637 could not run './exim', it was installing it with an empty version number,
4638 i.e. as "exim-". This error state is now caught, and the installation is
4641 38. An argument was missing from the function that creates an error message
4642 when Exim fails to connect to the socket for saslauthd authentication.
4643 This could cause Exim to crash, or give a corrupted message.
4645 39. Added isip, isip4, and isip6 to ${if conditions.
4647 40. The ACL variables $acl_xx are now saved with the message, and can be
4648 accessed later in routers, transports, and filters.
4650 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4651 strings in the file are not string-expanded.
4653 42. If a MAIL command specified a SIZE value that was too large to fit into an
4654 int variable, the check against message_size_limit failed. Such values are
4655 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4656 day this will have to be increased, but I don't think I want to be around
4657 when emails are that large.
4664 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4665 iconv() is not standard in FreeBSD.
4667 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4668 IPv6 enabled. The observed symptom was a segmentation fault on return from
4669 the function os_common_find_running_interfaces() in src/os.c.
4671 3. In the check_special_case() function in daemon.c I had used "errno" as an
4672 argument name, which causes warnings on some systems. This was basically a
4673 typo, since it was named "eno" in the comments!
4675 4. The code that waits for the clock to tick (at a resolution of some fraction
4676 of a second) so as to ensure message-id uniqueness was always waiting for
4677 at least one whole tick, when it could have waited for less. [This is
4678 almost certainly not relevant at current processor speeds, where it is
4679 unlikely to ever wait at all. But we try to future-proof.]
4681 5. The function that sleeps for a time interval that includes fractions of a
4682 second contained a race. It did not block SIGALRM between setting the
4683 timer, and suspending (a couple of lines later). If the interval was short
4684 and the sigsuspend() was delayed until after it had expired, the suspension
4685 never ended. On busy systems this could lead to processes getting stuck for
4688 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4689 process, before it forks any delivery processes. The open lookup caching
4690 mechanism meant that the open file or database connection was passed into
4691 the delivery process. The problem was that delivery processes always tidy
4692 up cached lookup data. This could cause a problem for the next delivery
4693 process started by the queue runner, because the external queue runner
4694 process does not know about the closure. So the next delivery process
4695 still has data in the lookup cache. In the case of a file lookup, there was
4696 no problem because closing a file descriptor in a subprocess doesn't affect
4697 the parent. However, if the lookup was caching a connection to a database,
4698 the connection was closed, and the second delivery process was likely to
4699 see errors such as "PGSQL: query failed: server closed the connection
4700 unexpectedly". The problem has been fixed by closing all cached lookups
4701 in a queue runner before running a delivery process.
4703 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4704 seem to have the "const" qualifier which it has on other OS. I've
4707 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4708 *of the same type* (client or server) with the same public name that an
4709 error should be diagnosed.
4711 9. When Exim looked up a host name for an IP address, but failed to find the
4712 original IP address when looking up the host name (a safety check), it
4713 output the message "<ip address> does not match any IP for NULL", which was
4714 confusing, to say the least. The bug was that the host name should have
4715 appeared instead of "NULL".
4717 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4718 user that is built into the binary, and the -C or -D options is used, root
4719 privilege is dropped before the configuration file is read. In addition,
4720 logging is switched to stderr instead of the normal log files. If the
4721 configuration then re-defines the Exim user, the unprivileged environment
4722 is probably not what is expected, so Exim logs a panic warning message (but
4725 However, if deliver_drop_privilege is set, the unprivileged state may well
4726 be exactly what is intended, so the warning has been cut out in that case,
4727 and Exim is allowed to try to write to its normal log files.
4733 1. smtp_return_error_details was not giving details for temporary sender
4734 or receiver verification errors.
4736 2. Diagnose a configuration error if two authenticators have the same public
4739 3. Exim used not to create the message log file for a message until the first
4740 delivery attempt. This could be confusing when incoming messages were held
4741 for policy or load reasons. The message log file is now created at the time
4742 the message is received, and an initial "Received" line is written to it.
4744 4. The automatically generated man page for command line options had a minor
4745 bug that caused no ill effects; however, a more serious problem was that
4746 the procedure for building the man page automatically didn't always
4747 operate. Consequently, release 4.20 contains an out-of-date version. This
4748 shouldn't happen again.
4750 5. When building Exim with embedded Perl support, the script that builds the
4751 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4752 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4754 6. The freeze_tell option was not being used for messages that were frozen on
4755 arrival, either by an ACL or by local_scan().
4757 7. Added the smtp_incomplete_transaction log selector.
4759 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4760 was accepting AUTH without a new EHLO.
4762 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4763 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4766 10. Logging of TCP/IP connections (when configured) now happens in the main
4767 daemon process instead of the child process, so that the TCP/IP connection
4768 count is more accurate (but it can never be perfect).
4770 11. The use of "drop" in a nested ACL was not being handled correctly in the
4771 outer ACL. Now, if condition failure induced by the nested "drop" causes
4772 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4773 or "require"), the connection is dropped.
4775 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4776 that yield "discard" can now be used with an "accept" or a "discard" verb,
4777 but an error is generated for any others (because I can't see a useful way
4778 to define what should happen).
4780 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4781 are now processed in the same way as lines in the Exim configuration file.
4782 In particular, continuation lines are supported.
4784 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4786 15. Added -ti meaning -t -i.
4788 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4789 domains, and warn by logging if others are found.
4791 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4792 returned from GIFCONF when Exim is trying to find the list of interfaces on
4793 a host. The code in os.c has been modified to copy each ifreq to an aligned
4794 structure in all cases.
4796 Also, in some cases, the returned ifreq's were being copied to a 'struct
4797 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4798 means the last couple of bytes of an IPv6 address could be chopped if the
4799 ifreq contained only a normal sockaddr (14 bytes storage).
4801 18. Named domain lists were not supported in the hosts_treat_as_local option.
4802 An entry such as +xxxx was not recognized, and was treated as a literal
4805 19. Ensure that header lines added by a DATA ACL are included in the reject log
4806 if the ACL subsequently rejects the message.
4808 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4809 MD5 (which is deprecated).
4811 21. When testing a filter file using -bf, Exim was writing a message when it
4812 took the sender from a "From " line in the message, but it was not doing so
4813 when it took $return_path from a Return-Path: header line. It now does.
4815 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4816 with a valid header line field name (a series of printing characters
4817 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4819 23. Changed "disc" in the source to "disk" to conform to the documentation and
4820 the book and for uniformity.
4822 24. Ignore Sendmail's -Ooption=value command line item.
4824 25. When execve() failed while trying to run a command in a pipe transport,
4825 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4826 could be confused with a return value of 69 from the command itself. This
4827 has been changed to 127, the value the shell returns if it is asked to run
4828 a non-existent command. The wording for the related log line suggests a
4829 non-existent command as the problem.
4831 26. If received_header_text expands to an empty string, do not add a Received:
4832 header line to the message. (Well, it adds a token one on the spool, but
4833 marks it "old" so that it doesn't get used or transmitted.)
4835 27. Installed eximstats 1.28 (addition of -nt option).
4837 28. There was no check for failure on the call to getsockname() in the daemon
4838 code. This can fail if there is a shortage of resources on the system, with
4839 ENOMEM, for example. A temporary error is now given on failure.
4841 29. Contrary to the C standard, it seems that in some environments, the
4842 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4843 program. Exim now does this explicitly; it affects the formatting of
4844 timestamps using strftime().
4846 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4847 complaints. I've now initialized all the variables, to avoid this.
4849 32. Header lines added by a system filter were not being "seen" during
4850 transport-time rewrites.
4852 33. The info_callback() function passed to OpenSSL is set up with type void
4853 (*)(SSL *, int, int), as described somewhere. However, when calling the
4854 function (actually a macro) that sets it up, the type void(*)() is
4855 expected. I've put in a cast to prevent warnings from picky compilers.
4857 34. If a DNS black list lookup found a CNAME record, but there were no A
4858 records associated with the domain it pointed at, Exim crashed.
4860 35. If a DNS black list lookup returned more than one A record, Exim ignored
4861 all but the first. It now scans all returned addresses if a particular IP
4862 value is being sought. In this situation, the contents of the
4863 $dnslist_value variable are a list of all the addresses, separated by a
4866 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4867 to accept a host name and all its aliases if the forward lookup for any of
4868 them yielded the IP address of the incoming connection. Now it accepts only
4869 those names whose forward lookup yields the correct IP address. Any other
4870 names are discarded. This closes a loophole whereby a rogue DNS
4871 administrator could create reverse DNS records to break through a
4872 wildcarded host restriction in an ACL.
4874 37. If a user filter or a system filter that ran in a subprocess used any of
4875 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4876 the wrong values were passed to the pipe command ($thisaddress had the
4877 value of $0, $0 had the value of $1, etc). This bug was introduced by
4878 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4880 38. Improved the line breaking for long SMTP error messages from ACLs.
4881 Previously, if there was no break point between 40 and 75 characters, Exim
4882 left the rest of the message alone. Two changes have been made: (a) I've
4883 reduced the minimum length to 35 characters; (b) if it can't find a break
4884 point between 35 and 75 characters, it looks ahead and uses the first one
4885 that it finds. This may give the occasional overlong line, but at least the
4886 remaining text gets split now.
4888 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4889 file descriptors might be low, and that setting 1000 would always raise it.
4890 It turns out that in some environments, the limit is already over 1000 and
4891 that lowering it causes trouble. So now Exim takes care not to decrease it.
4893 40. When delivering a message, the value of $return_path is set to $sender_
4894 address at the start of routing (routers may change the value). By an
4895 oversight, this default was not being set up when an address was tested by
4896 -bt or -bv, which affected the outcome if any router or filter referred to
4899 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4900 log only when "message" or "log_message" are set. However, if one of the
4901 conditions was an address verification, or a call to a nested ACL, the
4902 messages generated by the underlying test were being passed through. This
4903 no longer happens. The underlying message is available in $acl_verify_
4904 message for both "message" and "log_message" expansions, so it can be
4905 passed through if needed.
4907 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4908 new expansion $bh_ to give the encoded byte string without charset
4909 translation. Translation happens only if iconv() is available; HAVE_ICONV
4910 indicates this at build time. HEADERS_CHARSET gives the charset to
4911 translate to; headers_charset can change it in the configuration, and
4912 "headers charset" can change it in an individual filter file.
4914 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4915 that creates RFC 2047 encoded "words" labels them as that charset instead
4916 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4917 expansion operator; (ii) when Exim creates a From: line for a local
4918 message; (iii) when a header line is rewritten to include a "phrase" part.
4920 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4921 buggy, causing it to skip the first lines of messages whose message ID
4922 ended in 'D'. This would not have bitten before Exim release 4.14, because
4923 message IDs were unlikely to end in 'D' before then. The effect was to have
4924 incorrect size information for certain domains.
4926 45. #include "config.h" was missing at the start of the crypt16.c module. This
4927 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4930 46. If there was a timeout during a "random" callout check, Exim treated it as
4931 a failure of the random address, and carried on sending RSET and the real
4932 address. If the delay was just some slowness somewhere, the response to the
4933 original RCPT would be taken as a response to RSET and so on, causing
4934 mayhem of various kinds.
4936 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4937 when I implemented it. It didn't allow for the fact that some option values
4938 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4939 the right test for positive values.
4941 48. Domain names in DNS records are case-independent. Exim always looks them up
4942 in lower case. Some resolvers return domain names in exactly the case they
4943 appear in the zone file, that is, they may contain uppercase letters. Not
4944 all resolvers do this - some return always lower case. Exim was treating a
4945 change of case by a resolver as a change of domain, similar to a widening
4946 of a domain abbreviation. This triggered its re-routing code and so it was
4947 trying to route what was effectively the same domain again. This normally
4948 caused routing to fail (because the router wouldn't handle the domain
4949 twice). Now Exim checks for this case specially, and just changes the
4950 casing of the domain that it ultimately uses when it transmits the message
4953 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4956 50. If a filter generated a file delivery with a non-absolute name (possible if
4957 no home directory exists for the router), the forbid_file option was not
4960 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4961 the existing equality matching.
4963 52. Exim was using ints instead of ino_t variables in some places where it was
4964 dealing with inode numbers.
4966 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4967 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4968 called TMPDIR, and if it finds it is different, it changes its value.
4970 54. The smtp_printf() function is now made available to local_scan() so
4971 additional output lines can be written before returning. There is also an
4972 smtp_fflush() function to enable the detection of a dropped connection.
4973 The variables smtp_input and smtp_batched_input are exported to
4976 55. Changed the default runtime configuration: the message "Unknown user"
4977 has been removed from the ACL, and instead placed on the localuser router,
4978 using the cannot_route_message feature. This means that any verification
4979 failures that generate their own messages won't get overridden. Similarly,
4980 the "Unrouteable address" message that was in the ACL for unverifiable
4981 relay addresses has also been removed.
4983 56. Added hosts_avoid_esmtp to the smtp transport.
4985 57. The exicyclog script was not checking for the esoteric option
4986 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4987 will work only if exicyclog is run under the appropriate euid.
4989 58. Following a discussion on the list, the rules by which Exim recognises line
4990 endings on incoming messages have been changed. The -dropcr and drop_cr
4991 options are now no-ops, retained only for backwards compatibility. The
4992 following line terminators are recognized: LF CRLF CR. However, special
4993 processing applies to CR:
4995 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4996 nor a local message in the state where . is a terminator.
4998 (ii) If a bare CR is encountered in a header line, an extra space is added
4999 after the line terminator so as not to end the header. The reasoning
5000 behind this is that bare CRs in header lines are most likely either
5001 to be mistakes, or people trying to play silly games.
5003 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5004 was being incorrectly given as 18 bytes larger than it should have been.
5005 This is a VOB (very old bug).
5007 60. This may never have affected anything current, but just in case it has:
5008 When the local host is found other than at the start of a list of hosts,
5009 the local host, those with the same MX, and any that follow, are discarded.
5010 When the list in question was part of a longer list of hosts, the following
5011 hosts (not currently being processed) were also being discarded. This no
5012 longer happens. I'm not sure if this situation could ever has previously
5015 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5018 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5019 header. This is something that is recommended in a new Internet Draft, and
5020 is something that is documented as being done by Sendmail. There are two
5021 possible values. For messages generated by the autoreply transport, Exim
5024 Auto-Submitted: auto-replied
5026 whereas for all other generated messages (e.g. bounces) it adds
5028 Auto-Submitted: auto-generated
5030 63. The "personal" condition in filters now includes a test for the
5031 Auto-Submitted: header. If it contains the string "auto-" the message it
5032 not considered personal.
5034 64. Added rcpt_include_affixes as a generic transport option.
5036 65. Added queue_only_override (default true).
5038 66. Added the syslog_duplication option.
5040 67. If what should have been the first header line of a message consisted of
5041 a space followed by a colon, Exim was mis-interpreting it as a header line.
5042 It isn't of course - it is syntactically invalid and should therefore be
5043 treated as the start of the message body. The misbehaviour could have
5044 caused a number of strange effects, including loss of data in subsequent
5045 header lines, and spool format errors.
5047 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5048 client host had authenticated. This control can now be exercised by an ACL
5049 for more flexibility.
5051 69. By default, callouts do not happen when testing with -bh. There is now a
5052 variant, -bhc, which does actually run the callout code, including
5053 consulting and updating the callout cache.
5055 70. Added support for saslauthd authentication, courtesy of Alexander
5058 71. If statvfs() failed on the spool or log directories while checking their
5059 size for availability, Exim confusingly gave the error "space shortage".
5060 Furthermore, in debugging mode it crashed with a floating point exception.
5061 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5062 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5063 serious problem, Exim now writes to the main and panic logs when this
5064 happens, with details of the failure. It then refuses to accept the
5065 incoming message, giving the message "spool directory problem" or "log
5066 directory problem" with a 421 code for SMTP messages.
5068 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5069 0, 1, and 2 exist, because some OS complain for execs without them (see
5070 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5071 descriptors. However, the code omitted to check that the open succeeded,
5072 causing mysterious errors if for some reason the permissions on /dev/null
5073 got screwed. Now Exim writes a message to the main and panic logs, and
5074 bombs out if it can't open /dev/null.
5076 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5077 interact so that it is all more flexible. It is supposed to remain
5078 backwards compatible. Also added extra_local_interfaces.
5080 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5081 to bomb out with an assertion failure - to the client this appears as a
5082 connection drop. This problem occurs in the part of the code that was taken
5083 from the Samba project. Fortunately, the assertion is in a very simple
5084 function, so I have fixed this by reproducing the function inline in the
5085 one place where it is called, and arranging for authentication to fail
5086 instead of killing the process with assert().
5088 75. The SPA client code was not working when the server requested OEM rather
5089 than Unicode encoding.
5091 76. Added code to make require_files with a specific uid setting more usable in
5092 the case where statting the file as root fails - usually a non-root-mounted
5093 NFS file system. When this happens and the failure is EACCES, Exim now
5094 forks a subprocess and does the per-uid checking as the relevant uid.
5096 77. Added process_log_path.
5098 78. If log_file_path was not explicitly set, a setting of check_log_space or
5099 check_log_inodes was ignored.
5101 79. If a space check for the spool or log partitions fails, the incident is now
5102 logged. Of course, in the latter case the data may get lost...
5104 80. Added the %p formatting code to string_format() so that it can be used to
5105 print addresses in debug_print(). Adjusted all the address printing in the
5106 debugging in store.c to use %p rather than %d.
5108 81. There was a concern that a line of code in smtp_in.c could overflow a
5109 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5110 initially expressed, the concern was not well-founded, because trailing
5111 spaces are removed early. However, if the trailing spaces were followed by
5112 a NULL, they did not get removed, so the overflow was possible. Two fixes
5115 (a) I re-wrote the offending code in a cleaner fashion.
5116 (b) If an incoming SMTP command contains a NULL character, it is rejected
5119 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5120 runs initgroups(), so that if the Exim user is in any additional groups,
5121 they will be used during message reception.
5127 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob